Chemical profiling and pharmacological potential of myxobacteria isolated from Brazilian mangrove sediments

Myxobacteria are prolific producers of bioactive compounds with antibacterial, antifungal, anticancer, and antiparasitic activities. Traditionally associated with terrestrial soils, they have recently been found in marine and coastal ecosystems. Herein, we report the first occurrence and biotechnological potential of myxobacteria from Brazilian mangroves. Soil samples from the Portinho Mangrove (São Paulo, Brazil) yielded 18 isolates, mainly within the genus Myxococcus, including potentially novel lineages revealed by 16 S rDNA phylogeny. Organic extracts from liquid cultures were tested for cytotoxicity against tumor and non-tumor cell lines, as well as against the parasite Plasmodium falciparum. While non-tumor fibroblasts remained unaffected, several isolates exhibited marked cytotoxicity, with strains BRX-002 and BRX-014 showing IC50 values below 1 μg/mL against HT-1080 and HCT-116 cells. HRMS-based molecular networking suggests the presence of myxothiazol A, myxalamid B, and additional metabolites. These findings reveal Brazilian mangroves as a source of unexplored myxobacteria diversity with remarkable pharmacological potential.

Myxobacteria are Gram-negative bacteria commonly found in soil, characterized by their remarkable social behaviors and complex life cycle (Shimkets et al., 2006; Bhat et al., 2021). They move through gliding and form swarms during vegetative growth (Shimkets, 1990). Under harsh environmental conditions, the cells aggregate to produce fruiting bodies that contain dormant myxospores (Reichenbach, 2001; Bhat et al., 2021). Additionally, these bacteria stand out for producing a wide range of bioactive secondary metabolites, including compounds with antibacterial, antifungal, anticancer, and antiparasitic activities (Held et al., 2014; Bhat et al., 2021; Walesch et al., 2024). In this context, these compounds, and the producers thereof, have attracted growing interest due to their pharmacological potential.

Myxobacteria, traditionally linked to terrestrial habitats like organic-rich soils, have recently been found in marine and coastal environments (Saggu et al., 2023). Specific genera, such as Enhygromyxa, Pseudoenhygromyxa, and Haliangium, are characteristically found in these areas, suggesting that these ecosystems may promote strains with unique adaptive traits (Jiang et al., 2010; Garcia et al., 2018). Moreover, marine strains form a distinct phylogenetic clade separate from their terrestrial counterparts and have been detected across various oceanic regions (Brinkhoff et al., 2012). Additionally, metabolomic studies reveal considerable chemical diversity, with only 6%–11% of metabolites shared between species (Amiri Moghaddam et al., 2018). Consequently, recent research underscores marine myxobacteria as a promising source of novel bioactive compounds (Brinkhoff et al., 2012; Dávila-Céspedes et al., 2016).

Mangroves are unique environments characterized by variations in salinity, nutrients, oxygen, and organic compounds, which directly influence the microbial community. In these ecosystems, myxobacteria can adapt to a wide range of environmental conditions and available resources. Additionally, mangroves stand out for having the highest relative abundance of myxobacteria among saline environments (Zou et al., 2024). As an example, Indonesian mangroves have been found to harbor diverse myxobacterial communities, with isolates yielding compounds with promising antimicrobial activities, highlighting the potential of these ecosystems as sources of novel antibiotic compounds (Octaviana et al., 2023). Nonetheless, studies investigating the presence and biotechnological potential of these bacteria in Brazilian mangroves are practically nonexistent.

This study presents the first report on isolation of myxobacteria from Brazilian mangroves, highlighting the richness and potential of these ecosystems for bioprospecting bioactive compounds. Additionally, it describes anticancer and antimalarial activities for extracts produced by these bacteria, underscoring their biotechnological and pharmacological potential. Thus, this work aims to contribute to expanding knowledge about myxobacteria in tropical environments and their potential applications in medicine and industry.

