Rice (Oryza sativa L.) cultivar Kitaake was used as the wild type (WT) in this study. Seeds of Kitaake were obtained from the National Institute of Crop Science, Jeonju, Korea (http://www.nics.go.kr). Plants were raised in a growth chamber at 30°C under 60% humidity and 16 h light/8 h dark photoperiod.

M. oryzae strains RO1-1 and 007 were obtained from the Center for Fungal Genetic Resources, Seoul National University, Korea (http://genebank.snu.ac.kr). M. oryzae RO1-1 is virulent to rice cultivar Kitaake, whereas M. oryzae 007 is avirulent. Both M. oryzae strains were stored at −20°C and cultured on rice bran agar medium (20 g of rice bran, 20 g of sucrose, and 20 g of agar in 1 L of water) at 25°C in the dark for 10–14 days (Singh et al., 2016). Sporulation of M. oryzae cultures was induced by incubating the culture plates under continuous light for 3–4 days. M. oryzae conidia were harvested from the sporulated culture plates using 0.025% Tween20 (Sigma-Aldrich) in sterile water (Kankanala et al., 2007). A conidial suspension of M. oryzae in 0.025% (v/v) Tween 20 was adjusted to appropriate conidial concentrations using a hemacytometer.

GCaMP6f was cloned into the plant expression vector pGWB554 using standard molecular techniques, as described previously (Choi et al., 2014; Weigand et al., 2021). The resultant construct was transformed into rice cultivar Kitaake using Agrobacterium tumefaciens strain LBA4404. Briefly, 35S::GCaMP6f-mCherry (GCaMP6fmC) was delivered into rice calli using Agrobacterium-mediated transformation (Hiei et al., 1994). Transformed calli were selected on the half-strength Murashige and Skoog (1/2 MS; Sigma-Aldrich) medium (2.15 g of MS salts, 15 g of sucrose, and 3.5 g of Gelrite [Duchefa Biochemie] in 1 L of water) supplemented with 20 µg·mL^-1 hygromycin B (Duchefa Biochemie). After root and shoot formation, rice seedlings were transferred to water and acclimated for 2 days. Rice seedlings were then transferred into pots containing Baroker soil (Seoul Bio) and raised in a growth chamber.

The crystal structures of Ca²⁺-free GCaMP6 (ID: 3wlc), Ca²⁺-bound GCaMP6 (ID: 3wld), and mCherry (ID: 2h5q) were obtained from the RCSB Protein Data Bank (https://www.rcsb.org) (Shu et al., 2006; Ding et al., 2014). These three crystal structures were then used to construct crystal images of GCaMP6-mCherry using the PyMOL software (https://pymol.org) (Ding et al., 2014). The Ca²⁺ sensor construct GCaMP6f-mCherry (GCaMP6fmC) was cloned into the vector pGWB554 under the control of the Cauliflower mosaic virus (CaMV) 35S promoter ( Supplementary Figure 1A ).

The seeds of rice cultivar Kitaake and Ca²⁺ sensor (35S::GCaMP6fmC) transgenic rice lines were hulled, sterilized first with 100% ethanol for 1 min and then with 50% Clorox for 30 min, and washed three times with 3DW. The surface-sterilized seeds were then cultured on 1/2 MS medium (Sigma-Aldrich, St. Louis, MO) at 25°C under continuous light for 2 weeks. Leaves were collected from rice plants and subjected to genomic DNA extraction using the cetyltrimethylammonium bromide (CTAB) method. Plants were genotyped by PCR as described previously (Kim et al., 2011), and Ca²⁺ sensor (GCaMP6fmC) transgenic lines were identified using the hygromycin gene-specific and GCaMP6f-specific primers ( Supplementary Table 1 ). Ca²⁺ sensor (GCaMP6fmC) expression was verified in the transgenic lines by PCR using hygromycin resistance (HygR) and GCaMP6fmC primers ( Supplementary Figure 1B ). Without Ca²⁺ binding, the GCaMP6fmC has low intrinsic fluorescence. However, the green fluorescence of GCaMP6fmC increases after binding of Ca²⁺ to the calmodulin domain ( Figures 1A, B ).

Total RNA was extracted from rice plants using the TRIzol Reagent (Invitrogen) and used for cDNA synthesis. Transcript levels of rice Glutathione Reductase 1 (OsGR1), OsGR2, OsGR3, and Ubiquitin (OsUbi) genes were analyzed by reverse transcription polymerase chain reaction (RT-PCR) and real-time quantitative RT-PCR (real-time qRT-PCR) using gene-specific primers ( Supplementary Table 1 ). Transcript levels of OsGR1–3 were normalized relative to that of OsUbi and presented as mean ± standard deviation (SD) of three biological replicates. The experiments were repeated three times.

To investigate Ca²⁺-mediated iron- and ROS-dependent ferroptotic cell death in rice during M. oryzae infection, Ca²⁺ influx inhibitors, including ethylene glycol-bis(2-aminoethylether)-N, N, N′, N′-tetra-acetic acid (EGTA; Sigma-Aldrich) (Atkinson et al., 1990; Cessna and Low, 2001), verapamil hydrochloride (verapamil; Sigma-Aldrich) (Beneloujaephajri et al., 2013), N-acetyl-cysteine (NAC; Sigma-Aldrich) (Sun et al., 2012), neomycin sulfate (neomycin; ChemCruz) (Franklin-Tong et al., 1996), LiCl (Tokyo Chemical Industry) (Moyen et al., 1998), AlCl₃ (Sigma-Aldrich) (Kadota et al., 2005), and ruthenium red (RR; ChemCruz) (Bae et al., 2003), and Ca²⁺ influx enhancers, including acibenzolar-S-methyl (ASM; Sigma-Aldrich) (Brisset et al., 2000; Buonaurio et al., 2002), diamide (Sigma-Aldrich) (Kosower et al., 1969; Gilge et al., 2008), trifluoperazine hydrochloride (TFP; Sigma-Aldrich) (Kang et al., 2017), CaCl₂/calcimycin (C/C; Sigma-Aldrich) (Verma et al., 2011), CaCl₂/H₂O₂ (C/H; Sigma-Aldrich) (Rentel and Knight, 2004), rotenone (Sigma-Aldrich) (Li et al., 2003), and butylmalonic acid (BMA; Sigma-Aldrich) (Kamga et al., 2010), were applied onto rice leaf sheaths. All Ca²⁺ influx inhibitors and enhancers used in this study were treated on rice leaf sheaths with appropriate concentrations that never inhibited M. oryzae spore germination, appressorium formation and growth that are important for M. oryzae infection in rice. Among the Ca²⁺ influx enhancers, diamide and BMA were used to induce the depletion of reduced glutathione (GSH) (Gilge et al., 2008; Kamga et al., 2010), and rotenone was used to inhibit ROS burst by mitochondrial complex I (Li et al., 2003). Each Ca²⁺ influx inhibitor was applied onto a 5–7 cm long section of rice leaf sheath, together with avirulent M. oryzae 007 (5 × 10⁵ conidia·mL^–1) inoculation. Similarly, each Ca²⁺ influx enhancer (except TFP and rotenone) was applied onto a 5–7 cm long section of rice leaf sheath, together with virulent M. oryzae RO1-1 (5 × 10⁵ conidia·mL^–1) inoculation; TFP or rotenone was applied to rice sheath (5–7 cm long section) at 23 h post-inoculation (hpi) with virulent M. oryzae RO1-1. The Ca²⁺ influx inhibitor/enhancer treated and pathogen-inoculated rice leaf sheaths were incubated at 25°C under moist conditions. Rice leaf sheath samples were collected at different time points after inoculation with M. oryzae.

