Abstract
Translocation of metabolites between different plant species provides important hints in understanding the fate of bioactive root exudates. In the present study, targeted and untargeted mass spectrometry-based metabolomics was applied to elucidate the transfer of bioactive compounds between rye and several crops and weed species. Our results demonstrated that benzoxinoids (BXs) synthesized by rye were taken up by roots of neighboring plant species and translocated into their shoots. Furthermore, we showed roots of the rye plant took up compounds originating from neighboring plants. Among the compounds taken up by rye roots, wogonin was detected in the rye shoot, which indicates the root-to-shoot translocation of this compound. Elucidating the transfer of bioactive compounds between plants is essential for understanding plant-plant interactions, developing natural pesticides and understanding their modes of action.
Keywords: Keywords: Metabolomics, Mass spectrometry, Plant-plant interactions, Secondary metabolites, Bioactive compounds
Introduction
Plant roots exude a complex mixture of organic compounds, including secondary metabolites, which are essential in mediating plant interaction with other organisms (Bais et al., 2006). Root-exuded secondary metabolites can act as defense compounds against microbial pathogens and herbivores (Zhang et al., 2020, Hu et al., 2018) or shape the beneficial microbiome (Sikder et al., 2021, Kudjordjie et al., 2019). Moreover, root exudates containing allelochemicals can directly inhibit or reduce the germination or growth of other plant species (Weir et al., 2004, Hazrati et al., 2021). Other plants must absorb root-exuded metabolites or their breakdown products (Etzerodt et al., 2006, Etzerodt et al., 2008) before acting as growth inhibitors (Chiapusio et al., 2004). However, studies on the transfer of bioactive compounds between plants are scarce, specifically on root uptake from the soil and translocation into the shoots. Elucidation of the transfer of bioactive compounds between plant species is vital for developing commercial natural pesticides (Davies and Caseley, 1999).
Benzoxinoids (BXs) are tryptophan-derived heteroaromatic metabolites that act as natural pesticides and are mainly present in the Poaceae roots, including rye, maize, and wheat (Hazrati et al., 2019, Frey et al., 2009). They are essential in plant interactions with microorganisms (Cotton et al., 2019), herbivores (Wouters et al., 2016), and other plant species (Hazrati et al., 2020). The growth-inhibitory ability of BXs has been confirmed on several plant species (Schulz et al., 2013). Just recently, Hazrati et al. (2020) demonstrated that root exuded BXs can be taken up by hairy vetch plants and subsequently translocate into their shoots. Studying whether root uptake and translocation of BXs occur in the same manner in other plant species is of particular interest from an agronomic perspective because it may promote the development of selective natural herbicides. Here we are comparing the root uptake and translocation of BXs in several crop and weed species.
Metabolomics represents a field of research, which provides us a better understanding of complex biological systems (Cevallos-Cevallos et al., 2009). There are two distinct metabolomics approaches: targeted and untargeted. Targeted metabolomics refers to identifying and quantifying selected groups of metabolites or validation of biomarkers identified using non-targeted metabolomics (Lu et al., 2008, Roberts et al., 2012). Targeted metabolomics deals with the known metabolites of interest and, therefore, the coverage of detected metabolites is limited (Vrhovsek et al., 2012). Untargeted metabolomics aims to globally profile the metabolome and gather as much information on metabolites as possible. Untargeted metabolomics is often applied to generate hypotheses and discover biomarkers (Vinayavekhin and Saghatelian, 2010). High-resolution mass spectrometers (HRMS) are routinely used for untargeted metabolomics studies due to their high sensitivity and selectivity, which maximize metabolic coverage (Theodoridis et al., 2012). Ultra-high-performance liquid chromatography (UHPLC) hyphenated with an HRMS is currently the dominant technique for global metabolite profiling of complex biological samples, e.g. from plants, due to its versatility in separation and detection of compounds with a wide range of polarities, as well as its robustness (Pezzatti et al., 2019). Metabolite annotation and identification from the vast amount of generated features is the most significant bottleneck of untargeted mass spectrometry-based metabolomics (Misra and van der Hooft, 2016, Dunn et al., 2013). Therefore, in most cases, mass spectrometry-based metabolic profiling aims to identify known/unknown putative metabolites, which could be confirmed by MS experiments or by comparing with authentic standards (Lee et al., 2010, Krauss et al., 2010). There are several studies, which used targeted approach to show the root uptake and translocation into shoot of particular metabolites or a specific class of metabilites (Hazrati et al., 2020, Chiapusio et al., 2004, Lewerenz et al., 2020). Up to date, no studies have revealed two-way transfer of bioactive compounds between plant species using an untargeted metabolomics approach. In the present study , a combination of targeted and untargeted mass spectrometry-based metabolomics was applied to elucidate the extensive transfer of bioactive compounds between rye and several crops and weeds species.
