Description of Aliinostoc alkaliphilum sp. nov. (Nostocales, Cyanobacteria), a New Bioactive Metabolite-Producing Strain from Salina Verde (Pantanal, Brazil) and Taxonomic Distribution of Bioactive Metabolites in Nostoc and Nostoc-like Genera

Cyanobacteria are a group of oxygenic photosynthetic prokaryotes found in almost all habitats on earth including those characterized as extreme environments. It has been observed that the number of studies dealing with the biodiversity of extremophilic cyanobacteria is limited while studies exploring their bioactive potential are even scarcer. The taxonomy of three Nostoc-like cyanobacterial strains isolated from a shallow lake in Brazil was studied by applying a polyphasic approach. The bioactive potential of the strains was also evaluated using antimicrobial susceptibility testing. The metabolites present in the bioactive HPLC fractions were identified by UPLC/ESI/QTOF. Based on our phylogenetic inferences in combination with morphological and ecological information, we describe Aliinostoc alkaliphilum sp. nov., exhibiting antibacterial and antifungal activities. The main bioactive metabolite in all three strains was nocuolin A, which represents the first report of this metabolite in Aliinostoc. Our phylogenetic studies also revealed that many bioactive metabolite-producting strains that are currently assigned to Nostoc belong to other distinct evolutionary lineages. These findings highlight the importance of polyphasic approach studies in both cyanobacterial taxonomy and natural product discovery programs.


Introduction
Cyanobacteria represent a group of morphologically diverse oxyphototrophic bacteria found in almost all habitats on Earth including those which are considered hostile to life and are known as 'extreme' environments [1].
The immense diversity of cyanobacteria refers not only to their morphological variability and habitat preferences but also to a variety of functionally diverse and structurally complex metabolites they produce [2,3]. A result of this high degree of chemical diversity is the increased possibility of discovering substances with biomedically interesting properties in cyanobacteria [4]. Indeed, over 2000 metabolites have been identified from cyanobacteria thus far, most of which derive from genera such as Nostoc and Lyngbya [5]. This (Pantanal wetland, Brazil) based on a polyphasic approach. Furthermore, this study explores the bioactive potential of the new species against bacterial and fungal potential pathogens as part of our ongoing efforts to discover novel bioactive metabolites with pharmaceutical applications. Finally, this study utilizes all available genetic (16S rRNA gene; NCBI) and bioactive metabolite information (CyanoMetDB; [5]) to investigate whether genus Nostoc is indeed rich in bioactive compounds or whether this perceived chemical richness of Nostoc emerges from the lack of proper taxonomic studies based on the abovementioned polyphasic approach.

Cyanobacteria
Cyanobacterial strains CENA513, CENA514 and CENA524, initially classified as Nostoc sp. [37], were used in this study. These three strains were isolated from water samples collected from Salina Verde saline-alkaline lake (municipality of Aquidauana, Nhecolandia region, state of Mato Grosso do Sul) as previously described [37]. The lake is generally characterized by relatively high pH levels (8.4−9.7) and electrical conductivity values ranging between 1.98 and 15 mS cm −1 during the wet and dry seasons, respectively [36]. All strains were grown on solid and in liquid Z8 medium without nitrogen source [39]. Liquid cultures were maintained in an incubator shaker (Climo-Shaker ISF1-X, Kuhner) at 18 °C under constant illumination with white fluorescent light (18.65 μmol of photons s −1 m −2 ) and 100 rpm shaking speed or without shaking. Strain CENA513 was purified using previously described methods [40] and was used to confirm the cyanobacterial origin of the produced bioactive metabolites. For the bioactivity assays, strains were grown in 5L Erlenmeyer flasks containing 2.5 L of Z8 liquid media without nitrogen source under constant illumination of 10-15 μmol of photons s −1 m −2 and permanent sterile aeration at 18 °C for 30 days. Cells were harvested by centrifugation at 9000× g for 10 min at room temperature (RT) and were stored at −20 °C until lyophilization.

Microscopy
Three cyanobacterial strains, growing in liquid Z8 media without nitrogen source, were studied under a Leica MZ6 Stereomicroscope (Wetzlar, Germany) and a Zeiss Axioskop 2 plus Light Microscope (Jena, Germany). Light photomicrographs were acquired using an Axiocam 305 color digital camera and processed using ZEISS ZEN 2.6 (blue edition) software. Presence of sheath was confirmed by staining with India ink. We analyzed the shape and size of trichomes and vegetative cells in at least 20 trichomes from each strain, summarizing 50 cells. The shape, size and position of heterocytes and akinetes as well as the presence of hormogonia were also recorded and compared to data available for Aliinostoc species. Aliinostoc strains for which morphological assessment is missing were not included in the comparison.
Strain CENA513 was further studied by means of scanning electron microscopy (SEM) and transmission electron microscopy (TEM). For SEM observations, lyophilized cells were affixed to aluminum stubs using double-sided carbon tape, were spray coated in gold-palladium and observed under a JEOL JSM 35 Scanning Electron Microscope (Tokyo, Japan) operating at 19 kV. For TEM studies, cells were first immobilized in sodium alginate beads as described by Romo and Perez-Martinez [42] and were allowed to grow for 25 days in Z8 liquid media without nitrogen source as described above. Following this step, cells entrapped in alginate beads were fixed with 2.5% (v/v) glutaraldehyde solution in 0.1 M sodium cacodylate buffer for 3 h at RT and post-fixed in 1% (w/v) osmium tetroxide in the same buffer for 1 h. The specimen was dehydrated in a graded ethanol series (50,70,96 and 100%) and washed twice in propylenoxide. Sample was then embedded in two mixtures of low viscosity resin and propylenoxide (30% and 70%) and finally in low viscosity resin. Surface sections were stained with 2% uranyl acetate and lead citrate as described by Reynolds [43] and examined using a Jeol JEM-1400 Transmission Electron Microscope operating at 80 kV at the Electron Microscopy Unit of the Institute of Biotechnology (EMBI, University of Helsinki, Finland).
For the study of 16S-23S Internal Transcribed Spacer (ITS) region, total genomic DNA from all strains was extracted using the E.Z.N.A. Plant DNA Kit (Omega Bio-tek, Inc., Norcross, GA, USA) according to the manufacturer's protocol with some modifications that are presented below. During the first step of extraction, cyanobacterial biomass was transferred to a 2 mL screw cap tube together with SP1 buffer and sterile acid-washed glass beads (425-600 μm; Sigma-Aldrich). Cells were then disrupted using a FastPrep-24 ® cell disrupter (MP Biomedicals, Santa Ana, CA, USA) two times for 20 s at a speed of 6.5 m s −1 at RT. Following this step, RNase was added to the extracts, samples were mixed and incubated at 65 °C for 1 h. Partial 16S and the 16S-23S ITS region were amplified using the cyanobacteria-specific primers P5 5′-TGTACACACCGGCCCGTC-3′ sensu Boyer et al. [44] and 23S30R 5′-CTTCGCCTCTGTGTGCCTAGGT-3′ after Taton et al. [45]. The PCR reactions were prepared in 50 μL aliquots containing 1X Q5 reaction buffer (New England Biolabs), 0.2 mM of each deoxynucleotide triphosphates (dNTPs; Thermo Scientific, Waltham, MA, USA), 0.5 μΜ of each of the primers (Sigma-Aldrich), 0.02 U of Q5 High-Fidelity DNA Polymerase (New England Biolabs) and 90-100 ng of genomic DNA, determined using NanoDrop One/One Microvolume UV-Vis Spectrophotometer (Thermo Scientific). Amplifications were run in a BioRad iCycler thermal cycler. The profile used included an initial denaturation step at 98 °C for 30 s, followed by 35 cycles of denaturation at 98 °C for 10 s, annealing at 65 °C for 30 s and extension at 72 °C for 2 min and a final extension step of 2 min at 72 °C.
PCR products were analyzed on a 2.0% (w/v) agarose gel stained with EtBr. Two PCR fragments differing by approximately 300 bp were present in all reactions. Both fragments were gel-purified using the Nucleospin Gel and PCR clean up kit (Macherey-Nagel) and sequenced using BigDye Terminator v3.1 Cycle Sequencing Kit on an ABI 3730xl DNA analyser (Applied Biosystems) using the same primer pair as in PCR amplification. Sequencing was carried out at the Sequencing Unit of Institute for Molecular Medicine Finland (FIMM) Technology Center. The generated sequences were edited with CodonCode Aligner v.9.0.1 (CodonCode Corporation, Barnstable, MA, USA) and consensus sequences were obtained. The sequences were deposited in GenBank database under the accession numbers OK042916-OK042921.

