Multiquant version 3

Qualitative analysis of primary and secondary metabolites was carried out by comparing the accurate precursor ions (Q1), product ion (Q3) values, retention times (RT) and fragmentation patterns with those obtained by injecting standards under the same conditions when standards were available (Sigma-Aldrich, USA, http://www.sigmaaldrich.com/united-states.html). When standards were not available, qualitative analysis was conducted using the self-compiled database MWDB (MetWare Biological Science and Technology Co., Ltd, Wuhan, China) and publicly available metabolite databases. Repeated signals of K⁺, Na⁺, NH⁴⁺ and other high-molecular-weight substances were eliminated during the identification. Quantitative analysis of metabolites was performed in maximum reference match (MRM) mode. The characteristic ions of each metabolite were screened using a QQQ mass spectrometer to obtain signal strengths. Integration and correction of chromatographic peaks were performed using MultiQuant version 3.0.2 (AB SCIEX, Concord, Ontario, Canada). The corresponding relative metabolite contents were represented by chromatographic peak area integrals.

After gathering the metabolite spectrum analysis data of multiple samples and integrating the peak areas of all substance peaks, the mass spectra of the same metabolite in different samples were corrected for peak integration. Secondary MS data from the self-built database metware database (MWDB) were used to characterize substances with reference to public databases (METLIN, HMDB, ChemBank, PubChem, and MassBank). MRM mode of triple-quadrupole MS was used for the metabolite quantification. In MRM mode, the quadrupole initially selects the precursor ions (parent ions) of the target substances, thereby eliminating ions corresponding to other molecular weights to preliminarily remove interferences; the precursor ions, after being induced to ionize in the collision chamber, fragment into multiple fragment ions. These fragment ions are then filtered through the triple quadrupole to select a characteristic fragment ion, eliminating non-target ion interference, thus rendering the quantification more precise and with better repeatability. After obtaining the metabolite spectrum analysis data of multiple samples, the mass spectra peak areas of all compounds were integrated using the MultiQuant version 3.0.2 (AB Sciex, Framingham, MA, USA). Finally, the relative contents of the corresponding metabolites were expressed as the chromatographic peak area integrals.

Metabolite identification was conducted by the self-compiled MWDB database (MetWare biological science and Technology Co., Ltd., Wuhan, China) and publicly available metabolite databases. Quantitative analysis of metabolites was accomplished based on the MRM mode, and the characteristic ions of each metabolite were screened through the QQQ mass spectrometer to obtain signal strengths. Integration and correction of chromatographic peaks were performed using MultiQuant version 3.0.2 (AB SCIEX, Framingham, MA, USA). The corresponding relative metabolite contents were shown as chromatographic peak area integrals. Then, correlation analysis was performed to observe the reliability of metabolic composition within nine root samples. K-mean clustering analysis was applied to analyze the accumulation patterns of compounds and visualize the similarities and differences between WG, BG, and RG. The relevant metabolite contents were treated with centroid initialization and standardized processing, after which we performed K-mean analysis. Metabolites with significantly different contents were set with thresholds of variable importance in projection (VIP) ≥ 1 and fold change ≥2 or ≤0.5.

Metabolite identification was conducted by the self-compiled MWDB database (MetWare biological science and Technology Co., Ltd., Wuhan, China) and publicly available metabolite databases. Quantitative analysis of metabolites was done based on the MRM mode, and the characteristic ions of each metabolite were screened through the QQQ mass spectrometer to obtain signal strengths. Integration and correction of chromatographic peaks were performed using MultiQuant version 3.0.2 (AB SCIEX, Canada). The corresponding relative metabolite contents were shown as chromatographic peak area integrals. To investigate the metabolic differences and the degree of variation between samples and quality control sample (a mixer sample pooled together with equal mass from the aforementioned samples) with three replications, the unsupervised principal component analysis (PCA) was performed using the R package prcomp. To investigate metabolic changes in different samples, a hierarchical cluster analysis (HCA) was carried out using the R package heatmap, and the raw data was disposed of unit variance (UV) scaling before HCA analysis. The significantly differential metabolites were set with the thresholds of VIP (variable importance projection) ≥ 1 and Log₂(Fold change) ≥ 1 or ≤ − 1.

UPLC/ESI-MS/MS system equipped with ExionLC (SHIMADZU, Kyoto, Japan), QTRAP^TM 5500 MS/MS system (AB Sciex, Foster City, CA, USA) and a MultiQuant^TM Version 3.0.2 software (AB Sciex, Foster City, CA, USA) for data acquisition and analysis was used to quantify 5 mycotoxins in positive mode. Chromatographic separation was achieved using a C₁₈ column (2.1 mm × 50 mm, 1.7 μm bead diameter, Waters, Milford, MA, USA). Temperatures of the UPLC column and autosampler were set at 35 °C and 15 °C, respectively. The mobile phase A was water containing 2 mmol/L ammonium acetate, and the mobile phase B was acetonitrile. The program was starting from B/A (0/100, v/v) in the first 2 min, reached B/A (60/40, v/v) in 2 min to 3 min, and continued to decrease B/A (70/30, v/v) in 3 min to 19 min. Afterward, A was linearly increased to 100% (B/A, 0/100, v/v) within 2 min and maintained at this composition of the mobile phase for 0.1 min.