Arabidopsis thaliana ecotype Columbia was used in this study. The knockout mutant of CAT2 (cat2-1, SALK_076998) was previously reported (Zhang et al., 2020b), and the lap2-3 mutant (SAIL_192_A08) was obtained from the Arabidopsis Biological Resource Center (ABRC). Seeds were surface sterilized for 5 min in 5% commercial bleach, washed three times with sterile water, and plated on half-strength Murashige and Skoog (MS) medium (pH 5.8) (Sigma-Aldrich) containing 1% sucrose and 1% agar. Plants were stratified at 4°C for 3 days in the dark, and then transferred to chambers. The seedlings grown vertically at 22°C and 100 μmol m^–2 s^–1 illumination under 16 h light/8 h dark conditions for 5 days were transferred to half strength MS medium without or with 125 mM NaCl or 250 mM mannitol (Sigma-Aldrich) for another 5 days, and then the fresh weight was measured.

For germination assays, Arabidopsis seeds were sown on half-strength MS medium with 125 mM NaCl or 250 mM mannitol (Sigma-Aldrich), kept in the dark at 4°C for 3 days, and transferred to chambers for 5 days, and then the percentage of green cotyledons was calculated.

For glutathione (GSH) treatment, GSH was dissolved in ddH₂O to make a 100-mM stock solution and frozen at –20°C until used. When the autoclaved agar medium was cooled to below 50°C, the freshly prepared stock solution was directly added to the 1/2 MS medium to a final concentration of 100 μM, and then the medium was poured immediately into petri dishes and used to treat plant seedlings.

For cycloheximide (CHX) treatment, 5-days-old wild-type and lap2-3 mutant seedlings grown on 1/2 MS medium were transferred to liquid 1/2 MS medium containing 50 μM CHX and treated for 0, 2, 4, or 6 h. The treated seedlings were then harvested for immunoblot analysis of CAT2 protein accumulation using anti-CAT2 antibody.

Five-days-old seedlings were transferred to half-strength MS medium without or with 125 mM NaCl or 250 mM mannitol (Sigma-Aldrich) for 5 days, and then used for 3,3-diaminobenzidine (DAB) or nitroblue tetrazolium (NBT) staining to assay H₂O₂ or superoxide anion accumulation as described previously (Guan et al., 2013; Zhang et al., 2020b,c). For DAB staining, the seedlings were incubated in freshly prepared DAB staining solution (1 mg/ml DAB and 0.1% Tween 20 in 10 mM Na₂HPO₄) for 8 h, and then rinsed with 70% ethanol for several times to remove the chlorophyll. The images of the leaves were captured using a digital camera. For superoxide anion staining, the seedlings were vacuum infiltrated with 0.1 mg/ml NBT in 25 mM HEPES buffer (pH 7.6) for 2 h in darkness. Chlorophyll was removed by using 70% ethanol, and then the images of the leaves were captured using a digital camera. For DAB or NBT staining intensity quantification, images were first transformed into 8-bit type images and inverted, and the mean gray values of the image background and seedling leaves were collected using Photoshop CS5 (Adobe). The gray values were used to analyze the relative intensity of DAB staining or NBT staining. Three biological replicates were performed (a group of 10 seedlings was used as one biological sample), and thus, 30 seedlings were employed for the DAB or NBT staining intensity analysis.

Malondialdehyde (MDA) content was measured according to a previously described method (Zhang et al., 2020a). Briefly, 0.2 g of treated or untreated seedlings were ground in 1 ml of chilled reagent (0.25% thiobarbituric acid in 10% trichloroacetic acid). After incubation at 100°C for 30 min, the extracts were cooled at room temperature and centrifuged at 12,000g for 15 min. The absorbance of the supernatant was measured at 532, 450, and 600 nm. The MDA content was calculated based on the following equation: 6.45 × (OD₅₃₂ – OD₆₀₀) – 0.559 × OD₄₅₀.

