MdPP2C24/37, Protein Phosphatase Type 2Cs from Apple, Interact with MdPYL2/12 to Negatively Regulate ABA Signaling in Transgenic Arabidopsis

The phytohormone abscisic acid (ABA) plays an important role in the ability of plants to cope with drought stress. As core members of the ABA signaling pathway, protein phosphatase type 2Cs (PP2Cs) have been reported in many species. However, the functions of MdPP2Cs in apple (Malus domestica) are unclear. In this study, we identified two PP2C-encoding genes, MdPP2C24/37, with conserved PP2C catalytic domains, using sequence alignment. The nucleus-located MdPP2C24/37 genes were induced by ABA or mannitol in apple. Genetic analysis revealed that overexpression of MdPP2C24/37 in Arabidopsis thaliana led to plant insensitivity to ABA or mannitol treatment, in terms of inhibiting seed germination and overall seedling establishment. The expression of stress marker genes was upregulated in MdPP2C24/37 transgenic lines. At the same time, MdPP2C24/37 transgenic lines displayed inhibited ABA-mediated stomatal closure, which led to higher water loss rates. Moreover, when exposed to drought stress, chlorophyll levels decreased and MDA and H2O2 levels accumulated in the MdPP2C24/37 transgenic lines. Further, MdPP2C24/37 interacted with MdPYL2/12 in vitro and vivo. The results indicate that MdPP2C24/37 act as negative regulators in response to ABA-mediated drought resistance.


Introduction
The apple tree (Malus domestica) is one of the most important fruit trees in the world, and its yield is greatly affected by adverse environmental cues, such as drought stress [1,2]. As the world's largest apple producer, China accounts for 50% of the world's apple production, among which 'huashuo' is an elite hybrid cultivar [3]. Whole genome sequencing analysis has revealed accurate apple gene sequences and provides an effective method to study the function of stress-responsive genes [4,5]. Thus, the study of the apple's resistance to drought stress is urgently needed to cultivate various apple trees with excellent traits, such as 'huashuo'.
Unlike animals that can escape adverse environments, sessile plants have developed a series of defensive mechanisms to deal with various abiotic stresses, such as drought, cold, and heat [6]. As a major environmental factor, drought stress affects the geographical distribution of agricultural productivity and environmental sustainability, causing abscisic acid (ABA) accumulation in plants. Further, the phytohormone ABA regulates many physiological processes, including plant development, seed maturation, stomata movement, and signal transduction under abiotic stress [7,8].
Many key components in the ABA signaling pathway have been identified at the molecular level, including ABA receptors, group A type 2C protein phosphatases (PP2Cs), SnRKs (sucrose non-fermenting-1 related protein kinases), transcription factors, and ion

Structure and Homology Characterization of MdPP2C24/37
In apple, MdPP2C24 (LOC103407751, MDP0000883465) encodes a 408-amino acid protein on the third chromosome, while MdPP2C37 (LOC114826111, MDP0000203818) encodes a 424-amino acid protein on the seventh chromosome. In this study, based on the NCBI website, both MdPP2C24/37 were predicted to possess conserved PP2C catalytic domains ( Figure 1A). Using MEGA 7, a phylogenetic tree was generated according to the highly conserved PP2C homologous protein sequences ( Figure 1B, Supplementary Figure S1). In addition, the phylogenetic tree analysis revealed that MdPP2C24 shared 56.69% identity with AtPP2C3, and MdPP2C37 shared 60.55% identity with AtPP2CA ( Figure 1B). These results indicated that MdPP2C24/37 belong to the conserved PP2C gene family in plants. Figure 1A). Using MEGA 7, a phylogenetic tree was generated according to the highly conserved PP2C homologous protein sequences ( Figure 1B, Supplementary Figure  S1). In addition, the phylogenetic tree analysis revealed that MdPP2C24 shared 56.69% identity with AtPP2C3, and MdPP2C37 shared 60.55% identity with AtPP2CA ( Figure 1B). These results indicated that MdPP2C24/37 belong to the conserved PP2C gene family in plants.  To identify the cellular localization of MdPP2C24 and MdPP2C37, we performed in vivo transient expression analysis in tobacco protoplasts. 35S::MdPP2C24-GFP, 35S::MdPP2C37-GFP, and 35S::GFP constructs were transformed into tobacco leaf protoplasts using a polyethylene glycol (PEG)-mediated method [29]. The fluorescence sig-nal of MdPP2C24/37-GFP fusion proteins was observed in the nucleus, indicating that MdPP2C24/37 are nuclear proteins (Figure 2A).
To preliminarily study the MdPP2C24/37 response to drought stress in appl formed a 50 μM ABA or 200 mM mannitol treatment on apple leaves (Figure transcriptional analysis results using qRT-PCR supported the possib MdPP2C24/37 are involved in ABA or osmotic signaling in apple.

