Hordatines and Associated Precursors Dominate Metabolite Profiles of Barley (Hordeum vulgare L.) Seedlings: A Metabolomics Study of Five Cultivars

Abstract

In the process of enhancing crop potential, metabolomics offers a unique opportunity to biochemically describe plant metabolism and to elucidate metabolite profiles that govern specific phenotypic characteristics. In this study we report an untargeted metabolomic profiling of shoots and roots of barley seedlings performed to reveal the chemical makeup therein at an early growth stage. The study was conducted on five cultivars of barley: 'Overture', 'Cristalia', 'Deveron', 'LE7' and 'Genie'. Seedlings were grown for 16 days post germination under identical controlled conditions, and methanolic extracts were analysed on an ultra-high performance liquid chromatography coupled to high-resolution mass spectrometry (UHPLC-HRMS) system. In addition, an unsupervised pattern identification technique, principal component analysis (PCA), was performed to process the generated multidimensional data. Following annotation of specific metabolites, several classes were revealed, among which phenolic acids represented the largest group in extracts from both shoot and root tissues. Interestingly, hordatines, barley-specific metabolites, were not found in the root tissue. In addition, metabolomic profiling revealed metabolites potentially associated with the plants' natural protection system against potential pathogens. The study sheds light on the chemical composition of barley at a young developmental stage and the information gathered could be useful in plant research and biomarker-based breeding programs.

Keywords: Hordeum vulgare; anti-microbial metabolites; barley; liquid chromatography; mass spectrometry; metabolomics; multivariate data analysis; secondary metabolites.

Figure 1

Hordatine A biosynthesis. The first step is a coumaroyltransferase (ACT)-catalysed reaction of p-coumaroylCoA and agmatine resulting in the formation of p-coumaroylagmatine. The second step is the oxidative coupling of two molecules of p-coumaroylagmatine in the presence of peroxidase.

Figure 2

Principal components analysis (PCA) score plot models and hierarchical clustering analyses (HiCA) for shoot and root tissues of five cultivars of Hordeum vulgare (Northern Cape region of South Africa). The calculated Hoteling’s T2 with a 95% confidence interval is represented by the ellipses present in each PCA model. (A) Shoot tissue: five-component model explaining 63.2% variation (R²Xcum) and predicting 50.7% variation (Q²cum). (B) Root tissue: five-component model explaining 71.8% variation (R²Xcum) and predicting 61.2% variation (Q²cum). (C) HiCA dendrogram showing the hierarchical structure of shoot data and corresponding to the PCA model in (A). (D) HiCA dendrogram showing the hierarchical structure of root data and corresponding to the PCA model in (B). Data were acquired from hydromethanolic extracts and analysed by UHPLC–qTOF-MS in ESI(–) mode.

Figure 3

Mass fragmentation patterns of (A) p-coumaroylagmatine (m/z 277), (B) feruloylagmatine (m/z 307) and (C) sinapoylagmatine (m/z 337) characterised from barley root samples in the positive ionisation mode. The compounds exhibit identical neutral loss (m/z 130) and fragments corresponding to their dehydroxylated hydroxycinnamoyl moieties. The blue rectangles indicate the precursor ions, the orange arrows indicate the base peak fragment ions and the neutral loss fragments (m/z 130) are indicated in orange rectangles.

Figure 4

Mass fragmentation patterns of (A) hordatine A (m/z 551), (B) hordatine B (m/z 581), (C) hordatine C (m/z 611) and (D) hordatine D (m/z 641) characterised from barley shoot samples in the negative ionisation mode. The precursor ions are indicated with orange rectangles and the mass difference of 30 among the hordatines is shown by the double arrows. The compounds were all characterised by the presence of ions with m/z 131, 147 and 157 in each spectrum, and structures of these fragment ions generated on the ‘Massfrag’ tool of the MassLynx software are indicated with the orange arrow.

Figure 5

Structures of hordatines A, B, C and D as well as their corresponding glycosylated derivatives. Hordatine D is proposed by analogy with previously reported structures of A, B and C. Two chiral centres existing on carbon 2 and 3 of the dihydrobenzofuran moiety present in the core structures are indicated with black arrows. The red rectangle represents the cis/trans geometric isomer site.

Figure 6

Overview of pathway topology analysis: MetPA-computed metabolic pathways. A graphical depiction of data showing all matched pathways based on p-values and pathway impacts. Pathways with low impact to high impact (light yellow to bright red, respectively) active in barley shoots (A) and roots (B) at 16 days post germination are described according to their significance (pathway impact).

Figure 7

Interlinked pathway summary showing the biosynthesis and participation of all annotated metabolites in barley shoot and root tissues. Annotated metabolites are indicated in orange text, while the general pathway involved is indicated in red and highlighted with green boxes.

Figure 8

Multi-level 2D doughnut chart showing the classification of metabolites annotated in methanolic extracts from shoot and root tissues of five barley cultivars. The segments are representative of the number of metabolites in the class. The larger the segment, the more metabolites are present in the class. The additional layers represent the subclasses of metabolites present in the phenolic compounds class.

Figure 9

Bar graphs showing the occurrence of barley-specific hordatine metabolites and associated biosynthetic precursors across the cultivars ‘Cristalia’, ‘Deveron’, ‘Genie’, ‘LE7′ and ‘Overture’. (A) Hydroxycinnamic acid amides (HCAAs) in the shoot tissue. (B) HCAAs in the root tissue. (C) Hordatines in the shoot tissue. Each bar is representative of the average peak area corresponding to each metabolite and the error bars indicate standard deviations.