Mitogen‐activated protein kinase TaMPK3 suppresses ABA response by destabilising TaPYL4 receptor in wheat

Summary

• Abscisic acid (ABA) receptors are considered as the targeted manipulation of ABA sensitivity and water productivity in plants. Regulation of their stability or activity will directly affect ABA signalling. Mitogen‐activated protein kinase (MAPK) cascades link multiple environmental and plant developmental cues. However, the molecular mechanism of ABA signalling and MAPK cascade interaction remains largely elusive.

• TaMPK3 overexpression decreases drought tolerance and wheat sensitivity to ABA, significantly weakening ABA's inhibitory effects on growth. Under drought stress, overexpression lines show lower survival rates, shoot fresh weight and proline content, but higher malondialdehyde levels at seedling stage, as well as decreased grain width and 1000 grain weight in both glasshouse and field conditions at the adult stage. TaMPK3‐RNAi increases drought tolerance.

• TaMPK3 interaction with TaPYL4 leads to decreased TaPYL4 levels by promoting its ubiquitin‐mediated degradation, whereas ABA treatment diminishes TaMPK3–TaPYL interactions. In addition, the expression of ABA signalling proteins is impaired in TaMPK3‐overexpressing wheat plants under ABA treatment. The MPK3‐PYL interaction module was found to be conserved across monocots and dicots.

• Our results suggest that the MPK3‐PYL module could serve as a negative regulatory mechanism for balancing appropriate drought stress response with normal plant growth signalling in wheat.

Introduction

As one of the three major staple crops, wheat (Triticum aestivum) provides c. 19% of global dietary calories (Ray et al., ²⁰¹³). Due in part to climate change and increasing water scarcity, drought poses a substantial threat to agriculture worldwide, especially to the productivity of field crops (Lesk et al., ²⁰¹⁶). The time of onset, duration and intensity of drought stress can affect crop production to different degrees, and drought during the reproductive period can directly lead to > 50% losses in average yield (Hu & Xiong, 2014). The phytohormone abscisic acid (ABA) is rapidly produced under drought stress and plays a critical role in regulating a wide range of developmental processes (Zhu, 2016; Hauser et al., ²⁰¹⁷; Zhang et al., ²⁰²¹). For example, ABA can control stomatal closure to minimise water loss by transpiration or decrease photosynthesis (Roelfsema et al., ²⁰¹²; Munemasa et al., ²⁰¹⁵), reprogramme metabolic pathways to increase the accumulation of osmolytes and stress‐response proteins, arrest plant growth (Julkowska & Testerink, 2015), and promote leaf senescence to adapt and survive under extreme environmental stresses (Zhao et al., ²⁰¹⁶).

The proposed core ABA signalling pathway is comprised of pyrabactin resistance 1 (PYR1)/PYR1‐like/regulatory components of ABA receptors (PYR/PYL/RCAR) family proteins (from this point forwards referred to as PYLs) (Miyakawa et al., ²⁰¹³), a family of clade A protein phosphatases 2C (PP2Cs) and members of subgroup III sucrose nonfermenting‐1 (SNF1) related kinase 2s (SnRK2s) (Fujii et al., ²⁰⁰⁹; Cutler et al., ²⁰¹⁰). In response to abiotic stresses, ABA accumulates rapidly and binds PYL receptors (Cutler et al., ²⁰¹⁰). The ABA‐bound PYL receptors competitively bind to clade A PP2Cs (Soon et al., ²⁰¹²), including ABA‐insensitive 1 (ABI1), ABI2 (Ma et al., ²⁰⁰⁹), Hypersensitive to ABA1 (HAB1), HAB2, PP2CA and ABA hypersensitive germination 1 (AHG1) (Fujii et al., ²⁰⁰⁹; Park et al., ²⁰⁰⁹), which results in the release and subsequent activation of SnRK2s, which bind PP2Cs in the absence of ABA (Zhang et al., ²⁰¹⁵). SnRK2s can then interact with and phosphorylate downstream ABA‐responsive element‐binding protein/ABA‐responsive element‐binding factors (AREBs/ABFs), such as ABF2, ABF3, ABF4 and ABI5, to regulate the expression of ABA‐responsive genes (Hauser et al., ²⁰¹⁷).

There are 14 PYLs in Arabidopsis, including PYR1 and PYL1–13, which belong to the steroidogenic acute regulatory protein‐related lipid transfer (START) domain superfamily and have redundant functions (Gonzalez‐Guzman et al., ²⁰¹²). PYR1 and PYL1‐3 depend on ABA for inhibition of PP2Cs, whereas the other PYLs can interact with and inhibit PP2Cs in vitro to various extents in the absence of ABA (Fujii et al., ²⁰⁰⁹; Hao et al., ²⁰¹¹). Due to their central role in regulating transpiration, ABA receptors are well suited candidates for targeted manipulation of ABA sensitivity and water productivity (Helander et al., ²⁰¹⁶). Overexpression of ABA receptors has been extensively tested as a means of increasing ABA sensitivity and water‐use efficiency (WUE) in Arabidopsis (Santiago et al., ²⁰⁰⁹; Yang et al., ^{2016, 2019}), rice (Kim et al., ²⁰¹⁴), poplar (Yu et al., ²⁰¹⁷), and wheat (Mega et al., ²⁰¹⁹). In Arabidopsis, C‐terminally encoded peptide receptor 2 (CEPR2) promotes the phosphorylation and degradation of PYL4^Ser54 under normal conditions, whereas ABA inhibits this process to initiate ABA signalling (Yu et al., ²⁰¹⁹). The homologue of AtPYL4 in wheat has been demonstrated to modulate WUE and drought tolerance in wheat (Mega et al., ²⁰¹⁹). However, the phosphorylation of PYLs has been reported in relatively few studies (Chen et al., ²⁰¹⁸; L. Zhang et al., ²⁰¹⁸; Wang et al., ²⁰¹⁸; Li et al., ²⁰¹⁹; Yu et al., ²⁰¹⁹). From the limited available evidence, it can be inferred that the contributions of diverse protein kinases and protein phosphatases in regulating PYL‐mediated balance between plant growth and stress signalling remain largely unexplored.

Early studies have suggested that some components of the mitogen‐activated protein kinase (MAPK) cascade pathway are involved in ABA signalling and drought stress response. For example, MPK9 and MPK12, which tend to be highly expressed in guard cells, are involved in ABA and H₂O₂‐mediated stomatal closure (Jammes et al., ²⁰⁰⁹). MKK1–MPK6 positively regulates CAT1 expression and the production of H₂O₂, which depends on the ABA signalling pathway (Xing et al., ²⁰⁰⁸). ABA‐insensitive protein kinase 1 (AIK1), a member of the MAPKKK family, positively regulates ABA‐mediated root growth and stomatal closure through MKK5–MPK6 cascade (K. Li et al., ²⁰¹⁷). In Arabidopsis, MAPKKK17/18, MKK3 and MPK1/2/7/14 cascades can be activated by ABA signalling (Danquah et al., ²⁰¹⁵), and MAPKKK18–MKK3 cascades have been shown to positively regulate plant tolerance to drought stress (Y. Li et al., ²⁰¹⁷). In maize, ZmMPK5 phosphorylates ZmABA2, a short chain dehydrogenase/reductase that contributes to ABA biosynthesis, thereby enhancing the stability and activity of ZmABA2 and regulating ABA biosynthesis (Ma et al., ²⁰¹⁶). However, there is little direct evidence linking the MAPK cascade with ABA signalling and osmotic stress.

TaMPK3 is a member of the MAPK cascade, which are highly conserved signalling modules found in all eukaryotes that link environmental and developmental cues to a range of cellular responses (M. Zhang et al., ²⁰¹⁸). In this study, we identified TaMPK3 as an interaction partner of TaPYL4 in wheat by screening drought response‐related cDNA libraries in yeast. The interaction was confirmed through several reporter‐based assays in wheat and dicot hosts. Overexpression and RNAi silencing of TaMPK3 showed that it functions as a negative regulator of TaPYL‐mediated ABA responses by decreasing TaPYL levels, potentially through ubiquitin‐mediated degradation. Adult and seedling wheat with RNAi suppression of TaMPK3 exhibited increased tolerance to drought stress, while its overexpression decreased wheat tolerance to drought stress. In addition to showing a transient regulatory mechanism during the post‐drought recovery stage, our findings suggest that the negative regulation of TaPYLs by TaMPK3 contributes to maintaining balance between plant growth and response to drought stress in wheat.

