P. ternata (Thunb) Breit tubers were collected from the Experimental Farm of Huaibei University (E116.8, N34.0) and germinated in potted soil. The potted plants were kept at 25°C under a 16-h photoperiod and light intensity of 35 μmol m^-2s^-1. The leaves, roots, petioles, tubers and flowers were collected separately from the uniform-sized three-leaf stage seedlings treated with various abiotic stresses using previously published methods (Zhang et al., 2021). For temperature treatment, the potted plants were moved to RGC-1500C artificial climate incubators (Hefei Youke Instrument Equipment Co., Ltd, Hefei, China) at lower (4°C) and higher (40°C) temperatures, with those incubated at 25°C as control. The incubators had an illumination scheme of 16 h/8 h (light/dark) with a light intensity of 35 μmol m^-2s^-1. For other treatments, the three-leaf stage plants were rinsed in ddH₂O and cultured in half-strength MS liquid medium containing 150 mM NaCl, 25% polyethylene glycol (PEG6000), or 100 μM ABA for salt, drought and abscisic acid (ABA) treatments, respectively, with those maintained in solely half-strength MS liquid medium as control. Whole plants were collected at 0, 6, 12, 24 and 72 h of treatment and rinsed in ddH₂O. Three plants were bulked as one sample and three biological replicates were contained for each treatment. The sampled tissues or whole plants were frozen immediately in liquid nitrogen and preserved at –80°C for further analyses.

Arabidopsis Col-0 plants were grown in 2-L pots and cultured in RGC-1500C artificial climate incubators (Hefei Youke Instrument Equipment Co., Ltd, Hefei, China) at 22°C under a 16-h photoperiod and light intensity of 50 μmol m^-2s^-1.

A putative cDNA sequence of the PtCAT2 gene was obtained from our previously published full-length transcriptome data (Xue et al., 2019) and amplified using the primer pair PtCAT2-Full-F/PtCAT2-Full-R ( Table S1 ). The PCR products were purified and cloned into pEASY-T1 Simple Cloning Vector (Transgen, Beijing, China). The PtCAT2 gene was also subjected to RT-PCR analysis at 94°C for 2 min, followed by 35 cycles of 94°C for 30 s, 58°C for 30 s, 72°C for 2 min, and a final extension at 72°C for 5 min.

The open reading frame (ORF) of the PtCAT2 gene was obtained using the ORFfinder program (https://www.ncbi.nlm.nih.gov/orffinder/) and used to analyze the putative PtCAT2 protein sequence. The isoelectric point (pI), molecular weight (MW) and the grand average of hydropathicity (GRAVY) of the protein were then determined using the ExPASy ProtParam program (http://web.expasy.org/protparam/). Thereafter, the NCBI Conserved Domain (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and InterPro (http://www.ebi.ac.uk/interpro/) databases and the Motif Scan tool (http://myhits.isb-sib.ch/cgi-bin/motif_scan) were used to detect the catalytic active and heme-binding sites of PtCAT2. The subcellular localization of PtCAT2 was determined using the ProtComp (http://www.softberry.com/), CELLO (http://cello.life.nctu.edu.tw/) and PlantmPLoc (http://www.csbio.sjtu.edu.cn/bioinf/plant-multi/). Furthermore, SWISS-MODEL (http://swissmodel.expasy.org) was used to simulate the three-dimensional (3D) structure of PtCAT2. Functionally characterized CAT proteins (14) were used to construct the phylogenetic tree with the MEGA 7.0 program via the neighbor-joining method with 1000 bootstrap replicates.

The promoter sequence of PtCAT2 was obtained from our genomic data (unpublished data), and a 2 kb sequence upstream of its transcription initiation site (ATG) was analyzed using the PlantCARE program (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) to predict the promoter cis-elements.