Measurement of catalase activity was performed according to our previous report (Zhang et al., 2020b). Briefly, treated or untreated seedlings were ground with liquid nitrogen and then resuspended in cold extraction buffer (50 mM Tris-HCl, pH 7.4 and 10% glycerol). After centrifugation at 12,000g at 4°C for 15 min, the supernatant was used for assaying catalase activity using a Catalase Assay Kit (Beyotime Biotechnology) according to the manufacturer’s instructions.

The genomic sequence of 1,400 bp upstream of LAP2 translation start codon (ATG) and the full-length coding sequence (CDS) of LAP2 were amplified and cloned into pCAMBIA1300 between the EcoRI/KpnI and SacI/PstI restriction sites, respectively. The resulting plasmid was introduced into the lap2-3 mutant via Agrobacterium tumefaciens-mediated floral transformation. Com T4 transgenic plants were used for phenotypic analysis.

The full-length coding sequence of CAT2 was amplified using PCR and cloned into pEGAD, resulting in 35S:GFP-CAT2, and then introduced into the lap2-3 mutant using an A. tumefaciens (GV3101)-mediated transformation and the floral dip method. The T4 transgenic plants were used for phenotypic analysis.

Treated or untreated seedlings were collected for total RNA isolation, first-strand cDNA synthesis, and qRT-PCR as we described previously (Liu et al., 2018). The constitutively expressed ACTIN2/8 gene was used as an internal control. All experiments were repeated at least three times. The primer sequences are listed in Suppmentary Table 1.

To assay the interaction between CAT2 and LAP2 in yeast, the full-length CDS of LAP2 was cloned into pGADT7, and the plasmid pGBKT7–CAT2 was previously reported (Yuan et al., 2017). The yeast transformation and growth were carried out with the Matchmaker system (Clontech). Primer sequences are listed in Suppmentary Table 1.

The CDS of CAT2 and LAP2 were in cloned into the JW772 and JW771 vectors containing n-LUC and c-LUC, respectively. The resultant plasmids were transformed into A. tumefaciens (C58C1) and used to infiltrate Nicotiana benthamiana leaves. Three days after infiltration, the leaves were incubated in PBS solution containing 150 μg/ml D-luciferin potassium salt in the dark for 10 min before luminescence assay. The luciferase (LUC) image was captured by NightSHADE LB 985 according to the manufacturer’s instructions.

To assay the interaction between CAT2 and LAP2 in plants, the CDS of CAT2 was amplified using PCR with specific primers containing Flag sequence (GATTACAAGGATGACGACGATAAG), and then inserted into the pBARN vector at SmaI site behind the 35S promoter, resulting in 35S:CAT2-Flag. The CDS of LAP2 was cloned into the pEGAD vector at BamHI site, resulting in 35S:GFP–LAP2. The resultant plasmids were transformed into Agrobacterium GV3101 and used to infiltrate N. benthamiana leaves. Three days after infiltration, total proteins were extracted from the leaves and immunoprecipitated with gel-conjugated mouse anti-GFP mAb (ABclonal, #AE064, dilution at 1:1,000). The precipitants were resuspended, separated by SDS-PAGE gel, and then immunoblotted with anti-Flag.

For the bimolecular fluorescence complementation (BiFC) assays, the CDS of Arabidopsis CAT2 and LAP2 were cloned into pSP-YNE and pSP-YCE, respectively, resulting in 35S:CAT2-YFPn and 35S:LAP2-YFPc. The resultant plasmids were introduced into A. tumefaciens (GV3101) and used to infiltrate N. benthamiana leaves. Three days after infiltration, the leaves were used to observe the reconstituted YFP fluorescent signal under a confocal laser scanning microscope (Carl Zeiss LSM710 META laser scanning microscope).