Materials and Methods

Plant materials and growth conditions

Wheat plants were grown in a glasshouse controlled at 60% relative humidity, 25°C : 23°C, day : night temperatures and long‐day conditions (16 h : 8 h, light : dark) with a light intensity of c. 200 μmol m⁻² s⁻¹. The soil was imported nutrient soil of PINDSTRUP, which had a maximum water holding capacity of 5.88 (water/dry soil). Variety Jimai60 was used to analyse the expression level of TaMPK3 and construct a wheat cDNA library using RNA extracted from leaves of drought‐treated Jimai60 wheat plants. The wheat cultivar ‘Fielder’ (wild‐type, WT) was used to amplify gene sequences (TaMPK3 and TaPYLs) and generate transgenic wheat plants. The coding sequence (CDS) of TaPYL9 (TraesCS7A02G350800) was synthesised by BGI Co. (Beijing, China). Soybean (Glycine max, ‘Williams 82’) was used to amplify GmMPK3 and GmPYL4. Arabidopsis thaliana ‘Columbia‐0’ was used to amplify AtMPK3 and AtPYLs. Rice (Oryza sativa ssp. japonica ‘Nipponbare’) was used to amplify OsMPK3 and OsPYL4. Tobacco used in this study was Nicotiana benthamiana. Primers used in these studies are listed in Supporting Information Table S1.

Generation of transgenic wheat

To generate TaMPK3 transgenic wheat plants, the TaMPK3 CDS was cloned into the plant transformation vector pWMB110 driven by the maize (Zea mays) ubiquitin promoter. The 402 bp sense and 402 bp anti‐sense orientations of TaMPK3 specific fragments were linked by the 152 bp intron sequence of the maize alcohol dehydrogenase 1 (adh1) gene. The 956‐bp fragment was synthesised by Genewiz Co. (Suzhou, China). This recombinant DNA was then inserted into the pWMB110 vector to generate the pWMB110‐TaMPK3‐RNAi construct. Genetic transformations were performed using an Agrobacterium‐mediated transformation system. To isolate positive transgenic wheat lines, leaves of 10‐d‐old transgenic wheat seedlings grown in soil were used for DNA and RNA isolation and then PCR and quantitative reverse transcription PCR (qPCR) analyses were performed. Homozygous T3 seeds of wheat used for phenotypic analyses were stored for > 2 months after ripening. The T‐DNA insertion sites of six transgenic lines in the wheat genome were identified by the mhiTail‐PCR assays as previously reported (Tan et al., ²⁰¹⁹). Primers used in these studies are listed in Table S1.

ABA sensitivity assays

For germination assays, seeds of WT and TaMPK3‐overexpressing wheat plants were sterilised with 5% sodium hypochlorite for 15 min and then rinsed with water three times. After soaking in water for 12 h, seeds were placed on wet filter paper containing various concentrations of ABA (0, 1, 5 or 10 μM). After 2 d of growth in the glasshouse, images were taken and the shoot length for each seedling was evaluated.

For ABA sensitivity during seedling growth, seeds of WT and TaMPK3‐overexpressing wheat plants were sterilised with 5% sodium hypochlorite for 15 min and then rinsed with water three times. After soaking in water for 12 h, seeds were placed on wet filter paper. After 1 d of growth until the primary roots were c. 0.5 cm, the seedlings were grown in half‐strength Hoagland's liquid medium under glasshouse conditions for 9 d. Images were taken and the shoot and root lengths for each seedling were evaluated.

Subcellular localisation and bimolecular fluorescence complimentary (BiFC) assays in wheat mesophyll protoplasts

Transient expression assays were conducted as described previously (Liu et al., ²⁰¹³). TaMPK3 and TaPYLs were inserted into the subcellular localisation vector pJIT16318, which contains a CaMV 35S promoter and a C‐terminal GFP. The protoplasts were isolated from 7‐d‐old wheat seedlings and then transfected with pJIT16318‐TaMPK3/TaPYLs plasmids by PEG‐mediated transformation. The transfected protoplasts were then incubated at 23°C for 16 h. GFP fluorescence in the transformed protoplasts was imaged using a confocal laser scanning microscope (LSM700; Zeiss).

For BiFC assays, the open reading frame (ORF) of TaMPK3 was cloned into 35S::nYFP and the ORF of each TaPYL was cloned into 35S::cYFP to generate protein fusion constructs, as described previously (Wang et al., ²⁰²¹). Subsequently, appropriate pairs of plasmid DNAs were co‐transformed into wheat mesophyll protoplasts by PEG‐mediated transformation. Following incubation at 23°C for 16 h, YFP fluorescence in the transformed protoplasts was imaged using a LSM700 microscope.

Yeast‐two‐hybrid (Y2H) assays

The CDS of TaMPK3 or TaPYL4 was cloned into the Y2H ‘bait’ vector, pGBKT7 (630443; Clontech, Beijing, China). The CDS of each TaPYL was cloned into the Y2H ‘prey’ vector, pGADT7 (630442; Clontech). Bait and prey constructs were co‐transformed into yeast (Saccharomyces cerevisiae) strain AH109. Y2H assays were performed according to the manufacturer's instructions.

Luciferase complementary imaging (LCI) and BiFC assays in tobacco

Luciferase complementary imaging assays were performed as previously described (Fujikawa & Kato, 2007; Chen et al., ²⁰⁰⁸). Agrobacterium harbouring pCAMBIA1300‐nLUC, pCAMBIA1300‐cLUC and P19 were mixed to a final concentration of OD₆₀₀ = 0.2 and then incubated at room temperature for 4 h. Four different combinations of Agrobacterium were infiltrated into four different positions in the same leaves of tobacco and cultured for 36 h. At 7 min before detection, 0.2 mM d‐luciferin (7903; Biovision, Waltham, MA, USA) was uniformly coated on the same positions where Agrobacterium was injected. To test the effect of ABA on the interaction between TaMPK3 and TaPYL4 in tobacco leaves, equal amounts of TaMPK3‐nLUC/TaPYL4‐cLUC Agrobacterium were injected into same areas of tobacco leaves on the left and right and cultured for 30 h. Then, the left sides of TaMPK3‐nLUC/TaPYL4‐cLUC co‐injected tobacco leaves were treated with water and the right sides were treated with 10 μM ABA for 6 h. At least five leaves were injected in each experiment and each experiment was repeated at least three times. Subsequently, luciferase activity was measured using a low‐light cooled CCD imaging apparatus (Night SHADE LB 985; Berthold Technologies, Bad Wildbad, Germany).

For BiFC assays, TaMPK3, AtMPK3, GmMPK3 or OsMPK3 was cloned into pXY106 at the BamHI site to produce nYFP‐MPK3 fusion protein. TaPYL4, AtPYL4, GmPYL4 or OsPYL9 was cloned into pXY104 at the BamHI site to make the PYL‐cYFP fusion protein. Subsequent experimental methods were similar to LCI assays. YFP fluorescence in the leaves of tobacco was imaged using a LSM700 microscope.

In vitro pull‐down assays

To confirm the interaction between TaMPK3 and TaPYL4, the fusion proteins TaPYL4‐glutathione S‐transferase (GST) or TaMPK3‐maltose‐binding protein (MBP) were expressed in Escherichia coli BL21 (DE3) cells carrying a pGEX‐4T‐1‐TaPYL4 or pMAL‐c2x‐TaMPK3 construct. Then, TaPYL4‐GST proteins were purified with ProteinIso^® GST Resin (DP201‐01; TransGen, Beijing, China) and TaMPK3‐MBP proteins were purified with Amylose Resin (E8021S; NEB, Ipswich, MA, USA). For the pull‐down assay, TaMPK3‐MBP proteins were eluted with 0.4% maltose and TaPYL4‐GST proteins were left on the GST Resin. TaPYLs‐GST or GST (c. 3 μg) and TaMPK3‐MBP or MBP proteins (c. 5 μg) were incubated overnight at 4°C with constant rocking in 300 μl of pull‐down buffer (40 mM HEPES (pH 7.5), 10 mM KCl, 0.4 M sucrose, 3 mM MgCl₂, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM dithiothreitol (DTT), 0.2% Triton X‐100). Afterwards, the retained GST Resin was washed five times with 1 ml 1× phosphate‐buffered saline (PBS) buffer and analysed with anti‐MBP antibody (HT701‐01; TransGen) at a 1 : 4000 dilution.

