Discovery of novel glycerolated quinazolinones from Streptomyces sp. MBT27
- 84 Downloads
Actinobacteria are a major source of novel bioactive natural products. A challenge in the screening of these microorganisms lies in finding the favorable growth conditions for secondary metabolite production and dereplication of known molecules. Here, we report that Streptomyces sp. MBT27 produces 4-quinazolinone alkaloids in response to elevated levels of glycerol, whereby quinazolinones A (1) and B (2) form a new sub-class of this interesting family of natural products. Global Natural Product Social molecular networking (GNPS) resulted in a quinazolinone-related network that included anthranilic acid (3), anthranilamide (4), 4(3H)-quinazolinone (5), and 2,2-dimethyl-1,2-dihydroquinazolin-4(3H)-one (6). Actinomycins D (7) and X2 (8) were also identified in the extracts of Streptomyces sp. MBT27. The induction of quinazolinone production by glycerol combined with biosynthetic insights provide evidence that glycerol is integrated into the chemical scaffold. The unprecedented 1,4-dioxepane ring, that is spiro-fused into the quinazolinone backbone, is most likely formed by intermolecular etherification of two units of glycerol. Our work underlines the importance of varying the growth conditions for the discovery of novel natural products and for understanding their biosynthesis.
Keywords4(3H)-quinazolinones Molecular networking Metabolomics Carbon source
Actinobacteria are a major source of bioactive compounds, producing some two-thirds of all antibiotics as well as molecules with a wide variety of activities such as anticancer, antifungal and immunosuppressant [2, 4]. Traditionally, microbial natural product (NP) discovery has been done via high-throughput screening followed by iterative bioassay-guided fractionation and structure elucidation. While such pipelines were extremely successful and delivered a plethora of therapeutic agents, in the modern era the large pharmaceutical companies moved out of NP-discovery programs due to high cost and chemical redundancy [1, 9, 29]. At the same time, the power of genome sequencing brought the recognition that microorganisms harbor a vast and yet untapped biosynthetic potential, and it rapidly became clear that the potential for metabolic diversity of even the best-studied model organisms as producers of natural products had been grossly underestimated [3, 11, 19]. How could these compounds have been missed by the very extensive HT screening campaigns of the twentieth century? The answer is that many of the biosynthetic gene clusters (BGCs) discovered by genome mining are poorly expressed or cryptic under laboratory conditions [22, 28]. A drug-discovery pipeline that is rapidly gaining momentum involves combining genome mining with fluctuating the culturing conditions to achieve differential synthesis of NPs, followed by the metabolic profiling-based identification of the bioactivity of interest [13, 16, 48, 50]. A major challenge thereby lies in finding the appropriate chemical triggers or ecological cues to elicit the production of cryptic antibiotics (recently reviewed in [34, 55, 56]). The use of chemical elicitors is thereby a promising approach [10, 42, 56].
Manipulation of fermentation conditions for promising producer strains, known as the “one strain many compounds” (OSMAC) approach, is an effective way of enhancing the production of secondary metabolites [5, 33]. The regulatory networks that control the production of bacterial natural products respond strongly to changes in carbon, nitrogen or phosphate concentration [40, 41]. The re-isolation of known metabolites is a major bottleneck in the discovery of new bioactive natural products. A crucial step in this regard is the early identification of already known substances, to concentrate the efforts on the discovery of new ones, a process known as dereplication . Current dereplication strategies include hyphenated techniques, such as LC–MS, LC–NMR, LC–NMR–MS, and LC–SPE–NMR. Bioactivity fingerprinting has also been used to dereplicate natural products based on their biological modes of action , while molecular networking is a powerful tool for the visualization and dereplication of natural products [46, 54]. Mass spectrometry-based molecular networking relies on clustering of molecules based on similarities in their MS/MS fragmentation patterns, which depends on the structural features of the ionized molecules. The resulting clusters allow scientists to visually explore the metabolites produced by a given strain under a specific growth condition, allowing rapid dereplication of known compounds by automated spectral library searches, and to visualize their unknown structural analogues .
In this work we analyzed the potential of Streptomyces sp. MBT27 as a producer of natural products in response to changes in the carbon source. The strain had previously been identified as a promising producer of NPs . Extensive fluctuations in the secondary metabolite profiles were observed depending on the carbon source used, and statistical methods combined with GNPS molecular networking identified a family of known as well as novel quinazolinone compounds, in response to high concentrations of glycerol. Quinazolinones are heterocyclic compounds with a wide range of medical applications, such as antimicrobial, antiviral, antituberculosis, and as enzyme inhibitors (reviewed in [17, 20, 24]). Combination of MS and NMR methods identified the novel quinazolinones A (1) and B (2), which further expands the chemical space of this rich family of natural products.
