MdMYB46 could enhance salt and osmotic stress tolerance in apple by directly activating stress‐responsive signals

Summary To expand the cultivation area of apple (Malus×domestica Borkh.) and select resistant varieties by genetic engineering, it is necessary to clarify the mechanism of salt and osmotic stress tolerance in apple. The MdMYB46 transcription factor was identified, and the stress treatment test of MdMYB46‐overexpressing and MdMYB46‐RNAi apple lines indicated that MdMYB46 could enhance the salt and osmotic stress tolerance in apple. In transgenic Arabidopsis and apple, MdMYB46 promoted the biosynthesis of secondary cell wall and deposition of lignin by directly binding to the promoter of lignin biosynthesis‐related genes. To explore whether MdMYB46 could coordinate stress signal transduction pathways to cooperate with the formation of secondary walls to enhance the stress tolerance of plants, MdABRE1A, MdDREB2A and dehydration‐responsive genes MdRD22 and MdRD29A were screened out for their positive correlation with osmotic stress, salt stress and the transcriptional level of MdMYB46. The further verification test demonstrated that MdMYB46 could activate their transcription by directly binding to the promoters of these genes. The above results indicate that MdMYB46 could enhance the salt and osmotic stress tolerance in apple not only by activating secondary cell wall biosynthesis pathways, but also by directly activating stress‐responsive signals.


Introduction
Drought and salt stresses are the two of major environmental factors that affect the geographical distribution of plants in nature, limit plant productivity in agriculture and threaten food security (Zhu, 2001(Zhu, , 2002(Zhu, , 2016. As a consequence of these stresses, plants may suffer photosynthesis inhibition, metabolic dysfunction and cellular structure damage (Krasensky and Jonak, 2012). Apple is a worldwide fruit tree resource and an integral part of a human nutritious diet, but their global cultivation and promotion is limited by drought and high salt stress. So how apple plants sense stress signals and adapt to adverse environments are fundamental biological questions.
Drought mainly led to the osmotic stress while salt stress has both osmotic and ionic or ion-toxicity effects on cells (Zhu, 2016). The stress resistance of plants is closely related to the formation of secondary cell wall and the deposition of lignin. As the second largest biopolymer in vascular plants (Boerjan et al., 2003;Himmel, 2008), the main physiological function of lignin includes providing a structural barrier to the cell wall (Ferrer et al., 2008;Sykes et al., 2015;Weng and Chapple, 2010), performing basic biological functions such as mechanical support, impermeability and resistant biodegradation, and establishing biological defence mechanism (Bhuiyan et al., 2009;Xu et al., 2009). The lignin biosynthesis pathway is usually divided into two parts: the common phenylpropane pathway from phenylalanine to feruloyl-CoA and the specific pathway from feruloyl-CoA to monolignin units (Vanholme et al., 2012), which were regulated by a three-level regulatory network (Hussey et al., 2013;Kumar et al., 2015). The regulators of Tier 3 mainly include NST1 (NAC secondary wall thickening promoting factor 1), SND1 (secondary wall thickening promoting factor 1), NST2, VND6 (vascular-related NAC domain 6) and VND7 transcription factors (TFs), which can regulate the expression of Tier 2 or Tier 1. Tier 2 regulatory factors including MYB family TFs (MYB46, MYB83 and MYB55) and NAC TFs [SND3 and XND1 (xylem NAC domain 1)] (Yamaguchi et al., 2011;Zhong et al., 2010) can directly regulate the expression of structural genes and Tier 1, which can also directly bind to the cis-element of structural genes. Recent study found that MdMYB88 and MdMYB124 could regulate root xylem development by directly binding MdMYB46 promoters and influencing the expression of their target genes under drought condition (Geng et al., 2018). In addition to TFs, the metabolic regulatory network of lignin is also regulated by microRNAs (Zhao et al., 2015) and Kelch repeat F-box proteins . Although the regulatory network of lignin metabolism and its role in plant stress resistance have been paid more attention, the related research is mostly carried out in herbaceous plants such as Arabidopsis thaliana, and the related research on fruit trees, such as apple, is still lacking.
MYB46, as the direct target of SND1, plays a key regulator in the transcriptional network involved in the regulation of secondary wall biosynthesis in Arabidopsis (Zhong et al., 2007). Although its role in regulating secondary wall formation and its regulation mode of activating gene expression by binding elements such as SMRE, MYBCORE and AC-box in the promoter region have been elucidated (Nakano et al., 2015;Zhong and Ye, 2011), its function in other metabolic (such as disease susceptibility) and signal transduction pathways in plants is still being explored (Ram ırez et al., 2011). In recent years, MYB46 was found to be activated under stress (Geng et al., 2018;Shafi et al., 2019), and it was put forward that BplMYB46 might have the function of regulating the expression of genes involved in abiotic stress responses of Betula platyphylla (Guo et al., 2016). However, the genes and metabolic pathways that could be specifically regulated by MYB46 under stress remain to be further explored and determined.
Previous studies showed that several MYB proteins, such as AtMYB2, AtMYB14, AtMYB15 and AtMYBS3, could affect the stress tolerance of plants by regulating the expression of stressresponsive genes (Abe et al., 2003;Agarwal et al., 2006a;Chen et al., 2013;Su et al., 2010). Osmotic stress can induce a series of signal transduction and response reactions in plants (Zhu, 2002(Zhu, , 2016, some of which are controlled by abscisic acid (ABA) but some are not, in other words, both ABA-dependent and ABAindependent regulatory systems are involved in stress-responsive process (Yamaguchi-Shinozaki and Shinozaki, 2005). The cisacting element, ABA-responsive element (ABRE) and a group of TFs, ABRE-binding protein/ABRE-binding factors (AREB/ABFs) have pivotal functions in regulating ABA-dependent gene expression, while the cis-element, dehydration-responsive element/Crepeat (DRE/CRT) and DRE-/CRT-binding protein 2 (DREB2) TFs play key roles in regulating ABA-independent gene expression in response to osmotic stress (Abe et al., 2003;Agarwal et al., 2006a,b;Lata and Prasad, 2011;Thomashow, 2010;Yoshida et al., 2014). Using MYB46 as the entry point, we carried out research on the regulation network of secondary cell wall formation and the signal transduction mechanism to salt and osmotic stress in apple (Malus9domestica Borkh.). And we found that MdMYB46 can activate not only the expression of genes involved in the biosynthesis of secondary cell wall, but also ABAdependent and independent stress signals. In the aspect of improving plant stress tolerance, the biosynthesis of secondary cell wall and stress signal transduction pathways of plants were coordinated by MdMYB46.