Figure 2.
Nucleus-located genes MdPP2C24/37 were induced by ABA or mannitol. (A) Subcellular localization of MdPP2C24/37-GFP fusion proteins in tobacco protoplasts. 35S::GFP alone or MdPP2C24/37-GFP correspond to chlorophyll and bright field images, respectively, and the superposition of fluorescent illumination, chlorophyll, and bright field images is shown. (B) Real-time qPCR analysis of MdPP2C24/37 expression in apple leaves. Total RNA was isolated from the ABA-or mannitol-treated apple leaves and used for real-time qPCR. Results represent mean ± SE from three independent experiments, with similar results obtained. Values were significantly different from WT at * p < 0.05, ** p < 0.01 or *** p < 0.001.
To preliminarily study the MdPP2C24/37 response to drought stress in apple, we performed a 50 µM ABA or 200 mM mannitol treatment on apple leaves ( Figure 2B). The transcriptional analysis results using qRT-PCR supported the possibility that MdPP2C24/37 are involved in ABA or osmotic signaling in apple.

MdPP2C24/37 Regulate ABA Response in Germination and Seedling Establishment through Stress-Responsive Genes in Arabidopsis
To study the gene function of MdPP2C24/37 properly, we generated MdPP2C24/37 overexpression lines in Arabidopsis driven by the cauliflower mosaic virus (CaMV) 35S promoter. PCR analysis of DNA levels and real-time qPCR analysis of RNA transcript levels confirmed MdPP2C24/37 overexpression in these lines (Supplementary Figure S2). Three independent overexpression lines were obtained for each MdPP2C. Then, overexpression lines MdPP2C24 (24-2, 24-6) and MdPP2C37 (37-4, 37-7) were chosen as representative lines.
To study the roles of MdPP2Cs in the ABA response, we germinated the plants in MS media containing different concentrations of ABA. After treatment with 0.5 or 1 µM ABA, the seed germination rates of the MdPP2C24/37 overexpression lines were significantly higher than those of WT ( Figure 3A,B). In addition, the MdPP2C24/37 overexpression lines exhibited markedly higher cotyledon greening rates under the 0.5 µM ABA treatment than WT, including over 50% in the MdPP2C24 overexpression lines and over 20% in the MdPP2C37 overexpression lines. In contrast, only a 4% cotyledon greening rate was recorded in the wild type plants. The same cotyledon greening rate trend was observed after treatment with 1 µM ABA; namely almost 40% in the MdPP2C24 overexpression lines and 2% in the MdPP2C37 overexpression lines, whereas no cotyledon greening was observed in WT ( Figure 3C). In these results, no significant difference was observed among plants under control conditions ( Figure 3A-C, Supplemental Figure S3). Next, the 7-day-old seedlings were treated with liquid MS medium to which 50 µM ABA was added. A qPCR assay was used to detect the stress-responsive gene levels of DREB2A, KIN2, ABF4, and RD29A. The results showed that transcript levels of DREB2A and KIN2 were notably increased in the MdPP2C24/37 overexpression lines compared to WT under control conditions. After exogenous ABA treatment, transcript levels of DREB2A and KIN2 were almost unchanged in the MdPP2C24/37 overexpression lines; however, they were significantly increased in WT. Moreover, transcript levels of ABF4 and RD29A were upregulated after ABA treatment in all plants ( Figure 3D). These results indicated that MdPP2C24/37 decreased ABA sensitivity at the germination and seedling stages in Arabidopsis.

