Spatially patterned hydrogen peroxide orchestrates stomatal development in Arabidopsis

Abstract

Stomatal pores allow gas exchange between plant and atmosphere. Stomatal development is regulated by multiple intrinsic developmental and environmental signals. Here, we show that spatially patterned hydrogen peroxide (H₂O₂) plays an essential role in stomatal development. H₂O₂ is remarkably enriched in meristemoids, which is established by spatial expression patterns of H₂O₂-scavenging enzyme CAT2 and APX1. SPEECHLESS (SPCH), a master regulator of stomatal development, directly binds to the promoters of CAT2 and APX1 to repress their expression in meristemoid cells. Mutations in CAT2 or APX1 result in an increased stomatal index. Ectopic expression of CAT2 driven by SPCH promoter significantly inhibits the stomatal development. Furthermore, H₂O₂ activates the energy sensor SnRK1 by inducing the nuclear localization of the catalytic α-subunit KIN10, which stabilizes SPCH to promote stomatal development. Overall, these results demonstrate that the spatial pattern of H₂O₂ in epidermal leaves is critical for the optimal stomatal development in Arabidopsis.

Fig. 1. H₂O₂ induces the nuclear localization of KIN10.

a Diagram of progressive stomatal lineage in the Arabidopsis early leaf epidermis. M meristemoid cell, SLGC stomatal lineage ground cell, PC pavement cell, GC guard cell. b–d KI prevents DCMU-induced nuclear localization of KIN10-YFP in M and SLGC. Seedlings of pKIN10::KIN10-YFP transgenic plants were treated with or without 50 μM DCMU and/or 1 mM KI for 12 h. n = 113 (Mock), n = 107 (DCMU) and n = 108 (DCMU + KI) meristemoid or SLGC cells in 10 cotyledons were analyzed by ImageJ in d. e–g H₂O₂ induces the nuclear localization of KIN10-YFP in plants. Seedlings of pKIN10::KIN10-YFP transgenic plants were treated with or without 2 mM H₂O₂ and/or 50 μM CHX for 2 h. n = 102 (Mock), n = 110 (H₂O₂), n = 103 (CHX) and n = 101 (CHX + H₂O₂) meristemoid or SLGC cells in 10 cotyledons were analyzed by ImageJ in g. h–l Overexpression of CAT2 reduces the nuclear localization of KIN10-YFP. n = 102 (Col-0), n = 104 (p35S::CAT2-Myc) meristemoid or SLGC cells in 10 cotyledons were analyzed by ImageJ in l. Seedlings of pKIN10::KIN10-YFP (b–g) and pKIN10::KIN10-YFP/p35S::CAT2-Myc (h–l) transgenic plants were grown on ½ MS solid medium containing 1% sucrose under 16 h light/8 h dark photoperiod with 100 µMol/m²/s for 4 days. Serial Z-stack projection images were used for quantitative analysis. The white arrows inside the images show the areas used for line scan measurements that yielded plot profiles shown in the lower panels. C1, cytoplasmic signals of KIN10-YFP in meristemoid cells; N1, nuclear signals of KIN10-YFP in meristemoid cells; C2, cytoplasmic signals of KIN10-YFP in SLGC; N2, nuclear signals of KIN10-YFP in SLGC. Box plot shows maxima, first quartile, median, third quartile, minima. Different letters above the bars indicate statistically significant differences between the samples (Brown-Forsythe ANOVA analysis followed by Dunnett’s T3 multiple comparisons test, p < 0.05 (d, g); Two-way ANOVA analysis followed by Tukey’s multiple comparisons test, p < 0.05 (l)). Adjustments were made for multiple comparisons test. Scale bars in confocal images represent 10 µm.

Fig. 2. H₂O₂ specifically accumulates in meristemoid cells.