Collection and sediment processing

Six samples of sediment (approximately 20 g each) were collected from the surface at Portinho Mangrove, Praia Grande, São Paulo, Brazil (23°59′11.3″S 46°24′12.9″W), placed in sterile conical tubes, and transported on ice in a cooler box to the laboratory for immediate pre-treatment. Under a laminar flow hood, the samples were spread in sterile Petri dishes, air-dried, and subsequently stored in sterile glass vials at room temperature until isolation procedures.

Isolation, purification and storing of myxobacteria strains from sediment samples

For isolating myxobacteria, the baiting method was employed, where soil samples were distributed on WCX (w/v: 1.5% agar, 0.1% CaCl 2.2H 2 O, 0.01% cycloheximide, pH 7.2 KOH 1 M) agar plates with small spots of dead Escherichia coli BL21 as bait to attract predatory myxobacteria. The plates were incubated at 30 °C for up to 4 weeks. Swarming colonies or fruiting bodies were identified using a stereomicroscope, carefully collected with a sterile needle, and transferred to new Petri dishes containing Baker’s Yeast (VY/2) medium (w/v, agar 1.5%, baker’s yeast (Fermipan red) 0.5%, CaCl 2.2H 2 O 0.1%, pH 7.2 KOH 1 M). This transfer process was repeated until the complete purification of strains. The purified cultures were then transferred into 50 mL of Casitone-Yeast extract (CY) liquid medium (Bacto Casitone (Difco) 3%, Bacto Yeast Extract (Difco) 1%, CaCl 2.2H 2 O 0.1%, pH 7.2 KOH 1 M), and aliquots of well-grown cultures, totaling 1 mL in volume, were preserved in a solution containing 25% glycerol (v/v), designated by the letters BRX followed by three numbers determined by the order each strain was added to the MyxoMarin microorganism collection, and stored at −80 °C for future use.

Solvent extraction of myxobacteria culture to obtain a crude extract

The myxobacteria strains were cultured in Erlenmeyer flasks containing 50 mL of CY broth, maintained under continuous 200-rpm rotary agitation for approximately 5 days. Subsequently, a 10 mL sample from this culture was transferred to a new 250 mL Erlenmeyer flask containing 100 mL of VY/2 medium supplemented with 2% (w/v) adsorbent resin (Amberlite XAD-16) and subjected to agitation (Hoffmann et al., 2018). Cultures were harvested after 5–15 days, depending on the strain, and filtered through a 63 μm mesh sieve capable of retaining both biomass and resin. The collected material was then washed with distilled water and extracted with 75 mL of acetone for 2 h, followed by an additional 2-h extraction with 75 mL of methanol, both under agitation at 200 rpm. The solvent phases were pooled together, filtered to remove the resin and biomass, and evaporated under vacuum in a rotary evaporator. The resulting extract was transferred to a microtube and thoroughly dried in a rotary sample concentrator to obtain the crude extracts. Additionally, to serve as a negative control for subsequent analyses, an uninoculated “culture medium sample” (VY/2 medium supplemented with 2% w/v XAD-16) was incubated and extracted following this identical protocol.

Evaluation of anticancer activities

The tumor cell lines HCT-116 (ATCC CCL-247; human colon adenocarcinoma; cultured in RPMI-1640) and HT-1080 (ATCC CCL-121; human fibrosarcoma; cultured in EMEM) and a non-tumor cell line NIH/3T3 (ATCC CRL-1658; mouse embryonic fibroblasts; cultured in DMEM) were used as models for assessment of cytotoxicity of the crude extracts. All media were supplemented with 10% fetal bovine serum (FBS) and 1% antibiotics (penicillin and streptomycin). The cultures were incubated in a controlled environment with 5% CO 2, 95% relative humidity, and a temperature of 37 °C. Maintenance was performed every 3 days to ensure healthy cell growth, absence of contamination, and appropriate confluence in the culture flasks.

