Integration of biosynthesis and long-distance transport establish organ-specific glucosinolate profiles in vegetative Arabidopsis

Abstract

Although it is essential for plant survival to synthesize and transport defense compounds, little is known about the coordination of these processes. Here, we investigate the above- and belowground source-sink relationship of the defense compounds glucosinolates in vegetative Arabidopsis thaliana. In vivo feeding experiments demonstrate that the glucosinolate transporters1 and 2 (GTR1 and GTR2), which are essential for accumulation of glucosinolates in seeds, are likely to also be involved in bidirectional distribution of glucosinolates between the roots and rosettes, indicating phloem and xylem as their transport pathways. Grafting of wild-type, biosynthetic, and transport mutants show that both the rosette and roots are able to synthesize aliphatic and indole glucosinolates. While rosettes constitute the major source and storage site for short-chained aliphatic glucosinolates, long-chained aliphatic glucosinolates are synthesized both in roots and rosettes with roots as the major storage site. Our grafting experiments thus indicate that in vegetative Arabidopsis, GTR1 and GTR2 are involved in bidirectional long-distance transport of aliphatic but not indole glucosinolates. Our data further suggest that the distinct rosette and root glucosinolate profiles in Arabidopsis are shaped by long-distance transport and spatially separated biosynthesis, suggesting that integration of these processes is critical for plant fitness in complex natural environments.

Figure 1.

GLS Content in 3-Week-Old Hydroponically Grown Wild-Type and gtr1 gtr2 Mutants. HPLC analysis of GLS content in the rosette (A), roots (B), and in the entire plant (C) of 3-week-old hydroponically grown Col-0 wild-type and gtr1 gtr2 mutants (black bars are the wild type; gray bars are gtr1 gtr2). Bars indicate sd. *Two-tailed t test (P < 0.01) versus the wild type (n = 12 to 16).

Figure 2.

Translocation of the Aliphatic 2-Propenyl GLS Administered to Leaves and Roots of Arabidopsis in the Vegetative Growth Phase. GTR1- and GTR2-dependent translocation of the aliphatic 2-propenyl GLS in the vasculature was investigated by administering 2-propenyl GLS to a leaf and roots of hydroponically grown 3-week-old wild-type or gtr1 gtr2 plants. (A) 2-Propenyl GLS distribution measured 72 h after infiltration to a leaf. The maximum injected amount was ∼10 nmol. (B) 2-Propenyl GLS distribution after 72 h of incubation in medium containing 2-propenyl GLS; 100% represents total amount of 2-propenyl GLS detected. Numbers are average ± sd; *two-tailed t test (P < 0.01) versus corresponding wild-type organ (n = 13 to 17). nd, none detected.

Figure 3.

Expression of GTR1 and GTR2 at Root Branching Points of 3-Week-Old Plants. Transgenic Arabidopsis plants expressing a NLS and GFP-GUS fusion under control of 2-kb promoter regions upstream of either GTR1 or GTR2 were analyzed by GUS staining and counterstained with propidium iodide ([A],[C], and [E]). In similar sections, confocal laser scanning microscopy was used to confirm presence of GFP in the nuclei ([B], [D], and [F]). The wild type ([A] and [B]),GTR1_pro:NLS-GFP-GUS ([C] and [D]), and GTR2_pro:NLS-GFP-GUS ([E] and [F]). All plants were 3 weeks old and in the vegetative growth phase. LR, lateral root. Bars = 50 μm.

Figure 4.

Relative Content of GLS in Micrografted Arabidopsis Plants in the Vegetative Growth Phase. Rosettes and roots from the wild type, the GLS synthesis impaired myb28 myb29 cyp79b2 cyp79b3 quadruple mutant (qko), and the gtr1 gtr2 mutant were reciprocally grafted using 4-d-old seedlings. The grafts are depicted as the genotype of rosette and roots, respectively. GLS content in the rosette and roots was analyzed by LC-MS in 3-week-old plants and related to the total amount found in wild-type homografts. Numbers represent the relative amount in the entire plant. Bars and numbers are averages; error bars and parentheses are sd. Groups in subfigures are determined by one-way ANOVA (P < 0.05) versus the wild type (n = 3 to 7). nd, none detected. (A) SC aliphatic GLS. (B) LC aliphatic GLS. (C) Indole GLS.

Figure 5.

Uptake of 8MTO in GTR1- or GTR2-Expressing X. laevis Oocytes. 8MTO accumulation in X. laevis oocytes expressing either GTR1 or GTR2 after 1 h incubation in medium containing in 100 μM 8MTO. Uptake after 1 h of 100 μM of the previously characterized (Nour Eldin et al., 2012). GTR1 and GTR2 substrate 4MTB was set to 100%. Bars are averages; error bars indicate sd (n = 4).

Figure 6.

Localization of Fluorescent GTR1 and GTR2 Protein Fusions in Roots. Roots of 3-week-old transgenic lines expressing full-length genomic regions of GTR1 or GTR2 C-terminally fused to YFP or mOrange, respectively, under control of the 2-kb endogenous promoters used in Figure 3. (A), (D), and (G) show longitudinal optical sections along the xy axis of intact roots, while subfigures (B), (E), and (H) show cross sections of the main root in lateral root branching points. Subfigures (C), (F), and (I) show cross sections of the main root alone. LR, lateral root. Bars = 50 μm. (A) to (C) Wild-type overlay of YFP and mOrange channels. (D) to (F) YFP signal from a representative plant expressing GTR1-YFP. (G) to (I) GTR2-mOrange signal from a representative plant expressing GTR2-mOrange. Arrows point to areas with expression outside the stele.

Figure 7.

Model of GLS Transport between the Rosette and Roots in Vegetative Arabidopsis. (A) Indole, SC, and LC aliphatic (SC: C3-C5 and LC: C6-C8, respectively) GLS accumulate in the rosette, whereas in roots, the predominant GLS are indole and LC aliphatic. In this model, organ-specific GLS profiles are achieved through integration of in situ synthesis and long-distance bidirectional transport between the rosette and roots in phloem and xylem. GTR1 and GTR2 are essential for establishing these profiles as they are involved in retention and storage of LC aliphatic GLS in roots. GTR1- and GTR2-dependent storage of GLS is likely to occur in areas surrounding the lateral root branching points. (B) Phloem loading of GLS across the plasma membrane is facilitated through the proton-dependent GTR1 and GTR2, and root retention can be facilitated through the same mechanism. The transport mechanism(s) responsible for indole GLS transport and export of GLS across the plasma membrane to the apoplast is currently unknown. Apo, apoplast; CC, companion cell; Pd, plasmodesmata; SE, sieve element; Sym, symplast.