To confirm the interaction between TaMPK3 and TaPYLs, the fusion proteins TaPYLs‐His or TaMPK3‐GST were expressed in Escherichia coli BL21 (DE3) cells carrying a pCold‐TaPYLs or pGEX‐4T‐1‐TaMPK3 construct. Then, TaPYLs‐His proteins were purified using ProteinIso^® Ni‐NTA Resin (DP101‐01; TransGen) and TaMPK3‐GST proteins were purified using GST Resin. For pull‐down assays, TaMPK3‐GST proteins were eluted and TaPYLs‐His proteins were left on the Ni‐NTA Resin. TaMPK3‐GST (c. 5 μg) and TaPYLs‐His or His proteins (c. 3 μg) were incubated overnight at 4°C with constant rocking in 300 μl of pull‐down buffer as above. Afterwards, the retained Ni‐NTA Resin was washed five times with 1 ml 1× PBS buffer and analysed with anti‐GST antibody at a 1 : 4000 dilution.

Co‐IP assays

The ORF of TaMPK3 was cloned into the modified binary vector pTCK303‐Flag and The ORF of TaPYL4 was cloned into the modified binary vector pCAMBIA2300‐MYC to generate 35S::TaMPK3‐Flag and 35S::TaPYL4‐MYC. To assess the interaction between TaPYL4 and TaMPK3, the 35S::TaPYL4‐MYC vector was transiently expressed in tobacco leaves together with 35S::TaMPK3‐Flag or 35S::Flag. Anti‐Flag Agarose was used to immunoprecipitate proteins, which was further analysed using immunoblotting with anti‐MYC (M20019; Abmart, Shanghai, China) and anti‐Flag antibody (HT201; TransGen).

In vitro phosphorylation assay

Recombinant proteins were expressed in E. coli BL21 (DE3) cells and purified using affinity tags. For TaMPK3 autophosphorylation, 1 μg TaMPK3‐GST, 1 μg TaMPK3^K65R‐GST or 1 μg GST (empty vector) was incubated in 50‐μl reaction buffer (20 mM HEPES (pH 7.5), 10 mM MgCl₂, 1 mM DTT, 25 μM ATP) for 1 h at 25°C, and the reaction was stopped by adding sodium dodecyl sulphate (SDS) loading buffer. Autophosphorylation was detected using a 75 μM phos‐tag™ Acrylamide AAL‐107 assay (F4002; APExBIO, Houston, TX, USA) with anti‐GST monoclonal antibody (HT601‐02; TransGen) at 1 : 4000 dilution. For TaMPK3‐mediated phosphorylation, 2 μg TaMPK3‐GST was incubated with 1 μg TaPYLs‐His in 50 μl of the reaction buffer described above for 1 h at 25°C and the reaction was stopped by adding SDS loading buffer. Phosphorylation was detected using a 75 μM phos‐tag™ Acrylamide AAL‐107 assay with anti‐His monoclonal antibody at a 1 : 4000 dilution. In the TaMPK3^K65R mutation of TaMPK3 the 65^th amino acid was mutated from K to R for loss of autophosphorylation activity. λ protein phosphatase (λPPase) (P0753S; NEB) was used to verify that the delayed migration band of TaPYLs in phos‐tag gel was the result of its phosphorylation.

In vitro degradation assays of PYLs by cell‐free analysis

In vitro degradation assays were conducted as described previously (Wang et al., ²⁰⁰⁹). Here, 7‐d‐old seedlings (0.2 g) of WT, TaMPK3‐RNAi and TaMPK3‐OE were harvested and ground into fine powder in liquid nitrogen. Total protein was extracted in 500 μl NB1 buffer (50 mM Tris–2‐(N‐morpholino)ethanesulfonic acid (Tris–MES) (pH 8.0), 0.5 M sucrose, 1 mM MgCl₂, 10 mM EDTA, 5 mM DTT) and cell debris was removed by two centrifugation steps (c. 15 000  g , 10 min, 4°C) to collect supernatants. For degradation assays of TaPYL4, equal amounts (c. 1 μg) of purified TaPYL4‐His were incubated in 100 μl total proteins (c. 500 μg) of WT, TaMPK3‐OE or TaMPK3‐RNAi plants. Then, 100 μM MG132 (474790; Calbiochem, San Diego, CA, USA) was selectively added to the degradation assays at 60 min. The mixtures were incubated at room temperature for the whole assays and analysed at indicated intervals (0, 15, 30, or 60 min for degradation assays of TaPYL4) for protein blotting. Protein abundances of TaPYL4 were determined using immunoblot and anti‐His antibody (HT501‐01; TransGen) at a 1 : 4000 dilution.

Drought treatments

Details of drought experiments at the seedling stage and adult stage in glasshouse condition or in the field condition are available in Methods S1. Further information on soil water content (%) during the drought treatments at the seedling stage is listed in Table S2. Rainfall data in the field are listed in Tables S3 and S4. Soil water content data in the field are listed in Table S5.

Results

TaMPK3 is an interaction partner of TaPYL4

TaPYL4 was previously shown to participate in regulating WUE and drought tolerance in wheat (Mega et al., ²⁰¹⁹). To broaden our understanding of PYLs in wheat, TaPYL4 was used as a bait in Y2H screens of a wheat cDNA library constructed using RNA extracted from leaves of drought‐treated wheat variety Jimai60, to identify potential protein interaction partners. The interacting candidates of TaPYL4 are listed in Table S6. Sequence analysis of the candidate interactors showed that TaMPK3, a MAPK, exhibited potentially direct interactions with TaPYL4. TaMPK3 contained typical MPK signature motifs, including an ATP binding signature (IGRGAYGIVCSVMNFETREMVAIKK), a catalytic C‐loop (VIHRDLKPSNLLL), an activation T‐loop (TEY), a common docking (CD) domain (LHDVADEPIC) and an EF‐hand CBP (RMLTFNPLQRITVEEAL) domain (Fig. S1) (Goyal et al., ²⁰¹⁸). Amino acid sequence alignment against public databases (NCBI) revealed that it encoded a protein with high sequence similarity to Arabidopsis AtMPK3 (73.42%) and rice OsMPK3 (91.33%) (Table S7). A phylogenetic tree reconstructed using full‐length amino acid sequences of TaMPK3 and its orthologues from other plant species showed separate clustering between MPK3 from monocots and dicots, suggesting potential functional divergence between these lineages (Fig. S2).

Further evidence showed that yeast cells co‐expressing TaPYL4‐AD and TaMPK3‐BD were able to grow on synthetic drop‐out selection medium lacking Leu, Trp, Ade, and His (Fig. 1a). Using LCI assays in tobacco leaves, a strong luminescence signal could be observed in TaMPK3‐nLUC/TaPYL4‐cLUC co‐injection area, whereas no signal was detected in the co‐injection areas of negative controls (Fig. 1b). We then purified recombinant GST‐tagged PYL4 (TaPYL4‐GST) and MBP‐tagged TaMPK3 (TaMPK3‐MBP) for GST pull‐down assays, which showed that TaPYL4‐GST protein could pull down TaMPK3‐MBP protein (Fig. 1c). Further coimmunoprecipitation (Co‐IP) assays using tobacco leaves transiently expressing TaPYL4‐MYC together with Flag or TaMPK3‐Flag confirmed that TaPYL4‐MYC could be coimmunoprecipitated by TaMPK3‐Flag fusion protein but not the Flag control (Fig. 1d). These results showed that TaPYL4 did indeed physically interact with TaMPK3 in vitro and in vivo.

Fig. 1.