Phylogenetic Analysis of 16S rRNA
Sequences were evaluated for their homology to publicly available 16S rRNA genes of cyanobacteria through a BLASTn search in the National Center for Biotechnology Information (NCBI) database. The compiled dataset included taxa sharing ≥97% 16S rRNA sequence identity with the studied strains, bioactive metabolite-producing Nostoc strains and representatives of morphologically similar genera (231 taxa including the outgroup). All selected sequences were at least 1100 bp long. Synechococcus elongatus PCC 6301 (NR_074309) was used as the outgroup. Multiple sequence alignment was performed using ClustalX version 2.1 [46]. The Akaike Information Criterion [47] as implemented in jModeltest v.2.1.10 [48] was used to choose the best-fit model of DNA substitution. The model GTR + I + G was selected for Bayesian Inference (BI), which was carried out in MrBayes v.3.2.6 [49]. For this analysis, two runs with four chains each (one cold and three heated chains) were run simultaneously for 25 × 10 6 Markov Chain Monte Carlo (MCMC) generations starting with a random tree. Sampling frequency was every 1000th generation. The average standard deviation between the two MCMC runs was below 0.01, indicating convergence. The first 25% of trees was discarded as burn-in and a 50% majority rule consensus tree was calculated including posterior probabilities. ML analysis was performed using the web IQ-Tree (http://iqtree.cibiv.univie.ac.at/ (accessed on 15 April 2022) [50]), using automatic model selection (selected model; TVMe + I+G4) and 1000 bootstrap replicates. The phylogenetic trees were visualized using FigTree v1.4.2 (created by Andrew Rambaut, Edinburgh, UK, http://tree.bio.ed.ac.uk/software/figtree (accessed on 20 April 2022) and re-drawn in Inkscape v0.48.4 (created by Inkscape Developer Team, Boston, MA,USA. www.inkscape.org (accessed on 20 April 2022). All taxa used in the phylogenetic analyses and their GenBank accession numbers are presented in Table S2. A distance matrix was also calculated in MEGA 11 [51] using the Kimura-2 parameter model [52].

Determination of 16S-23S rRNA ITS Secondary Structures
Identification of conserved 16S-23S rRNA intergenic spacer (ITS) regions was carried out according to Iteman et al. [53]. Hypothetical secondary structures of D1-D1′ and Box B were predicted using Mfold web server [54] by applying the default conditions except from the draw mode that was set to 'Untangle with loop fix', choosing secondary structures with minimum free energy (ΔG). The obtained structures were compared with the available homologous structures of the known Aliinostoc species. Although we obtained operons with both tRNAs and with no tRNAs in all Aliinostoc alkaliphilum strains, comparison included operons with no tRNAs only. All figures were edited in Adobe Photoshop v.23.2.0 (Adobe Systems Inc., San Jose, CA, USA)

Preparation of Extracts
For antimicrobial susceptibility tests, 500 mg of freeze-dried biomass from each strain was transferred in five 2 mL screw-cap tubes (100 mg per tube) together with glass beads (0.5 mm diameter, Scientific Industries Inc., Bohemia, New York, NY, USA) and 1 mL of 100% LC/MS grade methanol (MeOH; Merck, Darmstadt, Germany). Cells were disrupted using a FastPrep-24 ® cell disrupter (MP Biomedicals, Santa Ana, CA, USA) two times for 20 s at a speed of 6.5 m s −1 at RT. Intracellular bioactive compounds were obtained after centrifugation at 10,000× g for 5 min at RT. Supernatants from each strain were pooled and solvent was evaporated under N2 flow. Extracts were re-solubilized in 5 mL of 100% MeOH and stored at −20 °C until further evaluation.

