Application of feature-based molecular networking and MassQL for the MS/MS fragmentation study of depsipeptides.

The Feature-based Molecular Networking (FBMN) is a well-known approach for mapping and identifying structures and analogues. However, in the absence of prior knowledge about the molecular class, assessing specific fragments and clusters requires time-consuming manual validation. This study demonstrates that combining FBMN and Mass Spec Query Language (MassQL) is an effective strategy for accelerating the decoding mass fragmentation pathways and identifying molecules with comparable fragmentation patterns, such as beauvericin and its analogues. To accomplish this objective, a spectral similarity network was built from ESI-MS/MS experiments of Fusarium oxysporum at various collision energies (CIDs) and paired with a MassQL search query for conserved beauvericin ions. FBMN analysis revealed that sodiated and protonated ions clustered differently, with sodiated adducts needing more collision energy and exhibiting a distinct fragmentation pattern. Based on this distinction, two sets of particular fragments were discovered for the identification of these hexadepsipeptides: ([M + H]+) m/z 134, 244, 262, and 362 and ([M + Na]+) m/z 266, 284 and 384. By using these fragments, MassQL accurately found other analogues of the same molecular class and annotated beauvericins that were not classified by FBMN alone. Furthermore, FBMN analysis of sodiated beauvericins at 70 eV revealed subclasses with distinct amino acid residues, allowing distinction between beauvericins (beauvericin and beauvericin D) and two previously unknown structural isomers with an unusual methionine sulfoxide residue. In summary, our integrated method revealed correlations between adduct types and fragmentation patterns, facilitated the detection of beauvericin clusters, including known and novel analogues, and allowed for the differentiation between structural isomers.

Enhancing chemical and biological diversity by co-cultivation.

In natural product research, microbial metabolites have tremendous potential to provide new therapeutic agents since extremely diverse chemical structures can be found in the nearly infinite microbial population. Conventionally, these specialized metabolites are screened by single-strain cultures. However, owing to the lack of biotic and abiotic interactions in monocultures, the growth conditions are significantly different from those encountered in a natural environment and result in less diversity and the frequent re-isolation of known compounds. In the last decade, several methods have been developed to eventually understand the physiological conditions under which cryptic microbial genes are activated in an attempt to stimulate their biosynthesis and elicit the production of hitherto unexpressed chemical diversity. Among those, co-cultivation is one of the most efficient ways to induce silenced pathways, mimicking the competitive microbial environment for the production and holistic regulation of metabolites, and has become a golden methodology for metabolome expansion. It does not require previous knowledge of the signaling mechanism and genome nor any special equipment for cultivation and data interpretation. Several reviews have shown the potential of co-cultivation to produce new biologically active leads. However, only a few studies have detailed experimental, analytical, and microbiological strategies for efficiently inducing bioactive molecules by co-culture. Therefore, we reviewed studies applying co-culture to induce secondary metabolite pathways to provide insights into experimental variables compatible with high-throughput analytical procedures. Mixed-fermentation publications from 1978 to 2022 were assessed regarding types of co-culture set-ups, metabolic induction, and interaction effects.

Time-related multivariate strategy for the comprehensive evaluation of microbial chemical data.

INTRODUCTION: In microbial metabolomics, the use of multivariate data analysis (MDVA) has not been comprehensively explored regarding the different techniques available and the information that each gives about the metabolome. To overcome these limitations, here we show the use of Fusarium oxysporum cultured in the presence of exogenous alkaloids as a model system to demonstrate a comprehensive strategy for metabolic profiling. MATHERIALS AND METHODS: F. oxysporum was harvested on different days of incubation after alkaloidal addition, and the chemical profiles were compared using LC-MS data and MDVA. We show significant innovation to evaluate the chemical production of microbes during their life cycle by utilizing the full capabilities of Partial Least Square (PLS) with microbial-specific modeling that considers incubation days, media culture availability, and growth rate in solid media. RESULTS AND DISCUSSCION: Results showed that the treatment of the Y-data and the use of both PLS regression and discrimination (PLSr and PLS-DA) inferred complemental chemical information. PLSr revealed the metabolites that are produced/consumed during fungal growth, whereas PLS-DA focused on metabolites that are only consumed/produced at a specific period. Both regression and classificatory analysis were equally important to identify compounds that are regulated and/or selectively produced as a response to the presence of the alkaloids. Lastly, we report the annotation of analogs from the piperidine alkaloids biotransformed by F. oxysporum as a defense response to the toxic plant metabolites. These molecules do not show the antimicrobial potential of their precursors in the fungal extracts and were rapidly produced and consumed within 4 days of microbial growth.

Dereplication of Natural Products Using GC-TOF Mass Spectrometry: Improved Metabolite Identification by Spectral Deconvolution Ratio Analysis.

Dereplication based on hyphenated techniques has been extensively applied in plant metabolomics, thereby avoiding re-isolation of known natural products. However, due to the complex nature of biological samples and their large concentration range, dereplication requires the use of chemometric tools to comprehensively extract information from the acquired data. In this work we developed a reliable GC-MS-based method for the identification of non-targeted plant metabolites by combining the Ratio Analysis of Mass Spectrometry deconvolution tool (RAMSY) with Automated Mass Spectral Deconvolution and Identification System software (AMDIS). Plants species from Solanaceae, Chrysobalanaceae and Euphorbiaceae were selected as model systems due to their molecular diversity, ethnopharmacological potential, and economical value. The samples were analyzed by GC-MS after methoximation and silylation reactions. Dereplication was initiated with the use of a factorial design of experiments to determine the best AMDIS configuration for each sample, considering linear retention indices and mass spectral data. A heuristic factor (CDF, compound detection factor) was developed and applied to the AMDIS results in order to decrease the false-positive rates. Despite the enhancement in deconvolution and peak identification, the empirical AMDIS method was not able to fully deconvolute all GC-peaks, leading to low MF values and/or missing metabolites. RAMSY was applied as a complementary deconvolution method to AMDIS to peaks exhibiting substantial overlap, resulting in recovery of low-intensity co-eluted ions. The results from this combination of optimized AMDIS with RAMSY attested to the ability of this approach as an improved dereplication method for complex biological samples such as plant extracts.