TaMPK3 interacted with TaPYL4. (a) Interaction between TaPYL4 and TaMPK3 in the yeast‐two‐hybrid (Y2H) assays. Images were taken after culturing at 30°C for 4 d on synthetic drop‐out selection medium lacking Leu, Trp, His and Ade. The numbers at the top indicate four serial dilutions. AD, GAL4 activation domain; BD, GAL4 DNA‐binding domain. (b) Luciferase complementary indexing (LCI) assays demonstrating that TaPYL4‐cLUC directly interacted with TaMPK3‐nLUC in tobacco (Nicotiana benthamiana). The dotted circles represent the injection areas. Empty vectors were used as negative controls. (c) Pull‐down assays showing that TaPYL4 physically interacted with TaMPK3 in vitro. Pull‐down assays were used to test TaMPK3‐MBP fusion protein or MBP control interaction with TaPYL4‐GST or GST using anti‐MBP immunoblotting. Coomassie brilliant blue image indicates the loading of proteins after pull‐down reactions (bottom). (d) Co‐IP assays showing the interaction between TaPYL4 and TaMPK3 in tobacco. TaPYL4‐MYC vector was co‐transfected into tobacco leaves with TaMPK3‐Flag or Flag vector. After 36–48 h of incubation, total proteins were used for immunoprecipitation with anti‐Flag agarose, and the eluted fractions were detected with anti‐MYC antibody.

TaMPK3 directly interacted with multiple TaPYLs in vitro

At least nine TaPYLs have been previously reported in wheat (Mega et al., ²⁰¹⁹). Given the high sequence similarity (Fig. S3), and similar subcellular locations among these nine proteins (Fig. S4), and therefore potential for redundant functions, we performed BiFC assays in wheat mesophyll protoplasts to determine whether TaPYL1‐9 could physically interact with TaMPK3. These assays showed clear YFP fluorescence in wheat mesophyll protoplasts co‐transformed with TaMPK3‐nYFP and TaPYLs‐cYFP, but not in the negative controls (Fig. 2a).

Fig. 2.

TaMPK3 interacted with multiple TaPYLs. (a) Bimolecular fluorescence complimentary (BiFC) assays showing the interactions between TaPYLs‐cYFP and TaMPK3‐nYFP in wheat (Triticum aestivum) mesophyll protoplasts. Images were observed under a confocal laser scanning microscope (LSM700; Zeiss). Bar, 10 μm. (b) Luciferase complementary indexing (LCI) assays were performed in tobacco (Nicotiana benthamiana) to further confirm the interactions between TaMPK3‐nLUC and TaPYLs‐cLUC. Empty vectors were used as negative controls. (c) Interaction between TaMPK3 and TaPYLs in the yeast‐two‐hybrid (Y2H) assays. Images were taken after culturing at 30°C for 4 d. The numbers at the top indicate three serial dilutions. AD, GAL4 activation domain; BD, GAL4 DNA‐binding domain. (d) Pull‐down assays showing the interactions between TaMPK3 and TaPYLs. Pull‐down assays were used to test TaMPK3‐GST fusion protein interaction with His or TaPYLs‐His using anti‐GST immunoblotting. CBB image indicates the loading of proteins after pull‐down reactions (bottom).

To verify these results, we then performed LCI, Y2H and pull‐down assays. In agreement with BiFC experiments, LCI assays showed strong luminescence signals in samples co‐expressing TaMPK3‐nLUC/TaPYLs‐cLUC, but not in the negative controls, suggesting that TaMPK3 could interact with TaPYL1‐9 (Fig. 2b). Y2H assays showed that the yeast cells co‐expressing TaMPK3‐BD and TaPYL4/5/6/7/8/9‐AD, but not TaPYL1/2/3‐AD, could grow on synthetic drop‐out selection medium lacking Leu, Trp, His and Ade (Fig. 2c), further suggesting interaction between TaMPK3 and a limited set of PYLs in yeast. We then performed pull‐down assays with purified recombinant GST‐tagged TaMPK3 (TaMPK3‐GST) and His‐tagged TaPYL1‐9 (TaPYLs‐His) and found that all nine TaPYL‐His proteins but not His tag could pull down TaMPK3‐GST, indicating that TaMPK3 physically interacted with TaPYLs in vitro (Fig. 2d). In summary, all PYL proteins reported in wheat (TaPYL1‐9) could interact with TaMPK3, suggesting the possibility that TaMPK3 had a conserved regulatory function for control of TaPYL receptor activity in wheat.

MPK3‐PYL interacting module is conserved between monocots and dicots

As interactions between MPK3 and PYLs had not been previously reported in plants, we sought to determine if this interaction module was wheat specific or widely distributed across plant families. To test this possibility, we cloned MPK3 orthologues from two dicots (Arabidopsis and soybean) and two monocots (rice and wheat) to test for interactions between MPK3 and PYLs from these respective species by BiFC and LCI assays in tobacco. BiFC assays showed strong YFP signals in samples co‐expressing TaMPK3‐nYFP/TaPYL4‐cYFP, AtMPK3‐nYFP/AtPYL4‐cYFP, GmMPK3‐nYFP/GmPYL4‐cYFP or OsMPK3‐nYFP/OsPYL9‐cYFP, but no detectable signal in the negative controls, suggesting that MPK3 could interact with PYL4 in Arabidopsis, soybean, and wheat, and PYL9 in rice (Fig. S5a). In agreement with our results in wheat (Fig. 1b), we also observed strong luminescence signals in LCI assays of samples co‐infiltrated with AtMPK3‐nLUC/AtPYL4‐cLUC, GmMPK3‐nLUC/GmPYL4‐cLUC or OsMPK3‐nLUC/OsPYL9‐cLUC, but not the negative controls. These results further verified that MPK3 could interact with PYL4 in Arabidopsis and soybean, and with PYL9 in rice (Fig. S5b). Taken together, these findings showed that the MPK3–PYL interaction module is conserved in both monocot and dicot lineages of higher plants.

ABA sensitivity was reduced in TaMPK3 ‐overexpressing wheat plants and enhanced in TaMPK3‐RNAi wheat plants

Given that the TaMPK3 sequence obtained by Y2H screening was located in subgenome D of wheat, we therefore selected TaMPK3‐4D (TraesCS4D02G198600) to generate TaMPK3‐overexpressing (TaMPK3‐OE) transgenic wheat lines and RNA interference (TaMPK3‐RNAi) plants in the ‘Fielder’ background driven by the maize (Z. mays) ubiquitin promoter. Quantification using qPCR relative expression assays verified that the TaMPK3‐OE lines (OE‐2, OE‐6 and OE‐11) exhibited obviously higher transcript levels (Fig. S6a), whereas the RNAi lines (i‐1, i‐3, and i‐10) had lower TaMPK3 expression (Fig. S6b). Genomic insertion sites in the transgenic lines were identified by Tail‐PCR (Fig. S6c,d).

After sterilisation in 5% sodium hypochlorite and soaking in water for 12 h before the radicle broke through the seed coat, seeds of WT and TaMPK3‐OE wheat plants were placed on wet filter paper containing 0, 1, 5 or 10 μM ABA. All seeds germinated within 12 h on wet filter paper with ABA treatments, with no apparent differences in seed germination observed between WT and TaMPK3‐OE wheat. However, TaMPK3‐OE wheat lines displayed variability in their growth phenotype under different concentrations of ABA (Fig. S7a–c). In particular, the shoot lengths of TaMPK3‐OE seedlings were significantly longer than those of WT under ABA treatments (Fig. S7d). Phenotypic analysis further indicated that all three TaMPK3‐OE lines exhibited decreased ABA sensitivity during the seedling stage compared with that of WT (Fig. 3). More specifically, the shoot and root lengths of TaMPK3‐OE seedlings were significantly longer than those of WT following 9 d of ABA treatment. Furthermore, no significant differences in seedling shoot length or root length were detected between TaMPK3‐OE and WT wheat under control conditions (Fig. 3a,b). The TaMPK3‐RNAi lines exhibited distinct growth patterns at 9 d of ABA treatment during the seedling stage (Fig. 3c,d). These results suggested that TaMPK3 overexpression weakened and TaMPK3‐RNAi enhanced the inhibitory effects of ABA on the wheat growth.

Fig. 3.

Abscisic acid (ABA) sensitivity was reduced in TaMPK3‐overexpressing wheat (Triticum aestivum) plants and enhanced in TaMPK3‐RNAi wheat plants. (a) Phenotypes of TaMPK3‐overexpressing (OE‐2, OE‐6, and OE‐11) and wild‐type (WT) wheat plants under control and different concentrations of ABA treatments for 9 d. (b) Bar graphs of seedling height (shoot and root length) under control and different concentrations of ABA treatments for 9 d. Each treatment had at least three independent replicates and each replicate contained eight plants. Each data point was the mean (±SD) of 15 seedlings. (c) Phenotypes of TaMPK3‐RNAi (i‐1, i‐3, and i‐10) and WT wheat plants under control and different concentrations of ABA treatments for 9 d. (d) Bar graphs of seedling height (shoot and root length) under control and different concentrations of ABA treatments for 9 d. Each treatment had at least three independent replicates and each replicate contained eight plants. Each data point was the mean (±SD) of 15 seedlings. Statistical analysis using one‐way ANOVA with multiple comparisons revealed the significant differences compared with WT (*, P < 0.05; **, P < 0.01).