Material and Methods
Chemicals
2-Benzoxazolinone (BOA) and 6-methoxy-benzoxazolin-2-one (MBOA) were purchased from Sigma-Aldrich. The following eight BXs standards were obtained as part of an on-going patenting process in our lab: 2-hydroxy-1,4-benzoxazin-3-one (HBOA), 2-hydroxy-7-methoxy-1,4-benzoxazin-3-one (HMBOA), 2-β-d-glucopyranosyloxy-1,4-benzoxazin-3-one (HBOA-glc), 2-β-d-glucopyranosyloxy-7-methoxy-1,4-benzoxazin-3-one (HMBOA-glc), 2,4-dihydroxy-1,4-benzoxazin-3-one (DIBOA), 2-β-d-glucopyranosyloxy-4-hydroxy-1,4-benzoxazin-3-one (DIBOA-glc), double hexose derivative of DIBOA (DIBOA-glc-hex) (structure not fully elucidated), and double hexose derivative of HBOA (HBOA-glc-hex) (structure not fully elucidated). 2-β-d-Glucopyranosyloxy-4-hydroxy-7-methoxy-1,4-benzoxazin-3-one (DIMBOA-glc) was obtained as described in a previous study (Pedersen et al., 2017). High-performance liquid chromatography (HPLC)-grade methanol and acetonitrile were purchased from Rathburn (Walkerburn, Scotland); MS-grade methanol, acetonitrile, and isopropanol from Fischer Scientific (Loughborough, UK); and acetic acid from Baker (Griesheim, Germany). HPLC-grade water was obtained from a Milli-Q system (Millipore, Billerica, MA) for the analysis using LC-MS/MS. HPLC-MS grade water was purchased from Fisher Scientific for analysis in UHPLC-QTOF-MS.
Pot experiment
One-liter pots filled with sandy loam field soil (2.8% organic matter, 11.5% clay, 28.4% silt, and 57.2% sand) were used as the growth medium. Plants were grown under controlled conditions in a greenhouse at a 16/8 h photoperiod with a temperature of 22/18 °C (day/night) and were watered through a sub-irrigation system. Six rye seeds were sown in the center of the pot as target species, and 12 seeds of a neighboring crop/weed species were sown in a ring around the rye plants. After germination of the seeds, the number of rye and neighboring seedlings were thinned to 4 and 8 plants per pot. Neighboring plant species included in the study was seven crop species: rye, Alexandrian clover (Trifolium alexandrinum ), hairy vetch (Vicia villosa ), fodder radish (Raphanus sativus ), oat (Avena sativa ), subterranean clover (Trifolium subterraneum ), Austrian pea (Pisum sativum ) and two weed species: Lolium (Lolium multiflorum ) and Sinapis Sinapis arvensis . Eight replicates were used for each treatment. Plants were harvested five weeks after sowing. At harvest, rye and neighboring plants were carefully separated from each other. Plant roots were slightly shaken to remove attached soil, washed with distilled water, and separated from their shoot. Harvested plant material was immediately transferred into the liquid nitrogen before transferring to a freezer at -80 ˚C. Samples were freeze-dried, and the lyophilized root and shoot material was grounded to a fine powder using a mechanical disrupter Genogrinder 2010 from Spex (Metuchen, NJ) at a speed of 1500 rpm for 90 s (repeated twice).
Metabolite Extraction from Plant Material
Homogenized ground plant material (20 mg) was transferred into an Eppendorf tube, and 1 ml of 80% methanol containing 1% acetic acid was added. Samples were ultrasonicated for 45 min and centrifuged (Sigma 1–14 K micro-centrifuge, Buch and Holm, Herlev, Denmark) at 4500 g and 21°C for 10 min. Subsequently, the supernatant (~ 0.9 ml) was collected in a new Eppendorf tube. One ml of extraction solvent was added to the pellet, and ultrasonication, centrifugation, and supernatant collection were repeated. Finally, ~ 1.8 ml of supernate was collected and stored at -20°C. Extracts were filtered through a 0.22 μ m PTFE syringe filter and transferred into glass vials to quantify BXs by HPLC-QqQ-MS. For qualitative analysis by UHPLC-QTOF-MS, 1ml of extraction was transferred into Eppendorf tubes, and the solvent was evaporated until dryness in a SpeedVac (SPD121P, Thermo Scientific, USA) for six hours. Then 0.2 ml of 80% MeOH/H2O was added to the Eppendorf tube, and it was centrifuged for 120 seconds. Extracts were transferred to vials for further qualitative analysis.