MdMYB46 protein has the R2 and R3 functional regions
The MYB transcription factor gene MdMYB46 [numbered XP_008363629 in National Center for Biotechnology Information or MD03G1176000 in the Apple Genome (GDDH13 V1.1) database] was isolated from the wild GL-3 apple (M. 9domestica Borkh). Multiple sequence alignments and phylogenetic analysis indicated that MdMYB46 was an R2R3-type protein and in the same subfamily with MtMYB46, GmMYB46A, GmMYB46B and VvMYB46A among those AtMYB46 orthologs ( Figure S1).

MdMYB46 with transcriptional activation activity can be activated by stress signals
To elucidate the localization pattern of MdMYB46 in cells, the CDS region (without stop codon) was inserted into the pRI-eGFP vector to construct a MdMYB46-eGFP fusion gene, which was then transiently expressed by infiltration in tobacco leaves. As shown in Figure 1a, the localization results showed that MdMYB46 localized to the nucleus. Using the yeast two-hybrid system, the transcriptional activity of MdMYB46 was validated and the specific activity region was analysed. It was found that MdMYB46 has transcriptional activation activity at the C-terminal position of 141-374 amino acids (Figure 1b,c).
In order to clarify the expression patterns of MdMYB46 under abiotic stress, non-transgenic GL-3 apple plants were treated with exogenous ABA, NaCl and mannitol. QRT-PCR results showed that the expression level of MdMYB46 was up-regulated under these stresses (Figure 1d).