ABA diminished interactions between TaMPK3 and TaPYL4 resulting inhibited the degradation of TaPYL4

Considering that TaPYL4 overexpression reportedly increases ABA sensitivity and drought tolerance of wheat (Mega et al., ²⁰¹⁹), while TaMPK3 overexpression resulted in decreased sensitivity to ABA, we therefore hypothesised that TaMPK3 could decrease the stability of TaPYLs. Western blot analysis of cell‐free lysates showed that TaPYL4 protein degradation was significantly suppressed in TaMPK3‐RNAi plants, but elevated in TaMPK3‐OE plants compared with WT. In addition, treatment with 26S proteasome inhibitor MG132 resulted in partial inhibition of TaPYL4 degradation (Fig. 4a), indicating that TaMPK3 performed a major role in regulating the degradation and stability of TaPYLs in wheat.

Fig. 4.

Abscisic acid (ABA) diminished the interaction between TaMPK3 and TaPYL4 and influenced the degradation of TaPYL4. (a) In vitro cell‐free protein degradation assays, showing degradation of TaPYL4‐His in extracts from TaMPK3‐RNAi, wild‐type (WT), and TaMPK3‐overexpressing wheat (Triticum aestivum) plants under control and 20 μM ABA treatments. Immunoblots were probed with anti‐His antibody. Rubisco was used as a loading control. (b) The interactions between TaMPK3‐BD and TaPYLs‐AD based on yeast‐two‐hybrid (Y2H) assays. ABA (5 μM) was added in the synthetic drop‐out selection medium lacking Leu, Trp, His and Ade to test the influence of ABA on the interactions of TaMPK3 and TaPYLs. Control was used for excluding the nonspecific effects of ABA. The numbers at the top indicate three serial dilutions. AD, GAL4 activation domain; BD, GAL4 DNA‐binding domain. (c) Luciferase complementary indexing (LCI) assays demonstrating that TaPYL4‐cLUC directly interacted with TaMPK3‐nLUC in tobacco (Nicotiana benthamiana) and ABA diminished the interaction. The dotted circles represent the injection areas. Three biological replications were performed with similar results. Each data point was the mean relative luciferase activity (±SD) of five leaves. Statistical analysis using Student's t‐test revealed the significant differences (**, P < 0.01) between control and 10 μM ABA treatments. (d) Expression of ABA signalling genes, including TaPP2C1, TaPP2C2, TaPP2C6, TaDHN3, and TaPOD21, in TaMPK3‐OE and WT plants under control (CK) and 20 μM ABA treatment conditions. Each data point was the mean (±SD) of three biological replicates. Statistical analysis using Student's t‐test revealed the significant differences of these genes (*, P < 0.05; **, P < 0.01) between WT and TaMPK3‐overexpressing wheat plants under CK and 20 μM ABA treatment conditions.

As ABA treatment diminished TaPYL4 degradation in the TaMPK3‐OE, WT and TaMPK3‐RNAi wheat plants (Fig. 4a), we therefore sought to determine whether ABA had any effect on interactions between TaMPK3 and TaPYLs. Using Y2H assays, we confirmed that exposure to 5 μM ABA diminished the interaction between TaMPK3 and TaPYLs in yeast, indicated by slower growth of yeast cells co‐transformed with TaMPK3‐BD and TaPYLs‐AD compared with that of untreated cells (Fig. 4b). In tobacco, we also detected a lower luminescence signal in leaves co‐infiltrated with TaMPK3‐nLUC and TaPYL4‐cLUC constructs under 10 μM ABA treatment compared with that of control leaves (Fig. 4c). YB2‐nLUC and YA16‐cLUC (Yu et al., ²⁰²¹) was used as the positive control (Fig. S8), the interaction of which was not influenced by ABA treatment. Consistent with a reduction in sensitivity to ABA, subsequent qPCR‐based quantification of ABA signalling pathway gene expression (i.e. TaPP2C‐1, TaPP2C‐2, TaPP2C‐6, TaDHN3 and TaPOD21) in ABA‐treated or ABA‐untreated TaMPK3‐OE and WT seedlings revealed a significant decrease in expression in TaMPK3‐OE plants compared with WT under ABA treatment. In addition, TaPP2C‐6, TaDHN3 and TaPOD21 exhibited significantly higher expression in TaMPK3‐OE wheat under untreated condition (Fig. 4d). These results indicated that TaMPK3 contributed to reducing the sensitivity to ABA through ubiquitination‐dependent inhibition of TaPYL4.

TaMPK3 could phosphorylate TaPYL4 in an autophosphorylation‐dependent manner and promote the degradation of TaPYL4 in vitro

Previous studies have indicated that TaMPK3 can autophosphorylate its Tyr‐196 residue, but not its Ser or Thr residues and that it exhibits autophosphorylation‐dependent protein kinase activity (Takezawa, 1999). To assess its autophosphorylation, we first generated a variant of TaMPK3 (TaMPK3^K65R) in which the lysine (K) residue at position 65 was converted to an arginine (R) to abolish its autophosphorylation activity. We then incubated purified TaMPK3‐GST, TaMPK3^K65R‐GST, or GST (empty vector) proteins with kinase reaction buffer and ATP, then quantified autophosphorylation by 75 μM phos‐tag™ Acrylamide AAL‐107 assays with anti‐GST monoclonal antibody. The results showed that WT kinase (TaMPK3‐GST) underwent autophosphorylation, indicated by hysteretic migration, whereas the mutant TaMPK3^K65R‐GST and the GST tag alone did not (Fig. 5a).

Fig. 5.

TaMPK3 could phosphorylate TaPYL4 and influence the degradation of TaPYL4. (a) Phos‐tag assay showing the autophosphorylation of TaMPK3. Phos‐tag™ acrylamide was used in this experiment. TaMPK3^K65R was used as the mutation of TaMPK3, in which the 65^th amino acid was mutated from K to R for loss of autophosphorylation activity. CBB image indicates the loading of proteins (bottom). (b) TaMPK3 phosphorylated TaPYL4 and TaPYL4^M1 but not TaPYL4^M2 in vitro. Phos‐tag™ acrylamide was used to analyse phosphorylation. Input or CBB indicates the loading of proteins (bottom). TaPYL4^M1 stands for TaPYL4^S58A; TaPYL4^M2 stands for TaPYL4^T3A/T177A. λ protein phosphatase (λPPase) was used to verify that the delayed migration band of TaPYL4 in phos‐tag gel was the result of its phosphorylation. (c) Cell‐free lysates via protoplasts transiently expressing TaMPK3‐Flag, TaMPK3^K65R‐Flag or TaMPK3^D191G/E195A‐Flag. Anti‐Flag antibody was used to test the protein level of TaMPK3‐Flag, TaMPK3^K65R‐Flag or TaMPK3^D191G/E195A‐Flag in 30 μg protoplasts total protein. (d–f) In vitro degradation assays the degradation of TaPYL4‐His (d), TaPYL4^M1 (e), and TaPYL4^M2 (f) in cell‐free lysates of protoplasts transiently expressing TaMPK3‐Flag, TaMPK3^K65R‐Flag, TaMPK3^D191G/E195A‐Flag or wild‐type (WT). Immunoblots were probed with anti‐His antibody. Rubisco was used as a loading control.