The ORF fragment of PtCAT2 was amplified from the pEASY-T1-PtCAT2 vector using primers PtCAT2-EcoRI-F/PtCAT2-SmaI-R ( Table S1 ). The amplified product was then transformed into the multiple cloning sites upstream of the green fluorescent protein (GFP) of the pCAMBIA1302 vector via double enzyme digestion and ligation. Thereafter, the obtained pCAMBIA1302-PtCAT2-GFP construct was verified by enzyme digestion, and the expression vector was transformed into onion cells according to a previously published method (Xu et al., 2014). The GFP fluorescence of the onion cells was then observed under a confocal microscope (FV1000SP, Olympus, Japan) 24 h later.

RNA isolation and cDNA synthesis were performed according to our previously published method (Zhang et al., 2021). Briefly, RNA from P. ternata and Arabidopsis plants was extracted using the standard TRIzol method. The obtained RNA was used as the template for cDNA synthesis, conducted using the 5×All-In-One MasterMix transcription and AccuRT Genomic DNA Removal Kits (Cat#G492) (ABM, Richmond, BC, Canada), following the manufacturer’s instructions.

Quantitative real-time RT-PCR (qRT-PCR) was performed on an ABI7500 fast Real-Time PCR system (Applied Biosystems, Foster City, CA, USA) using a 2×AceQ qPCR SYBR Green Master Mix (Vazyme, Piscataway, NJ, USA). The ribosomal 18S gene (Pt18S) and β-tubulin (AtTUB) were used as internal controls for P. ternata and Arabidopsis, respectively, and the relative expression of the target genes was calculated using the 2^{−ΔΔ CT} method (Livak and Schmittgen, 2001). Three technical replicates and three biological replicates were used for each treatment. The primer sequences used for the qRT-PCR are listed in Supplementary ( Table S1 ).

The PtCAT2 cDNA, amplified from the pEASY-T1-PtCAT2 vector using the primer pair PtCAT2-Smal I-F/PtCAT2-Xba I-R, was purified and inserted into the multiple cloning sites of the pCAMBIA2300-35S-GUS-CaMVterm vector (Duan et al., 2015). The obtained recombinant vector and the empty vector pCAMBIA2300-35S-GUS-CaMVterm were verified by double enzyme digestion and then transformed into Agrobacterium-tumefaciens strain EHA105. Thereafter, Arabidopsis (Col-0) was transformed with the Agrobacterium cells containing the vectors via the floral-dip method (Clough and Bent, 1998). The transgenic plants were screened by plating the surface-disinfected seeds onto half-strength MS media containing 10 mg·L^-1 kanamycin, and the positive plants were transplanted into pot soil for seed harvesting. The seeds obtained from individual plants were then screened on 10 mg·L^-1 kanamycin until the T₃ homozygous lines were obtained.

The T₃ transgenic seeds were surface-disinfected, vernalized and germinated on half-strength MS media containing 25% (m/v) polyethylene glycol (PEG)-6000. Their germination rates were recorded, and the images were captured two weeks after the treatments.

For the water deficit treatment, 2-week-old seedlings from T₃ homozygous transgenic Arabidopsis lines and WT were transplanted into six-hole plastic plates filled with a mixture of soil and vermiculite (3:1). Five plates containing 30 plants were included for each treatment in triplicate. The seedlings were maintained at 22°C under 70% humidity and a 16 h/8 h light/dark photoperiod and subjected to water deficit treatment after one week. Leaves were harvested after seven days of drought stress, frozen in liquid nitrogen, and preserved at −80°C for further analysis. After ten days of drought stress, the survival rates of the seedlings were recorded, and the images were captured.

The activities of antioxidant enzymes superoxide dismutase (SOD), peroxidase (POD), CAT and the content of malondialdehyde (MDA) were measured as described previously (Tian et al., 2022). Histochemical staining with diaminobezidin (DAB) and nitrotetrazolium blue chloride (NBT) was performed using the method by Zhang et al. (2021). H₂O₂ and O₂ ^·- levels were analyzed by using the detection kits (Nanjing Jiancheng Bioengineering Institute, Jiangsu, China).