Overexpression of MdPP2C24/37 Modulate Drought Tolerance
Next, the relationship between MdPP2C24/37 and ABA-dependent drought stress response was studied. A stomatal aperture assay revealed that the MdPP2C24/37 overexpression lines were insensitive to ABA compared with WT ( Figure 4A,B). Under normal conditions, however, no differences were observed among these plants. However, higher water loss rates were observed in the detached leaves of the MdPP2C24/37 overexpression lines at the indicated time points ( Figure 4D). Furthermore, when exposed to drought stress for 15 days, fewer MdPP2C24/37 overexpression lines survived than WT ( Figure 4C). The expression levels were based on total RNA extracted from WT and MdPP2C24/37 overexpression lines in liquid MS medium, or liquid MS medium supplemented with 50 μM ABA for 3 h. Results represent the mean ± SE from 3 independent experiments. The expression levels are presented as relative units, with levels under control conditions taken as 1. All experiments were replicated three times with similar results. Values were significantly different from WT at * p < 0.05, ** p < 0.01 or *** p < 0.001.   In this study, no significant differences were observed for chlorophyll, MDA, and H 2 O 2 levels among plants under control conditions. However, under drought stress, chlorophyll levels decreased more rapidly in all MdPP2C24/37 overexpression lines than in WT ( Figure 4E). Further, MDA and H 2 O 2 accumulation was significantly higher in MdPP2C24/37 overexpression lines than in WT ( Figure 4F,G). The drought stress method used in this study involved withholding water for 12 days. In general, the results presented above indicated that the MdPP2C24/37 overexpression lines showed decreased resistance ability to ABA-mediated drought stress in adult Arabidopsis.

MdPP2C24/37 Involved in Response to Osmotic Stress through Stress-Responsive Genes in Arabidopsis
Next, the MdPP2C24/37 overexpression lines were studied in response to osmotic stress. The germination and cotyledon greening rates of the MdPP2C24/37 overexpression lines were higher than those of WT under 200 or 300 mM mannitol treatment ( Figure 5A,B). All of the MdPP2C24/37 overexpression lines displayed over 80% cotyledon greening rates, while the wild type plants only showed 35% under the 200 mM mannitol treatment. Further, the MdPP2C24 overexpression lines displayed almost 70% cotyledon greening rates, the MdPP2C37 overexpression lines displayed 40%, and WT only displayed 7% in MS medium supplemented with 300 mM mannitol ( Figure 5C). In these results, no significant difference was observed among plants under control conditions ( Figure 5A-C, Supplementary Figure S3). Subsequently, the 7-day-old seedlings were treated with liquid MS medium containing 200 µM mannitol. A real-time qPCR assay was used to detect the stress-responsive gene levels of DREB2A, KIN2, ABF4, and RD29A. DREB2A and KIN2 expression levels in the MdPP2C24/37 overexpression lines were much higher than those in WT under control conditions. Under exogenous mannitol treatment, transcript levels of DREB2A and KIN2 were almost unchanged in the MdPP2C24/37 overexpression lines; however, most transcript levels were significantly increased in WT. Further, transcript levels of ABF4 and RD29A were upregulated after mannitol treatment in all plants ( Figure 5D). To summarize, MdPP2C24/37 decreased osmotic sensitivity at the germination and seedling stages in Arabidopsis.

Both MdPP2C24/37 Interacted with MdPYL2/12 In Vitro and Vivo
In this study, the coding sequences of MdPP2C24/37 were used as prey and several possible MdPYLs/PYR1 were used as bait in a yeast two-hybrid assay. Under the fourminus selection medium (SD/-Leu/-Trp/-His/-Ade) growth conditions, MdPYR1 was unable to interact with MdPP2Cs, while MdPYL2/9/12 interacted with MdPP2C24/37 in vitro ( Figure 6A).