Antimicrobial Disk Diffusion Susceptibility Testing
Sterile blank paper disks (Abtek Biologicals Ltd.; Liverpool, UK) were aseptically transferred in 96-well polystyrene microplates (F-bottom; Greiner Bio One, Kremsmünster, Austria). Following this step, 300 μL of crude extract (corresponding to 30 mg of freeze-dried biomass) was transferred on each disk and allowed to air-dry overnight. Disks containing 300 μL of 100% MeOH were used as negative controls in both bacterial and fungal bioassays. Kanamycin 1000 mg antibiotic disks (REF 03-KAN1000, Abtek Biologicals Ltd.; Liverpool, UK) were used as a positive control in bacterial bioassays, whereas disks containing 50 μL of nystatin solution (Sigma-Aldrich, St. Louis, MO, USA, stock solution; 1 mg mL −1 dissolved in 70% MeOH (v/v)) were used as positive controls in all fungal bioassays. All disk diffusion assays were performed following the CLSI guidelines as described in the CLSI document M02-A11 [55] and Leber [41]. Plates were incubated overnight (bacteria and yeast) or for 72 h (filamentous fungi) and inhibition zones (including paper disk diameter) were recorded.

Fractionation of Extracts by High-Performance Liquid Chromatography (HPLC)
Two grams of freeze-dried cell biomass were extracted twice with 60 mL of 100% MeOH using a Heidolph Silent Crusher M (Schwabach, Germany) at 20,000 rpm for 30 s. Samples were centrifuged at 10,000× g for 5 min at RT and supernatants from each strain were combined in a round bottom flask. Cell components and highly hydrophobic compounds were removed by solid phase extraction (SPE); the cartridges used were the commercially available C18 SPE cartridges (Phenomenex, 5 g, 20 mL Giga tubes). Following this step, samples were collected in round bottom flasks and cartridges were washed with 30 mL of MeOH 80% (v/v). Methanol was removed by evaporation using a Büchi rotary evaporator (Flawil, Switzerland) followed by lyophilization to remove water.
The lyophilized samples were subjected to reverse phase HPLC using an Agilent HP 1100 Series High-Performance Liquid Chromatography system (Agilent technologies, Palo Alto, CA, USA) equipped with diode array detector (DAD). Two solvents were used in the analysis: solvent A was 0.1% ammonium formate (Sigma-Aldrich, St. Louis, MO, USA) in MQ water and solvent B was HPLC grade acetonitrile (ACN) 100% (VWR Chemicals, Darmstadt, Germany). Prior to analysis, all lyophilized samples were dissolved in 4 mL of HPLC eluent mixture consisting of 50% of solvent A and 50% of solvent B. All samples were mixed well and then sonicated for 10 s in a Bandelin Sonorex super 10P ultrasonic bath (Berlin, Germany) at RT. Following this step, samples were centrifuged at full speed (16,100× g) for 5 min, supernatant from each sample was transferred to a 4 mL amber glass vial and stored at −20 °C until further analysis. Metabolites were separated on a Luna ® C18(2) column (150 × 4.60 mm, 5 μm, 100 Å, Phenomenex, Torrance, CA, USA). Separation was achieved using the two solvents (A and B) in a linear gradient at a temperature of 30 °C. In detail, concentration of solvent B was increased from 50% to 100% in 20 min, flow rate was 1.0 mL per minute and the elution strength of B was held at 100% for another 10 min to ensure that all compounds were eluted from the column. The entire amount of each sample was injected into the column in 50 μL batches, and each fraction was collected in separate glass test tubes. Fractions eluting at the same time were combined before solvent evaporation. Both the organic solvents and water were removed as described above. The lyophilized fractions were dissolved in 4 mL of methanol 100%, transferred to clean 4 mL amber glass vials and stored at −20 °C until further analysis. HPLC fractions were used in disk diffusion assays as described above. Crude extract, kanamycin (bacteria) and nystatin (fungi) were used as positive controls, whereas methanol 100% was used as a negative control.

Chemical Analysis and Identification of Bioactive Metabolites
Crude extracts and bioactive fractions were analyzed using an Acquity UPLC system (Waters, Manchester, UK), equipped with a Kinetex ® C8 LC column (50 × 2.1 mm, 1.7 μm, 100 Å, Phenomenex, Torrance, CA, USA). UPLC was operated with a flow rate of 0.3 mL min −1 in a gradient mode, at a temperature of 40 °C. The two solvents used in the analysis were 0.1% formic acid (Fluka, Sigma-Aldrich, St. Louis, MO, USA) in MQ water as solvent A and 0.1% formic acid in acetonitrile-isopropanol (1:1) as solvent B. In all cases, the strength of solvent B was increased from 5% to 100% in 5 min. Injection volume was 1.0 μL. Mass spectra were recorded with a Waters Synapt G2-Si mass spectrometer (Waters, Manchester, UK) and measurements were performed using negative and positive electrospray ionization (ESI) in resolution mode. Ions were scanned in a range from m/z 50 to 2000 and UV detector range from 210 to 800 nm. Mass spectrometry (MS) analyses were performed with scan times of 0.1 s. Capillary voltage was 2.5 kV, source temperature was 120 °C, sampling cone was 40.0, source offset was 80.0, desolvation temperature was 600 °C, desolvation gas flow was 1000 L h −1 and nebulizer gas flow was 6.5 bar. Leucine-encephalin was used as a lock mass reference compound and calibration was carried out with sodium formate and Ultramark 1621 ® . UV data were also collected from 210 to 800 nm. The elemental compositions of bioactive metabolites were used for conducting an extensive search in online databases such as SciFinder ® and identify the bioactive molecules. Chemical structures from all identified compounds were drawn using ChemDraw Professional v.17.0 software (PerkinElmer informatics Inc., Waltham, MA, USA)

Taxonomic Analysis
TEM observations: Cell ultrastructure is shown in Figure 2. Nucleoid region is scattered through the cytoplasm (Figure 2A-F). In this region, there are ribosomes, carboxysomes ( Figure 2F) and storage inclusions, such as cyanophycin granules and polyphosphate bodies, appearing as black or electron transparent reserve spaces ( Figure 2D). The thylakoidal system surrounds the nucleoid regions with the thylakoids arranged in small curved or whorled parallel groups ( Figure 2C-F). Cell division proceeds by the formation of a septum, which is continuous with the peptidoglycan layer ( Figure 2E). Plasmodesmata were also observed ( Figure 2D). Etymology: al.ka.li′phi.lum. N.L. n. alkali (from the Arabic word 'alqali', ashes of salt wort); N.L. adj. philus -a -um (from Gr. masc. n. philos, friend); N.L. neut. adj. alkaliphilum, friend of alkaline environments.
A. alkaliphilum can be distinguish from the remaining Aliinostoc species based on its habitat preferences and morphological features observed under light microscope (Table  S3). Unlike all other validly described Aliinostoc species (A. morphoplasticum NOS, A. catenatum SA24 and A. magnakinetifex SA18) that form macroscopic mats in slightly alkaline environments (pH 7-8) in India (A. morphoplasticum NOS) and Iran (A. catenatum SA24 and A. magnakinetifex SA18), A. alkaliphilum is a planktic species thriving in highly alkaline waters (pH 8.4-9.7) in Brazil. Furthermore, A. alkaliphilum is characterized by dark brown cells and the presence of two to three adjacent heterocytes, while the previously described Aliinostoc species were characterized by yellowish-brown colored cells (A. morphoplasticum NOS) or blue-green colored cells (A. catenatum SA24 and A. magnakinetifex SA18) and solitary heterocytes. The spherical to oval akinete shape of A. alkaliphilum set this species apart from A. morphoplasticum NOS and A. catenatum SA24 which possess oblong and oval akinetes, respectively.