In Arabidopsis, the phosphorylation status of PYL4 at Ser54 influenced its stability under normal conditions (Yu et al., ²⁰¹⁹), to verify whether the related site (Ser58 in TaPYL4) had an effect on the stability of TaPYL4, we created the mutant1 of TaPYL4 (TaPYL4^S58A). As proline‐guided MAP kinases could phosphorylate Ser or Thr residues followed by a C‐terminal proline residue (S/TP site), we speculated that Thr3 and Thr177 of TaPYL4 might be the phosphorylation sites of TaMPK3 (Fig. S9) and then created the mutant2 of TaPYL4 (TaPYL4^T3A/T177A). In addition, MPK3^D193G/E197A is constitutively active in Arabidopsis (Berriri et al., ²⁰¹²), we therefore created related mutant of TaMPK3 (TaMPK3^D191G/E195A) to further clarify the relationship between TaMPK3 and TaPYL4. The modified amino acid residues in TaMPK3 (Fig. S1) or TaPYL4 were conserved among A, B, and D subgenomes (Fig. S9b). To observe TaMPK3‐mediated substrate phosphorylation, we incubated TaMPK3‐GST with TaPYL4‐His, TaPYL4^S58A‐His and TaPYL4^T3A/T177A‐His in kinase reaction buffer, then conducted immunoblot assays using phos‐tag assays. The hysteretic migration band of TaPYL4 indicated that TaMPK3 could phosphorylate TaPYL4 and TaPYL4^S58A, but not TaPYL4^T3A/T177A. λPPase was used to verify that the delayed migration band of TaPYL4 in phos‐tag gel was the result of its phosphorylation (Fig. 5b). Taken together, these results indicated that TaMPK3 could function as a protein kinase targeting TaPYL4 in a manner dependent on its autophosphorylation.

We firstly verified that a strong luminescence signal could be observed in TaMPK3^K65R/TaPYL4, TaMPK3^D191G/E195A/TaPYL4, TaMPK3/TaPYL4^S58A or TaPYL4^T3A/T177A co‐injection area, whereas no signal was detected in the co‐injection areas of negative controls by LCI assays in tobacco, which indicated the interaction existed between them (Fig. S10). To verify the impact of TaMPK3, TaMPK3^K65R or TaMPK3^D191G/E195A on the stability of TaPYL4, TaPYL4^S58A or TaPYL4^T3A/T177A, we used western blot analysis of cell‐free lysates via protoplasts transiently expressing TaMPK3‐Flag, TaMPK3^K65R‐Flag or TaMPK3^D191G/E195A‐Flag. We first corrected the protein level of TaMPK3‐Flag in 30 μg protoplasts total protein (Fig. 5c). Then, in vitro degradation assays of PYL4 showed that the degradation of TaPYL4 and its two mutant proteins was significantly elevated in the protoplasts with TaMPK3‐Flag, but the TaMPK3^K65R or TaMPK3^D191G/E195A mutant had no discernible effect on PYL4 degradation compared with WT (Fig. 5d–f). These results indicated that TaMPK3 promoted degradation of TaPYL4 was not related to its phosphorylation status, and TaMPK3^K65R or TaMPK3^D191G/E195A impacted the function of TaMPK3 on TaPYL4 by some way unknown.

Exogenous ABA‐ and drought‐induced TaMPK3 transcription and protein accumulation

The expression of TaMPK3, also known as WCK‐1 (Takezawa, 1999), is reportedly induced by the fungal elicitor calcium ionophore A23187 as well as drought (Takezawa, 1999; Cevher‐Keskİn et al., ²⁰¹⁵). In the present study, we conducted qPCR‐based mRNA quantification and immunoblot assays to examine TaMPK3 expression patterns in different tissues and at seedling stage under PEG6000‐induced drought or exogenous ABA treatments. The results of qPCR assays indicated that TaMPK3 was most highly expressed in vegetative tissues, and especially in leaves (Fig. S11a), with mRNA levels increasing after PEG and ABA treatments, respectively peaking at 0.5 and 1 h in leaves (Fig. S11b). Compared with untreated controls (Fig. S11c), exogenous ABA (Fig. S11d) and PEG (Fig. S11e) treatments increased the abundance of TaMPK3 protein, and anti‐Erk1/2 antibody was used for the western blot (WB) assays (Fig. S12). Taken together, these results suggested that TaMPK3 expression was upregulated by exposure to drought conditions or exogenous ABA at both the transcript and protein levels.

TaMPK3 was rapidly induced during post‐drought recovery while TaPYL4 degradation surged and weakened concurrently with TaMPK3 protein levels

We analysed TaMPK3 expression levels in wheat under three different degree of drought treatments and during rehydration periods, and found that it was slightly induced in wheat plants that suffered drought stress with c. 47% soil water content (Fig. S13c) and rapidly induced by 0.5 h of rehydration (Fig. S13d,e). Although the expression levels of TaMPK3 started to decline after 1 h of rehydration, however it still maintained higher expression levels compared with that under drought conditions (Fig. S13d). To reinforce the evidence, we exposed the wheat to relatively severe drought condition with c. 31% soil water content, and then re‐watered for recovery to explore the expression level of TaMPK3. We found that it was strongly induced after 0.5 h of re‐watering and its transcripts still kept high expression levels after 1 and 2 h of re‐watering (Fig. S13e). Consistent with these findings, the degradation of TaPYL4 decreased concurrently with the diminishing TaMPK3 protein levels during rehydration stage in vitro (Fig. S13f,g). These results indicated that the inhibition of TaPYL4 by TaMPK3 remained stable under drought conditions and that plants need high, but transient, expression of TaMPK3 for rapid recovery after drought stress.

Wheat seedling drought tolerance was decreased by TaMPK3 overexpression and increased by TaMPK3 suppression

In light of our findings of the effects of TaMPK3 on ABA sensing, we next examined how its overexpression or suppression affected drought tolerance in wheat seedlings. To this end, we examined the phenotypes of TaMPK3‐OE, WT and TaMPK3‐RNAi seedlings that were germinated in well watered conditions, but then withheld irrigation for 18 d. Notably, the overexpression lines began to wilt on day 12 post germination, and became flaccid by day 15, compared with WT plants that wilted by day 13 and became flaccid by day 15 (Fig. 6a). Measurements of proline at day 12 post germination and malondialdehyde (MDA) at day 15 showed that proline contents were significantly lower in TaMPK3‐OE plants than WT, whereas MDA contents were significantly higher (Fig. 6b,c). After a 3‐d recovery period with full irrigation, we found that TaMPK3‐OE plants were severely dehydrated and largely unable to recover (Fig. 6a), displaying significantly lower survival rates and shoot fresh weights than WT (Fig. 6d,e).

Fig. 6.

Overexpression of TaMPK3 significantly decreased and RNAi of TaMPK3 significantly increased drought tolerance in wheat (Triticum aestivum) at seedling stage. (a) Phenotypes of TaMPK3‐RNAi (i‐1, i‐3, and i‐10), wild‐type (WT), and TaMPK3‐overexpressing (OE‐2, OE‐6, and OE‐11) wheat plants under withholding water treatments for water‐deficit stress. Follow‐up photographs were taken at 6, 13, 15, and 16 d after planting. After water was withheld for 18 d, all plants were re‐watered and photographs were taken after 3 d of rehydration. Each treatment had at least three independent replicates and each replicate contained 18 plants. (b) The survival rates of TaMPK3‐RNAi, WT and TaMPK3‐overexpressing wheat plants under water‐deficit stress. The wheat seedlings were re‐watered on the 18^th day and survival rate was calculated 3 d later. (c–e) The fresh weight of shoot after re‐watering for 3 d (c); the proline content at the 12^th day (d); and the malondialdehyde (MDA) content at the 15^th day (e) of TaMPK3‐RNAi, WT and TaMPK3‐overexpressing wheat plants under water‐deficit stress treatment. Each data point was the mean (±SD) of three biological replicates. Statistical analysis using one‐way ANOVA with multiple comparisons revealed the significant differences compared with WT (*, P < 0.05; **, P < 0.01).

By contrast, TaMPK3‐RNAi plants began to wilt on day 13 post germination and became flaccid by day 16 (Fig. 6a). However, these plants displayed a more drought tolerant phenotype than WT, with higher survival rates, shoot fresh weight and proline content, but lower MDA content compared with WT (Fig. 6b,d,e). Notably, WT and TaMPK3‐RNAi wheat seedlings showed similar survival rates (> 80%) after 3 d of recovery post‐drought treatment, whereas the TaMPK3‐OE lines had an average of 3% survival rates.

