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.