The pET28a-CAT2 plasmid and His-tagged CAT2 protein expression and purification were previously reported (Yuan et al., 2017). The CDS of LAP2 was cloned into pGEX4T-1 and transformed into Escherichia coli BL21 (DE3). The GST-fused LAP2 protein was expressed with 0.5 mM isopropyl β-D-thiogalactopyranoside at 22°C for 12 h and then purified using GST agarose (Thermo Scientific) according to the manufacturer’s instructions.

The LAP2 activity with the synthetic fluorogenic substrate leucine-4-methylcoumaryl-7-amides (Leu-MCA) was performed according to a previously reported method (Waditee-Sirisattha et al., 2011). Briefly, the reaction mixture (100 μM of Leu-MCA, 50 mM Tris-Cl buffer, pH 8.5, 2.0 mM MnSO₄) was incubated at 37°C. Once the purified LAP2 or CAT2 protein was added, the fluorescence signal of 7-amino-4-methylcoumarin released by LAP2 hydrolysis activity was measured at an excitation wavelength of 360 nm and an emission wavelength of 460 nm by BioTek Cytation 3 for 45 min. According to a previous report, the fluorescence intensities at 30 min were further analyzed (Waditee-Sirisattha et al., 2011). Three independent biological replicates were performed.

Detection of GABA content in the wild-type and mutant seedlings was modified from previously reported methods (Liu et al., 2018; Yue et al., 2018). Briefly, 0.1 g of the wild-type lap2-3 and cat2-1 mutant seedlings were ground with liquid nitrogen to a fine powder and extracted using 1 ml acetonitrile. After centrifugation at 12,000g for 15 min, the supernatant was transferred and dried under nitrogen gas in a new tube. The crude extract resuspended in 100 μl acetonitrile containing 1% 2,4-dinitrofuorobenzene was mixed with 100 μl of 0.1 M Na₂B₄O₇ and then incubated at 60°C for 1 h in the dark. After derivatization, GABA standards and the samples were, respectively, separated and monitored by Agilent 1100 HPLC (C18 reversed-phase column) with the mobile phase (acetonitrile and phosphate buffer) at a flow rate of 1.0 ml/min at 360 nm.

Data are means (±SD) of three biological replicates, and the asterisks indicate statistically significant differences (^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001, Student’s t test). Bars with different letters indicate significant differences at p < 0.05 by two-way ANOVA with Tukey’s multiple comparison test.

CAT2 plays a vital role in plant salt and drought stress tolerance by scavenging stress-induced H₂O₂; thus, identification of factors regulating CAT2 function will shed light on the mechanisms underlying plant responses to high salinity or drought-induced oxidative stress. In our previous report showing that CAT2 functions in plant resistance to the necrotrophic pathogen Botrytis cinerea infection (Yuan et al., 2017), we also identified LAP2 as a CAT2-interacting protein, but we did not show the screening result previously. We therefore cloned the full-length coding sequence of LAP2 and performed a yeast two-hybrid experiment to examine the interaction between CAT2 and LAP2. Our result showed that LAP2 interacts with CAT2 in yeast (Figure 1A). The interaction between CAT2 and LAP2 was confirmed by co-immunoprecipitation (co-IP) experiment. GFP-tagged LAP2 and Flag-tagged CAT2 were transiently co-expressed in N. benthamiana leaves, and CAT2-Flag could be immunoprecipitated using anti-GFP antibody, indicating that CAT2 and LAP2 form a complex in plant cells (Figure 1B). Also, the CAT2–LAP2 interaction was verified using a LUC complementation imaging assay in which nLUC-CAT2 and cLUC-LAP2 were co-expressed in N. benthamiana leaves. Luciferase fluorescence could be detected in N. benthamiana leaves after incubated with the substrate of luciferase–luciferin (Figure 1C). Additionally, we also examined the interaction of the two proteins with BiFC assays by expressing YFPn-tagged CAT2 and YFPc-tagged LAP2. Our confocal microscopy results showed that a strong florescence signal was detected in the cytosol (Figure 1D), suggesting that LAP2 interacts with CAT2 in the cytosol, similar with the previously reported localization of CAT2–NAC1 interaction or CAT2–CPK3 interaction (Li et al., 2015).