Phylogenetic Analyses and 16S-23S ITS Secondary Structures of Aliinostoc
A total of 231 taxa were used in the phylogenetic analysis. The 16S rRNA nucleotide sequences of the three studied strains, previously deposited in NCBI under the accession numbers KX458483 (CENA513), KX458484 (CENA514) and KX458485 (CENA524), were 1413 bp long and shared 99.56-99.83% sequence similarity with each other and 97.53-97.85% sequence similarity with the type strain of Aliinostoc, A. morphoplasticum NOS (Table S4). Both BI and ML analyses yielded nearly identical topologies and were mapped on the same phylogenetic tree (Figure 3). In both analyses, several distinct clades were supported by high bootstrap and posterior probability values. Similarly, genus Aliinostoc formed a well-supported monophyletic clade, which appeared to be more closely related to Purpureonostoc and other, morphologically distinct genera of Aphanizomenaceae and Nostocaceae (see Figure 3) than Nostoc sensu stricto clade. The same analyses showed that the three studied strains, CENA513 T , CENA514 and CENA524, were firmly placed within the Aliinostoc clade.
Two ITS regions, containing either both or no tRNAs, were obtained from each studied strain. D1-D1′ and box B regions from the non-tRNA containing operons of A. alkaliphilum strains were compared to the corresponding secondary structures of all known Aliinostoc (Figures 4 and 5 and Table 1). Secondary structures of D1 stem region and box B from tRNA-containing operons of A. alkaliphilum are presented in Figure S1, whereas a detailed description of all D1 helices and box B regions is presented in Table S5. For Aliinostoc alkaliphilum, D1-D1′ helix was identical in all operons lacking tRNAs ( Figure 4D) but had significant differences from the remaining Aliinostoc species in terms of length (Table 1), primary and secondary structures, except from the 6 bp-long basal stem (GAC-CUA-UAGGUC), which was common in all Aliinostoc species (Figure 4A-D). In detail, the D1 stem region of A. alkaliphilum is characterized by a 6 bp-long basal stem followed by a 5-residue right bulge, a 1-residue right bulge, a 3-residue asymmetrical internal loop and an 11-residue terminal hairpin ( Figure 4D and Table S5). On the other hand, D1-D1′ helices of A. morphoplasticum NOS and A. catenatum SA24 were characterized by the presence of a 6-residue basal stem followed by five internal loops and a 4 bp-long terminal hairpin ( Figure 4A,B and Table S5). In contrast to A. alkaliphilum, the D1 helix of A. magnakinetifex SA18 is characterized by an 8-residue asymmetrical internal loop followed by a 2-residue left bulge, a 7-residue asymmetrical internal loop and 5-residue terminal hairpin (Table  S5). The box B structures of all three A. alkaliphilum strains were identical to each other ( Figure 5D) and structurally identical to A. morphoplasticum NOS ( Figure 5A). Both A. alkaliphilum and A. morphoplasticum NOS box B folded structures were characterized by a 4 bp-long basal stem, followed by a 3-residue asymmetrical internal loop, a 5-residue stem region and a 4-residue terminal hairpin. The only differences observed between these two species were the presence of guanine (G) at position 9 of A. alkaliphilum instead of adenine (A) as in A. morphoplasticum NOS and the nucleotide sequence of their terminal hairpin (5′-GAAA-3′ in A. alkaliphilum and 5′-AAUU-3′ in A. morphoplasticum NOS). The box B regions of A. catenatum SA24 and A. magnakinetifex SA18 differ from A. alkaliphilum in respect to their sequence length (Table 1), primary and secondary structures. In detail, A. catenatum SA24 ( Figure 5B) differs from A. alkaliphilum ( Figure 5D) by having a 5 bp instead of a 4 bp-long basal stem followed by a 4-residue asymmetrical internal loop (3-residue in A. alkaliphilum) and a 4-residue stem region (5 bp-long stem region in A. alkaliphilum). Furthermore, the nucleotide sequence of the terminal hairpin in A. catenatum SA24 (5′-CGCU-3′) is significantly different from A. alkaliphilum (5′-GAAA-3′). In the box B structure of A. magnakinetifex SA18 (Figure 5C), the 5-residue basal stem is followed by a 4-residue asymmetrical internal loop, a 3-residue stem region, a 5-residue asymmetrical internal loop, a 2-residue stem region and a 4-residue terminal hairpin (5′-CGAG-3′). The presence of two internal loops as well as the nucleotide sequence of its terminal hairpin separates A. magnakinetifex SA18 from A. alkaliphilum and the remaining Aliinostoc species.