Materials and methods
Bacterial strains and growth conditions
Streptomyces sp. MBT27 was obtained from the Leiden University strain collection and had previously been isolated from the Qingling Mountains, Shanxi province, China . Cultures were grown in triplicate in 100 mL Erlenmeyer flasks with 30 mL of liquid minimal medium (MM; ), supplemented with various carbon sources, and inoculated with 10 µL of 109/mL spore suspension. The carbon sources (percentages in w/v) were: 1% mannitol + 1% glycerol, 1% mannitol, 2% mannitol, 1% glycerol, 2% glycerol, 1% glucose, 2% glucose, 1% fructose, 1% arabinose or 1% N-acetylglucosamine (GlcNAc). The cultures were incubated in a rotary shaker at 30 °C at 220 rpm for 7 days.
General experimental procedures
NMR spectra were recorded in deuterated methanol (CD3OD) on a Bruker 600 MHz and were referenced using the residual 1H signal of deuterated solvent at 3.30 ppm [49, 53]. FT-IR was measured on Perkin–Elmer FT-IR Spectrometer Paragon 1000. UV measurements were performed using a Shimadzu UV mini-1240. Optical rotations were measured on a JASCO P-1010 polarimeter. HPLC purification was performed on Waters preparative HPLC system comprised of 1525 pump, 2707 autosampler, 2998 PDA detector, and Water fraction collector III. The columns used were SunFire C18 column (5 µm, 100 Å, 10 × 250 mm) and SunFire C18 column (10 µm, 100 Å, 19 × 150 mm). TLC was performed using aluminum plates coated with silica gel 60 F254 (Merck). All organic solvents and chemicals were of analytical or LCMS grade, depending on the experiment.
Following fermentation, culture supernatants were extracted with ethyl acetate (EtOAc) and evaporated under reduced pressure. For LC-ESI-QTOFMS analyses, extracts were dissolved in MeOH to a final concentration of 1 mg/mL, and 1 μL was injected into Waters Acquity UPLC system equipped with Waters Acquity HSS C18 column (1.8 μm, 100 Å, 2.1 × 100 mm), which is coupled to Agilent 6530 QTOF MS equipped with Agilent Jet Stream ESI source (Agilent Technologies, Inc., Palo Alto, CA, USA). For the LC, solvent A was 95% H2O, 5% acetonitrile (ACN) and 0.1% formic acid; solvent B was 100% ACN and 0.1% formic acid. The gradient used was 2% B for 1 min, 2–85% for 9 min, 85–100% for 1 min, and 100% for 3 min. The flow rate used was 0.5 mL/min. As for the MS, the following ESI source parameters were used: capillary voltage 3 kV, source temperature 325 °C, drying gas flow rate 10 L/min, and fragmentor 175 V. Full MS spectra were acquired in positive mode in the range of 100–1700 m/z, in the extended dynamic range mode. Internal reference masses of purine and Agilent HP-921 were continuously delivered to the ESI source through an Agilent 1260 isocratic pump.
Thermo Instruments MS system (LTQ Orbitrap XL, Bremen, Germany) equipped with an electrospray ionization source (ESI) was used for LC–MS/MS analysis. The Waters Acquity UPLC system equipped with Waters Acquity PDA was run using a SunFire Waters C18 column (3.5 µm, 100 Å, 4.6 × 150 mm), at a flow rate of 0.9 mL/min. Solvent A was 95% H2O, 5% acetonitrile (ACN) and 0.1% formic acid; solvent B was 100% ACN and 0.1% formic acid. The gradient used was 2% B for 1 min, 2–85% for 15 min, 85–100% for 3 min, and 100% for 3 min. As for the MS, the following ESI parameters were used: capillary voltage 5 V, spray voltage 3.5 kV, capillary temperature 300 °C, auxiliary gas flow rate 10 arbitrary units, and sheath gas flow rate 50 arbitrary units. Full MS spectra were acquired in the Orbitrap in positive mode at a mass range of 100–2000 m/z, and FT resolution of 30,000. Data-dependent MS2 spectra were acquired in the ion trap for the three most intense ions using collision induced dissociation (CID). The resulting chemical data were compared with those in SciFinder and Antibase .