The epidermal cell layers excised from the inoculated rice leaf sheaths were observed under a microscope (Zeiss equipped with Axioplan 2; Campbell, CA), as described previously (Kankanala et al., 2007). The infected epidermal cells were counted and categorized into two types: cells with invasive hyphae (IH) and cells with hypersensitive response (HR) cell death. Infected cells of each infection type were quantified. The experiment was repeated three times.

Images of Ca²⁺ influx in rice sheath cells were taken using a microscope (Zeiss equipped with Axioplan2) with bright field filter, green fluorescence filter (excitation [Ex]/emission [Em] wavelengths: 450–490/515–565 nm), and red fluorescence filter (Ex/Em: 546/590 nm). At least three regions of interest (ROIs) were selected to quantify changes in Ca²⁺ levels by calculating the GCaMP6f/mCherry ratio at different time points or at 36 hpi using the ImageJ software. Green and red fluorescence signal intensities were measured using ImageJ installed with the Macro (Weigand et al., 2021). Data were exported into an Excel file, and the ratio of green to red fluorescence signal intensities was calculated. The values were compared with those of the mock control and expressed as relative fluorescence intensities (RFI). Images were taken using Leica TCS SP5 confocal microscope (Leica, Mannheim, Germany), with bright field, EGFP (Ex/Em: 488/500–540 nm), and mCherry (Ex/Em: 587/600–680 nm) filters, and merged.

Fluo-5F AM (Invitrogen) is a low affinity intracellular Ca²⁺ indicator suitable for detecting high intracellular Ca²⁺ levels ranging from 1 µM to 1 mM. Binding of Ca²⁺ to Fluo-5F is catalyzed by cellular esterases that break the ester bonds of Fluo-5F AM ( Figure 2B ). Intracellular Ca²⁺ dynamics in M. oryzae-infected rice sheaths were determined by the Ca²⁺ indicator Fluo-5F AM (Invitrogen) (Eaddy and Schnellmann, 2011). Fluo-5F AM was used to visualize intracellular Ca²⁺ ions during M. oryzae infection. Briefly, the epidermal layer recovered from the rice leaf sheaths was immersed in 0.5 M sucrose for 10 min for rapid plasmolysis. The leaf sheath epidermis was then incubated in a solution of Fluo-5F AM at a final concentration of 50 µM at 37°C for 1 h, followed by a rapid washing using 3DW. The epidermis was kept at room temperature for another 30 min to allow the reaction to occur. To visualize both Ca²⁺ and ROS simultaneously inside the same rice cell, a mixture of 50 µM Fluo-5F AM and 5 µM PO1 was used as described above. The above steps were performed under light-blocking conditions. Samples were observed under a fluorescence microscope (Zeiss equipped with an Axioplan 2; Campbell, CA) with a bright field (BF) filter and/or a green fluorescence (GF) filter. Fluorescence intensities of at least three regions of interest (ROIs) were selected to quantify changes in Ca²⁺ levels by measuring green fluorescence intensities at different time points using ImageJ software (Grossi et al., 2016). Corrected total cell fluorescence (CTCF) values were calculated as previously described (Jakic et al., 2017; Bora et al., 2021): CTCF = integrated fluorescence density – (ROI area × mean fluorescence of background readings).

Rice leaf sheaths were inoculated with avirulent M. oryzae 007 (4 × 10⁵ conidia·mL^-1) in 10 mM EGTA, and apoplastic washing fluid was prepared at 12, 24, 36, and 48 hpi as described previously (Rohringer et al., 1983), with some modifications. The M. oryzae-inoculated and/or EGTA-treated rice leaf sheaths were vacuum-infiltrated with 3DW (distilled water) and then centrifuged at 4,000× g. The supernatants were completely evaporated to dryness using an evaporator, and the residue was resuspended in a certain amount of 3DW. Apoplastic Ca²⁺ concentration in rice leaf sheaths was measured using the o-CPC method (Corns and Ludman, 1987), which is based on the reaction of Ca²⁺ with o-cresolphthalein complexone (o-CPC), resulting in the formation of an intense violet-colored complex. Briefly, 100 μL of intercellular fluid prepared from rice leaf sheath was added to a reaction solution containing 375 μL of 1 M ethanolamine (pH 10.6; Sigma-Aldrich), 71.6 μL of 100 mM 8-hydroxyquinoline (Sigma-Aldrich), 8.2 μL of 10 mM o-CPC (Sigma-Aldrich), 2.32 μL of 37% HCl (Samchun), and 3DW (up to a total volume of 1 mL). The absorbance of the Ca²⁺-o-CPC complex was measured at 575 nm using SP-2000 UV spectrophotometer (SmartPlus) (Corns and Ludman, 1987). Apoplastic Ca²⁺ concentration in rice leaf sheaths was calculated based on a standard curve obtained using 0–25 μM standard CaCl₂ (Sigma-Aldrich).

ROS (H₂O₂) localization in M. oryzae-infected rice sheath cells was determined by Peroxy Orange 1 (PO1; Sigma-Aldrich) staining. The red fluorescent ROS indicator PO1 simultaneously visualizes Ca²⁺ (green fluorescence) and ROS accumulation inside the same infected rice cell through its distinct red fluorescence emission. Briefly, the epidermal cell layer peeled off from rice leaf sheaths was soaked in 5 µM PO1 for 40 min in the dark at room temperature (Muhlemann et al., 2018; Samuel et al., 2022). The samples were washed three times with 1× phosphate-buffered saline (PBS) and observed under a fluorescence microscope (Zeiss equipped with Axioplan 2; Campbell, CA) (Ex/Em: 546/590 nm).

A chemiluminescence assay (Singh et al., 2016; Chen et al., 2022) was performed to measure ROS levels in M. oryzae-infected rice leaf sheath cells. Briefly, small sections of epidermal cell layer (0.5 cm × 0.2 cm) were transferred into individual wells of a 96-well plate, with each well containing 100 µL of chemiluminescent solution (30 µL of luminol [Bio-Rad, Hercules, CA], 1 µL of 1 mg·mL^–1 horseradish peroxidase [HRP; Jackson Immunoresearch, West Grove, PA], and 69 µL of Milli-Q water), and incubated in the dark for 5 min at room temperature. ROS chemiluminescence was detected using GloMax 96 Microplate Luminometer (Promega, Madison, WI) and expressed as relative luminescence units (RLU). Experiments were independently repeated three times.