Quantification of BXs in Plant Material by LC-QqQ-MS
The plant extracts were analyzed by LC-MS/MS using an Agilent 1100 HPLC system coupled with a 3200 QTRAP mass spectrometer (AB SCIEX, Foster City, CA). The extracts were diluted to fit the signals from the analytes into the range of the standard curve. Negative electrospray ionization (ESI-) was used, and the mass spectrometry was operated in the multiple reaction monitoring mode (MRM). Analyst Software (version 1.6.2) was used for instrument control, data acquisition, and subsequent quantifications. The chromatography was performed using a 250 mm × 2 mm id 4 μ m Synergi Polar RP-80Å column (Phenomenex, Macclesfield, U.K.) with a flow rate of 300 μ L/min and an injection volume of 25 μ L. The temperatures of the column oven and autosampler were set at 30 and 10 °C, respectively. Two mobile phases (A: 7% acetonitrile in water and B: 78% acetonitrile in water, each containing 20 mM acetic acid) were used in a linear gradient system. The binary gradient was as following: 100 % A at 0-1 min, 92 % A at 1-3 min, 90 % A at 3-13 min, 30 % A at 13-14 min, 10 % A at 14-17 min, 100 % A at 17-25 min for equilibration. Standard compounds were used to optimize the compound‐dependent parameters (Table 1). The most intense mass transition was considered as a quantifier and the second mass transition as a qualifier. Quantifications were done based on standard curves prepared in the range of 0.39-400 ng/mL. Data points of the standard curves were weighted according to x–1.
Untargeted metabolomic profiling of plant samples using UHPLC-QTOF-MS
The metabolite profiling of roots and shoots was performed using an Agilent Infinity 1290 UHPLC system coupled to Agilent 6545 quadrupole time of flight (QTOF) mass spectrometer equipped with an Agilent dual-jet stream electrospray ion source. The whole system was controlled by Masshunter software. An HSS T3 (C18) 1.8 μ m, 2.1 mm × 150 mm was used for chromatographic separation of non-targeted compounds with a flow rate of 0.5 mL/min and a sample volume of 10 μ L was injected for each run. The temperature of the column oven was set at 40 °C. The mobile phases were (A): 100% LC-MS grade water with 0.02 M formic and (B): acetonitrile with 0.02 M formic acid. The gradient was as follows: 0 min, 0% B; 1 min, 0% B; 3 min, 10% B; 10 min, 70% B; 15 min, 100% B; 17.5 min, 100% B; 18 min, 0% B; 21 min, 0% B. Spectra were acquired in full scan MS1 and data-dependent MS2. Data were collected in ESI(±) modes with a mass range of m/z 75–1500 Da. QTOF-MS parameters were set as follows: Fragmentor voltage at 120 V, capillary voltage at 4000 V, skimmer voltage at 65 V, collision energy at 30 eV, drying gas temperature at 325 °C (8 L/min), nebulising gas pressure at 40 psi, sheath gas temperature and flow at 300 °C and 12 L/min, respectively.
Data processing, statistics, and visualization
To compare the biomass and quantification of BXs in the plants, a one-way ANOVA with post‐hoc Tukey’s test was applied to compare each group with the other groups. Comparisons were performed using GraphPad Prism version 8.4.2 (La Jolla California, USA). Acquired raw metabolomics data were converted into ABF files using the freely available converter (https://www.reifycs.com/AbfConverter/index.html). Spectra deconvolution, peak alignment, gap filling, and metabolite annotation were performed by MS-DIAL software (ver. 4.38). NIST17 (https://chemdata.nist.gov) and MoNA (https://mona.fiehnlab.ucdavis.edu) were used for mass spectral library searches. MSP file containing MS/MS spectra, MS1 isotopic spectrum, metabolite name, adducts, reverse-dot score, and tentative formula were exported from MS-DIAL for further data processing (Tsugawa et al., 2015). Detailed parameters of MS-DIAL can be found in the supplementary materials (Table S1). MS-FLO was utilized to improve the feature list’s quality by curating the features for duplicates and adducts (DeFelice et al., 2017). The data file was filtered by removing duplicate reported metabolite before submitted to statistical analysis. The untargeted metabolomics data were submitted to the log-transformation and Pareto method. Principle component analysis (PCA) was performed for all the annotated compounds with SIMCA version 17.0 (Umea, Sweden). The hierarchical clustering heatmaps of normalized metabolites (peak intensities were log-transformed and pareto scaled) were made by MetaboAnalyst version 5.0 (Chong et al., 2018). ClassyFire was used to classify the annotated compounds based on their reported InChIKey from MS-DIAL (Djoumbou Feunang et al., 2016). All the raw data from untargeted metabolomics analysis can be found in the supplementary materials.
Results
Quantification of BXs in roots and shoots of neighboring plants
Six aglycone BXs (DIBOA, DIMBOA, BOA, HBOA, HMBOA and MBOA) and six glycoside BXs (DIBOA-glc, DIMBOA-glc, HMBOA-glc, HBOA-glc, HBOA-glc-hex and DIBOA-glc-hex) were identified and quantified in the roots of eight neighboring plants species (Figure 1). DIBOA, DIBOA-glc, DIMBOA and DIMBOA-glc were found at the highest concentrations in neighboring plants. The concentration of BXs was significantly lower in Sinapis and subterranean clover compared to oat. Among the 12 BXs absorbed by the roots of neighboring plants, four (DIBOA, DIBOA-glc, DIBOA-glc-hex and DIMBOA-glc) were also present in the shoots. DIBOA-glc was the dominant BX present in the shoots of neighboring plants. The concentrations BXs in the shoots of Lolium were significantly higher than other plant species. Figure S1 shows the different concentrations of each BX in the shoot of the neighboring plant.