Antimicrobial Susceptibility Testing and Chemical Analysis
Aliinostoc alkaliphilum strains CENA513 T , CENA514 and CENA524 inhibited the growth of S. aureus, A. flavus and Mucor sp. In addition, strain CENA513 T also inhibited the growth of B. cereus (Table 2 and Figures S2 and S3). None of the studied A. alkaliphilum strains affected the growth of the remaining microbial strains, i.e., E. faecium, M. luteus, P. aeruginosa, A. baumannii, E. aerogenes, S. enterica, C. albicans, C. krusei, C. parapsilosis and C. neoformans. Antimicrobial tests carried out using HPLC fractions of CENA514 and 524 revealed the presence of antimicrobial metabolites in one HPLC fraction of CENA514 and one of CENA524. In both cases, the bioactive fractions exhibited bactericidal activity against of S. aureus, fungicidal activity A. flavus and Mucor sp. and bacteriostatic activity against B. cereus. In strain CENA513 T , bioactive metabolites were present in two HPLC fractions. One fraction exhibited the same bioactivities as in CENA514 and 524, whereas the second fraction exhibited bactericidal activity against B. cereus (Figures S4-S7). LC-MS analysis of crude cyanobacterial extracts and bioactive fractions from all three strains was carried out to identify the dominant bioactive metabolite(s). The bioactive metabolite found in crude extracts of all three strains had m/z 299.23 and its elemental composition was C16H30N2O3. As seen in Figure S9 and Table S6, mass spectra and product ion spectra of crude extracts from all three CENA strains matched well with the nocuolin A reference material. Furthermore, strain CENA513 T produced three to eleven times more nocuolin A (Figure 6) compared to the other two strains ( Figure S8). Analysis of HPLC fractions showed that the total amount of nocuolin A present in each strain was only found in the bioactive fractions of CENA514 and 524 (99% and 97% of total amount of nocuolin A, respectively). Similarly, 97% of total amount of nocuolin A of CENA513 T was found in one out of two bioactive fractions of CENA513 T that inhibited the growth of S. aureus, A. flavus and Mucor sp. (Figures S10-S12).

Taxonomic Distribution of Bioactive Metabolite-Producing Nostoc-like Strains
A total of 67 strains ascribed to Nostoc (including the three strains studied herein) produced 267 bioactive metabolites with a wide range of activities (Table S7). Nucleotide sequences of 16S rRNA gene were available for 43 strains (≈64%) and were included in our phylogenetic analysis. Sequencing data from Nostoc sp. BEA-0956 were also available but sequence was too short to be included in the analysis (≈800 bp). As seen in Figure 7 as well as Figure S13 and Table S7, only 17 out of 43 'Nostoc' strains, producing 102 out of 267 bioactive compounds (≈38%), fall into the Nostoc sensu stricto clade. The remaining 26 strains (Figures 7 and S13 and Table S7), which produce approximately 37% (ncomp = 98) of the bioactive compounds that are attributed to Nostoc, are firmly placed within the recently established genera Aliinostoc (nstrains = 9; ncomp = 30), Pseudoaliinostoc (nstrains = 6; ncomp = 36), Dendronalium (nstrains = 1; ncomp = 2) and the well-known genus Desmonostoc (nstrains = 4; ncomp = 6). In addition, six strains belonging to five distinct genetic lineages within Nostocales (clades A-E; Figure S13) were responsible for the production of 24 bioactive metabolites, 15 of which were microcystin variants (Table S7 and Figure S13). Finally, 24 strains producing 67 compounds (≈25%), although assigned to Nostoc have no morphological nor molecular data available to support this assignment, making their further study impossible (Table S7).

Discussion
A new Aliinostoc species, A. alkaliphilum, exhibiting antibacterial and antifungal properties is described herein using a combination of cytomorphological, molecular and ecological criteria. Furthermore, the main bioactive metabolite identified in A. alkaliphilum is the recently described bioactive molecule nocuolin A [56]. Finally, this study shows that a great number of bioactive metabolite-producing strains, which are ascribed to Nostoc, belong to phylogenetically distant cyanobacterial lineages.
The genus Aliinostoc, with A. morphoplasticum NOS as the type strain, has been described from benthic rocks of a eutrophic pond in Sihora (Jabalpur, India) by Bagchi et al. [21]. Although Aliinostoc is morphologically indistinguishable from Nostoc, 16S rRNA phylogenetic analysis carried therein revealed a clear genetic separation of Aliinostoc from Nostoc sensu stricto and other nostocalean genera, thus allowing the establishment of this new genus. Differences observed in D1 stem and box B secondary structures of Aliinostoc and previously established genera further supported the abovementioned findings. Phylogenetic analyses of 16S rRNA also led Saraf et al. [57] and Kabirnataj et al. [38] to the establishment of new Aliinostoc species (A. soli, A. tiwarii and A. constrictum, A. catenatum, A. magnakinetifex, respectively). However, recent studies of Lee et al. [24] supported the transfer of A. soli, A. tiwarii and A. constrictum to the new genus Pseudoaliinostoc, a genus that is morphologically similar but phylogenetically distant from Aliinostoc [24]. The 16S rRNA phylogenetic studies presented in our work are in line with the abovementioned findings.

Molecular Evaluation
Herein, information deriving from 16S rRNA gene and ITS analysis in conjunction with light microscopy and ecological data support the placement of our strains under the recently described genus Aliinostoc. In detail, the three studied strains, i.e., CENA513 T , CENA514 and CENA524, shared 97.53-97.85% 16S rRNA sequence similarity with the type strain of Aliinostoc, A. morphoplasticum NOS, which is higher than the 95.0% cut-off point proposed by Stackebrandt and Goebel [58] for genus delimitation. Our strains shared 99.56-99.83% 16S rRNA sequence similarity with each other; these values are higher than the recommended threshold of 98.65% for bacterial species demarcation [59] and strongly suggest the establishment of a new Aliinostoc species.
It has been previously shown that the 16S-23S ITS region is more variable compared to 16S ribosomal gene [60,61] and a growing number of studies have used the information provided by 16S-23S ITS region as an additional tool in polyphasic approach studies of cyanobacterial strains. Specifically, the folded D1-D1′, box B regions and, in some cases, V3 structures have been successfully used to support the establishment of new genera and species [21,38,[62][63][64][65][66][67]. Herein, the differences observed between D1-D1′ and box B regions of A. alkaliphilum and the remaining Aliinostoc further support the findings of 16S rRNA phylogenetic analysis. Unlike box B, which appears to be more conserved, the secondary structures of D1-D1′ helices exhibit significant differences between different Aliinostoc species, except from A. morphoplasticum and A. catenatum that only have minor differences in their primary structure. High interspecies variability of the D1-D1′ region has been previously observed in other genera such as Potamolinea [66], Wilmottia [68], Ancylothrix [65], Macrochaete [63] and many others.
The presence of multiple ribosomal operons containing one tRNA gene, both tRNA genes or no tRNA genes has been previously reported in many cyanobacterial taxa [44,69]. Unlike the previously described Aliinostoc species, which only have one operon lacking both tRNAs, the three A. alkaliphilum strains have two different operons. One operon contains both tRNAs (for isoleucine and alanine), whereas the second lacks both tRNAs. Although secondary structures of the tRNA-containing operons cannot be compared with the remaining Aliinostoc due to their lack of tRNAs, it is worth noting that the study of D1 stem in the tRNA-containing operons of our strains revealed the presence of two distinct patterns (see supplementary material). Interestingly one of the D1-D1′ patterns observed in the tRNA-containing operons of A. alkaliphilum is structurally similar to the D1-D1′ regions of A. morphoplasticum NOS and A. catenatum SA24. At the same time, the box B regions of the tRNA-containing operons of A. alkaliphilum were identical to the ones recovered from operons lacking tRNAs. Intraspecies variability of different ITS regions has been previously reported in different cyanobacterial taxa. For example, highly variable D1-D1′, box B and V3 helices were observed in strains of Ancylothrix terrestris [65], whereas the D1-D1′and box B secondary structures of four Koinonema pervagatum strains exhibited three and two distinct patterns, respectively [70]. Variable D1-D1′, box B and V3 regions were also observed in several strains of Scytonema hyalinum [69].