Fe³⁺ accumulation in rice leaf sheath cells was detected and visualized by Prussian blue staining (Liu et al., 2007). Briefly, the epidermal layer of rice leaf sheath cells was isolated and then stained with a Prussian blue solution (7% potassium ferrocyanide [Sigma-Aldrich] and 2% HCl [Samchun], v:v = 1:1) for 15 h at room temperature and washed three times with 3DW. The stained epidermal cells were observed under a microscope (Zeiss equipped with Axioplan 2, Campbell, CA). Fe³⁺ were detected as a bright blue signal in sheath epidermal cells, because of their binding to ferric ferrocyanide in cells.

Lipid peroxidation in rice leaf sheath samples was determined by measuring the level of malondialdehyde (MDA), a product of unsaturated fatty acid peroxidation, using thiobarbituric acid (TBA). Rice leaf sheath tissue was ground in liquid nitrogen, and the powdered tissue was mixed with an equal amount of reaction solution (0.5% [w/v] TBA [Sigma-Aldrich], 20% [v/v] trichloroacetic acid [TCA; Sigma-Aldrich], and 0.25 mL of 175 mM NaCl in a total of 2 mL of 50 mM Tris-Cl [pH 8.0]). Samples were incubated in boiling water for 5 min and then centrifuged at 14,000 × g for 5 min at 4°C. The absorbance (optical density [OD]) of each supernatant was measured at 450, 532, and 600 nm with the SP-2000UV spectrophotometer (Woongki Science, Seoul), as previously described (Zhang et al., 2009). MDA concentration was calculated using the following equation (Dangol et al., 2019).

where C_MDA is the concentration of MDA, and OD₄₅₀ , OD₅₃₂ , and OD₆₀₀ represent the OD of the supernatant at 450, 532, and 600 nm, respectively.

The content of reduced glutathione (GSH) in rice leaf sheaths was measured spectrophotometrically. Freshly harvested conidial suspensions (4 × 10⁵ conidia·mL^-1) of M. oryzae strains, with or without Ca²⁺ influx inhibitor or enhancer, were used to inoculate rice leaf sheaths. The inoculated leaf sheaths were incubated for 48 h in the dark in a moistened box at 25°C and then ground in liquid nitrogen. Equal amounts of the powdered sample and 5% (w/v) metaphosphoric acid (Sigma-Aldrich) were mixed, and the homogenates were centrifuged at 21,000 × g for 20 min at 4°C. The supernatants were collected, and each supernatant was passed through a 0.45 µm nylon filter (Sigma-Aldrich). GSH quantification was performed as described previously (Griffith, 1980; Airaki et al., 2011). Briefly, 100 µL of the filtered supernatant was added to 600 µL of reaction buffer (100 mM sodium phosphate buffer [pH 7.5] and 1 mM EDTA [Sigma-Aldrich]). Then, 40 µL of 0.4% (w/v) 5,5’-dithiobis (2-nitrobenzoic acid) (DTNB, Sigma-Aldrich) and 350 µL of Milli-Q water were added to each sample, and the mixtures were incubated at room temperature for 5 min. To detect GSH, the absorbance of each mixture was measured at a wavelength of 412 nm using a spectrophotometer. The GSH content of each rice leaf sheath sample was quantified by constructing a calibration curve using a wide range of concentrations (0–25 µM) of standard GSH (Sigma-Aldrich).

To investigate the effect of Ca²⁺ influx on the conversion of glutathione disulfide (GSSG; oxidized glutathione) into GSH (reduced glutathione) by GR, 30 µL of 1 mM GSSG (Sigma-Aldrich) and 20 µL of 4.8 mM NADPH (nicotinamide adenine dinucleotide phosphate, reduced form; Sigma-Aldrich) were added to 50 mM Tris-Cl (pH 7.5). Then, 10 µL (0.06 U) of rice GR (Koma Biotech) and yeast GR (Sigma-Aldrich) each was added to the mixture to initiate the reduction of GSSG, and the total volume of the reaction mixture was increased to 1 mL. According to the Beer-Lambert Law, 0.1 mM NADPH has an optical density which is equal to 0.622 through a 1 cm light path. Therefore, because of the consumption of NADPH during the reduction of GSSG, the OD₃₄₀ of the sample was expected to decrease with a molar extinction coefficient of 6.22 mmol^-1·cm^-1 at 340 nm. The decrease in the absorbance of each sample was measured at 340 nm using SP-2000 UV spectrophotometer (SmartPlus). The reactions were monitored for 5 min at room temperature.

To perform the inhibition assay, different volumes of 100 mM CaCl₂ were added to the reaction system so that the final concentrations of CaCl₂ were 0, 10, 20, 30, and 50 mM. To investigate the inhibitory effect of different cations on GSSG reduction in vitro, 20 mM CaCl₂ (Sigma-Aldrich), MgCl₂ (Sigma-Aldrich), NaCl (Samchun), and KCl (Duksan) were added to the reaction system. The reaction was then monitored for 5 min by detecting the decrease in OD at 340 nm.

All the results are expressed as the mean ± standard deviation (SD). Statistical comparisons were performed by the least significant difference (LSD) test and Student’s t-test using GraphPad Prism 8 software (GraphPad Software, Inc., San Diego, CA, USA).