Large-scale fermentation and isolation of metabolites (1) and (2)
For large-scale fermentation, Streptomyces sp. MBT27 was grown in eight 2 L Erlenmeyer flasks, each containing 500 mL liquid MM supplemented with 2% w/v glycerol at 30 °C for 7 days. The metabolites were extracted from the spent media of culture filtrates using EtOAc, and the solvent was subsequently evaporated under reduced pressure at 40 °C. The crude extract (1.4 g) was adsorbed onto 1.4 g silica gel (pore size 60 Å, 70–230 mesh, Sigma Aldrich), and loaded on silica column, which was eluted using gradient mixtures of n-hexane, acetone, and MeOH. The fractions eluted with acetone 100% were combined, reconstituted in MeOH, and injected into the preparative SunFire column (19 × 150 mm), which was eluted with H2O:MeOH gradient of 7–100% in 30 min, at a flow rate of 15 mL/min, to yield five fractions. Quinazolinones A and B were further purified on the semi-preparative SunFire column (10 × 250 mm), run at 3 mL/min. Fraction 2 was eluted using H2O:MeOH gradient of 20–40% in 20 min, to yield Quinazolinone A (1, 1 mg). On the other hand, Fraction 3 was eluted using H2O:MeOH gradient of 30–40% in 20 min, to yield Quinazolinone B (2, 1.5 mg).
Quinazolinone A (1)
1H and 13C NMR data for compounds 1 and 2
δH (mult., J in Hz)
δH (mult., J in Hz)
7.88 (dd, J = 7.8, 1.2)
7.66 (dd, J = 7.8, 1.2)
7.07 (t, J = 7.8)
6.69 (td, J = 7.8, 1.2)
7.46 (td, J = 7.8, 1.2)
7.26 (td, J = 7.8, 1.2)
7.23 (brd, J = 7.8)
6.73 (brd, J = 7.8)
4.43 (d, J = 9.0); 4.24 (d, J = 9.0)
3.67 (d, J = 10.8); 3.62 (d, J = 10.8)
3.59 (d, J = 11.4); 3.47 (d, J = 11.4)
3.67 (d, J = 10.8); 3.62 (d, J = 10.8)
3.86 (d, J = 12.6); 3.80 (d, J = 12.6)
3.66 (d, J = 3.0)
Quinazolinone B (2)
colorless, amorphous powder, UV (MeOH) λmax (log ε) 226 (4.18), 347 (3.03) nm; IR νmax 3350, 2950, 1649, 1526, 1049, 751 cm−1; 1H and 13C NMR data, see Table 1; HRESIMS (positive mode) m/z 209.0915 [M + H]+ (calcd. for C10H13N2O3, 209.0921).
Computation of mass spectral networks
MS/MS raw data were converted to a 32-bit mzXML file using MSConvert (ProteoWizard)  and spectral networks were assembled using Global Natural Product Social molecular networking (GNPS) (https://gnps.ucsd.edu) as described . For both parent and MS/MS fragment ions, the mass tolerance was set to 0.5 Da, while the minimum cosine score was set to 0.7. The data were clustered using MSCluster with a minimum cluster size of three spectra. The spectra in the network were also searched against GNPS spectral libraries. A minimum score of 0.5 was set for spectral library search, with at least two fragment peaks matching. Cytoscape 3.5.1 was used for visualization of the generated molecular networks . The edge thickness was set to represent the cosine score, with thicker lines indicating higher similarity between nodes. LC–MS/MS data were deposited in the MassIVE Public GNPS data set (MSV000082988). The molecular networking job in GNPS can be found at https://gnps.ucsd.edu/ProteoSAFe/status.jsp?task=8fd1fcfa0ff744a9808e80bc7be12115. The annotated MS/MS spectra were deposited in the GNPS spectral library for quinazolinone A (CCMSLIB00004684355), and B (CCMSLIB00004684354).
Prior to statistical analysis, mzXML files were imported into Mzmine 2.31  for data processing. Mass ion peaks were detected using the exact mass algorithm with a noise level set to 1.0 × 104. Afterwards, chromatograms were built for the detected masses with a minimum time span of 0.05 min, m/z tolerance of 0.001 m/z and minimum height of 1.0 × 104. Chromatogram deconvolution was then performed using local minimum search algorithm (search minimum in RT range 0.1 min, chromatographic threshold 90%, minimum relative height 1%, minimum absolute height 1.0 × 104, minimum ratio of peak top/edge two and peak duration range 0.05–3 min). In the generated peak lists, isotopes were identified using isotopic peaks grouper (m/z tolerance 0.001 m/z and retention time tolerance 0.1 min), and variations in retention time were reduced using retention time normalizer (m/z tolerance 0.001 m/z and retention time tolerance 1 min). All the peak lists were subsequently aligned using join aligner (m/z tolerance 0.001 m/z, m/z weight 20, retention time tolerance 0.1 min, and retention time weight 20), and missing peaks were detected through gap filling using peak finder (intensity tolerance 1.0%, m/z tolerance 0.001 m/z, and retention time tolerance of 0.2 min). Finally, the aligned peak list was exported as a comma-separated file for statistical analysis.