The electrolyte leakage (El) was measured following a previously published method (Zhang et al., 2021). Briefly, the leaves were sliced, immersed in deionized water, and incubated at room temperature on a shaker (20×g) for 1 h. The initial conductivities of the treatment (C1) and blank (CK1) samples were measured by a conductivity meter (DDSJ-318, INESA Scientific Instrument Co., Ltd., China) before boiling. Thereafter, the samples were boiled for 10 min and cooled down to room temperature, and their second conductivities (C2 and CK2) were recorded. The El was presented as the relative conductance (C) calculated using the following equation: C (%) = (C1 − CK1)/(C2 − CK2) × 100.

The water loss rate was measured by weighing the leaf samples at 0, 1, 2, 3, 4 and 5 h after their detachment from the stems, and the relative water loss rate was presented as the ratio of leaf weight at each time point relative to that at 0 h.

The mitochondria isolation and respiratory measurements were performed following a previously published method (Lyu et al., 2018). The leaves of two-week-old wild-type and transgenic Arabidopsis seedlings grown at 22°C were used to isolate the mitochondria. The respiration of the isolated mitochondria was determined using a Clark-type oxygen electrode (Hansatech, King’s Lynn, England). The reaction buffer contained 0.3 M mannitol, 10 mM TES-KOH (pH 7.5), 3 mM MgSO₄, 10 mM NaCl, 5 mM KH₂PO₄ and 0.1% BSA. Succinate (SA) at 5 mM and NADH at 1 mM were the respiratory substrates. The buffer was added with 5 mM SA and 1 mM NADH, and then incubated at 25°C for reaction.

The exogenous H₂O₂-caused oxidative damage on chlorophyll was measured following the method by Tian et al. (2022). Briefly, whole leaves obtained from the same positions on 2-week-old transgenic and WT Arabidopsis plants were incubated in 10 mL solution containing 0, 0.5% and 1% H₂O₂ for 24 h at 25°C under a 16 h/8 h light/dark photoperiod. Thereafter, the images were captured, and the chlorophyll content of the leaves was determined, as described previously (Lichtenthaler, 1987).

Data were analyzed using the Statistical Analysis System software (IBM SPSS Statistics 21.0, IBM Corp., NY, USA), and the significant differences among treatments were determined using the Students’ t-test or Duncan’s multiple range test (P < 0.05). The data are presented as means ± SD values of three biological replicates (n=3).

From a P. ternata transcriptome data (Xue et al., 2019), 13 CAT homologs were identified. Of them, i1_HQ_Pts1_c90512/f3p3/1940 (annotated as PtCAT2), showed a remarkable drought-inducible expression profile ( Figure S1 ). PtCAT2 cDNA consisted of 1479 bp long ORF ( Figure S2 ) encoding 492 amino acids. Its molecular weight and isoelectric points were 57 KDa and 6.91, respectively. Moreover, the protein had an aliphatic index of 37.65 and a GRAVY of –0.570, suggesting that PtCAT2 is a hydrophilic protein. The sequence alignment analysis identified the CAT activity motif (FARERIPERVVHARGAS) and heme-binding site (RIFAYADTQ) of PtCAT2 at positions 54–70 and 344–352, respectively, consistent with the other analyzed CATs ( Figure S3 ). This implies that PtCAT2 is a typical CAT.

Furthermore, the SWISS-MODEL analysis showed that the three-dimensional (3D) structure of PtCAT2 has four typical subunits ( Figure S4 ). The PtCAT2 aligned with the template protein 4qol.1.A from amino acids 11 to 490, covering 97.36% of the template. Moreover, the PtCAT2 protein shared a high homology (48.54%) with the template protein, and it had a GMQE value and QMEAN Z-Score of 0.81 and –1.08, respectively, confirming the high reliability of the predicted structure.