We first examined time-course images and levels of Ca²⁺ influx in the leaf sheaths of rice Kitaake during M. oryzae RO1-1 (virulent) and 007 (avirulent) infection ( Figure 2 ). During virulent (compatible) M. oryzae RO1-1 infection, primary hyphae grew from the appressorium, differentiated into thicker, bulbous invasive hyphae (IH) in the invaded rice cell to spread into neighboring cells ( Figure 2A ). However, avirulent (incompatible) M. oryzae 007 infection induced severe HR death response with dark brown cellular aggregates at 48 h post-inoculation (hpi) in rice leaf sheath cells. We have used the Ca²⁺ indicator Fluo-5F AM to monitor time-course changes in cytoplasmic Ca²⁺ influx by visualizing and quantifying Ca²⁺ fluorophores in living rice cells (Eaddy and Schnellmann, 2011). Fluo-5F is initially non-fluorescent. However, it becomes fluorescent when it binds to free Ca²⁺ in cells. During avirulent M. oryzae 007 infection, cytoplasmic Ca²⁺ influx began to appear at 12 hpi, increased markedly at 24 hpi, and peaked at 36 hpi ( Figures 2A, C ). At 48 hpi, extensive HR cell death occurred with dark brown cellular aggregates; however, Ca²⁺ influx levels decreased rapidly. By contrast, cytoplasmic Ca²⁺ levels were rarely or not detected in rice cells during virulent M. oryzae RO1-1 infection. We also stained rice leaf sheaths with a mixture of Fluo-5F AM and Peroxy Orange 1 (PO1) to visualize localization of Ca²⁺ and ROS (H₂O₂) accumulation (36 hpi) during infection, respectively ( Figure 3A ). The interplay of Ca²⁺ and ROS during plant immunity is a well-known phenomenon, but its mechanisms are not fully understood. Interestingly, we observed a marked co-localization of Ca²⁺ and ROS accumulation around invasive hyphae (IH) as well as inside the invaded and neighboring cells of rice leaf sheaths at 36 hpi with avirulent M. oryzae 007 ( Figure 3A ). To determine the role of ferric ions (Fe³⁺) in HR cell death during M. oryzae infection, we further stained rice leaf sheath tissues with Prussian blue solution to detect Fe³⁺, as described previously (Dangol et al., 2019). Unlike Ca²⁺ or ROS accumulation, Fe³⁺ was observed (blue color) at hyphal invasion sites in the HR cell death response of rice leaf sheath cells at 48 hpi with avirulent M. oryzae 007 ( Figure 3A ). By contrast, Ca²⁺, ROS and iron (Fe³⁺) accumulation was not detected in the rice sheath cells infected by virulent M. oryzae RO1-1 ( Figure 3A ). Avirulent M. oryzae 007 infection, but not virulent M. oryzae RO1-1 infection, significantly induced Ca²⁺ and ROS accumulation (36 hpi), iron accumulation (48 hpi), HR cell death (48 hpi), reduced glutathione (GSH, γ-L-glutamyl-L-cysteinylglycine) depletion (48 hpi), and lipid peroxidation (48 hpi) in rice leaf sheath cells ( Figures 3A, D ).

The Ca²⁺ sensor was designed as the mCherry fused to the N-terminus of GCaMP6f and consisted of two tandem fluorescent proteins (Weigand et al., 2021). The protein-based Ca²⁺ sensor (GCaMP6fmC) was used as an alternative tool to compare two separate cytoplasmic Ca²⁺ influxes in the wild type and Ca²⁺ sensor transgenic rice lines. To investigate if Ca²⁺ mediates ferroptotic cell death in rice immunity, we next generated Ca²⁺ sensor rice (cultivar Kitaake) transgenic lines as an alternative Ca²⁺ visualization tool, in which GCaMP6f-mCherry fusion Ca²⁺ reporter (GCaMP6fmC) can be expressed to emit strong green fluorescence by binding of Ca²⁺ to the calmodulin domain ( Figure 1A ; Supplementary Figure 1 ). The Ca²⁺ sensor (GCaMP6fmC) made it possible to specifically detect and visualize intracellular Ca²⁺ changes inside living rice cells without any additional calcium indicator staining. Avirulent (incompatible) M. oryzae 007 grew poorly, with only a few invasive hyphae (IH) leading to HR cell death in rice leaf sheaths ( Figures 2A , 1B, D ). Ca²⁺ influx levels in the rice leaf sheaths were measured by confocal and fluorescence microscopy ( Figure 2A , green fluorescence in the GF images; Figure 1B ; Supplementary Figures 1C , 2 ). Peroxy Orange 1 and Prussian blue staining showed the accumulation of ROS (H₂O₂) and Fe³⁺, respectively, in leaf sheath cells ( Figure 1B , red fluorescence in the RF images for ROS and blue-colored deposits for Fe³⁺). Ca²⁺ and ROS accumulation was significantly induced around the IH and within the invaded rice cells at 36-48 hpi with avirulent M. oryzae 007 ( Figures 1B, D ). By contrast, ROS accumulation was not detected in rice cells infected with virulent M. oryzae RO1-1. Avirulent M. oryzae 007, but not virulent M. oryzae RO1-1, significantly induced Fe³⁺ accumulation inside and around the IH in rice sheath cells at 36-48 hpi ( Figure 1B ). Fe³⁺ was observed as a blue color inside the IH and at the hyphal invasion sites in the HR cell death response where ROS accumulated. HR cell death was detected as vesicle-containing dark brown cellular aggregates inside the infected rice cells at 48 hpi ( Figure 1B ). During avirulent M. oryzae infection, abundant ROS bursts and HR cell death at 48 hpi were accompanied by significant depletion of reduced glutathione (GSH, γ-L-glutamyl-L-cysteinylglycine) and significant increase in lipid peroxidation [malondialdehyde (MDA) levels] ( Figures 1B, D ). GSH is one of the major water-soluble small molecule antioxidants that protect plant cells from oxidative damage (Airaki et al., 2011).

Ca²⁺ influx into the leaf sheath cells of Ca²⁺ sensor (GCaMP6fmC) transgenic line B was quantified over time during M. oryzae RO1-1 (virulent) and 007 (avirulent) infection ( Figure 1C ). During avirulent M. oryzae 007 infection, we observed M. oryzae appressorium formation at 12 hpi, hyphal invasion initiation at 24 hpi, strong Ca²⁺ influx at 18-36 hpi, abundant HR cell death at 48 hpi, along with ROS and iron accumulation, GSH depletion, and lipid peroxidation ( Figures 1B, D ). Overall, our data suggest that avirulent M. oryzae 007 infection induces higher levels of Ca²⁺ influx to mediate ROS- and iron-dependent ferroptotic cell death compared with virulent M. oryzae RO1-1 infection ( Figures 2 - 1 ; Supplementary Figure 2 ).

During avirulent M. oryzae 007 infection, Ca²⁺ -mediated ROS and iron accumulation in cells induced ferroptotic HR cell death ( Figure 4 ). Avirulent M. oryzae infection induced high levels of Ca²⁺ influx and ROS and iron accumulation in rice leaf sheath cells, as visualized by confocal and fluorescence microscopy ( Figure 4A ). We applied the membrane-nonpermeable calcium chelator ethylene glycol-bis(2-aminoethylether)-N, N, N′, N′-tetra-acetic acid (EGTA) (Ellis-Davies and Kaplan, 1994) onto rice leaf sheaths to investigate whether EGTA blocks Ca²⁺ influx from the apoplast to the cytoplasm during rice HR cell death. EGTA (10 mM) treatment inhibited the accumulation of Ca²⁺, ROS and iron, and the induction of HR cell death by avirulent M. oryzae infection, which ultimately allowed the fungal hyphae to grow normally inside the leaf sheath cells ( Figure 4A ). We also quantified the infection phenotypes (IH and HR), Ca²⁺ influx, ROS production, and GSH and MDA levels in mock (water)- and EGTA-treated rice Kitaake and Ca²⁺ sensor (GCaMP6fmC) transgenic line B at different times after inoculation with avirulent M. oryzae 007 ( Figures 4B, C ). EGTA-treated leaf sheaths contained more pathogen-infected cells but fewer HR cells, compared to the mock (water)-treated leaf sheaths during avirulent M. oryzae 007 infection ( Figure 4B ). EGTA treatment significantly inhibited Ca²⁺ influx and ROS production, correlated with strong GSH production and reduced lipid peroxidation in avirulent M. oryzae 007-infected rice cells ( Figure 4C ).