Discussion
Adverse environmental cues severely affect plant growth and development. To deal with stress, plants have developed a series of strategies, such as ABA signal conduction [14,[30][31][32]. PP2Cs, important signal transductors in ABA signaling, have been reported in
Considerable evidence indicates that PP2C localization is required for different functions. In Arabidopsis, plasma membrane-localized members of PP2Cs, namely PP2C.D2, PP2C.D5, and PP2C.D6, are major regulators of cell expansion through their physical interaction with SAUR19 and PM H + -ATPases, and they inhibit cell expansion by dephosphorylating the penultimate threonine of PM H + -ATPases [40]. AtABI1, a nuclear protein, is essential to confer insensitivity towards ABA [41]. In this study, MdPP2C24/37 were nucleus-located and induced by ABA (Figure 2). The results indicated that both MdPP2Cs are probably involved in nuclear signal transduction, which prompted us to analyze their functions in ABA responsiveness. Further results revealed that the expression of MdPP2C24/37 in Arabidopsis led to plant insensitivity to ABA during seed germination and seedling establishment and to ABA-induced drought stress (Figures 3 and 4, Supplementary  Figure S3). Previous reports revealed that PP2Cs are involved in ABA signaling through their interaction with ABA receptors PYR/PYLs [32,[42][43][44][45]. However, whether MdPP2Cs associate with MdPYLs/PYR1 in apple is not clear. In this study, MdPP2C24/37 were found to interact with ABA receptors MdPYL2/12 in the nucleus through a yeast two-hybrid assay and BiFC assay ( Figure 6). The results proved that MdPP2C24/37 are negative regulators, as typical PP2Cs, in ABA signaling. However, MdPYL9, a homologue to PYL9 in Arabidopsis, Zea mays, and M. domestica [46][47][48], was not observed to interact with MdPP2C24/37 in vivo ( Figure 6, Supplementary Figure S4). In summary, these results suggest that MdPP2C24/37 interact with MdPYL2/12, specifically in the nucleus, and play negative roles in ABA signaling.
In higher plants, ABA signaling can be triggered by multiple stresses, such as drought and osmotic stress [27]. In addition, some PP2Cs play pivotal roles in osmotic stress [49]. For instance, ZmPP2C-As in maize has been found to be dramatically induced by osmotic stress, resulting in a higher seed germination rate under osmotic stress and higher death rate when plants are exposed to drought stress [24]. OsPP108 and OsSIPP2C1 in rice, which are upregulated by high salt, exogenous ABA, and drought treatment, act as negative regulators in ABA signaling [50,51]. In this study, mannitol was used to reveal the roles played by MdPP2C24/37 in the relationship between ABA and osmotic stress signaling. The results indicated that transcript levels of MdPP2C24/37 were induced by ABA or mannitol ( Figure 2). The MdPP2C24/37 overexpression lines were hyposensitive at the germination and seedling establishment stages under ABA or mannitol treatment in Arabidopsis (Figures 3 and 5, Supplementary Figure S3). Further, stress-related genes, namely DREB2A, KIN2, ABF4, and RD29A, exhibited almost the same expression tendencies under ABA or mannitol treatment in all MdPP2C24/37 overexpression plants (Figures 3 and 5). Plants induce stomatal movement through ABA signaling [32] and closely guard cells to maintain water balance when they are exposed to drought stress [52]. In this study, all of the MdPP2C24/37 overexpression lines displayed inhibited stomatal closure in response to exogenous ABA treatment ( Figure 4). Moreover, osmotic stress, which accompanies water deficit, can be regulated through ABA-dependent signaling [28,53]. Overall, this study comprehensively reported the roles of MdPP2C24/37 at different stages. At the seed germination and seedling establishment stages, MdPP2C24/37 overexpression plants were hyposensitive to both ABA and mannitol. At the adult stage, overexpression of MdPP2C24/37 in Arabidopsis led to a lower survival rate through its inhibition of stomatal closure (Figures 3-5). These results are consistent with those found for PtNF-YA9 (NUCLEAR FACTOR Y9) in Populus trichocarpa [54]. In summary, these observations demonstrate that MdPP2C24/37 participate in multiple signaling pathways at different growth stages and function as connecting links in response to ABA signaling and osmotic stress.
Current evidence suggests that redundancies in PP2Cs create major obstacles to studying the functions of PP2Cs genes using genetic approaches [55]. For instance, ABI1 and ABI2 phosphatases play overlapping roles in controlling ABA actions [10]. It is noteworthy that both MdPP2C24/37 have highly conserved PP2C catalytic domains but share only 54% homologous similarity (Figure 1). This analysis suggests that MdPP2C24/37 may have different functions, as well as redundant functions, in ABA signaling in apple. That is, the functions of MdPP2C24/37 may diverge at some point in response to ABA and osmotic stress, and the functions of their potential targets may be different from those of MdPYL2/12 ( Figure 6). Therefore, these synergistic effects further suggest that MdPP2C24/37 may participate in other signaling pathways.

Real-Time qPCR Analysis
Total RNA of fresh leaves from four-week-old subculture apple shoots were extracted using the CTAB assay [56]. RNA of A. thaliana was isolated from 7-day-old seedlings using RNAiso Plus (Cat. #9109, Takara, Japan). Reverse transcription was performed using the cDNA Synthesis Kit (Cat. #RR047A, Takara, Japan). Real-Time qPCR was performed using SYBR Premix Ex Taq (Cat. # RR430A, Takara, Japan). ACTIN2 (At3g18780) was used as an internal control for Arabidopsis, while MdEF-1α (LOC103443462) was used as an internal control for apple. All gene-specific primer pairs are listed in Supplementary Table S1.

