Metabolomics characterization of maize responses to MCB attack
Metabolomics profile was obtained in five biological replicates per each genotype-treatment combination. A 3 cm portion at the bottom of the stem internode bellow the ear was taken from each plant. Frozen samples were lyophilized and ground to a fine powder using an electric mill. Extraction was performed using 50 mg of this powder through 500 μL of 80% aqueous methanol by sonication for 15 min. Samples were centrifuged for 10 min in order to remove plant debris (16000 × g, at room temperature). Supernatant was filtered through a 0.20 µm micropore PTFE membrane and placed in vials for further analysis. Five μL of each sample were injected into an ultra‐high‐performance liquid chromatography (UHPLC) system (Thermo Dionex Ultimate 3000 LC) connected to a QTOF detector (Bruker Compact™) with a heated electrospray ionization (ESI) source. Chromatographic separation was performed in a Intensity Solo 2 C18 column (2.1 × 100 mm 1.7 µm pore size) (Bruker Daltonics, Germany) using a binary gradient solvent mode consisting of 0.1% formic acid in water (solvent A) and acetonitrile (solvent B). The following gradient was used: 3% B (0–3 min), from 3% to 25% B (3–10 min), from 25% to 80% B (10–18 min), from 80% to 100% B (18–22 min), and held at 100% B until 24 min. The flow rate was established at 0.4 mL min−1 and column temperature was controlled at 35 °C.
MS data were acquired using an acquisition rate of 2 Hz over the mass range of 50–1200 m/z. Both polarities (±) of ESI mode were used under the following specific conditions: gas flow 9 L min−1; nebulizer pressure 2.6 bar; dry gas 9 L min−1; dry temperature 220 °C. Capillary and endplate offset were set to 4500 and 500 V, respectively. To monitor the performance of data acquisition, the run sequence was started with three blanks (methanol, the solvent used in sample extraction) and a standard compound (triphenyl phosphate in positive ionization mode and chloramphenicol in negative ionization mode). Auto MS/MS fragmentation in pooled samples was performed in order to facilitate compound identification. For MS/MS analysis data was acquire using an acquisition rate of 8 Hz and precursor ions collected using an absolute threshold of 1500 counts and a cycle time of 1.0 s.
The algorithm T-Rex 3D from the MetaboScape 4.0 software (Bruker Daltoniks, Germany) was used for peak alignment and detection. The generated dataset was imported into Metaboanalyst (Chong, Wishart & Xia, 2019) to perform statistical analyses. In order to remove non‐informative variables, data was filtered using the interquantile range filter (IQR). Moreover, Pareto variance scaling was used to remove the offsets and adjust the importance of high‐ and low‐abundance ions to an equal level. The resulting three‐dimensional matrix (peak indices, samples and variables) was further subjected to multivariate data analysis. Within each inbred, partial least squares discriminant analysis (PLS‐DA) was carried out to investigate and visualize the pattern of metabolite changes between the control and each infestation treatment. This analysis was applied to obtain an overview of the complete dataset and discriminate those variables that are responsible for variation between groups (control vs 48-hours feeding to characterize the short-term response to MCB attack and control vs9-days feeding to characterize the long-term response). The PLS‐DA model was evaluated through a cross‐validation (R2 and Q2 parameters). The quality assessment (Q2) and R2 statistics provide a qualitative measure of consistency between the predicted and original data or, in other words, estimates the predictive ability of the model.
For each inbred and infestation treatment, features with a VIP (variable importance in projection) score >2 in the PLS‐DA model (control vs infestation treatment) were selected and considered as the most influential features in that inbred response to MCB attack. In many studies, a VIP value >2 is a correct threshold for feature selection, but this cut‐off depends on the number of variables used. For tentative identification, accurate mass data and isotopic pattern distributions for the precursor and product ions (if available) were studied and compared to the spectral data of reference compounds with a mass tolerance of 5 ppm. PubChem(Kim, Chen, Cheng, Gindulyte, He, He, Li, Shoemaker, Thiessen, Yu, Zaslavsky, Zhang & Bolton, 2019), MassBank (Horai, Arita, Kanaya, Nihei, Ikeda, Suwa, Ojima, Tanaka, Tanaka, Aoshima, Oda, Kakazu, Kusano, Tohge, Matsuda, Sawada, Hirai, Nakanishi, Ikeda, Akimoto, Maoka, Takahashi, Ara, Sakurai, Suzuki, Shibata, Neumann, Iida, Funatsu, Matsuura, Soga, Taguchi, Saito & Nishioka, 2010), KEGG(Kanehisa, Sato, Kawashima, Furumichi & Tanabe, 2016), KNApSAcK (Afendi, Okada, Yamazaki, Hirai-Morita, Nakamura, Nakamura, Ikeda, Takahashi, Altaf-Ul-Amin, Darusman, Saito & Kanaya, 2012), Metlin (Guijas, Montenegro-Burke, Domingo-Almenara, Palermo, Warth, Hermann, Koellensperger, Huan, Uritboonthai, Aisporna, Wolan, Spilker, Benton & Siuzdak, 2018), and Chemspider (Ayers, 2012) databases were used.