Binding of Ca²⁺ to EGTA forms the EGTA-Ca complex ( Figure 5A ). To investigate if EGTA causes chelation of apoplastic Ca²⁺ in rice cells, we prepared apoplastic washing fluid from M. oryzae-inoculated and/or EGTA-treated rice leaf sheaths ( Figure 5 ; Supplementary Figure 3 ). Ca²⁺ in the intercellular (apoplastic) fluid of rice leaf sheaths was detected using the calcium-o-cresolphthalein complexone (o-CPC) method ( Supplementary Figure 3A ). EGTA treatment markedly lowered apoplastic Ca²⁺ levels to block Ca²⁺ influx into the cytoplasm of rice leaf sheaths throughout avirulent M. oryzae 007 infection, ultimately suppressing ferroptotic HR cell death, but leading to normal fungal growth and susceptible disease ( Figures 5B, C ; Supplementary Figure 3B ). Avirulent M. oryzae 007 infection resulted in 83.93% of HR cell death in rice sheaths. However, EGTA treatment significantly reduced HR cell death to 16.60%, which is comparable to virulent M. oryzae RO1-1 infection ( Figure 5C ). Together, our data suggest that the membrane-nonpermeable calcium chelator EGTA effectively blocks apoplastic Ca²⁺ influx into the cytoplasm to inhibit Ca²⁺-mediated iron-dependent ROS accumulation leading to the formation of normal M. oryzae hyphal structure and susceptible blast disease.

Acibenzolar-S-methyl (ASM) is one of the most effective plant activators that can induce systemic acquired resistance (SAR) against a broad range of plant pathogens (Matsuo et al., 2019). During virulent M. oryzae RO1-1 infection, fungal hyphae grew well inside the rice leaf sheath epidermal cells ( Figure 6A ). However, compared with the mock (water) treatment, ASM treatment effectively induced HR cell death and ROS and iron accumulation in virulent M. oryzae-infected rice cells ( Figures 6A, B ). ASM treatment reduced the number of cells with IH, but increased the number of cells with HR, in susceptible rice leaf sheaths infected ( Figure 6B ). ROS and iron accumulation in rice cells was detected by Peroxy Orange 1 and Prussian blue staining, respectively ( Figure 6A ). We also visualized Ca²⁺ influx under confocal and fluorescence microscopy ( Figure 6A ). The quantification of Ca²⁺ influx revealed that ASM treatment induced Ca²⁺ influx compared with the mock treatment ( Figure 6C ). A chemiluminescent assay showed that ASM treatment significantly induced the accumulation of ROS in rice leaf sheath cells at 36 hpi ( Figure 6C ). However, ASM treatment inhibited GSH accumulation, but enhanced lipid peroxidation (MDA levels), in rice leaf sheath cells at 48 h after inoculation with virulent M. oryzae RO1-1 ( Figure 6C ). Collectively, these results suggest that ROS and iron accumulation, GSH depletion and lipid peroxidation are required for ASM-induced Ca²⁺-mediated ferroptotic cell death during rice-M. oryzae interactions.

Based on the knowledge of pharmacological Ca²⁺ channel modulation identified in animal systems, we selected some cytoplasmic Ca²⁺ influx regulators to investigate whether exogenous chemicals control Ca²⁺-mediated iron- and ROS-dependent ferroptotic cell death in the rice immune response. Different Ca²⁺ influx inhibitors, including EGTA (Atkinson et al., 1990; Cessna and Low, 2001), verapamil hydrochloride (verapamil) (Beneloujaephajri et al., 2013), N-acetyl-cysteine (NAC) (Sun et al., 2012), neomycin sulfate (neomycin) (Franklin-Tong et al., 1996), lithium chloride (LiCl) (Moyen et al., 1998), aluminum chloride (AlCl₃) (Kadota et al., 2005), and ruthenium red (RR) (Bae et al., 2003) were applied onto the rice leaf sheaths infected with avirulent M. oryzae 007 to compare their effects on Ca²⁺ influx, ROS accumulation, lipid peroxidation, and HR cell death ( Figures 4 , 7A ; Supplementary Figure 4 ).

Treatment with the calcium chelator EGTA dramatically inhibited Ca²⁺ influx, ROS and iron accumulation and HR cell death in rice cells compared with the other tested Ca²⁺ influx inhibitors during avirulent M. oryzae infection ( Figures 4 , 7A ). However, other Ca²⁺ influx inhibitors showed variation in the suppression of Ca²⁺ influx, ROS accumulation, lipid peroxidation (MDA level) and HR cell death ( Figure 7A ; Supplementary Figure 4 ). Verapamil treatment is known to reduce cytoplasmic Ca²⁺ levels with attenuated ROS bursts in plant cells (Beneloujaephajri et al., 2013). Inhibition of Ca²⁺ influx by verapamil reduced cytoplasmic Ca²⁺ and ROS accumulation, lipid peroxidation and HR ferroptotic cell death, as detected by Prussian blue staining ( Figure 7A ; Supplementary Figure 4 ). NAC treatment increased GSH concentration and consequently reduced ROS accumulation, lipid peroxidation, and HR cell death, accompanied by decreased Ca²⁺ levels and increased M. oryzae infection ( Figures 7A , 8D ; Supplementary Figure 4 ). LiCl inhibits the release of intracellular Ca²⁺ from vacuoles (Moyen et al., 1998). The two-pore channel 1 (TPC1) channel family is a ROS-responsive Ca²⁺ channel, and aluminum is a specific blocker for TPC1, a voltage-dependent Ca²⁺ permeable channel (Kawano et al., 2004). TPC1 plays an important role in inducing ROS-dependent cytoplasmic Ca²⁺ influx during oxidative stress (Kadota et al., 2005). Neomycin, LiCl, AlCl₃, and RR significantly inhibited Ca²⁺ influx during avirulent M. oryzae 007 infection ( Figure 7A ; Supplementary Figure 4 ). They also significantly reduced the ROS accumulation during avirulent M. oryzae 007 infection ( Figure 7A ; Supplementary Figure 4 ). They were relatively effective in limiting lipid peroxidation (MDA level) and HR cell death ( Figure 7A ; Supplementary Figure 4 ). The Ca²⁺ influx inhibitors also suppressed Fe³⁺ accumulation and HR cell death induced by avirulent M. oryzae 007 infection, which led to the successful colonization of IH as detected by Prussian blue staining ( Supplementary Figure 4 ). These data indicate that different Ca²⁺ influx inhibitors significantly limit Ca²⁺-mediated ROS and iron accumulation, and lipid peroxidation, allowing the formation of normal hyphal structures of avirulent M. oryzae 007 in rice leaf sheaths and leading to blast disease ( Figures 4 , 7A ; Supplementary Figure 4 ).