For the evaluation of cytotoxicity of crude extracts, HCT-116, HT-1080 (1 × 10 4 cells/mL), and NIH/3T3 (3 × 10 4 cells/mL) cells were seeded in 96-well plates at 200 µL per well and left to adhere for 24 h. After this period, extracts diluted in DMSO were added to the wells to achieve duplicates at concentrations of 0.003, 0.016, 0.008, 0.4, 2, 10 and 50 μg/mL. Doxorubicin was used as positive control, in a concentration range between 0.003 and 10 μM, and DMSO was used as a negative control at a final concentration of 0.5%. Cells were exposed to extracts and controls for 72 h and, thereafter, cytotoxicity was accessed through the MTT assay (Mosmann, 1983). The medium in each well was replaced with 150 µL of medium containing 0.5 mg/mL MTT [3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyl tetrazolium bromide], and the plate was incubated for an additional 2 h. The spent media was removed, the formazan precipitate was dissolved in 150 µL of DMSO, and absorbance of each well was measured at 570 nm using a spectrophotometer. The percentage of viable cells following treatments was calculated in relation to absorbance values obtained in non-treated wells, and the mean inhibitory concentration (IC 50) along other parameters was determined through a concentration-response curve obtained by nonlinear regression analysis using GraphPad Prism 10.0.

In vitro anti Plasmodium falciparum activity

In vitro assays were conducted to evaluate the activity of various extracts against the blood-stage of Plasmodium falciparum (3D7 strain, obtained from Bei Sources), a drug-sensitive lineage. Parasites were cultured in human erythrocytes at 3% hematocrit in RPMI 1640 medium (Sigma) supplemented with 25 mM HEPES, 21 mM sodium bicarbonate, 11 mM glucose, 0.5% Albumax II (Thermo Fisher), and 100 μg/mL gentamicin (Sigma). For culture preparation, 10 mL of peripheral blood was collected in a heparinized tube. The blood was centrifuged at 800 g for 7 min, then the supernatant containing white cells was discarded. It was washed with RPMI culture medium to completely remove the white cells, and then the blood was kept in 50% RPMI medium at a temperature of 2 °C–8 °C. The parasites were kept in an oven at 37 °C. The culture medium was changed daily and parasitemia was monitored in methanol-fixed smears, stained with Giemsa and viewed under an optical microscope at 1,000x (Trager and Jensen, 1976).

Cultures were synchronized using a 0.5% D-sorbitol solution at 5%. Briefly, the culture was centrifuged at 800 g for 5 min, the supernatant was removed and 10x the volume of the sorbitol blood pellet was added. The culture was incubated at 37 °C for 5 min and after that the culture was centrifuged again, the supernatant was removed and then RPMI medium was added to obtain a well-synchronized parasitemia. In this assay, parasitemia was calculated from 1,000 cells, and the cultures were diluted to 0.5% parasitemia and 2% hematocrit by adding appropriate volumes of erythrocytes maintained at a 50% hematocrit in RPMI-1640 culture medium. An aliquot of 180 µL of parasites was distributed in 96-well plates pre-prepared with 20 µL of the myxobacterial crude extracts, at a concentration of 100 μg/mL to 0.09 μg/mL, negative control wells corresponding to non-parasitized erythrocytes and absent antimalarials were set up simultaneously where they were incubated for 72 h at 37 °C. After the 72-h incubation period with the compounds, the culture medium was removed from the plates and the cells were resuspended with 100 µL of PBS buffer (NaCl 116 mM, Na 2 HPO 4 10 mM, KH 2 PO 4 3 mM) and lysed with 100 µL of lysis buffer (Tris-Base 20 mM, EDTA 5 mM, 0.0008% (v/v) Triton X-100, 0.008% (m/v) saponin, pH 8.0), plus 0.002% (v/v) SYBR green I. The plates were then incubated at room temperature for 30 min, and the fluorescence was read (Hitachi F-7000) with excitation of 480 nm and emission 520 nm. The 50% inhibitory concentration of parasite growth (IC 50) was determined using concentration-response curves (Platon et al., 2023).