Morphology and Ecology of Aliinostoc
In addition to molecular data, macroscopic and microscopic characteristics as well as ecology further corroborate the establishment of A. alkaliphilum. Specifically, A. magnakinetifex SA18 and A. catenatum SA24 were isolated from slightly alkaline soil in Iran where they form greenish-blue colonies with tough mucilaginous texture and thick bluish-green macroscopic mats, respectively [38]. On the other hand, A. morphoplasticum NOS forms macroscopic mats on benthic rocks in eutrophic waters with slightly alkaline pH in Sihora, Jabalpur, India. The new species described herein (A. alkaliphilum) is planktic, does not form macroscopic mats in nature and was isolated from a Brazilian alkaline lake with pH levels that range between 8.4 and 9.7. In terms of morphology, cell width alone cannot be used for species discrimination in Aliinostoc since all species have overlapping minimum and maximum cell width dimensions. However, the darkbrown-colored cells of A. alkaliphilum and the presence of two to three adjacent heterocytes, separate our strains from the previously described Aliinostoc species that had cells with yellowish-brown (A. morphoplasticum NOS) or blue-green color (A. catenatum SA24 and A. magnakinetifex SA18) and solitary heterocytes. The spherical to oval akinete shape of A. alkaliphilum separates it from A. morphoplasticum NOS and A. catenatum SA24 that possess oblong and oval akinetes, respectively [21,38]. Lastly, A. alkaliphilum is characterized by motile hormogonia with terminal heterocytes and gas vesicles. In a previous study, Bagchi and co-workers [21] reported that the presence of motile hormogonia characterizes all Aliinostoc strains and should be considered as the morphological autapomorphic diacritical character for the genus Aliinostoc. However, the only Aliinostoc species that share this characteristic are A. morphoplasticum and A. alkaliphilum. Based on these findings, this feature cannot be considered as an autapomorphic characteristic for this genus anymore.

Bioactive Metabolites Produced by Aliinostoc alkaliphilum
As discussed by Genuário et al. [35], the number of studies dealing with the biodiversity of extremophilic cyanobacteria is limited, while studies exploring their bioactive potential are even scarcer. In this study, strains belonging to the extremophilic cyanobacterium A. alkaliphilum inhibited the growth of S. aureus, Mucor sp. and Aspergillus flavus. Chemical analysis of crude extracts revealed the presence of nocuolin A in different concentrations, which could explain the differences observed in inhibition zones produced by the three different CENA strains.
Nocuolin A is the first and thus far the only naturally occurring oxadiazine discovered from living organisms [56]. Specifically, nocuolin A was discovered by Voráčová and co-workers [56] while screening cyanobacterial strains for apoptosis inducers. Apart from anticancer activity, studies by González-Fuente et al. [71] revealed the antimicrobial potential of an unnamed compound, which is structurally identical to nocuolin A, as well as its isomers against pathogenic strains of Nocardia, Tsarkamurella, Mycobacterium, Staphylococcus, Acinetobacter, Candida and Aspergillus.
Although the bioactive fractions that inhibit the growth of S. aureus, A. flavus and Mucor sp. were characterized by impurities, the only common compound in all fractions was nocuolin A. The inhibitory activity of nocuolin A against S. aureus and A. flavus as well as the bacteriostatic activity of the compound against B. cereus is consistent with the previous findings of González-Fuente et al. [71], whereas the antifungal activity of nocuolin A against Mucor is reported here for the first time. No inhibitory activity was observed against A. baumannii and C. parapsilosis. These findings contradict the previous results of González-Fuente et al. [71] who showed that pure nocuolin A (100 μg per disk) inhibits the growth of A. baumannii (10-12 mm diameter of inhibition zone) and C. parapsilosis. These contradictory results could be explained by the different concentrations of nocuolin A in our cyanobacterial samples. However, purification of nocuolin A and additional antimicrobial susceptibility assays are required to confirm the abovementioned findings. To date, nocuolin A has been observed in heterocytous cyanobacteria belonging to Nostoc, Dolichospermum, Anabaena, Nodularia and Trichormus [56]. The discovery of nocuolin A in A. alkaliphilum expands the taxonomic distribution of this compound, which was thus far restricted to the abovementioned heterocyte-forming genera.
In addition to S. aureus, Mucor sp. and Aspergillus flavus, strain CENA513 T exhibited anti-Bacillus cereus activity too. Bioactivity-guided fractionation of CENA513 T crude extract revealed the presence of three lysoglycerolipids (DGMG 16:1/0:0, SQMG 16:0/0:0 and LPG 16:0/0:0) and five other unidentified molecules in fractions that inhibited the growth of this potential pathogen. The antimicrobial potential of SQMGs has been previously demonstrated [72,73] but there are no reports regarding the antibacterial potential of the remaining lysoglycerolipids. B. cereus is an opportunistic bacterial pathogen found in soil and food products [74]. Apart from the obvious pharmaceutical potential, the identification of metabolites that could specifically target B. cereus cells without affecting the quality of food products or having any cytotoxic effects may also have applications in the food industry. Therefore, further studies are required to isolate and characterize the interesting anti-Bacillus cereus compound(s).