Different Ca²⁺ influx enhancers, including ASM (Brisset et al., 2000; Buonaurio et al., 2002), diamide (Kosower et al., 1969; Gilge et al., 2008), trifluoperazine dihydrochloride (TFP) (Kang et al., 2017), CaCl₂/calcimycin (C/C) (Verma et al., 2011), CaCl₂/H₂O₂ (C/H) (Rentel and Knight, 2004), rotenone (Li et al., 2003), and butylmalonic acid (BMA) (Kamga et al., 2010) were applied onto rice leaf sheaths infected with virulent M. oryzae RO1-1 to compare their effects on Ca²⁺ influx, ROS accumulation, lipid peroxidation, and HR cell death ( Figures 6 , 7B ; Supplementary Figure 5 ). ASM treatment significantly enhanced Ca²⁺ influx, ROS and iron accumulation and HR cell death in rice leaf sheath cells during virulent M. oryzae infection compared with the other tested Ca²⁺ influx enhancers ( Figures 6 , 7B ).

Other Ca²⁺ influx enhancers differentially induced Ca²⁺ influx, ROS accumulation, lipid peroxidation (MDA level) and HR cell death during virulent M. oryzae infection ( Figure 7B ; Supplementary Figure 5 ). Diamide, TFP, C/C, C/H, and BMA differentially induced Ca²⁺ influx; however, the induction of Ca²⁺ influx by rotenone was not significant. Diamide, TFP, C/C, C/H, and rotenone increased ROS levels by at least 15.98%. However, BMA did not affect ROS levels during infection. These Ca²⁺ influx enhancers also increased only slightly MDA levels in rice leaf sheath cells. However, diamide, TFP, C/C, C/H, rotenone, and BMA increased HR cell death up to 22.06-26.23% ( Figure 7B ). Prussian blue staining showed that Ca²⁺ influx enhancers effectively stimulated Fe³⁺ accumulation inside and around IH in rice cells during virulent M. oryzae infection ( Supplementary Figure 5 ). These data indicate that Ca²⁺ influx enhancers trigger robust Ca²⁺ influx to varying degrees in compatible (susceptible) rice cells due to their different mode of actions, leading to iron- and ROS-dependent ferroptotic HR cell death in response to virulent M. oryzae infection ( Figures 6 , 7B ; Supplementary Figure 5 ).

Glutathione reductase (GR) reduces glutathione disulfide (GSSG; oxidized glutathione) to produce reduced glutathione (GSH) in the presence of NADPH which can be converted to NADP⁺ in plant cells ( Figure 8A ). In vitro, increasing Ca²⁺ concentration from 10 mM to 50 mM gradually inhibited the activity of rice GR or yeast GR ( Figure 8B ; Supplementary Figures 6A, D ), indicating that higher Ca²⁺ concentration more effectively inhibits the reduction of GSSG to GSH by GR. Lineweaver-Burt plot showed non-competitive inhibition of GSSG reduction to GSH through rice GR or yeast GR by increasing Ca²⁺ concentration from 10 mM to 50 mM ( Supplementary Figures 6B, E ). Several metal ions, including Ca²⁺, Mg²⁺, K⁺, and Na⁺, which are abundant intracellular cations in plant cells, exhibited different inhibitory effects on the reduction of GSSG to GSH by rice GR or yeast GR in vitro (Tandoğan and Ulusu, 2007); however, Ca²⁺ was the most effective in inhibiting the reduction of GSSG to GSH ( Figure 8B ; Supplementary Figures 6C, F ). This suggests that Ca²⁺ may more specifically inhibit OsGR activity and deplete GSH production than the other ions Mg²⁺, K⁺, and Na⁺ in rice cells.

The antioxidant enzyme glutathione reductase (GR) is responsible for maintaining the supply of reduced glutathione (GSH) for the cellular control of ROS inside cells (Couto et al., 2016). GR plays an important role in scavenging ROS to regulate the redox state of glutathione in plants (Wang et al., 2018). During avirulent M. oryzae infection, the expression of rice cytoplasmic OsGR2 was effectively reduced relative to chloroplast OsGR1 and OsGR3 in leaf sheath cells compared with virulent M. oryzae infection ( Figure 8C ; Supplementary Figure 7 ). This indicates that avirulent M. oryzae infection significantly downregulates cytoplasmic OsGR expression to inhibit the reduction of GSSG to GSH. GSH depletion and lipid peroxidation are essential signaling events leading to iron- and lipid ROS-dependent ferroptotic cell death during avirulent M. oryzae infection (Dangol et al., 2019). During avirulent M. oryzae infection, treatment with the Ca²⁺ influx inhibitors EGTA and verapamil effectively rescued GSH depletion, increasing GSH content by 32.2%, and 28.3%, respectively ( Figure 8D ). The other Ca²⁺ influx inhibitors also more significantly increased GSH content compared with that of the control M. oryzae infection alone. However, during virulent M. oryzae infection, the Ca²⁺ influx enhancer ASM significantly decreased GSH production by 30.2% ( Figure 8D ). The other Ca²⁺ influx enhancers also more significantly reduced GSH production compared with that of the control M. oryzae infection alone.

To study the role of cytoplasmic Ca²⁺ influx in plant cells, it is absolutely necessary to accurately monitor fine spatial and temporal changes in intracellular Ca²⁺ concentration (Choi et al., 2012). The discovery and application of the Ca²⁺-sensitive bioluminescent photoprotein aequorin, made it possible to detect changes in cytoplasmic Ca²⁺ in the submicromolar range (Shimomura et al., 1963). However, despite advances in the Ca²⁺ detection techniques, the difficult detection and imaging of the weak luminescence signal of aequorin have been a limiting factor for its use in Ca²⁺ research (Hepler, 2005). Ca²⁺ indicators used for intracellular Ca²⁺ monitoring are mainly small molecules or proteins with highly specific sensitivity and responsiveness to intracellular Ca²⁺ (Kanchiswamy et al., 2014). The emerging role of resistosomes as Ca²⁺-permeable channels has stimulated the study of intracellular Ca²⁺ in plant immunity (Bi et al., 2021; Jacob et al., 2021). Accurate spatial and temporal Ca²⁺ imaging approaches are required to monitor resistsome-mediated Ca²⁺ influx into living rice cells during M. oryzae infection. Genetically encoded Ca²⁺ sensors that can be expressed inside the transgenic plant cells are preferably used because they do not require a staining step for Ca²⁺ detection.