Identification of isolates and phylogenetic analysis

Genomic DNA from pure strains grown in CY broth was extracted using the phenol:chloroform method (Green and Sambrook, 2017). Amplification of the 16 S rRNA gene was carried out by PCR with the universal primers F27 (5′-AGA​GTT​TGA​TCC​TGG​CTC​AG-3′) and R1429 (5′- TAC​GGC​TAC​CTT​GTT​ACG​ACT​T-3′). The forward and reverse sequences were analyzed in silico using BioEdit and closely related type strains were identified using the EzBioCloud (Yoon et al., 2017). The sequences were deposited in the NCBI database, and the accession numbers are provided in Supplementary Material. The phylogenetic relationships between the myxobacterial strains isolated in this study and other reference strains deposited in GenBank were established using their partial 16 S rRNA gene sequences. The sequences were aligned using MEGA X (Kumar et al., 2018) and analyzed through different phylogenetic methods. The tree based on the neighbor-joining (NJ) method was constructed in MEGA X with 1,000 bootstrap replicates.

HPLC-MS/MS and HPLC-DAD-MS analyses

The dried samples of BRX-002, BRX-008, and BRX-014 and a culture medium sample were solubilized under vortex mixing in methanol: water (1:1) at a final concentration of 1 mg/mL and subsequently filtered with 0.22 µm PTFE syringe filters to vials of 1 mL. The diluent methanol: water (1:1) was used as a blank. A volume of 3 μL of each sample was analyzed by HPLC-MS/MS in timsTOF fleX MALDI-2 system. This equipment has both MALDI (Matrix-Assisted Laser Desorption/Ionization) and electrospray ionization (ESI) sources. For this study, the data were acquired with ESI. The chromatographic process was performed in a C18 column (Kinetex XBTM, Phenomenex, 100 × 2.1 mm; particle size 2.6 μm) connected to a guard cartridge with the same material. Elution was performed with the water (A) and acetonitrile (B), both with formic acid 0.1% (v/v), at flow rate of 0.35 mL min−1, using the following gradient: 0–23 min, 5%–100% B; 23–27 min, 100% B; 27–28 min, 100%–5% B; 28–33 min, 5% B. The oven temperature was set at 40 °C. MS and MS/MS spectra were acquired between m/z 100 and 1,000 in positive mode. The ESI-MS parameters were capillary voltage, 3.5 kV; End Plate Offset, 500 V; dry temperature, 250 °C; gas flow 9.0 L/min; pressure, 4.0 bar N 2 was used as drying and nebulizer gas. MS/MS was acquired in Auto MS/MS mode.

In addition to mass spectrometry data, we obtained the chromatographic profile of extracts using liquid chromatography coupled to a Diode-Array Detector (DAD), aiming to verify the majoritarian constituents. Analyses by HPLC-DAD-MS were performed in a UFLC LC-6AD chromatograph (Shimadzu) coupled to a Diode Array Detector (DAD) and the micrOTOF-QII mass spectrometer (Bruker Daltonics), using the same column, mobile phase, gradient of elution and chromatographic parameters of HPLC-MS/MS analyses. Absorptions in UV-visible were recorded in the range of 200–800 nm. The ESI-MS were acquired between m/z 100 and 1,000 in positive mode with the following parameters: capillary voltage, 3.5 kV; End Plate Offset, 500 V; dry temperature, 200 °C; gas flow, 9.0 L/min; pressure, 4.5 bar N 2 was used as drying and nebulizer gas. In this second analysis, the precursor ions of annotated compounds from HPLC-MS/MS analyses were also observed in the mass spectra (MS 1 spectra). Thus, HPLC-DAD-MS analyses provided complementary chemical data to the results initially obtained by tandem mass spectrometry in HPLC-MS/MS analyses.