Transient Expression Assay and Transgenic Arabidopsis Constructs
To achieve transient expression in tobacco mesophyll cell protoplasts, the MdPP2C24/37 CDSs were cloned into the pBI221-eGFP vector. Analysis of transient expression in protoplasts was performed as previously described [29].
To generate transgenic Arabidopsis, MdPP2C24/37 CDSs were cloned into the pCambia2300 vector. The plasmids were introduced into Arabidopsis Col-0 by Agrobacterium tumefaciens (GV3101)-mediated transformation. The floral dip method was adopted in this study [58]. The seeds of the transgenic plants were screened in MS medium supplemented with 35 mg/L kanamycin. The T3 homozygous seeds were used for phenotypic analysis. The primers used in this assay are listed in Supplementary Table S1.

Phenotype Analysis
For germination and seedling establishment assays, approximately 150 sterilized seeds of each line were sown in MS medium with various concentrations of ABA (Cat. #90769, Sigma-Aldrich, America) or mannitol (Cat. #M8140, Solarbio, China). The germination rate was calculated every 12 h 6 days after sowing. The cotyledon greening rates were calculated on the 6th day after sowing.
For the seedling establishment assay, seeds grew vertically in 1/2 MS medium for 7 days with 0 or 0.5 µM ABA or 300 mM mannitol. The root lengths were measured using ImageJ software (https://imagej.en.softonic.com/, accessed on 1 January 2020).
For the stomatal aperture assay, rosette leaves of 3-week-old plants were detached and floated in stomatal opening solution (5 mM KCl, 10 mM MES (2-(N-morpholino) ethanesulfonic acid), 50 µM CaCl2, pH 6.15) in the dark for 2 h, then exposed to continuous light for 2 h, followed by the addition of 10 µM (±) ABA. In order to estimate ABAinduced stomatal closure, apertures were recorded on epidermal strips after 2 h incubation.
For the water loss assay, rosette leaves of 3-week-old Arabidopsis plants were detached. The water loss rate was monitored at indicated times. For drought stress analysis, 3-week-old plants were not watered for 15 days and then rewatered. The survival rate was calculated.

Determination of Drought-Responsive Physiological Indices
The 3-week-old Arabidopsis plants were subjected to normal conditions or exposure to drought stress. Then, leaves were collected for chlorophyll, malondialdehyde (MDA), and H 2 O 2 content determination according to a previously described protocol [59].

Bimolecular Fluorescence Complementation (BiFC) Assay
The BiFC assay was performed according to a previously described protocol [62]. The BiFC vectors pSPYNE (YFP N )/pSPYCE (YFP C ) harbor either the N-terminus or C-terminus of eYFP. YFP N was fused to the C termini of MdPYL2/9/12, and PYR1 and YFP C were fused to the C termini of MdPP2C24/37. The vectors were transformed into A. tumefaciens GV3101. Pairwise construct combinations were transiently expressed in N. benthamiana epidermal cells. The previously described YFPC-AtCARK3 and YFPN-AtPYL1 constructs were used as positive controls [34]. The fluorescence signal of YFP was detected after infiltration for 3 days using an Olympus optical microscope at an excitation wavelength of 560 nm. The primers used for these constructs are listed in Supplementary Table S1.

Statistical Analysis
Data are represented as the mean ± SD and ± SEM. Statistical analysis was performed using Student's t-test. Values were significantly different from WT at * p < 0.05, ** p < 0.01 or *** p < 0.001.

Conclusions
In conclusion, this study demonstrated that MdPP2C24/37, isolated from 'huashuo', are nucleus-located and negatively regulate seed germination, seedling establishment, stomatal aperture, and stress-related gene expression in response to drought stress. The interactions between MdPP2C24/37 and ABA receptors MdPYL2/12 provide new a direction for studying the ABA transduction mechanism in apple. Overall, these results provide evidence that MdPP2C2C24/37 respond to ABA and osmotic stress, thus enriching the ABA signal regulation network in apple.

Conflicts of Interest:
The authors declare no conflict of interest.