Ascorbic acid deficiency activates cell death and disease resistance responses in Arabidopsis

Abstract

Programmed cell death, developmental senescence, and responses to pathogens are linked through complex genetic controls that are influenced by redox regulation. Here we show that the Arabidopsis (Arabidopsis thaliana) low vitamin C mutants, vtc1 and vtc2, which have between 10% and 25% of wild-type ascorbic acid, exhibit microlesions, express pathogenesis-related (PR) proteins, and have enhanced basal resistance against infections caused by Pseudomonas syringae. The mutants have a delayed senescence phenotype with smaller leaf cells than the wild type at maturity. The vtc leaves have more glutathione than the wild type, with higher ratios of reduced glutathione to glutathione disulfide. Expression of green fluorescence protein (GFP) fused to the nonexpressor of PR protein 1 (GFP-NPR1) was used to detect the presence of NPR1 in the nuclei of transformed plants. Fluorescence was observed in the nuclei of 6- to 8-week-old GFP-NPR1 vtc1 plants, but not in the nuclei of transformed GFP-NPR1 wild-type plants at any developmental stage. The absence of senescence-associated gene 12 (SAG12) mRNA at the time when constitutive cell death and basal resistance were detected confirms that elaboration of innate immune responses in vtc plants does not result from activation of early senescence. Moreover, H2O2-sensitive genes are not induced at the time of systemic acquired resistance execution. These results demonstrate that ascorbic acid abundance modifies the threshold for activation of plant innate defense responses via redox mechanisms that are independent of the natural senescence program.

Figure 1.

The involvement of key redox couples in the expression of PR proteins in plant cells. The major soluble reductant couples are arranged according to their midpoint potentials, with principal protein components with which they interact indicated. DHAR, Dehydroascorbate reductase; Fd_ox, oxidized ferredoxin; Fd_red, reduced ferredoxin; FTR, ferredoxin-thioredoxin reductase; GR, glutathione reductase. Solid arrows indicate known interactions, while broken arrows indicate putative components or redox couples implicated in the signal transduction cascade.

Figure 2.

The delayed development phenotype of the vtc1 (vtc1) and vtc2 (vtc2) mutants compared to the wild type (WT). The number of weeks after sowing is indicated.

Figure 3.

The appearance of the apical meristem is delayed in the vtc1 mutant compared to the wild type. The days to flowering were compared in vtc1 (▪) and wild-type plants (▴). The experiment was performed using 100 plants and repeated three times with similar results.

Figure 4.

The appearance of individual dead cells (microlesions) in the rosette leaves of wild-type (A), vtc1 (B), and vtc2 (C) plants during development. Cell death was monitored using autofluorescence (top) or lactophenol blue staining (bottom) of leaf tissues. To aid clarity, examples of individual dead cells are marked with fine arrowheads, while large patches of dead cells are marked by thick arrowheads in the bottom images.

Figure 5.

The number of dead cells present in vtc1 and wild-type leaves through development. Values were calculated from leaves stained with lactophenol blue as shown in Figure 4, and data represent the means ± se of five separate experiments involving 25 samples per line.

Figure 6.

Pathogen-induced macroscopic and microscopic cell death symptoms are greatly decreased in Pst-infected vtc1 leaves compared to Pst-infected wild-type leaves. A and B, Lactophenol blue staining of 7-week-old wild-type and vtc1 leaves before (A) and after (B) pathogen inoculation (5 × 10⁶ cfu/mL). The black arrows in A indicate the presence of individual dead cells in vtc1 but not wild type (WT) in the absence of inoculation. Attached leaves were infiltrated with Pst at single-point sites (black arrowheads; B [top]), and pathogen-induced cell death was analyzed at 5 d postinfiltration. Middle section, The whole leaves show that patches of dead cells occur around inoculation sites in wild type but not in vtc1. Magnification of the infiltration sites (B [bottom]) shows massive cellular collapse (indicated by fat arrowheads) in the wild-type plants, whereas only a few small groups of dead cells in vtc1 (thin arrows). Bars, 50 μm.

Figure 7.

The vtc1 and vtc2 mutants display enhanced resistance to infection by Pst. Leaves from 6-week-old plants were locally infiltrated with Pst (5 × 10⁶ cfu/mL), and symptoms were analyzed at the infiltration sites (indicated by yellow arrows) at 4 d postinfection. Visible disease symptoms, displayed by Pst-infected wild-type (Col-0) leaves, were much reduced in vtc1 and vtc2 plants (A). Moreover, Pst growth curves in planta, quantified in leaf discs excised at the indicated times from the infiltration sites inoculated with a low titer of Pst (10⁵ cfu/mL), show that bacterial growth is restricted in vtc1 and vtc2 leaves compared with wild-type (Col-0) controls. Data represent the mean ± se of three independent experiments (B).

Figure 8.

Growth of Pst is increasingly restricted as vtc1 rosettes develop. Leaves from 2-, 4-, 6-, 8-, and 11-week-old plants were infiltrated with Pst (10⁵ cfu/mL). Pst growth in planta was quantified in leaf discs excised from infiltration sites at 5 d postinfiltration. Data represent the mean ± se of three independent experiments.

Figure 9.

Six-week-old vtc1 (left) and vtc2 (right) rosette leaves constitutively express PR transcripts. RNA-blot analyses of wild-type, vtc1, and vtc2 naive leaf tissues were hybridized with Arabidopsis probes (Uknes et al., 1992). The 5S rRNA probe was used as the gel-loading control.

Figure 10.

Accumulation of defense-related markers. The abundance of PR5 and SAG12 mRNA was determined on RNA blots prepared from naive leaves of 2-, 4-, 6-, 8-, and 11-week-old vtc1 and wild-type (Col-0) plants. Gel loading was controlled by hybridization with the 5S rRNA probe.

Figure 11.

Light microscopy sections of naive leaves showing the comparative cell structure of 10-week-old plants. Wild-type leaves (A) have larger cells than vtc1 (B) or vtc2 (C) leaves. Bar, 100 μm.

Figure 12.

The effect of rosette age on cell area in vtc1 (□), vtc2 (▪), and wild-type (▴) leaves.

Figure 13.

Detection of the cellular localization of GFP fused to NPR1 in transformed 4-week-old wild-type (Col-0-NPR1; top) and vtc1 plants (vtc1-NPR1; middle). Top and middle sections show fluorescence only in stomatal guard cells. The bottom section shows that SA causes fluorescence to appear in mesophyll nuclei as well as in guard cells. The bottom image shows fluorescence in 4-week-old Col-0-NPR1 24 h after spraying with 1 mm SA.

Figure 14.

The effect of rosette age on the cellular localization of GFP fused to NPR1 in transformed Col-0 (Col-0-NPR1) and vtc1 (vtc1-NPR1) plants. Fluorescence in the nuclei of mesophyll (arrowheads) and stomatal (fine arrows) guard cells is indicated for leaf samples of 6- and 8-week-old plants. For the vtc1-NPR1 8-week-old sample, the bright-field image (bottom left) of the leaves has been overlaid with the GFP fluorescence (bottom right) to illustrate the cellular localization of the fluorescence. The inset shows a higher focus image of an individual stomatal guard cell pair in the epidermis in the overlaid image.