TaMPK3 negatively regulated drought resistance of wheat at the vegetative stage

In consideration of these effects in wheat seedlings, we next examined the drought tolerance and agronomic trait phenotypes of adult (jointing) stage TaMPK3‐OE, WT and TaMPK3‐RNAi plants under glasshouse conditions. Except for that TaMPK3‐OE had shorter plant heights than WT and TaMPK3‐RNAi wheat plants under well watered condition, there were no significant phenotypic differences among them (Fig. S14). After withholding water for 15 d at the heading stage, followed by full irrigation until ripe, we observed significant differences in plant growth, and spike and seed development among TaMPK3‐OE, WT and TaMPK3‐RNAi plants under drought conditions (Fig. 7a), with TaMPK3 overexpression incurring obvious negative impacts on agronomic traits. For example, spikelet number, tiller number, panicle length, plant height, grain number per plant, grain weight per plant, 1000 grain weight, grain width and grain length (Fig. 7b–j) were all significantly lower in the OE lines compared with WT following drought treatments. By contrast, TaMPK3‐RNAi plants exhibited higher tiller number (Fig. 7c), longer panicle length (Fig. 7d), higher grain number per plant (Fig. 7f) and higher grain weight per plant (Fig. 7g) than WT. However, no significant differences were identified between TaMPK3‐RNAi and WT plants for spikelet number (Fig. 7b), plant height (Fig. 7e), 1000 grain weight, grain width or grain length (Fig. 7h–j).

Fig. 7.

Overexpression of TaMPK3 significantly decreased and RNAi of TaMPK3 significantly increased drought tolerance in wheat (Triticum aestivum) in glasshouse conditions. (a) The phenotypic analysis of TaMPK3‐RNAi (i‐1, i‐3, and i‐10), wild‐type (WT), and TaMPK3‐overexpressing (OE‐2, OE‐6, and OE‐11) wheat plants under drought stress in glasshouse conditions. (b–j) The spikelet number (b); the tiller number (c); the panicle length (d); the plant height (e); the grain number per plant (f); the grain weight per plant (g); the 1000 grain weight (h); the grain width (i); and the grain length (j) of TaMPK3‐RNAi, WT, and TaMPK3‐overexpressing wheat lines under drought stress in glasshouse conditions. Each treatment had four independent replicates and each replicate contained six plants. Each data point was the mean (±SD) of 10 independent samples. Statistical analysis using one‐way ANOVA with multiple comparisons revealed the significant differences compared with WT (*, P < 0.05; **, P < 0.01).

We then investigated the drought tolerance of TaMPK3‐OE, WT and TaMPK3‐RNAi wheat plants at the flowering stage in field conditions and found that the results were consistent with that of drought tolerance assays conducted at the seedling (Fig. 6) and jointing stages (Fig. 7) under glasshouse conditions. In the field, TaMPK3‐OE, WT and TaMPK3‐RNAi wheat plants showed no significant phenotypic differences under well watered conditions with the exception of shorter plant heights for TaMPK3‐OE plants than that of WT and TaMPK3‐RNAi lines (Fig. S15). However, when water was withheld for 15 d after flowering, the leaves of TaMPK3‐OE wheat plants became yellow earlier than those of WT, which in turn yellowed slightly earlier than that of TaMPK3‐RNAi plants (Fig. 8a). Notably, under drought conditions in the field, no significant differences were detected in tiller number, effective tiller number, grain number per plant, panicle length, spikelet number, or grain length (Fig. 8) among TaMPK3‐OE, WT and TaMPK3‐RNAi wheat plants. With increasing duration of drought treatment, TaMPK3‐OE lines eventually displayed lower 1000 grain weight (Fig. 8j) and grain width (Fig. 8k) compared with WT and TaMPK3‐RNAi plants. The TaMPK3‐RNAi wheat lines showed no significant difference in agronomic traits relative to WT under drought conditions in the field (Fig. 8). These results indicated that TaMPK3 overexpression conferred negative effects on drought tolerance and yield, while TaMPK3 suppression resulted in positive or neutral impacts depending on growth stage and conditions.

Fig. 8.

Overexpression of TaMPK3 significantly decreased drought tolerance in wheat (Triticum aestivum) in the field. (a, b) Phenotypic analysis (a) and the seeds (b) of TaMPK3‐RNAi (i‐1 and i‐3), wild‐type (WT), and TaMPK3‐overexpressing (OE‐2 and OE‐11) wheat plants under drought condition in the field. (c–k) The plant height (c); the tiller number (d); the effective tiller number (e); the grain number per plant (f); the panicle length (g); the spikelet number (h); the grain length (i); the grain width (j); and the 1000 grain weight (k) of TaMPK3‐RNAi, WT and TaMPK3‐overexpressing wheat lines under drought stress in the field. Each treatment had three independent replicates and each replicate contained 15 plants. Each data point was the mean (±SD) of 10 independent samples. Statistical analysis using one‐way ANOVA with multiple comparisons revealed the significant differences compared with WT (**, P < 0.01).

Discussion

Abscisic acid functions as the major hormone for regulating plant stress response and adaptation for survival in harsh and variable environments. When challenged by drought stress, plants increase their synthesis of ABA to reduce transpiration and photosynthesis, reprogramme metabolism to produce osmoprotectants, inhibit growth and promote senescence to adapt and to survive in severe conditions (Zhu, 2016; Zhang et al., ²⁰²¹). The ABA signalling pathway is conserved across all plants and has been proposed as an essential factor in the early adaptation to terrestrial growth (Chen et al., ²⁰²¹). The main ABA receptors in plants are members of the PYL protein family (Park et al., ²⁰⁰⁹; Miyakawa et al., ²⁰¹³). A recent study suggested that the earliest common ancestor of terrestrial plants may have diverged from unicellular algae after acquiring PYL genes from soil bacteria (Cheng et al., ²⁰¹⁹). These genes may have later undergone neofunctionalisation to adapt to limited water conditions during the colonisation of land (Hauser et al., ²⁰¹¹). Recent studies suggest that PYL phosphorylation at different sites by various protein kinases may play distinct roles in different stress‐response pathways (Chen et al., ²⁰¹⁸; L. Zhang et al., ²⁰¹⁸; Wang et al., ²⁰¹⁸; Li et al., ²⁰¹⁹; Yu et al., ²⁰¹⁹). It can be inferred from the limited evidence that the involvement of different protein kinases and protein phosphatases in regulating PYL control of balance between plant growth and stress responses should arouse wide concerns. To better understand the regulatory mechanisms controlling the PYL‐mediated ABA signalling pathway, TaPYL4, which has been reported to contribute to drought tolerance in wheat (Mega et al., ²⁰¹⁹), was selected for further study. Screening for interaction partners of TaPYL4 under drought stress revealed TaMPK3 as a candidate protein (Fig. 1), suggesting that TaMPK3 also participated in TaPYL4‐mediated ABA signalling.

Moreover, our understanding of the relationship between MAPK signal cascades and ABA signalling is still fragmented in wheat. It is well known that ABA modulates the transcription, protein accumulation, and kinase activity of several MAPK signalling pathway components in diverse plant species (Liu, 2012; Danquah et al., ²⁰¹⁴; Wimalasekera & Scherer, ²⁰¹⁸). Similarly, our study shows that the ABA signalling pathway could regulate TaMPK3 expression at both the transcriptional and protein levels in wheat. In particular, the finding of ABRE response elements in the promoter region of TaMPK3 (Table S8) suggests direct transcriptional regulation by the ABA signalling pathway. In addition to transcriptional modulation by ABRE elements, earlier research revealed that ABA can mediate the induction of Ca²⁺ production under stress conditions (Hu et al., ²⁰⁰⁶), which contributes to Ca²⁺ accumulation, resulting in the upregulation of TaMPK3 (Takezawa, 1999). This finding indicates that ABA can also indirectly regulate MPK3 accumulation by mediating the induction of Ca²⁺ production. Notably, TaMPK3 activity can also be modulated at the protein level. Specifically, pharmacological evidence suggests that ABA can indirectly activate kinase activity through the Ca²⁺–CaM‐mediated pathway in Arabidopsis (de Zelicourt et al., ²⁰¹⁶). Here, we confirmed that TaMPK3 could interact with CaM (Fig. S16), which is proposed to participate in activating multiple protein kinases, protein phosphatases and channel proteins (Bouché et al., ²⁰⁰⁵). This cumulative evidence strongly suggests that TaMPK3 activation might be directly regulated through Ca²⁺–CaM signalling in wheat. Taken together, our results therefore suggest that ABA signalling and TaMPK3 protein appear to form an ABA‐TaMPK3‐TaPYL4 loop, which could play an important role in regulating of tolerance to drought or other abiotic stresses in plants.

