GC-MS analysis
The plant primary metabolites were analysed by GC-MS as described by Roessner et al., (2000). 100 mg of freshly collected leaves and roots were ground finely with liquid nitrogen and extracted with 1.4 ml of 100% methanol containing ribitol as internal standard. The mixture was kept at 70°C for 15 min and then mixed vigorously with 1.4 ml of water. Further, the sample was transferred in glass vial (GL-14 Schott Duran) and centrifuged at 4°C for 15 min at 542 rpm. The upper phase was taken in a fresh centrifuge tube and evaporated by using Speed Vac drying for 45 min. Finally, the samples were kept in -80°C for storage until analysis. For derivatization, the residue was dissolved in 80 µ l of methoxyamine hydrochloride with pyridine at 30°C for 90 min, followed by 80 µ l of MSTFA (N-methyl-N-(trimethylsilyl) trifluoroacetamine) and 20 µ l of FAME mix (Sigma, 1 µ g.µ l-1 in hexane) at 30°C for 30 min.
The derivatized sample was analysed on a system LECO-PEGASUS GCXGC-TOF-MS system (LECO Corporation, USA) equipped with 30 m Rxi-5ms column with 0.25 mm internal diameter and 0.25 μ m film thickness (Restek, USA). The ion, interference and source of injection temperatures were maintained at 200°C, 225°C and, 250°C, respectively. Chromatographic separation was carried out under following conditions: isothermal heating at 70˚C for 5 min, followed by 5˚C min-1 oven temperature ramp to 290˚C and final heating at 290˚C for 5 min. The flow rate of carrier gas (helium gas) was adjusted to 1.5 ml. min-1. A volume of 1 µ l sample was used for analysis and injected in split less mode, and scan mass range 70 to 600 at 2 scans/ s.
All data sets were generated in the form of NETCDF files from ChromaTOF software 4.50.8.0 chromatography version (LECO Corporation, USA) GC-MS and were further exported to MetAlign 3.0 (www.metalign.nl). The MSClust software was used to adjust signal to noise ratio of ≥2, for baseline correction, noise estimation and identification of mass peak (ion-wise mass alignment). Non-repetitive (<3 samples) mass signals were discarded. The MSClust analysis was performed on the above results for the reduction of data and compound mass extraction. Further, the MSClust files were further exported to NISTMS Search v2.2 software for identifying compounds with the NIST (National Institute of Standard and Technology) Library and Gӧlm Metabolome Database Library (http://gmd.mpimp-golm.mpg.de/). Metabolite identity was given based on the maximum matching factor (MF) value (>700) and least retention index (RI) value. The intensity of each metabolite value normalized with internal standard ribitol value. Finally, only annotated metabolites were taken into consideration for further analysis. Lists of detected metabolites were given in the supplementary data.