Feature-based molecular networking (FBMN)

Raw data from LC-MS/MS analyses, derived from a single biological replicate, were converted to mzML format using the TIMSCONVERT workflow (Luu et al., 2022). Converted files were preprocessed in MZmine version 4.4.3 (BMC Bioinformatics, United Kingdom) with the following parameters: (i) Mass detection (MS level 1): Mass detector - centroid; Noise level: 1.0 × 10 3; (ii) Mass detection (MS level 2): Mass detector - centroid; Noise level: 2.0 × 10 2; (iii) Chromatogram builder: Minimum consecutive scans - 3; Minimum intensity for consecutive scans - 1.0 × 10 3; Minimum absolute height - 1.5 × 10 3; m/z tolerance (scan-to-scan) - 0.02 m/z or 5 ppm; (iv) Local minimum feature resolver: Chromatographic threshold - 70%; Minimum search range RT/Mobility (absolute) - 0.02; Minimum relative height - 1.5 × 10 3; Min ratio of peak/top edge - 1.70; Peak duration range (min/mobility) - 0.00–1.50; Minimum scans (data points) - 10. (v) 13C isotope filter (formerly: isotope grouper): m/z tolerance (intra-sample) - 0.002 m/z or 0.00 ppm; Retention time tolerance – 0.2 min; maximum charge – 2. (vi) Join aligner: m/z tolerance (sample-to-sample) - 0.02 m/z or 5 ppm; Weight for m/z – 3; Retention time tolerance - 0.3 min; Weight for m/z – 1; (vii) Feature list row filter: Keep rows that match all criteria - Feature with MS2 scan. After preprocessing, the feature list was exported as. mgf (spectral data) and. CSV (data table with RT, precursor ions and Peak area) files. The. mgf and. CSV files were uploaded to GNPS2 platform (https://gnps2.org/homepage) and analyzed using the FBMN workflow. Both the Precursor Ion Tolerance and Fragment Ion tolerance were set to 0.02 Da. Connections between consensus spectra were set to cosine score above than 0.7 and minimum of 6 matched fragment ions. Spectral matches between experimental data and libraries were set to cosine scores above 0.7 (“Library Minimum Cosine”) and minimum of 5 matched fragment ions (“Library Minimum Matched Peaks”). The metadata of this molecular network is publicly available at https://gnps2.org/status?task=dd9fad94eaaf470d9da6fc0d69690a75. Given the acquisition of high-resolution mass spectrometry data (HRMS), metabolites were considered annotated only if they exhibited ions corresponding to their respective protonated molecular formulas in the MS spectra with mass errors below 10 ppm.

In recent decades, halophilic and halotolerant myxobacteria have gained attention as valuable sources of diverse secondary metabolites from marine environments (Zou et al., 2024). However, little is known about their occurrence in Brazilian marine environments and the biotechnological potential of isolates derived from these habitats. In this context, eighteen strains displaying characteristic myxobacteria features, such as fruiting body formation and swarming motility (Figure 1), were isolated from different samples collected at the Portinho Mangrove, São Paulo (Supplementary Table S1).

For taxonomic characterization of the isolates, the 16 S rRNA gene was amplified, while successful amplification was achieved for 10 strains. This partial success may reflect intrinsic challenges in lysing certain myxobacteria cells, as some genera are known to resist common DNA extraction protocols. Members of the genus Corallococcus have been reported to exhibit such resistance (Mohr et al., 2017). It is also possible that the strains were predominantly present in the form of myxospores, which are notoriously difficult to lyse. Alternatively, vegetative cells or myxospores have been underrepresented in the cultures, limiting both DNA recovery and subsequent PCR amplification (Mohr et al., 2017). These findings also suggest that further methodological adaptations may be necessary to fully access their complete taxonomic diversity. Nevertheless, analysis of the 16 S rRNA gene sequences identified all 10 isolates as belonging to the genus Myxococcus, with similarity values ranging from 99.43% to 99.79% with known species (Supplementary Table S2). The presence of Myxococcus, a genus typically found in terrestrial environments, highlights the diversity of these communities. This distribution likely reflects the role of mangroves as dynamic ecotones at the land-sea interface; the constant input from river runoff and tidal effects facilitates the recruitment of microbial lineages from both terrestrial and freshwater sources, favoring the establishment of generalist Myxococcota (Octaviana et al., 2023; Zou et al., 2024).