Statistical analysis was performed using MetaboAnalyst , where log transformation and pareto scaling was initially applied to the data. A heat map of all detected masses, among the different growth conditions, was generated in MetaboAnalyst, to which additional hierarchical clustering analysis (HCA) was performed using Euclidean distance measure and Ward clustering algorithm. Student’s t-tests with multiple testing correction (Benjamini–Hochberg-false discovery rate or FDR) were used to determine significant differences in the intensities of the metabolites, under two different growth conditions. The thresholds set for statistically significant differences were a fold change ≥ 4, together with FDR corrected p value ≤ 0.05. Based on these criteria, a volcano plot was generated. To identify the difference in intensity of a single mass feature among multiple growth conditions, one-way ANOVA was performed, followed by a post hoc Tukey’s honest significant difference (HSD) test.
Results and discussion
The influence of carbon sources on secondary metabolite production
Besides known molecules, the network also contained many mass features that could not be assigned to any of the previously identified metabolites. One of these was an ion with an m/z value of 281.1151 (1), which was very highly increased in intensity (7000-fold, p value = 0.002) when the glycerol concentration was increased from 1 to 2%. It was closely connected (cosine score 0.91) to another upregulated ion with an m/z value of 209.0915 (2), which was also not previously reported. Both ions were part of the large spectral family comprising the annotated anthranilic acid (3), anthranilamide (4), and the 4-quinazolinones 5 and 6, suggesting that they are structurally related metabolites. It is important to note that the concentrations of quinazolinones A and B did not change significantly when the concentration of either mannitol or glucose was doubled from 1 to 2%, or when mannitol was added as additional carbon source to glycerol (Fig. S1). This strongly suggests that the effect of glycerol is not due to increased availability of the carbon source.
The relationship between the increase in glycerol concentration and the strong increase in the production of the compounds 1 and 2, was investigated further by expanding the range of glycerol concentrations and analyzing the metabolic profiles. For this, Streptomyces sp. MBT27 was cultured in glycerol concentrations ranging from 1 to 4% (w/v). ANOVA, followed by a post hoc Tukey’s HSD test, was performed to trace the variation in the production of 1 and 2 among the different culturing conditions. A box plot was used to visualize such variation (Fig. S2). As observed earlier, the production of 1 and 2 was significantly increased when the glycerol concentration was increased from 1 to 2%. However, further increase in glycerol concentration (3% and 4%) did not lead to any significant increase in the production of 1 and 2. Accordingly, MM with 2% glycerol was used to culture the bacteria, for the purpose of purification and identification of the new metabolites 1 and 2.
Isolation and structure elucidation of novel quinazolinones
The [M + H]+ molecular ion at m/z 209.0932 in the ESI-HRMS spectrum resulted in a molecular formula of C10H12N2O3 for 2 (yield 0.3 mg/L). The 13C NMR spectrum presented 10 signals, which was consistent with the molecular formula (Table 1). The 1H NMR spectrum of 2 resembled that of 1 in the downfield aromatic region, and the major difference is the absence of a set of signals for the additional glycerol unit. Further HMBC experiment confirmed 2 is a mono-glycerolated 4-quinazolinone (Fig. 5). Compound 2 was named quinazolinone B.
The 4-quinazoline alkaloids represent an important class of nitrogen-containing heterocyclic compounds , which have highly diverse biological activities, such as antitumor, anti-inflammatory, antihypertensive, antimicrobial, anticonvulsant and antifungal activities . Though the basic bicyclic core (the fused benzene ring and pyrimidine ring) is normally conserved in most of naturally occurring 4-quinazoline derivatives , a variety of structural substitutions have oftentimes been found on the pyrimidine ring of the 4-quinazolinone. The chemical structures of quinazolinones A and B and the glycerol-dependent production suggest that glycerol participates in the construction of the ring system. We propose that quinazolinones A and B form a new sub-branch in the family of the quinazolines. Particularly, quinazolinone A contains an exciting seven-membered ring that we believe may be formed by intermolecular etherification of two units of glycerol, and this ring is further linked with the quinazolinone backbone by a spiro atom at C-2, which is unprecedented. Antimicrobial activities of compounds 1 and 2 were tested against B. subtilis and E. coli. None of the compounds showed antimicrobial activity against these indictor strains (data not shown).
