Maturation of arabidopsis seeds is dependent on glutathione biosynthesis within the embryo

Abstract

Glutathione (GSH) has been implicated in maintaining the cell cycle within plant meristems and protecting proteins during seed dehydration. To assess the role of GSH during development of Arabidopsis (Arabidopsis thaliana [L.] Heynh.) embryos, we characterized T-DNA insertion mutants of GSH1, encoding the first enzyme of GSH biosynthesis, gamma-glutamyl-cysteine synthetase. These gsh1 mutants confer a recessive embryo-lethal phenotype, in contrast to the previously described GSH1 mutant, root meristemless 1(rml1), which is able to germinate, but is deficient in postembryonic root development. Homozygous mutant embryos show normal morphogenesis until the seed maturation stage. The only visible phenotype in comparison to wild type was progressive bleaching of the mutant embryos from the torpedo stage onward. Confocal imaging of GSH in isolated mutant and wild-type embryos after fluorescent labeling with monochlorobimane detected residual amounts of GSH in rml1 embryos. In contrast, gsh1 T-DNA insertion mutant embryos could not be labeled with monochlorobimane from the torpedo stage onward, indicating the absence of GSH. By using high-performance liquid chromatography, however, GSH was detected in extracts of mutant ovules and imaging of intact ovules revealed a high concentration of GSH in the funiculus, within the phloem unloading zone, and in the outer integument. The observation of high GSH in the funiculus is consistent with a high GSH1-promoterbeta-glucuronidase reporter activity in this tissue. Development of mutant embryos could be partially rescued by exogenous GSH in vitro. These data show that at least a small amount of GSH synthesized autonomously within the developing embryo is essential for embryo development and proper seed maturation.

Figure 1.

In situ imaging of GSH using the MCB reagent in the rml1 GSH-deficient mutant. A, Wild-type seedling (scale bar = 5 mm). B, Homozygous rml1 seedling (scale bar = 1 mm) 10 d after germination. C to G, In situ labeling of root tips of wild-type and mutant seedlings with 100 μm MCB (green) and 50 μm PI (red) for 30 min. C, Maximal projection of an intact wild-type root tip (scale bar = 50 μm). D, Median optical section of a labeled wild-type root (scale bar = 20 μm). E, Maximal projection of an intact rml1 root tip (scale bar = 20 μm). F, Median optical section of a labeled rml1 root imaged under the same conditions used for the wild-type root tip shown in D (scale bar = 20 μm). G, The same median optical section of a labeled rml1 root as in F, imaged with increased detection sensitivity to visualize residual amounts of GSH (scale bar = 20 μm). Intensity bars for estimation of GSB concentrations in D, F, and G give GSB concentrations in millimolars.

Figure 2.

Characterization of T-DNA insertion alleles of Arabidopsis GSH1. A, GSH1 gene structure and location of three independent T-DNA insertions in GSH1. B, Opened silique from self-fertilized gsh1-T1/+ heterozygote showing 25% embryos with lethal phenotype. C, Wild-type control. Scale bars in B and C = 1 mm.

Figure 3.

Partial rescue of gsh1-T1 homozygous plants by GSH. Green (A and B) and white (C and D) embryos collected from siliques of a heterozygous gsh1-T1/+ plant were plated onto either nutrient medium (A and C) or nutrient medium containing 1 mm GSH (B and D) and allowed to grow for a further 24 d.

Figure 4.

Low-M_r thiols in developing wild-type and gsh1-T1 mutant seeds. Ovules were harvested 9 to 13 d after self pollination when homozygous gsh1-T1 mutants could easily be identified by their white color. Black bars = ecotype Columbia; white bars = homozygous gsh1-T1. Values are mean ± sd; n = 3.

Figure 5.

GSH imaging in intact ovules and isolated embryos. Ovules and isolated embryos were incubated in 100 μm MCB for 30 min prior to imaging. A, Superimposed transmission image (gray) and fluorescence image (green) of an ovule of a wild-type plant. The globular-stage embryo is fluorescently labeled (arrowhead). B, Ovule from a heterozygous rml1 plant after self pollination. The superimposed transmission image (gray) and fluorescence image (green) indicate the absence of GSH labeling in the globular-stage embryo (arrowhead). The presented image is representative for about 25% of the ovules in siliques of a rml1/+ heterozygote plant. C to H, Isolated embryos of wild type (C and D), rml1 (E and F), and gsh1-T1 (G and H) labeled with 100 μm MCB for GSH (green) and 50 μm PI for cell walls and viability of the cells (red). Images C, E, and G show merged green and red signals, including red autofluorescence from chloroplasts. Images D, F, and H show only the green signal for GSH labeled with MCB. The labeling pattern depicted for the rml1 and gsh1-T1 embryos is representative for 25% of the embryos of the respective self-pollinated heterozygous lines. All other embryos of these lines label like wild-type embryos. Scale bars in A and B = 50 μm; scale bars in C to H = 100 μm.

Figure 6.

In situ labeling of GSH in intact ovules. Siliques were opened without detaching the ovules from the false septum of the ovary and then placed in 100 μm MCB for 30 min. A, Superimposed transmission image (gray) and fluorescence image (green) of an opened silique with several attached ovules, all of which show strong fluorescence (scale bar = 500 μm). B, Detail of a single ovule (scale bar = 20 μm). C, Transmission image of an ovule with a globular-stage embryo (arrowhead; scale bar = 100 μm). D, Optical section through the ovule shown in C after labeling with 100 μm MCB and 50 μm PI (scale bar = 100 μm). E and F, Histochemical assay for detection of GUS activity in siliques (E) and embryos (F) of Arabidopsis plants expressing an AtGSH1-promoter∷GUS fusion (scale bars = 250 μm).