Arabidopsis wild-type (Col-0) and acd5 mutant plants were used in this study. Seeds were surface sterilized with 70% ethanol for 7 min, washed three times with sterile water, and sown on half-strength Murashige and Skoog (MS) medium plates supplemented with 1% sucrose and 0.6% agar. After 2 days of stratification at 4°C, plates were kept in the greenhouse at 22°C and a 16-h light/8-h dark cycle for 8 days. Then the seedlings were transplanted onto soil and grown in the same greenhouse.

Three-week-old soil-grown plants were irrigated with water, 300 mM chloride salt solution (composed of 60 mM MgCl₂, 60 mM CaCl₂, and 60 mM KCl), or other salt solutions as specified in the figure legends once to soil capacity, typically, 15 mL of salt solution was irrigated into one well of the 32-well seedling-planting tray. Thereafter, the plants were irrigated with water as required (Yang et al., 2019). For salt spraying treatment, chloride salt solutions of different concentrations containing 0.02% Silwet-77 were sprayed onto plants. The phenotype of plants was recorded as indicated in each figure legend.

Bacterial infection was conducted as previously described (Yang et al., 2019). Briefly, after 24 h of Cl^- treatment, the 3^rd and 4^th leaves of three-week–old plants were infiltrated with 10 mM MgSO₄ (mock), Pseudomonas syringae pv maculicola (Pma) DG3, DG6, or Pseudomonas syringae pv tomato (Pst) DC3000-GFP at OD₆₀₀ as described in the corresponding figure legend. The virulent Pma DG3 is a recA derivative of Pma ES4326, and the avirulent Pma DG6 is a recA derivative of Pma ES4326 expressing the type-III effector avrRpt2 (Yao and Greenberg, 2006). These two pathogens have been maintained in Dr. Yao’s lab for years after being gifted by Dr. Greenberg. For the bacterial growth assay, leaf discs were collected at indicated time points to determine the population of bacteria as colony forming units. The colonization of DC3000-GFP was visualized under a fluorescence microscope (DM5000B, Leica).

Botrytis cinerea infection was performed as previously described (Zeng et al., 2021) with a minor modification. B. cinerea was cultured on V8 medium according to Zhang et al. (2015b) for 7 days. B. cinerea spores were suspended in potato dextrose broth (PDB) and adjusted to 1 × 10⁵ spores mL^-1. Five μL spore suspension was dropped onto the 3^rd and 4^th leaf of the plants. Then plants were covered with a dome and kept in the dark. The Spodoptera exigua larvae (2nd instar) were purchased from Henan Jiyuan Baiyun Industry (China). The feeding assay was performed as previously described (Huang et al., 2021).

Detection of the ROS induced by flg22 was performed according to Yang et al. (2019). Leaf discs were incubated with distilled water overnight in the dark. Then distilled water was substituted with 34 μg/mL L-012 (Wako), 20 μg/mL peroxidase, and 100 nM Flg22 or a Mock treatment without Flg22. Flg22-induced ROS was detected with a microplate detector (Spark, Tecan) by recording the relative light units (RLU) of luminescence.

The expression of FRK1 (At2g19190) was detected as described previously (Yang et al., 2019). Three hours post Flg22 infiltration, the samples for total RNA isolation were collected. Total RNA was extracted using the E.Z.N.A. plant RNA Kit (R6827-01, Omega) according to the manufacturer’s instructions. One µg RNA was reverse-transcribed into cDNA using the PrimeScript RT kit with genomic DNA Eraser (Takara). Ten µL reaction system containing 1 µL of diluted cDNA sample (1:20), 5 µL SYBR Premix Ex Taq II (Takara), and 0.4 µL gene-specific primers was mixed in each 96-well plate well and then quantitatively analyzed by quantitative PCR (Step One plus, ABI). ACTIN2 (At3g18780) was analyzed as the internal reference gene to normalize values for transcript abundance. The 2^-△△CT method (Livak and Schmittgen, 2001) was used to calculate the relative expression level of FRK1. Three technical replicates were performed for each template and primer combination. The primers for amplification are as follows: FRK1-F, 5′-GCCAACGGAGACATTAGAG-3′; FRK1-R, 5′-CCATAACGACCTGACTCATC-3′; ACTIN2-F, 5′-GGTAACATTGTGCTCAGTGGTGG-3′; ACTIN2-R, 5′-GGTGCAACGACCTTAATCTTCAT-3′.