Is genus Nostoc Rich in Bioactive Compounds?
Cyanobacteria are an attractive source of bioactive metabolites with biotechnological and pharmaceutical applications [4]. Interestingly, most of the bioactive metabolite-producing strains are ascribed to genera which are polyphyletic [6,9]. A recent example derives from the polyphyletic genus Lyngbya. According to Tidgewell and co-workers [75], 240 or 35% of marine cyanobacterial natural products derive from genus Lyngbya. Within Lyngbya, 183 out of 240 produced metabolites (76%) are attributed to Lyngbya majuscula [75]. This profound chemical prolificacy of L. majuscula morphotypes led Engene et al. [9] to a polyphasic approach study of Lyngbya populations aiming to provide a better understanding of the relationship between bioactive metabolite production and biological diversity of this chemically rich genus. Although all studied strains were morphologically similar to Lyngbya, phylogenetic analysis suggested that these Lyngbya-like strains fell outside the Lyngbya sensu stricto clade. This resulted in the establishment of the genus Moorena [8,76] followed by Okeania [11] and Dapis [10], which are nowadays considered responsible for the production of several bioactive metabolites previously assigned to Lyngbya. Similar to Lyngbya, the phylogenetic analysis presented herein revealed that several bioactive compounds ascribed to Nostoc belong to different cyanobacterial lineages. In fact, members of Nostoc produce less than 40% of the bioactive metabolites currently assigned to Nostoc. Most bioactive compound families such as microcystins, anabaenapeptins, microviridins, mycosporine-like amino acids and many others produced by Nostoc strains seem to be randomly distributed among different cyanobacterial lineages [5], but metabolites such as cryptophycins [77,78], merocyclophanes [79,80] and nosperin [81] appear to be exclusively produced by Nostoc. Considering that a total of 67 compounds are produced by strains with unclear taxonomic position, the possibility of identifying more 'true Nostoc' strains after the application of polyphasic approach studies is still possible. However, even if all these unclassified strains are assigned to Nostoc, the percentage of bioactive compounds produced by true Nostoc species is less than 64%.

Bioactive Metabolites from Nostoc-like Genera
Based on data collected from CyanoMetDB [5] and our phylogenetic analysis, the recently established genera Aliinostoc and Pseudoaliinostoc produce 30 and 36 bioactive metabolites, respectively. Apart from nocuolin A produced by A. alkaliphilum, five more strains of the Aliinostoc clade produce 29 functionally and structurally diverse bioactive metabolites. These include Aliinostoc spp. CENA535 and CENA548 which produce the toxic lipopeptides puwainaphycins [82], the microcystin-producing Aliinostoc sp. strain CENA88 [83] and Aliinostoc sp. CENA175, which produces volatile compounds [84]. The abovementioned metabolites are produced by other cyanobacterial genera too [29]. The production of the trypsin inhibitors nostosins identified from Aliinostoc sp. strain FSN-E [27] and pseudospumigins, a new family of linear tetrapeptides that show protease inhibitory activity, reported from Aliinostoc sp. CENA543 [85], is thus far restricted to Aliinostoc. In addition to pseudospumigins, Jokela and co-workers [85] also report that Aliinostoc sp. CENA543 is the first free-living strain that produces very large amounts of the hepatotoxic metabolite nodularin-R. Furthermore, studies by Shishido et al. [86] revealed the production of new and known variants of namalides and anabaenapeptins by Aliinostoc sp. CENA543. The production of natural products in Pseudoaliinostoc is limited to [7.7] paracyclophanes, a family of aromatic polyketides exhibiting a broad spectrum of bioactivities ranging from antimicrobial to cytotoxic activities [87][88][89][90][91]. Specifically, Pseudoaliinostoc strains CAVN2, CAVN10 and UIC 10274 are the only known producers of carbamidocyclophanes [87,89,92,93], whereas ribocyclophanes have only been reported from Pseudoaliinostoc spp. UIC 10366 and UIC 10279 [90]. Lastly, Pseudoaliinostoc sp. UIC 10022A together with Cylindrospermum licheniforme and Cylindrospermum stagnale are the only known producers of cylindrocyclophanes [88,91,94,95]. The remaining groups of [7.7] paracyclophanes, i.e., merocyclophanes and nostocyclophanes, are produced by members of Nostoc, i.e., Nostoc spp. UIC 10100 and UIC 10062 [79,80] and Nostoc linckia UTEX B1932, respectively [94,96,97]; the latter species has been transferred to the genus Desmonostoc. According to our phylogenetic analysis, muscoride A and B as well as deprenylmuscoride A and B [98,99] are produced by members of Desmonostoc. Furthermore, the production of these alkaloids together with muscotoxins produced by D. muscorum CCALA 125 [100] appears to be restricted to Desmonostoc. Lastly, the bioactive metabolites produced by Dendronalium and the remaining lineages A-E include anabaenapeptins and mycosporine-like amino acids which, as mentioned earlier, are widely distributed among cyanobacteria. Taking into consideration the above-mentioned findings, screening Aliinostoc, Pseudoaliinostoc as well as Desmonostoc strains for antimicrobial and/or anticancer substances may lead to the discovery of potentially new bioactive metabolites with potentially interesting biomedical and/or biotechnological applications in these relatively new cyanobacterial genera.