In this study, we used fluorescence-based chemical Ca²⁺ indicator (Fluo-5F AM) and protein Ca²⁺ sensor (GCaMP6fmC), which are relatively more reliable compared to luminescence, to detect Ca²⁺ changes in living cells during cell death in rice immunity. The non-ratiometric Fluo Ca²⁺ indicator was first introduced to monitor Ca²⁺ influx in Lima bean leaf cells in response to herbivore attack (Maffei et al., 2004). Here, we selected Fluo-5F AM as a low-affinity intracellular Ca²⁺ indicator suitable for detecting high intracellular Ca²⁺ levels in the range of 1 μM to 1 mM in rice cells. We next used a genetically encoded ratiometric Ca²⁺ sensor (GCaMP6fmC) (Weigand et al., 2021) to investigate cytoplasmic Ca²⁺ influx in rice sheath cell during M. oryzae infection. The two different fluorescence-based Ca²⁺ detection tools were first shown to be highly reliable for detecting and quantifying changes in cytoplasmic Ca²⁺ levels in living rice cells during M. oryzae infection. In parallel with Ca²⁺ imaging, staining with Peroxy Orange 1 (PO1), a red fluorescent ROS indicator, revealed that the ROS burst occurs in the same region of the rice cell where Ca²⁺ influx occurs during avirulent M. oryzae infection. The red fluorescent ROS indicator PO1 could simultaneously visualize Ca²⁺ (green fluorescence) and ROS accumulation inside the same rice cell through its distinct red fluorescence emission.

In the current study, we found that avirulent M. oryzae infection triggers a robust Ca²⁺ influx, which mediates iron- and ROS-dependent ferroptotic cell death in rice. The strong ROS burst is a cellular signaling event that occurs in HR cell death in plant immunity (Van Breusegem and Dat, 2006; Jwa and Hwang, 2017). Pathogen effectors can interact with NLR receptors of host plants, leading to ROS burst and HR cell death in plant immune responses (Jones and Dangl, 2006; Dangl et al., 2013; Han and Hwang, 2017). The ZAR1 resistosome has recently been proposed to be a calcium-permeable channel triggering plant immunity and cell death (Bi et al., 2021). Ca²⁺ influx from the apoplast to the cytoplasm may mediate iron- and ROS-dependent ferroptotic HR cell death in rice immunity. Plant cells maintain a steady state Ca²⁺ concentration of 100-200 nM in contrast to the Ca²⁺ concentration in the extracellular space at 1-10 mM (Martins et al., 2013). The steep Ca²⁺ concentration gradient between the cytoplasm (~100 nM) and apoplast (~1 mM) (Stael et al., 2012) may rapidly increase cytoplasmic Ca²⁺ influx through the resistosome, a membrane-localized Ca²⁺-permeable channel, which triggers calcium signaling of ROS- and iron-dependent ferroptotic cell death in rice immunity. Avirulent M. oryzae infection sustained high iron and ROS accumulation in ferroptotic HR cell death, with a layer of high Ca²⁺ concentration near the cell membrane, GSH depletion, and lipid peroxidation. The maintenance of a persistently high Ca²⁺ concentration gradient near the cell membrane is an ionic stress, which can have a significant impact on ambient ROS burst, iron accumulation, and antioxidant defense machineries. Ferroptotic HR cell death may only be caused by prolonged Ca²⁺/ROS bursts during ETI, but not transient Ca²⁺/ROS bursts during PTI (Grant and Loake, 2000). It is thus likely that continuous Ca²⁺ influx into the cytoplasm through the calcium-permeable cation channels, including the putative NLR resistosomes, mediates iron- and ROS-dependent ferroptotic HR cell death in the rice immune response.

Different Ca²⁺ influx inhibitors such as EGTA, verapamil, NAC, neomycin, LiCl, AlCl3, and RR significantly suppressed cytoplasmic Ca²⁺ influx, ROS accumulation, glutathione (GSH) depletion, and lipid peroxidation, leading to reduced iron- and ROS-dependent ferroptotic HR cell death during avirulent M. oryzae infection. Notably, the membrane-impermeable Ca²⁺ chelator EGTA (Ellis-Davies and Kaplan, 1994; Cessna and Low, 2001) and the Ca²⁺ channel blocker verapamil (Beneloujaephajri et al., 2013) more effectively suppressed GSH depletion in rice sheaths than other Ca²⁺ influx inhibitors during avirulent M. oryzae infection. Ca²⁺ chelation by EGTA blocked Ca²⁺ influx from the apoplastic environment into the cytosol of rice cells during infection. EGTA can deplete extracellular free Ca²⁺ sources through high-affinity Ca²⁺ chelation, which essentially blocks Ca²⁺ influx from the apoplast. Resistosomes are thought to act as Ca²⁺-permeable channels that connect the cytoplasm to a high-level extracellular Ca²⁺ pool (Bi et al., 2021). The binding of Ca²⁺ to EGTA may prevent Ca²⁺ from freely diffusing across the cell membrane and consequently lower cytoplasmic Ca²⁺ levels.

Verapamil limits the transport of extracellular Ca²⁺ across the plasma membrane into the cytosol (Bergson et al., 2011). However, the inhibitory effect of verapamil on ferroptotic cell death was significantly lower than that of Ca²⁺ chelation with EGTA. NAC is a precursor of the antioxidant glutathione, which inhibits ROS levels and the oxidation of cellular glutathione (Sun et al., 2012). Neomycin acts as an inhibitor of polyphosphoinositide hydrolysis in animals and also inhibits Ca²⁺-dependent polyphosphoinositide-specific phospholipase C (PLC) activity in plants (Franklin-Tong et al., 1996). RR, a non-competitive inhibitor of the mitochondrial uniporter, is responsible for Ca²⁺ uptake into mitochondria in animal cells (Bae et al., 2003), and it also inhibits plant cell cation channels in a voltage-independent manner (White, 1996). Taken together, the seven different Ca²⁺ influx inhibitors tested in this study significantly inhibit Ca²⁺-mediated ROS, iron accumulation, GSH depletion, and lipid peroxidation, which ultimately increase the growth of invasive hyphae (IH) inside rice cells during avirulent M. oryzae infection leading to blast disease.

Plant cells have high Ca²⁺ levels in both the apoplast and internal stores, and the cytoplasmic Ca²⁺ influx can increase intracellular Ca²⁺ concentration (Stael et al., 2012). Different Ca²⁺ influx enhancers ASM, diamide, TFP, CaCl₂/calcimycin (C/C), CaCl₂/H₂O₂ (C/H), rotenone, and BMA differentially induced Ca²⁺ influx, ROS accumulation, GSH depletion, lipid peroxidation (MDA level) leading to iron- and ROS-dependent ferroptotic HR cell death during virulent M. oryzae infection. The plant defense activator ASM, a salicylic acid (SA) analogue, induces systemic acquired resistance (SAR) against a broad spectrum of plant pathogens and is widely used for crop protection (Buonaurio et al., 2002; Matsuo et al., 2019). ASM significantly enhanced Ca²⁺ influx, as well as iron and ROS accumulation leading to ferroptotic cell death in rice even during virulent M. oryzae infection. ASM induced the highest increase in cytoplasmic Ca²⁺ level among the seven Ca²⁺ enhancers tested and effectively inhibited virulent M. oryzae infection. These results suggest that the enhancement of Ca²⁺ influx by ASM may activate SAR via resistosomes in plants, thereby contributing to disease control.

