Unveiling the root-rhizosphere environment of perennial wheat: a metabolomic perspective

Abstract

Background: Perennial grain roots grow continuously, enhancing soil carbon sequestration and forming a "holobiont" with the microbiome, essential for nutrient acquisition and stress resilience. Consequently, perennial grains serve as ideal models for investigating long-term dynamics between root systems and the rhizosphere environment. Despite their potential, the rhizosphere environment of perennial grains remains underexplored. This research utilizes an untargeted metabolomic approach to characterize the root-rhizosphere molecular signals in four new perennial grain (NPGs) lines named 235a, 280b, 11,955, and OK72, across four years of growth.

Results: Metabolomic analysis annotated 2,527 metabolites, most of which originated from fungi (30.3%), bacteria (23%), and plants (15.5%). Principal component analysis explained 54.8% of the variation between rhizosphere and root metabolites, with 8.7% variation separating 1st and 4th year root metabolites, while rhizosphere metabolites showed less variation between years. The comparison between the annual durum wheat variety and NPGs revealed 616 differentially abundant metabolites in roots and 15 in the rhizosphere, already at the 1st year of growth. In the 4th year, NPGs metabolomes diverged significantly from Thinopyrum intermedium, which stood in the soil for 11 years, with 184 root and 138 rhizosphere differentially abundant metabolites. Comparison between genotypes diversified NPGs in the 1st year, showing a higher abundance of root metabolites for OK72 compared to the other lines, including key modulators of root architecture like glutathione and serotonin, and compounds from α-linoleic acid metabolism, which are known to induce systemic resistance against pathogens and herbivore defense. Differences among NPGs also emerged in the 4th year, with OK72 separating from the other three, sharing with Thinopyrum intermedium a higher abundance of purine nucleosides and diazanaphthalenes.

Conclusions: The metabolomic analysis revealed that starting from the 1st year, the roots of NPGs produce a set of metabolites distinct from those of the annual durum species, many of which are defense molecules against biotic and abiotic stresses (e.g., syringic acid, glutathione, and α-linoleic acid pathway compounds). The OK72 genotype, which exhibits below-ground traits more aligned with perennialism, differs from the other lines in the abundance of several interesting metabolites, confirming it as an ideal parental candidate for developing new perennial wheat lines.

Keywords: Crop resilience; Perennialism trait; Rhizosphere environment; Root metabolomics.

Fig. 1

Overview of the 2527 unique metabolites annotated, subdivided according to their superclass (ClassyFire classification) (A) and their origin according to Lotus (B). The values in brackets represent the number of metabolites annotated for each superclass and their relative percentage, respectively

Fig. 2

Principal component plots (PCA) of metabolite profiles in 11955, 235a, 280b, and OK72 root or rhizosphere metabolome collected after one year (Y1) or four years (Y4) of growth. Metabolites in cv Ardente (Triticum durum) and Tpi (Thinopyrum intermedium) root and rhizosphere were collected after one and eleven years of growth, respectively. Ellipse displays 95% confidence regions for each cluster. Dotted ellipses highlight the time- and root/rhizosphere-specific clustering of NPGs

Fig. 3

Graphical representation of DAMs identified in the different comparisons (ANOVA and Fisher’s LSD post hoc p < 0.05). A Number of DAMs showing statistical differences in at least one NPG genotype compared to T. durum cv Ardente (Ard) or Thinopyrum intermedium (Tpi) in root and rhizosphere in the 1 st year (Y1) and in the 4th year (Y4), respectively. B Number of metabolites showing statistical differences within the four NPG genotypes in the 1 st year of growth in root and rhizosphere and the 4th year of growth in root and rhizosphere

Fig. 4

Comparison of the root metabolomic profiles between NPGs in the 1 st year of growth and T. durum cv Ardente. A Principal component plot (PCA) of metabolite profiles from root and grouped according to their genotype (11955, 235a, 280b, OK72, and durum wheat cv Ardente) after one year of growth. Ellipse displays 95% confidence regions for each cluster. B Hierarchical clustering analysis and heatmap visualization of root metabolites grouped according to their classes (ClassyFire classification). Each class is associated with its own Superclass by a different color. Significant metabolites in the comparison of NPGs vs. durum cv Ardente were determined using One-way ANOVA and Fisher’s post hoc (p > 0.05). Higher concentrations are shown in red, while lower concentrations are shown in blue. C Box plots showing abundance of root metabolites adenine, dalpatein-apiofuranosyl-glucopyranoside, pseudouridine, and syringic acid in NPGs at the1st year and cv Ardente. Different letters represent statistically significant differences (One-way ANOVA and post hoc Fisher’s LSD (p < 0.05, FDR < 0.05)). Means and medians are represented by black dots and lines, respectively

Fig. 5

Comparison of the rhizosphere metabolomic profiles between NPGs in the 1 st year of growth and T. durum cv Ardente. A Principal component plot (PCA) of metabolite profiles from rhizosphere and grouped according to their genotype (11955, 235a, 280b, OK72 and durum wheat cv Ardente) after one year of growth. Ellipse displays 95% confidence regions for each cluster. B Hierarchical clustering analysis and heatmap visualization of rhizosphere metabolites grouped according to their classes (ClassyFire classification). Each class is associated with its own Superclass by a different color. Significant metabolites in the comparison NPGs vs. cultivar Ardente were determined using One-way ANOVA and Fisher’s post hoc (p > 0.05). Higher concentrations are shown in red, while lower concentrations are shown in blue. C Box plots showing abundance of rhizosphere metabolites malonyldaidzin, dechloro-dehydrogriseofulvin, Cyclo (D-Tyr- L-Leu) and neoarctin A, among NPGs. Different letters represent statistically significant differences (One-way ANOVA and post hoc Fisher’s LSD (p < 0.05, FDR < 0.05)). Means and medians are represented by black dots and lines, respectively

Fig. 6

Comparison of the root metabolomic profiles between NPGs. A Principal component plot (PCA) of metabolite profiles from root and grouped according to their genotype (11955, 235a, 280b, OK72) after one year of growth. Ellipse displays 95% confidence regions for each cluster. B Hierarchical clustering analysis and heatmap visualization of root metabolites grouped according to their classes (ClassyFire classification). Each class is associated with its own Superclass by a different color. Significant metabolites in the comparison between NPGs were determined using One-way ANOVA and Fisher’s post hoc (p > 0.05). Higher concentrations are shown in red, while lower concentrations are shown in blue. C Box plots showing abundance of root metabolites glutathione, kaempferol-glucosyl-glucosyl-glucoside, serotonin, trans-Ferulic acid, hydroxy-oxo-octadecenoic acid, hydroperoxyoctadeca-trienoic acid (HpOTrE), and oxo-[pent-enyl]cyclopentenyl]octanoic acid (OPDA) among NPGs. Different letters represent statistically significant differences (One-way ANOVA and post hoc Fisher’s LSD (p < 0.05, FDR < 0.05)). Means and medians are represented by black dots and lines, respectively

Fig. 7

Comparison of the metabolomic profiles between Tpi and NPGs at the 4th year. Principal component plot (PCA) biplot showing the loadings of the top 10 metabolite classes of DAMs annotated from the comparison between the root (A) and rhizosphere (B) of the 11-year old Thinopyrum intermedium (Tpi) and 4-year-old NPGs. Group centroids are represented by larger circles