Phylogenetic analysis suggested that CATs from Triticum monococcum, Triticum aestivum, Brassica oleracea, Saccharum hybrid, Zea mays, Oryza sativa, Setaria italica, Nelumbo nucifera, Selenicereus undatus, Arabidopsis thaliana, Hordeum vulgare, Sorghum bicolor and Aeluropus littoralis could be divided into two classes, consistent with previous publications (Sun et al., 2018). PtCAT2 clustered with class II CATs, though it did not have a high similarity with other CATs (the highest homology was 82% homology with the T. aestivum TaCAT) ( Figure S5 ), demonstrating that PtCAT2 is a hypothetical class II CAT.

Quantitative real-time RT-PCR indicated that PtCAT2 was ubiquitously expressed in P. ternata ( Figure 1A ) and had a 5.5-fold higher expression in the leaves, followed by flowers (4.9-fold), tubers (4.1-fold) and petioles (2.8-fold), compared with that in roots.

To identify the abiotic stress response of PtCAT2, we exposed the P. ternata plants to various abiotic stresses and used qRT-PCR to analyze the relative expression level of PtCAT2. When stressed with 25% PEG, we observed that the PtCAT2 mRNA level significantly increased within 72 h of the treatment, with the expression peak at 12 h (a 51-fold induction) ( Figure 1B ). PtCAT2 was also remarkably induced by 100 μM ABA, with a 25-fold expression at 12 h. The PtCAT2 mRNA abundance also increased to some extent when treated with salt (150 mM NaCl), low temperature (4°C) and heat (40°C); however, the inductions were less than 3.5-fold ( Figure 1B ). Thus, PtCAT2 could be remarkably inducible by drought and ABA but not by the other abiotic stresses tested in this study.

To further analyze the drought-induced expression of PtCAT2, we used the PlantCARE software to predict the possible acting cis-elements in its promoter. We identified various cis-elements in the PtCAT2 promoter ( Figure S6 ), which regulate its response to abiotic stresses. Among these were three MYCCONSENSUSAT (CANNTG), two MYCATRD22 (CACATG) and one ABRE, which may be involved in drought and ABA response. There were also GT, ACGTATERD1 and GAATTC cis-elements, which may be associated with the weak response of PtCAT2 to salt, cold and heat stresses. These results imply that the significant induction of PtCAT2 under drought could be attributed to MYCCONSENSUSAT or MYCATRD22 and probably mediated by ABA.

The plant expression vector harboring the fused PtCAT2-GFP cassete ( Figure S7 ) was transformed into onion epidermal cells. The fluorescence results showed the presence of GFP signals in the cell membrane and cytoplasm, not in the nucleus ( Figure 2 ). This is consistent with the results of the protein subcellular localization prediction tool (PSORT), which showed a 78.7% probability of localization in the cytoplasm and 4.3% of localization in the plasma membrane.

The T₃ homozygous lines of transgenic Arabidopsis were identified with RT-PCR, and three lines (OE1, OE2 and OE3) were selected for further analysis ( Figure 3A ). The seeds of the three lines were subjected to drought stress via germination on a half-strength MS medium containing 25% PEG. We found that the PtCAT2-transgenic and WT seeds had no significant difference during germination under normal conditions. However, the germination rate of the PtCAT2-transgenic Arabidopsis seeds was remarkably higher than that of the WT under PEG-simulated drought conditions ( Figure 3B ).

The drought tolerance of the transgenic Arabidopsis plants was further evaluated through water deficit treatment ( Figure 3C ). We found that most WT plants withered or even died after 10 days of water deficit treatment, while the transgenic plants maintained normal growth. Following a 5 d recovery after rewatering, over 95% of the transgenic plants recovered, compared to the 41% survival of the WT ( Figure 3D ). Thus, these results indicate that overexpressing PtCAT2 significantly enhanced the drought tolerance of Arabidopsis plants.