In light of the regulatory interactions between TaMPK3 and TaPYL4, we wondered if TaMPK3 functioned through conserved interaction modules with other PYLs in wheat. It is well known that proline‐guided serine/threonine MAPKs harbour a CD site that mediates substrate binding through a complementary binding motif (i.e. the D‐site) that is recognised by its interaction partners (Cargnello & Roux, 2011; Doczi & Bogre, ²⁰¹⁸). Analysis of PYL homologue protein sequences from wheat and Arabidopsis surprisingly revealed that TaPYLs had no canonical D‐site, which could explain why MPK3–PYL interactions have never been identified by high‐throughput sequence‐based screening in any species, to our knowledge. However, we verified that TaMPK3 could interact with all TaPYLs using BiFC assays in wheat protoplasts, LCI assays in tobacco leaves, and in pull‐down assays (Fig. 2). Furthermore, our findings confirmed that a conserved MPK3–PYL interaction module is present in both monocots and dicots (Fig. S5). Previous studies have shown that TaPYL4 overexpression can increase ABA sensitivity and drought tolerance in wheat (Mega et al., ²⁰¹⁹). Here, overexpression of TaMPK3 in wheat conferred hyposensitivity to ABA, leading to the attenuation of ABA inhibition of shoot elongation and decreased the expression of ABA‐responsive genes (Figs 3, 4, S7), as well as a significant decrease in wheat drought tolerance. Conversely, RNAi suppression of TaMPK3 significantly increased drought tolerance in wheat (Figs 6, 7, 8). These data led us to hypothesise that the effects of conserved interactions between TaMPK3 and different TaPYLs were related to TaPYL stability. Indeed, we found that TaMPK3 could promote TaPYL4 degradation, whereas ABA treatment diminished the interaction between TaMPK3 and TaPYL4, thereby inhibiting TaPYL4 degradation (Fig. 4). As proline‐guided MAP kinases phosphorylate serine or threonine residues followed by a C‐terminal proline residue (S/TP site), we then searched for this target site in TaPYLs and discovered that S/TP sites were absent in some (Fig. S9). These results suggested that, although the mechanism of TaMPK3–TaPYL interaction is conserved, TaMPK3‐mediated degradation of TaPYL ABA receptors may not depend on their phosphorylation by TaMPK3.

Based on these collective results, we therefore proposed that the interaction with TaMPK3 can inhibit TaPYL activity by mediating their degradation, therefore negatively regulating ABA response. We speculated that inhibition TaPYL activity by TaMPK3 could initiate a return to normal growth during rehydration after drought. In this period, the ABA signal must be decreased for regulators of growth under normal conditions to function correctly. To check this hypothesis, we analysed TaMPK3 expression levels in wheat during both drought treatment and re‐watering, and found that it was rapidly induced by 0.5 h of re‐watering, but declined after 2 h (Fig. S13d,e). Consistent with this finding, TaPYL4 degradation decreased concurrently with diminishing TaMPK3 protein levels during post‐drought recovery in vitro (Fig. S13f,g). These results indicated that TaPYL inhibition by TaMPK3 remained stable under drought conditions and that plants required high, but transient, expression of TaMPK3 for a rapid post‐drought recovery process.

Previous studies have revealed that ABA levels are elevated under drought, which can improve plant stress tolerance (Zhu, 2016; Hauser et al., ²⁰¹⁷; Zhang et al., ²⁰²¹). In addition, high ABA levels can inhibit plant growth and induce plant transition from vegetative to reproductive growth (Du et al., ²⁰¹⁸). As the surrounding environment returns to nonstress conditions, plants should make appropriate adjustments to function optimally. Based on this reasonable presumption in conjunction with our experimental findings, we proposed that TaMPK3 suppresses TaPYL activity to weaken ABA signalling as a crucial step in inhibiting the stress response during post‐drought recovery and facilitating activation of normal growth pathways. Considering that TaMPK3‐RNAi could significantly improve wheat drought tolerance without obviously affecting other agronomic traits, we propose that modulating TaMPK3 expression can improve plant drought tolerance by weakening its inhibition of ABA signalling. This potential application is promising for breeding programmes focused on stress tolerance in crops and provides valuable reference data for TaMPK3 expression under multiple growth conditions. However, it remains unclear what triggers high expression of TaMPK3 under post‐drought recovery. Collectively, our research revealed a dynamic molecular regulatory mechanism for normal, drought and post‐drought recovery periods via TaMPK3‐mediated inhibition of ABA signalling via the interaction and targeted degradation of TaPYL (Fig. 9).

Fig. 9.

The negative regulation model of the MPK3‐PYL module in abscisic acid (ABA) signalling with or without drought stress in wheat (Triticum aestivum). Interaction with TaMPK3 may inhibit TaPYL4 activity under both normal and stress conditions to negatively regulate ABA response via degradation of PYL4. Under normal conditions, TaPYL4 suppression by TaMPK3 may inhibit stress signalling and allow normal growth to proceed, while rapid production of ABA under drought stress can reduce TaMPK3‐mediated PYL degradation to promote ABA signalling, consequently activating a response to stress responses. To limit the inhibition of plant growth by stress signals and/or rapidly return to normal growth after the stress stimulus disappears, plants then increase TaMPK3 levels to downregulate TaPYL function in ABA signalling. We therefore propose that a regulatory module between the ABA core signalling components and MPK3 represent a critical mechanism for balancing stress and growth responses, allowing wheat plants to adapt to continuously changing environments. Arrows indicate promotion, bars (T) indicate inhibition, and dashed lines indicate postulated regulation.

Competing interests

None declared.

Author contributions

Z‐SX and Y‐ZM coordinated the project, conceived and designed experiments and edited the manuscript; YL and T‐FY performed the experiments and wrote the first draft; YL, T‐FY and Y‐TL generated and analysed data; LZ, Z‐WL, J‐PZ, G‐ZS, X‐YC, Y‐WL and MC contributed valuable discussions; Y‐BZ and JC contributed to the generation of the experiment materials. All authors have read and approved the final manuscript. YL, T‐FY and Y‐TL contributed equally to this work.

Supporting information

Fig. S1 Alignment and domain analysis of MPK3.

Fig. S2 Molecular phylogenetic analysis of plant MPK3 genes.

Fig. S3 Amino acid sequence alignment of TaPYLs.

Fig. S4 The subcellular location of nine TaPYLs.

Fig. S5 MPK3‐PYL module exists widely in monocots and dicots.

Fig. S6 Overexpression and RNAi of TaMPK3 in wheat (Triticum aestivum).

Fig. S7 Wheat (Triticum aestivum) plants overexpressing TaMPK3 showed reduced abscisic acid sensitivity.

Fig. S8 The impact of abscisic acid treatment on the interaction between TaMPK3 or TaMPK3^K65R and TaPYL4.

Fig. S9 The S/TP site analysis of TaPYLs.

Fig. S10 The interaction between different mutant combination of TaMPK3 and TaPYL4.

Fig. S11 Abscisic acid and polyethylene glycol induced the accumulation of TaMPK3 at both transcriptional and protein levels.

Fig. S12 The specificity of anti‐Erk1/2 on TaMPK3 protein.

Fig. S13 The responses of TaMPK3 and TaPYL4 during drought and/or post‐drought recovery stage.

Fig. S14 The agronomic traits of TaMPK3‐RNAi, wild‐type and TaMPK3‐overexpressing wheat (Triticum aestivum) in a glasshouse under watered condition.

Fig. S15 The agronomic traits of TaMPK3‐RNAi, wild‐type and TaMPK3‐overexpressing wheat (Triticum aestivum) in the field under watered condition.

Fig. S16 TaMPK3 interacted with TaCaM.

Methods S1 Details of drought experiments.

Table S1 Summary of primers used in this study.

Table S2 Soil water content (%) during the drought treatments at the seedling stage.

Table S3 Annual rainfall data from 2010 to 2020 at the experimental station (40°13′52″N, 116°33′52″E) of the Institute of Crop Sciences, CAAS, Beijing.

Table S4 Monthly rainfall data from 2018 to 2020 at the experimental station (40°13′52″N, 116°33′52″E) of the Institute of Crop Sciences, CAAS, Beijing.

Table S5 Soil water content (%) in the field.

Table S6 The list of the interacting candidates of TaPYL4.

Table S7 Sequence information of MPK3 in different species.

Table S8 The promoter of TaMPK3 gene.

Please note: Wiley Blackwell are not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

Data availability

The data that support the findings of this study are available from the corresponding author Z‐SX, upon reasonable request.