Biosynthesis of quinazolinones A and B
In conclusion, the metabolic potential of Streptomyces sp. MBT27 was analyzed under different nutritional conditions, showing major changes in the metabolome, depending on the carbon source used. In particular, the relatively unspectacular change from 1% to 2% glycerol resulted in a surprisingly global change in the secondary metabolome, with some compounds changing by almost four orders of magnitude. The use of GNPS molecular networking, together with LC–MS based dereplication, allowed us to annotate a cluster of quinazolinone-family molecules that connects to anthranilic acid and anthranilamide. Isolation and structure elucidation revealed the novel alkaloid natural products quinazolinone A and B. Based on the structures, the biosynthesis of the compounds most likely involves the conversion of anthranilic acid to anthranilamide, which is subsequently attached to two glycerol units, to produce compounds 1 and 2. Identification of the biosynthetic gene cluster and subsequent analysis of the individual enzymatic reactions should reveal the precise biosynthetic pathway for these exciting novel molecules.
- 3.Bentley SD, Chater KF, Cerdeno-Tarraga AM, Challis GL, Thomson NR, James KD, Harris DE, Quail MA, Kieser H, Harper D, Bateman A, Brown S, Chandra G, Chen CW, Collins M, Cronin A, Fraser A, Goble A, Hidalgo J, Hornsby T, Howarth S, Huang CH, Kieser T, Larke L, Murphy L, Oliver K, O’Neil S, Rabbinowitsch E, Rajandream MA, Rutherford K, Rutter S, Seeger K, Saunders D, Sharp S, Squares R, Squares S, Taylor K, Warren T, Wietzorrek A, Woodward J, Barrell BG, Parkhill J, Hopwood DA (2002) Complete genome sequence of the model actinomycete Streptomyces coelicolor A3(2). Nature 417:141–147PubMedGoogle Scholar
- 4.Bérdy J (2005) Bioactive microbial metabolites. J Antibiot (Tokyo) 58:1–26Google Scholar
- 6.Chambers MC, Maclean B, Burke R, Amodei D, Ruderman DL, Neumann S, Gatto L, Fischer B, Pratt B, Egertson J, Hoff K, Kessner D, Tasman N, Shulman N, Frewen B, Baker TA, Brusniak MY, Paulse C, Creasy D, Flashner L, Kani K, Moulding C, Seymour SL, Nuwaysir LM, Lefebvre B, Kuhlmann F, Roark J, Rainer P, Detlev S, Hemenway T, Huhmer A, Langridge J, Connolly B, Chadick T, Holly K, Eckels J, Deutsch EW, Moritz RL, Katz JE, Agus DB, MacCoss M, Tabb DL, Mallick P (2012) A cross-platform toolkit for mass spectrometry and proteomics. Nat Biotechnol 30:918–920PubMedGoogle Scholar
- 8.Conti R, Chagas FO, Caraballo-Rodriguez AM, Melo WG, do Nascimento AM, Cavalcanti BC, de Moraes MO, Pessoa C, Costa-Lotufo LV, Krogh R, Andricopulo AD, Lopes NP, Pupo MT (2016) Endophytic actinobacteria from the brazilian medicinal plant Lychnophora ericoides Mart. and the biological potential of their secondary metabolites. Chem Biodivers 13:727–736PubMedGoogle Scholar
- 11.Cruz-Morales P, Vijgenboom E, Iruegas-Bocardo F, Girard G, Yanez-Guerra LA, Ramos-Aboites HE, Pernodet JL, Anne J, van Wezel GP, Barona-Gomez F (2013) The genome sequence of Streptomyces lividans 66 reveals a novel tRNA-dependent peptide biosynthetic system within a metal-related genomic island. Genome Biol Evol 5:1165–1175PubMedGoogle Scholar
- 12.Dewick PM (2009) Medicinal natural products: a biosynthetic approach, 3rd edn. Ohn Wiley & Sons, ChichesterGoogle Scholar