A neighbor-joining analysis using the 16 S rRNA gene sequences was conducted to visualize the phylogenetic relationship of the isolates with known Myxococcus strains. This analysis supported their classification within the genus and revealed two main clusters: one closely related to M. fulvus and a smaller cluster related to M. virescens (Figure 2). This phylogenetic structure is mirrored by distinct pigmentation patterns: the strains related to M. fulvus consistently display a pinkish-orange coloration (Figure 1; e.g., BRX-002 and BRX-014), whereas those in the smaller cluster related to M. virescens are characterized by yellow fruiting bodies (Figure 1; e.g., BRX-008). The limited resolution of 16 S rRNA often masks fine-scale genomic and phenotypic differences that can be critical for species delimitation (Hassler et al., 2022). Furthermore, the well-supported separation of the BRX isolates into two distinct clades in the phylogenetic tree suggests intra-generic diversity. This phylogenetic structure indicates the possibility of novel lineages or species-level variants that are not distinguishable by 16 S rRNA. To determine species novelty, further genomic analyses and detailed phenotypic characterization are required.

Cytotoxicity assays are widely employed in natural product research to evaluate the anticancer potential of microbial extracts, particularly those derived from rare or underexplored taxa such as myxobacteria. Studies have shown that crude extracts from myxobacteria can affect cancer cell lines, often related to their production of bioactive secondary metabolites (Park et al., 2021; Quang et al., 2025; Sharma et al., 2022). Among the bacterial extracts evaluated, six demonstrated potent cytotoxic activity (IC 50 < 1 μg/mL) against at least one human cancer cell line. From this group, three extracts in particular, derived from strains BRX-002, BRX-008, and BRX-014, stood out due to their favorable selectivity profiles, as these showed significant activity against cancer cells (HCT-116 and HT-1080) while displaying no detectable toxicity against the non-tumoral NIH/3T3 cell line (IC 50 > 50 μg/mL), as detailed in Figure 3 and Supplementary Table S3. The selection of three extracts for chemical investigation was guided by their bioactivities and phylogenetic positions. Strains BRX-014 and BRX-002, which form a phylogenetic cluster near M. fulvus, were distinguished by their potent cytotoxicity against HCT-116 and HT-1080 cells. Strain BRX-008 was prioritized for its distinct profile: it was selectively cytotoxic towards HCT-116 (IC 50 < 0.003 μg/mL), exhibited antiplasmodial activity (IC 50 = 2.9 μg/mL), and formed a separate phylogenetic clade near M. virescens. This combination of factors made all three strains valuable candidates to further chemical assessments aiming at novel natural product discovery.