Previous reports showed that NAC1 and CPK3 interact with CAT2 and positively regulate plant tolerance to salt and drought stress, leading us to further examine the role of LAP2 in the regulation of CAT2 activity in abiotic stresses. A previous report showed that LAP2 mutants, lap2-1 and lap2-2, are sensitive to salt and drought stresses when grown in soil, while the underlying mechanism remains largely unknown (Waditee-Sirisattha et al., 2011). To confirm these results, we identified another LAP2 T-DNA insertion mutant, SAIL_192_A08 (named as lap2-3, Supplementary Figure 1A), and assayed its sensitivity to salt and osmotic stresses by adding high concentrations of NaCl or mannitol in half-strength MS medium. Consistent with previous reports, our results showed that the lap2-3 mutant as well as cat2-1, the knockout mutant of CAT2, was indeed sensitive to high salinity and osmotic stress in terms of green cotyledon rate and changes of fresh weight (Figures 2A–C). We also generated complementation lines (Com#1 and Com#2) expressing the full-length coding sequence of LAP2 driven by its native promoter in the lap2-3 mutant (Supplementary Figure 1B) and tested their stress sensitivity phenotype. Our results showed that the LAP2 complementation lines recovered reduced tolerance to high salinity or osmotic stress (Figures 2A–C), verifying that the hypersensitivity of the lap2-3 mutant to salt or osmotic stress is due to the LAP2 gene knockout.

The interaction between LAP2 and CAT2 and the hypersensitivity of their mutants to salt or osmotic stress suggested an increase of ROS accumulation in the lap2-3 mutant. As expected, significantly higher H₂O₂ accumulation visualized by DAB staining in the leaves was detected in both lap2-3 and cat2-1 mutants compared with the wild-type plants (Figures 3A,B). H₂O₂ overaccumulation in plants may also disturb superoxide anion metabolism. Therefore, we examined the accumulation of superoxide anion levels by NBT staining in the wild-type lap2-3 and cat2-1 mutant leaves and found that superoxide anion accumulation in both mutants was also substantially higher than that in the wild type (Figures 3C,D). MDA content is an indicator for the extent of cellular lipid peroxidation caused by excess ROS (Shi et al., 2012; Zhang et al., 2020a). In line with the results of the H₂O₂ and superoxide anion accumulation, the lap2-3 and cat2-1 mutants indeed had more MDA content than the wild type after treatment with high salinity or mannitol-induced osmotic stress (Figure 3E). Furthermore, we directly determined catalase activity in the wild-type and mutant plants. Our data showed that the catalase activity in the lap2-3 and cat2-1 mutants was significantly lower than that in the wild type either in the presence or absence of high salinity or osmotic stress treatment (Figure 3F), indicating that LAP2 is required for catalase activity in plants under stress conditions. In addition, the LAP2 complementation lines had similar accumulation of H₂O₂ and superoxide anion, MDA content, and catalase activity to the wild type (Figure 3). Together, these results revealed that LAP2 confers plant salt or osmotic stress tolerance by promoting catalase activity and, thus, reducing ROS accumulation.

Our above results showed a role of LAP2 in promoting plant catalase activity, raising a possibility that LAP2 may affect CAT2 expression. Thus, we first assayed the transcripts of CAT2 in the wild-type and lap2-3 mutant seedlings treated with high concentrations of salt or mannitol using qRT-PCR. Our results showed that high salinity or osmotic stress-induced CAT2 expression in the lap2-3 mutant was not lower than that in the wild type (Figure 4A), excluding the possibility of LAP2-mediated regulation of CAT2 transcription (Figure 4A). Next, we detected the CAT2 protein levels in the wild-type and lap2-3 mutant plants using anti-CAT2 antibody which was previously reported (Zhang et al., 2020b), and our results showed that CAT2 protein accumulation in the mutant was lower than that in the wild type either treated without or with salt or osmotic stress (Figure 4B), suggesting a role of LAP2 in the regulation of CAT2 protein stability. To do this, we examined the changes of CAT2 protein accumulation in the wild-type and lap2-3 mutant plants treated with the protein synthesis inhibitor CHX and the 26S proteasome inhibitor MG132. Our results showed that CAT2 protein level in the lap2-3 mutant seedlings was dramatically decreased in the presence of CHX compared with that in the wild type, while MG132 significantly represses the degradation of CAT2 protein both in the mutant and wild-type plants (Figure 4C), revealing that LAP2 is necessary for CAT2 protein stability in plants.