Transcriptome sequencing was conducted on the Illumina PE150 platform by Verygenome Technology (Guangzhou, China). The clean reads were aligned to the reference genome using HISAT2 (Sirén et al., 2014). Differential expression analysis was performed using the ‘DESeq2’ R package (1.18.0) (Wang et al., 2010). Differentially expressed genes were defined by log₂(fold change) ≥ 1 and false discovery rate (FDR) < 0.05. Gene Ontology analysis was performed using the clusterProfiler R package (Yu et al., 2012) with FDR < 0.05. We have deposited the full RNA sequencing data in NCBI (PRJNA1073023, https://www.ncbi.nlm.nih.gov/sra/PRJNA1073023).

Cell death was visualized by trypan blue staining as previously described (Yang et al., 2019; Zeng et al., 2021). Plant leaves collected at the indicated time points were submerged in staining buffer (lactic acid: water-saturated phenol: glycerol: distilled water = 1:1:1:1, V:V:V:V) containing 0.25 mg/mL trypan blue, and incubated in boiling water for 2 min, then decolorized in chloral hydrate solution. The blue cell death lesions were recorded by a stereo microscope with the ×20 objective (SteREO Lumar.V12, Carl Zeiss), and analyzed by ImageJ software (ImageJ 1.48v).

Callose deposition was detected by aniline blue staining (Bi et al., 2014). At 12 hours post infiltration with 100 nM Flg22 or Mock (10 mM MgSO₄), leaf samples were fixed in FAA (formaldehyde: acetic acid: ethanol: water=2:1:9:8, V:V:V:V) for 5 min. Then the samples were transferred into alcoholic lactophenol (95% ethanol:lactophenol=2:1, V:V), and incubated in boiling water for 2 min. The leaves were washed twice in 50% ethanol for 2 min. After being rinsed in distilled water, the samples were transferred into aniline blue staining buffer (0.01% aniline blue in 150 mM K₂HPO₄, pH 9.5) overnight, and visualized under a fluorescence microscope with the ×5 objective (DM5000B, Leica) and analyzed by ImageJ software.

Data are presented as means ± SE. Significant differences were determined by ANOVA post hoc tests (P < 0.05) using different letters or determined by Student’s t-test with P-values presented in each figure legend. Numbers of biological replicates or technical repeats are also given in each figure legend.

To decipher the mechanism by which 300 mM NaCl inhibits cell death in acd5 mutants, we tested salts with different anions or cations. As shown in Supplementary Figure S1A , the cell death lesions on acd5 were thoroughly suppressed by irrigation with 300 mM NaCl, 300 mM KCl, and 150 mM CaCl₂, but were not suppressed by 300 mM NaNO₃, 300 mM KNO₃, 150 mM Na₂SO₄, or 150 mM K₂SO₄. This result suggests that the major ion suppressing cell death in acd5 is chloride (Cl^-).

Furthermore, we tested a compound salt containing MgCl₂, CaCl₂, and KCl (Zhang et al., 2015a) to confirm the effect of Cl^- on cell death. Indeed, 300 mM Cl^- (60 mM MgCl₂, 60 mM CaCl₂, 60 mM KCl) also suppressed cell death in acd5 ( Supplementary Figure S1B ). Together, these results show that a certain concentration of Cl^- can suppress the cell death caused by interrupted sphingolipid metabolism. To limit the influence of individual cations, we used the MgCl₂–CaCl₂–KCl compound solution for Cl^- treatment in subsequent experiments described below.

Since NaCl also suppresses cell death lesions caused by pathogenic bacteria and enhances plant resistance (Singh et al., 2014; Yang et al., 2019), we next tested whether Cl^- could recapitulate this effect. To this end, plants pretreated with Cl^- were challenged with virulent Pseudomonas syringae pv. maculicola (Pma) DG3 and avirulent Pma DG6 (carrying the effector protein avrRpt2). Pma DG3 produces cell death induced by infection and Pma DG6 produces cell death as part of the hypersensitive response and disease resistance. As shown Figure 1 , Cl^- irrigation suppressed the cell death induced by Pma DG3 and Pma DG6, indicating Cl^- could suppress cell death that resulted from susceptibility in response to virulent pathogen infection, and to resistance (hypersensitive response-related cell-death) in response to avirulent pathogen infection. We also tested whether the cations introduced by the chloride salt altered plant response to pathogenic bacteria. As shown in Supplementary Figure S2 , although Cl^- salt could effectively suppress the cell death induced by Pma DG3, the solution containing equivalent cations, such as 60 mM K⁺, 60 mM Mg²⁺, and 60 mM Ca²⁺ released by KNO₃, 60 mM MgSO₄, and 60 mM Ca(NO₃)₂, respectively, was ineffective to confine the cell death caused by DG3 infection. This result further confirmed that chloride ion irrigation could substantially alter plant cell fate confronting pathogen challenge. Moreover, spraying with Cl^- suppressed the lesion development caused by spray infection with DG3 ( Supplementaryray Figure S3 ). Thus, these results suggested that Cl^- treatment could not only suppress the cell death in the lesion mimic mutant acd5, but could also suppress cell death induced by bacterial pathogens.