Diamide is a cell-permeable chemical oxidant that can specifically target thiols of GSH and free SH groups in proteins (Kosower et al., 1969; Gilge et al., 2008). Diamide treatment reduced intracellular GSH concentration as expected. The increase in Ca²⁺ levels induced a decrease in GSH, which increased ROS accumulation, lipid peroxidation, and HR cell death, resulting in a decrease in M. oryzae growth. These results suggest that reduced cytoplasmic GSH levels play a crucial role in iron- and lipid ROS-dependent HR cell death in rice immunity. TFP directly dissociates calmodulin (CaM) from IP₃R [IP₃ (inositol 1,4,5-trisphosphate) receptor] by interacting at the TFP-binding site of CaM and opens IP₃R to release large amounts of Ca²⁺ from intracellular stores such as the endoplasmic reticulum (ER) in animals (Kang et al., 2017). Calcium ionophore A23187 (calcimycin), a hydrophobic molecule, can selectively bind to Ca²⁺ and permeate the hydrophobic interior of lipid bilayers, increasing cell permeability to Ca²⁺ (Verma et al., 2011). H₂O₂ is known to induce a biphasic cytoplasmic Ca²⁺ response in Arabidopsis seedling plants (Rentel and Knight, 2004). The recent discovery of a novel plant receptor that covalently regulates Ca²⁺ channel activity by H₂O₂ is a good example of direct activation of Ca²⁺ influx by H₂O₂ (Wu et al., 2020). Rotenone specifically targets mitochondrial complex 1 of the electron transport chain and causes impaired ATP production and oxidative stress, leading to cellular dysfunction and ultimately cell death (Li et al., 2003). BMA inhibits the dicarboxylate transporter (DIC), which transports glutathione into mitochondria in animal cells (Kamga et al., 2010). Taken together, Ca²⁺ influx enhancers could effectively promote cytoplasmic Ca²⁺ influx to trigger iron- and ROS-dependent ferroptotic cell death in rice plants during virulent M. oryzae infection.

Ca²⁺ may operate more like an essential switch to signal plant cell death and immunity (Scrase-Field and Knight, 2003). Glutathione reductase (GR) catalyzes the NADPH-dependent reduction of oxidized glutathione (GSSG) to reduced glutathione (GSH) (Tandoğan and Ulusu, 2007). Robust Ca²⁺ influx into the cytoplasm triggered reduced OsGR expression and GSH depletion during avirulent M. oryzae infection. Inhibition of cytoplasmic OsGR2 expression by avirulent M. oryzae infection may cause GSH depletion, leading to rice ferroptotic cell death. The intracellular concentration of glutathione can indicate the condition of oxidative stress in cells (Pastore et al., 2001). Within the cell, glutathione exists in a reduced state (GSH) and an oxidized state (GSSG). Glutathione, an important cellular antioxidant, plays a crucial role in disease resistance in plants by helping to regulate intracellular ROS homeostasis (Hwang et al., 1992; Parisy et al., 2007). GSH depletion disrupts intracellular ROS homeostasis and leads to iron- and ROS-dependent ferroptotic cell death (Dangol et al., 2019). GSH depletion is a common phenomenon observed during both plant and animal ferroptosis (Stockwell et al., 2017; Dangol et al., 2019).

An increase in free Ca²⁺ levels in apoplasts during avirulent M. oryzae infection may cause Ca²⁺ influx into the cytoplasm via calcium-permeable cation channels, including the putative NLR-activated resistosomes. High cytoplasmic Ca²⁺ influx can inhibit the reduction of oxidized glutathione (GSSG) to GSH by GR, leading to GSH depletion in vitro. Cytoplasmic Ca²⁺ influx from the apoplast through the putative resistosomes induces GSH depletion, which leads to the production of lipid ROS, thus promoting iron-dependent ferroptotic cell death in rice. Robust Ca²⁺, ROS and iron accumulation occurred together in the vicinity of the plasma membrane, where resistosomes are likely localized, during avirulent M. oryzae infection. This highlights the importance of GSH depletion for the initiation of Ca²⁺-mediated ferroptotic cell death in rice during M. oryzae infection. Strong Ca²⁺ influx into the cytoplasm leading to GSH depletion may be due primarily to the irreversible influx of apoplastic Ca²⁺ through the putative resistosomes, Ca²⁺-permeable selective channels (Bi et al., 2021), in the plasma membrane. It is thus likely that the irreversible Ca²⁺ influx acts as a driving force for lipid ROS production that induces ferroptotic cell death.

Our data collectively support a model of Ca²⁺-mediated iron- and lipid ROS-dependent ferroptotic cell death by calcium-permeable cation channels, including the putative NLR-activated resistosomes and consequent GSH depletion during M. oryzae infection in rice ( Figure 9 ). Recently, multiple major NLR genes, including PigmR, which confer broad-spectrum resistance to M. oryzae have been identified in the rice genome (Deng et al., 2017; Wang L. et al., 2019; Zhai et al., 2019). Plant NLRs form resistosomes upon the recognition of pathogen effectors (Bi et al., 2021); however, there is still no experimental evidence that rice NLRs such as PigmR form resistosomes that recognize M. oryzae effectors. The NLR resistosomes (Wang J. et al., 2019) act as irreversible Ca²⁺-permeable channels in the plasma membrane to trigger HR cell death in plant immunity (Bi et al., 2021). Cytoplasmic Ca²⁺ influx by calcium-permeable cation channels, including the putative resistosome induces GSH depletion and eventually triggers a ROS burst at the same site where Ca²⁺ accumulation occurs. During virulent M. oryzae infection, NLR-activated resistosome formation and GSH depletion do not occur due to increased expression of rice glutathione reductase (OsGR), and M. oryzae hyphae grow invasively inside rice cells, causing disease without ferroptotic cell death. Blast disease (susceptibility)-related cell death is ROS-dependent but iron-independent in the compatible rice–M. oryzae interaction (Dangol et al., 2021). However, during avirulent M. oryzae infection, strong cytoplasmic Ca²⁺ influx through calcium-permeable cation channels, including the putative NLR-activated resistosomes triggers iron- and ROS-dependent ferroptotic cell death owing to the increase in Fe²⁺, ROS and lipid ROS levels as well as GSH depletion by repressed OsGR2 expression. Overall, our results suggest that cytoplasmic Ca²⁺ influx from the apoplast through calcium-permeable cation channels, including the putative NLR resistosomes (Wang J. et al., 2019) inhibits the reduction of GSSG to GSH under reduced OsGR expression levels, leading to iron- and lipid ROS-dependent ferroptotic cell death in the rice immune response.