If reduced stress tolerance of the lap2-3 mutant was really due to ROS overaccumulation and catalase activity deficiency, scavenging the excess ROS or overexpressing CAT2 to increase CAT2 catalase activity should rescue the phenotype of the mutant. To test this, GSH, a widely used reducing reagent that can lower down cellular ROS accumulation, was employed (Qin et al., 2020; Zhang et al., 2020b). We found that exogenously applied GSH significantly increased the tolerance of the lap2-3 mutant plant to salt or osmotic stress in terms of green cotyledon rate and fresh weight (Figures 5A,B). Furthermore, we generated 35S:GFP–CAT2 lap2-3 transgenic plants by overexpressing CAT2 in the mutant (Supplementary Figure 1C) and tested their stress tolerance. Our results showed that 35S:GFP–CAT2 lap2-3 displayed a more tolerant phenotype under salt or osmotic stress conditions compared with the lap2-3 mutant (Figures 5C,D). Together, these above results indicated that LAP2 functions in salt- or mannitol-induced oxidative stress through increasing CAT2 protein stability.

The above results indicate that LAP2, as a molecular chaperone, promotes CAT2 protein stability so as to confer plant-enhanced salt or osmotic stress. Considering their physical interaction, we wondered whether CAT2 affects LAP2 activity in turn. Therefore, His-tagged CAT2 and GST-tagged LAP2 proteins were expressed and purified from E. coli BL21 (DE3), and Leu-MCA was used as a synthetic substrate of LAP2 according to a previous report where Leu-MCA was the optimal substrate among various synthetic aminoacyl-MCAs (Waditee-Sirisattha et al., 2011). Our enzyme kinetics experiment showed that no significant changes of fluorescence signal were detected in the reaction mixture containing Leu-MCA when in the absence of LAP2 or in the presence of CAT2 alone, while fluorescence signal was strongly induced over time when LAP2 was added (Figures 6A,B). Interestingly, fluorescence signal was much higher in the presence of both CAT2 and LAP2, revealing that CAT2 stimulates LAP2 hydrolysis activity. Additionally, when more CAT2 protein was added, fluorescence signal elicited by LAP2 activity with Leu-MCA as the substrate was further increased (Figures 6A,B), indicating that CAT2 promotes LAP2 activity in a dosage-dependent manner in vitro.

The effect of CAT2 on LAP2 activity was further verified by monitoring the leucine aminopeptidase activities of the wild-type lap2-3 and cat2-1 mutant plants. Our results showed that both lap2-3 and cat2-1 harbored a lower leucine aminopeptidase activity toward the substrate Leu-MCA compared with the wild type (Figure 6C), supporting the stimulative effect of CAT2 on LAP2 activity in vivo. It is also reported that LAP2 regulates amino acid turnover and the metabolism of a vital signaling molecule, GABA, in Arabidopsis (Waditee-Sirisattha et al., 2011). Thus, we further examined the GABA content in the wild-type lap2-3 and cat2-1 mutant plants and found that, similar to the lap2-3 mutant, the cat2-1 mutant also had lower GABA content than the wild type (Figure 6D).

In summary, our study revealed that LAP2 promotes CAT2 protein stability, while CAT2 stimulates LAP2 hydrolysis activity in turn, together contributing to plant tolerance to salt or osmotic stress-induced oxidative damages.