Conclusions
The new species Aliinostoc alkaliphilum from Salina Verde (Pantanal, Brazil) exhibiting antibacterial and antifungal properties is described in this study using a polyphasic approach. The identification of the bioactive metabolite nocuolin A in Aliinostoc alkaliphilum represents the first report of this metabolite in Aliinostoc. Furthermore, the utilization of all 16S rRNA sequencing data of bioactive metabolite-producing strains assigned to Nostoc contributed to a better understanding of the biodiversity and production of bioactive metabolites in this polyphyletic genus and other Nostoc-like genera. Results obtained herein also highlight the importance of polyphasic taxonomic studies in natural product discovery efforts.
Supplementary Materials: The following supporting information can be downloaded at: www.mdpi.com/article/10.3390/w14162470/s1, Figure S1: D1 stem region in A. alkaliphilum strains in tRNA-containing operons (A-B) and operons without tRNA (C). A: CENA513 and CENA514 (OK042916 and OK042918, respectively); B: A. alkaliphilum CENA524 (OK042920); C: A. alkaliphilum CENA513, CENA514 and CENA524 (OK042917, OK042919 and OK042921, respectively). Figure S2: Antibacterial susceptibility tests performed using crude cyanobacterial extracts. The amount of sample per disk was equivalent to 30 mg of freeze-dried biomass. (a-b) Staphylococcus aureus HAMBI 66. (c-d) Bacillus cereus HAMBI 1881. Note the bacteriostatic activity of CENA514 and 524 on B. cereus in contrast to bactericidal activity of CENA513 on the same pathogen. pos: positive control (Kanamycin 1000 μg per disk). neg: negative control (300 μL of 100% MeOH per disk). Disk diameter = 6 mm, Figure S3: Antifungal activity of Aliinostoc alkaliphilum. CENA513 (left), CENA514 (middle), CENA524 (right) on Aspergillus flavus HAMBI 829 (a-c) and Mucor sp. HAMBI 831 (d-f). The amount of sample per disk was equivalent to 30 mg of freeze-dried biomass. Zones of inhibition in CENA513 are clearly larger compared to CENA514 and CENA524. Disk diameter = 6 mm. Figure S4: Antimicrobial susceptibility tests performed using HPLC fractions and crude extract of Aliinostoc alkaliphilum CENA513 (a-c), CENA514 (d-f) and CENA524 (g-i) against S. aureus HAMBI 66. The concentration of metabolites in both HPLC fractions and crude extract corresponds to 30 mg of freeze-dried biomass. Numbers on disks correspond to HPLC fractions of CENA513, CENA514 and CENA524. Note that HPLC fraction 3 of CENA513 is present in figures a and b for comparison reasons. cr: crude extract. pos: positive control (kanamycin 1000 μg per disk). neg: negative control (300 μL of 100% MeOH per disk). Disk diameter = 6 mm. Figure S5: Antimicrobial susceptibility tests performed using HPLC fractions of Aliinostoc alkaliphilum. CENA513 (a,b) against B. cereus HAMBI 1881. Note the bactericidal activity of HPLC fraction 513#2 and bacteriostatic activity of HPLC fractions 513#3 and 513#4 on B. cereus. Crude extract (cr) was also used as a positive control. The concentration of metabolites in both HPLC fractions and crude extract corresponds to 30 mg of freezedried biomass. 1-4: HPLC fractions. pos: positive control (kanamycin 1000 μg per disk). neg: negative control (300 μL of 100% MeOH per disk). Disk diameter = 6 mm. Figure S6: Antifungal susceptibility tests performed using HPLC fractions and crude extract of Aliinostoc alkaliphilum CENA513 (a-c), CENA514 (d-f) and CENA524 (g-i) against Mucor sp. HAMBI 831. The concentration of metabolites in both HPLC fractions and crude extract corresponds to 10 mg of freeze-dried biomass. (c) Positive and negative controls. Numbers correspond to HPLC fractions. cr: crude extract. pos: positive control [50 μL of nystatin solution (1 mg mL −1 )]. neg: negative control (300 μL of 100% MeOH per disk). Disk diameter = 6 mm. Figure S7: Antifungal susceptibility tests performed using HPLC fractions and crude extract of Aliinostoc alkaliphilum CENA513 (a-c), CENA514 (d-f) and CENA524 (g-i) against Aspergillus flavus. HAMBI 829. The concentration of metabolites in both HPLC fractions and crude extract corresponds to 10 mg of freeze-dried biomass. (c) Positive and negative controls. Numbers correspond to HPLC fractions. cr: crude extract. pos: positive control [50 μL of nystatin solution (1 mg mL −1 )]. neg: negative control (300 μL of 100% MeOH per disk). Disk diameter = 6 mm. Figure S8: Total ion current (black) and extracted ion (m/z 299.23; red) chromatograms (EIC) of CENA513, 514 and 524 strain crude extracts. Retention time in black font color and EIC areas in red font color. Figure S9: Mass spectra (A-C) of EIC peaks presented in Figure S8 and product ion spectra from m/z 299.23 (E-G) of CENA513, 514 and 524 extracts and nocuolin A reference spectra (D, H) by Voráčová et al. [60]. Figure S10: Total ion chromatograms (TIC) and extracted ion m/z 299.23 (protonated nocuolin A) chromatograms from Aliinostoc alkaliphilum CENA513 extract fractionated into 4 fractions by liquid chromatography. Peaks (TIC) eluting after 4.2 min are similarly present in all fractions and hence represent irrelevant compounds causing no bioactivity., Figure S11: Total ion chromatograms (TIC) and extracted ion m/z 299.23 (protonated nocuolin A) chromatograms from Aliinostoc alkaliphilum CENA514 extract fractionated into 5 fractions by liquid chromatography. Figure S12: Total ion chromatograms (TIC) and extracted ion m/z 299.23 (protonated nocuolin A) chromatograms from Aliinostoc alkaliphilum. CENA524 extract fractionated into 8 fractions by liquid chromatography. Figure S13: Phylogenetic relationships inferred from maximum likelihood analysis based on 16S rRNA sequences of Aliinostoc and related genera of Nostocales sensu Komarek et al. [13]. Clades hosting NP-producing strains (red font color) are shown in different font colors. All bioactive compounds produced by each strain are also shown. Numbers on nodes correspond to bootstrap values (≥50%) and posterior probabilities (≥ 0.50) obtained from maximum likelihood and Bayesian analyses, respectively. Asterisk (*) represents posterior probability of 1.0 for Bayesian analysis and bootstrap values of 100% for maximum likelihood analysis. Synechococcus elongatus PCC 6301 was the designated outgroup. GenBank accession numbers of sequences are given in brackets. The scale corresponds to substitutions/site. Table S1: Bacterial and fungal strains used in this study. Growth temperature and agar-solidified growth media used for strain propagation and disk diffusion assays are also presented. MH: Mueller-Hinton; MH-GMB: Mueller-Hinton with 2% glucose and 0.5 μg ml −1 methylene blue; PDA: Potato Dextrose Agar. All bacterial and yeast strains were incubated overnight. Strains of filamentous fungi were incubated for 3 days. Table S2: Taxa included in the phylogenetic analyses and references for bioactive metabolite-producing strains. Table S3: Morphological comparison and habitat preferences of all known Aliinostoc species. Table S4: Comparison of the 16S rRNA gene sequence identity among Aliinostoc species and other Nostocalean taxa. Similarity = 100 * [1 − (p-distance)]. Table S5: Detailed description of the ITS secondary structures of Aliinostoc species. Table S6: Nocuolin A specific ion m/z values and formulas by Voráčová et al. [60] and difference (Δ) compared to calculated (Calc) values from MS spectra of CENA513, 514 and 524 strains. Mass of ions 14 and 16 were inaccurate but accurate in product ion spectra from m/z 299.23. Table S7. Bioactive metabolite-producing strains and their GenBank accession numbers.