Initially, the chemical relationships among the samples were visualized using a Venn diagram, which illustrated the distribution of shared and unique chemical features. Extracts obtained from BRX-002 and BRX-014 share a greater number of features compared to the intersections of either sample with BRX-008. Furthermore, the number of features shared between BRX-002 and BRX-008 (90), as well as between BRX-014 and BRX-008 (87) is similar (Supplementary Figure S1). Thus, this chemical data corroborates the phylogenetic tree data of bacteria analyzed in this study (Figure 2). Subsequently, the molecular network based on HPLC-MS/MS data allowed observation of the distribution of compounds among the samples and annotations of potentially cytotoxic metabolites on tumor cell lines. Annotations were based on spectral libraries and literature data. Given the acquisition of high-resolution mass spectrometry data (HRMS), it was considered as annotated only metabolites that exhibited ions corresponding to their respective protonated molecular formulas in the MS spectra with errors below 10 ppm (Supplementary Table S4, Supplementary Material). The MS spectrum of compound 1 showed the precursor ion m/z 488 [M + H]+, as well as the product ion at m/z 456 [M + H-CH 4 O] +, corresponding to the loss of methanol (32 Da). Another key fragment at m/z 439 [M + H-CH 4 O-NH 3 ] + indicates a subsequent loss of ammonia (17 Da). These fragmentations suggested the annotation of compound 1 as myxothiazol A (Figure 4) (Perlova et al., 2006). Bithiazole derivatives, such as myxothiazol A, were previously isolated from the myxobacteria Myxococcus fulvus (Gerth et al., 1980; Ahn et al., 2007; Schieferdecker et al., 2014). The cluster with myxothiazol A (1) and other possible derivatives occurs only in extracts from strains BRX-002 and BRX-014. Furthermore, analyses by HPLC-DAD-MS suggested that myxothiazol A is the major constituent in these extracts, while it was not found in BRX-008 (Figure 4; Supplementary Figure S2).

The antifungal effect of myxothiazol A was previously described (Gerth et al., 1980) and this bioactivity may be related to the inhibition of mitochondrial respiration (Esser and Xia, 2024). As mentioned, samples BRX-014 and BRX-002 were highly cytotoxic against HCT-116 and HT-1080 cell lines. On the other hand, the extract of BRX-008 showed a distinct profile, with a low IC 50 against HCT-116 and moderate cytotoxicity against HT-1080 (Figure 3). These findings suggest that the presence of myxothiazol A (1) may be related to the cytotoxic effect presented by extracts obtained from strains BRX-002 and BRX-014. Moreover, these results opened perspectives for isolation and further investigation of the mechanism of action of 1. MS/MS spectrum of compound 2 corresponded to N-acetyl-2-phenylethylamine (Figure 4), which is the major constituent in HPLC-DAD-MS chromatogram of BRX-008 (Supplementary Figure S2). This metabolite was also previously described in an extract of M. fulvus (Daoud and Foster, 1993). N-acetyl-2-phenylethylamine isolated from marine bacteria Aquimarina sp. MC085 suppressed the metastasis of A549 human lung cancer cells (Lee et al., 2021). These findings suggest that cytotoxic effects of extract from BRX-008 may be related to the presence of compound 2.

Compounds 7 and 3 were annotated as indole derivatives tryptamine and N-acetyl-tryptamine, respectively (Figure 4). Indole alkaloids are extensively found in marine organisms, including bacteria, and have exhibited cytotoxic and antineoplastic activities (Netz and Opatz, 2015). As an example, fatty acyl derivatives of tryptamine isolated from the soft coral Eunicella granulata have shown cytotoxic effects on different human tumor cell lines (Reyes et al., 2006). N-acetyl-tryptamine (3) was previously isolated from the marine actinobacteria Micromonospora sp. A258 (Van Nam et al., 2019). Data from the search in microbeMASST also reported the presence of compound 3 in actinobacteria species that can be found in marine environments, such as Micromonospora aurantiaca and Salinispora aurenicola, but did not list myxobacteria among the sources of this metabolite. Compounds 4 and 9 were annotated as cyclo (Tyr-Pro) and cyclo (Pro-Trp), respectively, cyclic dipeptides containing a 2,5-diketopiperazine ring (Figure 4). Diketopiperazine peptides have demonstrated antimutagenic effects, suggesting the potential of these compounds as cancer chemopreventive agents (Rhee, 2004). Nevertheless, cyclic dipeptides promoted cell death in cultures of the HeLa cervical adenocarcinoma and Caco-2 colorectal adenocarcinoma (Vázquez-Rivera et al., 2015). In a previous evaluation, it was observed that cyclo (Tyr-Pro) is responsible for an oxidative burst in fungal cells. This effect is related to the chemical deterioration of lipids or proteins and loss of cell membrane functionality. The same mode of action in nematodes suggests that the activity may be universal in eukaryotic membranes (Vela-Corcia et al., 2024). MicrobeMASST searches showed a broad distribution of spectra of 4 and 9 among bacterial phyla, including samples of myxobacteria belonging to the order Myxococcales (Supplementary Figure S3).