Plant cell death is tightly related to resistance to pathogens. Therefore, we analyzed the interactions between plants and different pathogens to decipher the impact of Cl^- treatment on pathogenesis. GFP-tagged Pseudomonas syringae pv. tomato (Pst) DC3000 (DC3000-GFP) was inoculated to visualize Pst colonization of the phyllosphere in situ (Kong et al., 2020). As shown in Figures 2A, B , DC3000-GFP remained pathogenic to Arabidopsis, and multiplied in plants irrigated with H₂O, but not in plants irrigated with Cl^-.

Salicylic acid (SA) is the major phytohormone supporting plant resistance to bacterial pathogens. The Arabidopsis transgenic line NahG, which overexpresses a bacterial SA hydroxylase, accumulates little or no SA, leading to plant susceptibility to pathogens. In the present study, consistent with previous reports, NahG increased plant susceptibility to DG3, as indicated by much more severe cell death and more robust bacterial proliferation compared with control plants ( Figure 2C ). Intriguingly, Cl^- pretreatment also suppressed the cell death induced by DG3 and dampened bacterial growth in NahG plants ( Figure 2C ). However, the multiplication of DG3 in NahG-Cl^- was still significantly higher than that in WT-Cl^- treatment, indicating that the SA pathway played a role in Cl^–mediated enhancement of plant resistance to DG3.

Previously, Singh et al. (2014) discovered that certain abiotic stresses prime PTI, thus enhancing plant resistance to a subsequent biotic challenge. To check whether Cl^- treatment also changed the properties of PTI, we examined responses to a molecular pattern, the Flg22 epitope of flagellin. As shown in Figure 2D , Flg22 induced a rapid reactive oxygen species (ROS) burst in plants treated with Cl^- and control plants treated with H₂O. Cl^- treatment did not significantly alter the relative expression level of the PTI marker gene FRK1 and Flg22-induced callose deposition ( Figures 2E, F ). These results suggested that, in the present study, the enhancement of plant resistance to bacterial infection was partially dependent on the SA pathway, but not PTI responses.

To decipher the effect of Cl^- treatment on plant resistance, we next conducted transcriptome analysis. To this end, samples collected 9 hours post bacterial inoculation (hpi) were sequenced and analyzed ( Supplementary Figure S4 ). Compared with H₂O-Mock, C1^–Mock treatment downregulated biotic resistance–related genes ( Supplementary Figures S4B, D ). This is consistent with a report that salt treatment can suppress the SA pathway (Yang et al., 2019). Consistent with the observed course of DG3 infection, the genes related to immune response, such as SA, jasmonate, and ethylene pathways, were strongly induced in plants infected with DG3 after H₂O and Cl^- pretreatment compared to H₂O-Mock, Cl^–Mock, respectively ( Supplementary Figures S4B, D ). Genes related to response to toxic substance and benzene-containing compound metabolic process were upregulated in Cl^–DG3 compared with H₂O-DG3, suggesting these two processes might take part in the mechanism by which Cl^- enhances plant disease resistance.

Next, we investigated whether Cl^- irrigation could enhance plant resistance to a necrotrophic pathogen, the fungi Botrytis cinerea. At 24 h after Cl^- irrigation, a suspension of B. cinerea spores was dropped onto the leaves. As shown in Figures 3A, B , Cl^- irrigation greatly suppressed the lesions induced by B. cinerea infection. These results show that Cl^- irrigation can efficiently enhance plant resistance to necrotrophic pathogen infection.

We also tested that if Cl^- treatment could change plant resistance to herbivorous insect attack. At 24 h after Cl^- irrigation, second-instar larvae of Spodoptera exigua were placed on WT plants. Intriguingly, larval growth was greatly stunted in WT plants irrigated with Cl^- ( Figures 3C, D ). These results indicated that Cl^- treatment also enhanced plant resistance to an herbivorous insect.