- 13.Du C, van Wezel GP (2018) Mining for microbial gems: integrating proteomics in the postgenomics natural product discovery pipeline. Proteomics 18:1700332Google Scholar
- 16.Gubbens J, Zhu H, Girard G, Song L, Florea BI, Aston P, Ichinose K, Filippov DV, Choi YH, Overkleeft HS, Challis GL, van Wezel GP (2014) Natural product proteomining, a quantitative proteomics platform, allows rapid discovery of biosynthetic gene clusters for different classes of natural products. Chem Biol 21:707–718PubMedGoogle Scholar
- 21.Kieser T, Bibb MJ, Buttner MJ, Chater KF, Hopwood DA (2000) Practical Streptomyces genetics. John Innes Foundation, NorwichGoogle Scholar
- 23.Kornsakulkarn J, Saepua S, Srijomthong K, Rachtawee P, Thongpanchang C (2015) Quinazolinone alkaloids from actinomycete Streptomyces sp. BCC 21795. Phytochem Lett 12:6–8Google Scholar
- 25.Laatsch H (2012) Antibase 2012, the natural compound Identifier. Wiley-VCH Weinheim, GermanyGoogle Scholar
- 33.Romano S, Jackson SA, Patry S, Dobson ADW (2018) Extending the “one strain many compounds” (OSMAC) principle to marine microorganisms. Mar Drugs 16:244Google Scholar
- 35.Sanchez S, Chavez A, Forero A, Garcia-Huante Y, Romero A, Sanchez M, Rocha D, Sanchez B, Avalos M, Guzman-Trampe S, Rodriguez-Sanoja R, Langley E, Ruiz B (2010) Carbon source regulation of antibiotic production. J Antibiot (Tokyo) 63:442–459Google Scholar
- 37.Shaaban KA, Shepherd MD, Ahmed TA, Nybo SE, Leggas M, Rohr J (2012) Pyramidamycins A–D and 3-hydroxyquinoline-2-carboxamide; cytotoxic benzamides from Streptomyces sp. DGC1. J Antibiot (Tokyo) 65:615–622Google Scholar
- 45.Wang M, Carver JJ, Phelan VV, Sanchez LM, Garg N, Peng Y, Nguyen DD, Watrous J, Kapono CA, Luzzatto-Knaan T, Porto C, Bouslimani A, Melnik AV, Meehan MJ, Liu WT, Crusemann M, Boudreau PD, Esquenazi E, Sandoval-Calderon M, Kersten RD, Pace LA, Quinn RA, Duncan KR, Hsu CC, Floros DJ, Gavilan RG, Kleigrewe K, Northen T, Dutton RJ, Parrot D, Carlson EE, Aigle B, Michelsen CF, Jelsbak L, Sohlenkamp C, Pevzner P, Edlund A, McLean J, Piel J, Murphy BT, Gerwick L, Liaw CC, Yang YL, Humpf HU, Maansson M, Keyzers RA, Sims AC, Johnson AR, Sidebottom AM, Sedio BE, Klitgaard A, Larson CB, Boya PC, Torres-Mendoza D, Gonzalez DJ, Silva DB, Marques LM, Demarque DP, Pociute E, O’Neill EC, Briand E, Helfrich EJ, Granatosky EA, Glukhov E, Ryffel F, Houson H, Mohimani H, Kharbush JJ, Zeng Y, Vorholt JA, Kurita KL, Charusanti P, McPhail KL, Nielsen KF, Vuong L, Elfeki M, Traxler MF, Engene N, Koyama N, Vining OB, Baric R, Silva RR, Mascuch SJ, Tomasi S, Jenkins S, Macherla V, Hoffman T, Agarwal V, Williams PG, Dai J, Neupane R, Gurr J, Rodriguez AM, Lamsa A, Zhang C, Dorrestein K, Duggan BM, Almaliti J, Allard PM, Phapale P, Nothias LF, Alexandrov T, Litaudon M, Wolfender JL, Kyle JE, Metz TO, Peryea T, Nguyen DT, VanLeer D, Shinn P, Jadhav A, Muller R, Waters KM, Shi W, Liu X, Zhang L, Knight R, Jensen PR, Palsson BO, Pogliano K, Linington RG, Gutierrez M, Lopes NP, Gerwick WH, Moore BS, Dorrestein PC, Bandeira N (2016) Sharing and community curation of mass spectrometry data with Global Natural Products Social Molecular Networking. Nat Biotechnol 34:828–837PubMedGoogle Scholar
- 46.Winnikoff JR, Glukhov E, Watrous J, Dorrestein PC, Gerwick WH (2014) Quantitative molecular networking to profile marine cyanobacterial metabolomes. J Antibiot (Tokyo) 67:105–112Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.