Compounds 2 – 4, and 7 – 11 were found in all the samples. Differently, metabolites 5 and 6 were exclusively found in the extract of BRX-008 (Figure 4). MS/MS data of compound 5 corresponded to the homospermidine lipid Cmp-552 (Supplementary Figure S3), which has been previously isolated from the fruiting body of Myxococcus xanthu s (Hoffmann et al., 2018). Literature data reported moderate activity (14–28 μM) of homospermidine lipids against some Gram-positive bacteria (Hoffmann et al., 2018). The fragmentation profile of 6 corresponds to myxalamid B, which was also previously isolated from M. xanthus and found in different species of myxobacteria. In addition, microbeMASST showed that the spectra of myxalamid B (6) and other exclusive nodes of BRX-008 were previously described only in M. xanthus samples (Supplementary Figure S4). Hence, the chemical data is consistent with the genetic data, since BRX-008 presents phylogenetic proximity to the M. xanthus species (Figure 2). The antibiotic spectrum of myxalamid B covers molds, yeasts, and Gram-positive bacteria. An evaluation of mechanism of action showed that myxalamid B specifically blocked the electron flow at complex I in submitochondrial particles (Gerth et al., 1983), suggesting potential cytotoxicity of this metabolite. Spectrometric data of compound 8 corresponded to N-acetyl-tyramine (Figure 4). The cytotoxic properties of 8 includes reversal of resistance of P388 leukemia cells to doxorubicin (Kunimoto et al., 1987), and antitumoral activity against human melanoma A375 and human leukemia K562 cell lines (Zhao et al., 2010). Compound 10 was found in all the samples and annotated as lumichrome (Figure 4), a degradation product of riboflavin. In a previous work, lumichrome isolated from the culture broth of bacteria Streptomyces pseudovenezuelae SKH1-2 suppressed growth and induced apoptosis of human non-small lung cancer cells (Chantarawong et al., 2019). The occurrence of the spectrum corresponding to lumichrome (10) in different bacterial taxa was verified in microbeMASST and retrieved datasets from the order Myxococcales. Compound 11 was annotated as an isoquinolinic derivative (Figure 4). According to the microbeMASST data, spectrum of 11 is present in different bacteria phyla, but was not previously detected in myxobacteria datasets (Supplementary Figure S4).

In conclusion, the metabolomic profiling of the BRX myxobacterial strains revealed a diverse and promising chemical landscape. The annotation of specialized metabolites such as Myxothiazol A and Myxalamid B, both recognized for their potent cytotoxic activities, provides a robust chemical rationale for the biological effects observed in the crude extracts. However, it is noteworthy that these annotations account for only 1.61% of the total molecular features detected. Most of the registered metabolites (98.39%) remain unassigned in public spectral libraries, suggesting that the metabolome of this group offers a fertile ground for the discovery of novel chemical scaffolds. This substantial ‘chemical dark matter' underscores the potential for isolating unique compounds and highlights the BRX strains as a significant reservoir for future drug discovery efforts. It is important to acknowledge, however, that the low annotation rate also reflects a well-recognized limitation of untargeted metabolomics: the incompleteness of public spectral libraries, particularly for understudied microbial taxa such as myxobacteria. As a result, the chemical identity and biological relevance of the vast majority of detected features remain unknown, which constrains the mechanistic interpretation of the observed bioactivities. These limitations reinforce the need for complementary approaches, including whole-genome sequencing to identify biosynthetic gene clusters (BGCs), as well as targeted isolation and structure elucidation efforts, to fully characterize the chemical diversity and, specifically, the compounds responsible for the robust antimalarial and anticancer properties demonstrated by these myxobacterial extracts.