As the main ions of salinity, Na⁺ and Cl^- have been long considered toxic ions that impair plant growth and development in the field (Munns and Tester, 2008; Tavakkoli et al., 2011). However, exploration of the crosstalk between abiotic and biotic stresses reveals that certain abiotic stresses or inorganic ions can improve plant growth, nutrient utilization, and plant responses to biotic stresses (Wang et al., 2013; Singh et al., 2014; Geilfus et al., 2015; Wege et al., 2017; Zhang et al., 2017; Franco-Navarro et al., 2019; Yang et al., 2019; Rosales et al., 2020). Cl^- salt treatment can change the morphology of leaf cells and reduce stomatal density, thus increasing water use efficiency, and improve photosynthetic performance, possibly by stimulating chloroplast biogenesis (Franco-Navarro et al., 2019). Cl^- also facilitates the utilization of NO₃ ^- in plants, which provides a new guideline for developing agricultural fertilizers with an optimal NO₃ ^-/Cl^- ratio to reduce excessive NO₃ ^- use (Rosales et al., 2020). Here, we discovered that Cl^- can also suppress the accelerated cell death in acd5 ( Supplementary Figure S1 ). This indicates that, besides promoting plant growth and nitrogen use, Cl^- also participates in cell death regulation in plants, possibly by provoking antagonistic crosstalk between the phytohormone abscisic acid and the SA pathway, as observed for NaCl treatment (Yang et al., 2019). Since Cl^- channels and Cl^- itself as a signaling ion are tightly related to cell proliferation and innate immunity in animal cells (Valdivieso and Santa-Coloma, 2019; Lüscher et al., 2020; Chen et al., 2022), we also suspect that Cl^- may inhibit cell death in acd5 in a phytohormone-independent manner. Therefore, further research to decipher the molecular mechanism, for example by screening for suppressors of Cl^–mediated suppression of cell death in acd5, should provide interesting insight.

Compatible or incompatible plant–microbe interactions show differing extents of host cell death. We found that Cl^- salt treatment effectively suppressed the cell death caused by not only the hemibiotrophic pathogen DG3, which undergoes a compatible interaction with Arabidopsis, but also DG6, in which recognition of the effector protein avrRpt2 by plant defenses induces an incompatible interaction ( Figure 1 ; Yao and Greenberg, 2006). Cl^- also enhanced plant resistance to pathogen infection by suppressing bacterial multiplication in intercellular spaces, and this enhancement is partially dependent on the SA pathway but not PTI responses, especially those mediated by FLS2 ( Figure 2 ). This result indicated that the mechanism by which Cl^- treatment enhanced disease resistance differed from that resulting from anion channel mutants or stimulated by channel inhibitors (Guo et al., 2014a, 2014b). Moreover, in the absence the pathogen infection, the genes related to immune responses and defense metabolites were downregulated ( Supplementary Figure S4D ), which is consistent with the observation that salt treatment suppresses the SA pathway. In addition, this inhibitory effect ceased when plants encountered a pathogen infection. These data suggested that the effect of chloride salt on biotic stress-responsive genes can also be regulated by upstream signaling pathways that are triggered by other environmental clues. RNA sequencing data showed the genes related to responses to toxic substance and benzene-containing compound metabolic process were enhanced in Cl^–DG3 compared with H₂O-DG3 treatment ( Supplementary Figure S4 ), indicating these two processes play a role in Cl^- salt–induced plant biotic resistance response, although the molecular mechanism needs further elucidation.

Cl^- treatment also protected plants from the necrotrophic fungi B. cinerea and the herbivore S. exigua, which involve jasmonic acid (JA) signaling ( Figure 3 ). This broad-spectrum resistance and the observation that the SA and JA pathways undergo antagonistic crosstalk (Pieterse et al., 2012), suggest that Cl^- treatment does not merely exploit phytohormone pathways to enhance plant biotic resistance. We suspect Cl^- could stimulate the accumulation of defense metabolites that are toxic to necrotrophic fungi and insects, a hypothesis that will need further validation.

Screening for suppressors of Cl^–suppressed cell death in acd5 may enable us to uncover the effect of Cl^- on plant cell death, and we will also learn from experiments in mammalian and yeast systems. Identifying the defense metabolites that are differentially abundant in plants under the combined treatment of Cl^- irrigation and biotic stress may also be helpful and intriguing. Taken together, this research provides new clues to study the crosstalk between abiotic and biotic responses and underscores the influences and biological functions of the inorganic ion Cl^- on plant cell death and defensive mechanisms.