Since the Arabidopsis plants overexpressing PtCAT2 had enhanced drought tolerance, we subjected their leaves and those of the WT to DAB and NBT staining. In DAB staining, the transgenic leaves had lighter stains than the WT under drought stress but did not differ from the WT under normal conditions ( Figure 4A ). However, the NBT-stained leaves showed no much difference between transgenics and WT ( Figure 4B ). The H₂O₂ content was lower in transgenic leaves than in WT under normal and drought conditions (P<0.05) ( Figure 4C ), though O₂ ^·- content did not show significant difference between transgenics and WT ( Figure 4D ) except OE1. The results further confirm that overexpressing PtCAT2 efficiently degrades H₂O₂ in Arabidopsis under drought stress but not significantly affects O₂ ^·- production.

Malonaldehyde (MDA) is an important indicator of the degree of lipid peroxidation, commonly used to estimate membrane damage under oxidative stress (Ashraf et al., 2017). Although there was no difference between the MDA content of the WT and transgenic plants under normal conditions, the transgenic plants had significantly lower MDA content than WT plants under drought stress (P< 0.05) ( Figure 4E ). Further, the oxygen consumption rate was investigated to test the effect of PtCAT2 expression on the mitochondrial function. Three transgenic lines maintained higher oxygen consumption rate under drought stress in comparison with WT (P< 0.05) ( Figure 4F ). These data indicate that overexpression of PtCAT2 alleviates the membrane damage and maintain the mitochondrial function under drought stress.

We also determined the CAT activity of the transgenic lines and WT, and found that similar to the mRNA level, the CAT activities of the three transgenic lines were significantly higher than WT under normal and drought stress conditions ( Figure 5A ). This showed the enzymatic correlation between PtCAT2 and drought tolerance. We also quantified the activities of SOD and POD. The result showed that the POD and SOD activities were similar between WT and transgenic plants (P>0.05) under normal conditions. However, when stressed with the water deficit treatment, the POD and SOD activities increased in WT or transgenic plants, but the increment was significantly higher in transgenic plants compared to WT plants (P<0.05) ( Figures 5B, C ). The El was measured to determine the membrane integrity of the plants under water deficit stress. We found no significant differences between the El values of the WT and transgenic leaves under normal conditions; however, the transgenic leaves presented significantly lower El values under drought treatment (P< 0.05) ( Figure 5D ). These data suggest that PtCAT2 overexpression could alleviate the membrane damage by enhancing the free radical scavenging ability of Arabidopsis, thus improving its drought stress tolerance.

To reveal the effect of PtCAT2 overexpression on H₂O₂ accumulation under oxidative stress, we incubated the leaves of WT and transgenic Arabidopsis plants in different concentrations of H₂O₂ solution. When exposed to either 0.5% or 1.0% H₂O₂, the leaves of three transgenic lines were lusher than WT ( Figure 6A ) and had higher chlorophyll content ( Figure 6B ), indicating that PtCAT2 overexpression enables plants to degrade H₂O₂ and maintain their chlorophyll content.

We also monitored the water loss rate of the WT and transgenic leaves within 5 h after their detachment from the stems. The results showed that though the water content decreased in the WT and transgenic leaves, the transgenic lines maintained higher water content than the WT (P<0.05). This suggested that the PtCAT2- transgenic plants exhibited an improved water retention capacity ( Figure 6C ).

Since PtCAT2 overexpression enhanced drought tolerance in transgenic Arabidopsis plants, expression patterns of the drought stress-related genes and endogenous CATs were quantified in the WT and transgenic Arabidopsis lines under normal and drought conditions. The expression levels of drought stress-related genes, AtRAB18, AtRD22, AtRD29B and AtRD29A, were similar between WT and transgenic lines under normal conditions but were significantly upregulated in the transgenic lines compared to the WT under drought treatment (P<0.05) ( Figure 7A ). Moreover, the expression levels of the Arabidopsis endogenous CATs (AtCAT1, AtCAT2 and AtCAT3) were also significantly upregulated in the transgenic lines than the WT under drought stress. In particular, PtCAT3 had >10 folds upregulation in the three transgenic lines compared to the WT ( Figure 7B ). These results suggest that PtCAT2 overexpression activates the expressions of drought stress-related genes and endogenous CATs to confer drought tolerance in plants.