Measurement of H₂O₂ in the epidermal cells of wild-type cotyledon using H₂DCFDA (a–d) and BES-H₂O₂-Ac (e–h). Magnifications of the meristemoid and SLGC are shown in b and f. The white arrows inside the images show the areas used for line scan measurements that yielded plot profiles shown in c and g. The arrow labeling with A1 and A2 represents the fluorescent signals in meristemoid cells, and the arrow labeling with B1 and B2 represents the fluorescent signals in SLGC. i–l, Quantification of pHyPer fluorescent signals in cotyledon epidermal cells. Ratio imaging of Arabidopsis epidermal cells expressing HyPer (k). Panels i, j, k represent the same region excited at 488 and 405 nm and the corresponding ratio image, respectively. Emission was collected at 530 nm with a bandpass of 30 nm. Note that the red color indicates a higher ROS concentration and blue color indicates a lower level. m, DAB staining in the epidermal cells of leaves. Arrow indicates the meristemoid cell. n, o, Co-localization of H₂DCFDA and RFP signals from pSPCH::SPCH-RFP in cotyledon epidermal cells. Seedlings of Col-0, pHyPer or pSPCH::SPCH-RFP transgenic plants were grown on ½ MS solid medium containing 1% sucrose under 16 h light/8 h dark photoperiod with 100 µMol/m²/s for 4 days (a–m) or 3 days (n, o), and then stained with H₂DCFDA, BES-H₂O₂-Ac or DAB. Fluorescent signals were taken using LSM700 microscope from Zeiss. n = 104, n = 108 and n = 103 meristemoid or SLGC cells from 10 cotyledons were examined in d, h, and l, respectively. The fluorescent signals of H₂DCFDA, BES-H₂O₂-Ac or RFP were determined along a line drawn on the confocal images using ImageJ software. Scale bars in confocal images represent 20 μm. The box plot shows maxima, first quartile, median, third quartile, minima. Asterisk between the bars indicated statistically significant differences between the samples (Two-tailed student’s t test, ****p < 0.0001, ***p < 0.001).

Fig. 3. SPCH directly represses the expression of CAT2.

a–d The expression pattern of CAT2 in the cotyledon epidermal cells. Magnifications of the epidermal cells are shown in b. The pCAT2::GFP fluorescent signals were in green, PI-marked cell outlines were in red. Merged images showed the higher CAT2 expression levels in the pavement cells and the lower levels in the smaller cells where cell division occurs. n = 110 (Pavement), n = 102 (SLGC) and n = 102 (meristemoid) cells from 10 cotyledons were examined in d. e–g Co-localization of pCAT2::GFP and pSPCH::SPCH-RFP in cotyledon epidermal cells. n = 30 (Pavement), n = 30 (SLGC), and n = 30 (meristemoid) cells were examined in g. h Quantitative ChIP-PCR showed that SPCH binds to CAT2 promoter. Seedlings of p35S::YFP and p35S::SPCH-YFP were used to performed ChIP assays. The levels of SPCH binding were calculated as the ratio between p35S::SPCH-YFP and p35S::YFP, and then normalized to that of control gene PP2A. Error bars indicate standard deviation (S.D.) (n = 3 biologically independent samples). i SPCH directly binds to the promoter of CAT2 in vitro. MBP or MBP-SPCH were incubated with biotinylated DNA fragments from the PP2A and CAT2 promoters immobilized on streptavidin beads. The DNA-bound proteins were immunoblotted using anti-MBP antibody. j RT-qPCR analysis of the expression of CAT2 and BASL in wild type, spch-4 mutant and p35S::SPCH-Myc transgenic plants. PP2A gene was used as an internal control. Error bars indicate standard deviation (S.D.) (n = 3 biologically independent samples). Seedlings of pCAT2::GFP (a–d), pCAT2::GFP/pSPCH::SPCH-RFP (e–g), Col-0, spch-4 and p35S::SPCH-Myc transgenic plants (j) were grown on ½ MS solid medium containing 1% sucrose under 16 h light/8 h dark photoperiod with 100 µMol/m²/s for 4 days (a–d) or 3 days (e–g) or 6 days (j).The fluorescent signals of GFP were determined along a line drawn on the confocal images using ImageJ software. Fluorescent signals were taken using LSM700 microscope from Zeiss. The box plot shows maxima, first quartile, median, third quartile, minima. Different letters above the bars indicated statistically significant differences between the samples (Brown-Forsythe ANOVA analysis followed by Dunnett’s T3 multiple comparisons test, p < 0.05 (d, g); Adjustments were made for multiple comparisons test; Two-tailed student’s t-test, *p < 0.05 (h); One-way ANOVA analysis followed by Tukey’s multiple comparisons test, p < 0.05 (j). Adjustments were made for multiple comparisons test.). Scale bars in confocal images represent 20 μm.

Fig. 4. H₂O₂ promotes stomatal development.

a, b Quantification of epidermal cell types of Col-0 and indicated plants, expressed as percentage of total cells. GMC, guard mother cell; M, meristemoid. n = 13, 14, 14, 14, 14, 14, 15, 15, 12 independent cotyledons were examined in b. Seedlings of wild-type Col-0 and indicated plants were grown on ½ MS solid medium containing 1% sucrose under 16 h light/8 h dark photoperiod with 100 µMol/m²/s for 5 days. c, Quantification of the effects of H₂O₂ on stomatal index. n = 12, 15, 14, 12 (Col-0), n = 12 (kin10) independent plants were examined in c. Seedlings of wild-type Col-0 and kin10 mutant were grown on ½ MS solid medium containing 1% sucrose and different concentrations of H₂O₂ under 16 h light/8 h dark photoperiod with 100 µMol/m²/s for 8 days. d–g CAT mutations altered cell fate in Arabidopsis epidermis. n = 14, 15, 16, 15 independent cotyledons were examined in e. n = 13, 16, 13, 16 independent cotyledons were examined in g. Seedlings of pSPCH::nucGFP, pSPCH::nucGFP/cat2 cat3, pSPCH::SPCH-GFP, and pSPCH::SPCH-GFP/cat2 cat3 were grown on ½ MS solid medium containing 1% sucrose with or without 1 mM KI under 16 h light/8 h dark photoperiod with 100 μMol/m²/s for 3 days. Error bars indicate standard deviation (S.D.). Different letters above the bars indicated statistically significant differences between the samples (One-way ANOVA analysis (b), Two-way ANOVA analysis (c) followed by Uncorrected Fisher’s LSD multiple comparisons test, p < 0.05; No adjustments were made for multiple comparisons test; Brown-Forsythe ANOVA analysis followed by Dunnett’s T3 multiple comparisons test, p < 0.05 (e, g). Adjustments were made for multiple comparisons test). Scale bars in confocal images represent 20 µm.

Fig. 5. H₂O₂ induces KIN10 nuclear translocation through reducing the interaction between KIN10 and KINβ2.

a–d H₂O₂ reduced the interaction between KIN10 and KINβ2 in vitro and in vivo. Error bars represent the SD of three independent experiments. e–f H₂O₂ reduces the effects of KINβ2 on the KIN10-cytoplasmic retention in tobacco leaves. The tobacco leaves were transformed with KIN10-GFP or co-transformed with KIN10-GFP and KINβ2-RFP. After 2 days, the leaves were treated with or without 2 mM H₂O₂ for 6 h. n = 31, 34, 33, 31 independent cells were examined in f. The fluorescent signals of GFP (KIN10) were determined using ImageJ software. g, h H₂O₂ reduces the effects of KINβ2 on the KIN10-cytoplasmic retention in Arabidopsis leaves. n = 108 (Col-0-Mock or H₂O₂), n = 110 (p35S::KINβ2-RFP-Mock or H₂O₂), and n = 108 (kinβ2-Mock or H₂O₂) meristemoid or SLGC cells in 10 cotyledons were examined in h. The box plot shows maxima, first quartile, median, third quartile, minima. Seedlings of pKIN10::KIN10-YFP, pKIN10::KIN10-YFP/kinβ2 and pKIN10::KIN10-YFP/p35S::KINβ2-RFP were grown on ½ MS solid medium containing 1% sucrose under 16 h light/8 h dark photoperiod with 100 µMol/m²/s for 4 days, and then treated with H₂O₂ (Mock) or 2 mM H₂O₂ for 3 h. Serial Z-stack projection images were used for quantitative analysis. i, j Overexpression of KINβ2 inhibits stomatal development. n = 16, 20, 18, 16 independent cotyledons were examined in j. Seedlings of wild type, kinβ2, kinβ1 kinβ2, and p35S::KINβ2-YFP were grown on ½ MS solid medium containing 1% sucrose under 16 h light/8 h dark photoperiod with 100 µMol/m²/s for 5 days (i) or 8 days (j). Error bars indicate standard deviation (SD). k A working model for the function of hydrogen peroxide on stomatal development in Arabidopsis. SPCH directly binds the promoters of CAT2 and APX1 to reduce their expression, resulting in the high level of H₂O₂ in the meristemoids. The higher H₂O₂ in meristemoids induced the nuclear localization of KIN10 by reducing the interaction between KIN10 and KINβ2, thereby stabilized SPCH to promote stomatal development. Different letters above the bars indicated statistically significant differences between the samples (One-way ANOVA analysis followed by Tukey’s multiple comparisons test, p < 0.05 (b, d, j); Brown-Forsythe ANOVA analysis followed by Dunnett’s T3 multiple comparisons test, p < 0.05 (f, h)). Adjustments were made for multiple comparisons test. Scale bars in confocal images represent 10 µm (g) or 20 µm (e, i).