MdMYB46 positively regulates salt and osmotic stress tolerance in apple
In order to verify the function of MdMYB46 in apple, we constructed the overexpression and RNAi vectors of MdMYB46 with the CaMV 35S fragment in the promoter ( Figure S2A). Four MdMYB46-overexpressing transgenic lines (OE-MdMYB46-3, OE-MdMYB46-9, OE-MdMYB46-14 and OE-MdMYB46-25) and four MdMYB46-RNAi lines  were obtained with Agrobacterium-mediated transformation. Compared with the non-transgenic apple plants, the expression levels of MdMYB46 were significantly increased in the four MdMYB46overexpressing lines ( Figure S2B), while those in four MdMYB46-RNAi lines were significantly decreased ( Figure S2B).
The transgenic apple lines of OE-MdMYB46-3, OE-MdMYB46-9, RNAi-MdMYB46-7, RNAi-MdMYB46-8 and the non-transgenic apple plants (GL-3) were chosen for long-time stress treatment. It was found that MdMYB46-overexpressing lines were more resistant to salt and osmotic stress than the non-transgenic plants, while MdMYB46-RNAi lines were less resistant ( Figure 2a) (Figure 2d). The relative water content of the leaves can be used as an indicator of plant water loss. This study also examined the relative water content in leaves of transgenic and non-transgenic apple plants under salt and osmotic stress (Figure 2e). The results showed that under salt and osmotic stress, the water loss in MdMYB46-overexpressing apple leaves was the least, while that in the RNAi plants was the highest. Proline content could reflect the stress resistance of plants to a certain extent. With the highly hydrophilic, proline can stabilize the metabolic processes in cells and tissues, thus reducing the freezing point and effectively preventing cell dehydration. Plants with strong osmotic stress resistance often accumulate more proline. In this study, under the stress of salt and osmotic, the content of proline in the leaves of MdMYB46-overexpressing transgenic apple plants was higher than that in non-transgenic plants, while that in the RNAi plants was the lowest (Figure 2f).
Under adverse conditions, plant membrane lipid peroxidation often occurs. MDA is one of its products. It is usually used as an indicator of lipid peroxidation, indicating the degree of cell membrane lipid peroxidation and the strength of plant response to stress conditions. In this study, under the stress of salt and osmotic, the content of MDA in the leaves of MdMYB46overexpressing transgenic apple plants was lower than that of non-transgenic plants, while that in the RNAi plants was the highest (Figure 2g).
The content of lignin in the transgenic plants showed significant changes (Figure 2h). The content of lignin in the MdMYB46overexpressing plants was significantly higher than that in nontransgenic and MdMYB46-RNAi apple plants, indicated that MdMYB46 could significantly regulate the accumulation of lignin. Under osmotic stress, the lignin content in apple plants increased, but the increasing amount was less in non-transgenic and MdMYB46-RNAi plants than that in the overexpressing plants.
The above experimental results indicated that MdMYB46overexpressing apple plants had higher salt and osmotic stress tolerance than the non-transgenic plants, while MdMYB46-RNAi transgenic apples had weaker salt and osmotic stress resistance.

The transcript levels of secondary cell wall biosynthesisrelated genes in apple and arabidopsis are induced by MdMYB46
The results of stress test showed that MdMYB46-overexpressing apple plants were more resistant to osmotic stress. Previous studies have shown that MYB46 could enhance the resistance of plants to abiotic stress by regulating the formation of secondary walls of plant cells (Guo et al., 2016;Zhong et al., 2007). Therefore, we explored the function of MdMYB46 in the regulation of secondary cell wall, in which the transgenic Arabidopsis and transgenic apples were used.
With overexpressing MdMYB46 ( Figure S4), Arabidopsis plants showed obviously change in phenotype. Compared with the wild type, MdMYB46-overexpressing Arabidopsis plants not only showed short stature, but also showed the characteristics of blade curling ( Figure S3A) and floral organ development abnormalities ( Figure S3B), which may be caused by the ectopic lignin deposition into epidermal or mesophyll cells (Zhong et al., 2007). However, it was so different in MdMYB46-overexpressing apple plants, of which the size was similar to the non-transgenic plants ( Figure S3C).
To further investigate the regulation network of MdMYB46, the transcriptional levels of MYB transcription factors MYB58, MYB63 and structural genes related to lignin and cellulose biosynthesis in transgenic plants were examined. The expression levels of MdMYB58, MdMYB63, lignin biosynthesis genes (MdC4H, MdC3H, MdCAD, MdF5H, MdHCT, Md4CL, MdCOMT and MdCCR) and cellulose biosynthesis genes (MdCESA4 and MdCESA8) were significantly up-regulated and down-regulated in the MdMYB46-overexpressing and -RNAi lines, respectively (Figures 3 and S5A). In MdMYB46-overexpressing lines of Arabidopsis, the expression levels of AtMYB63, AtMYB58, At4CL1, AtCCoAOMT, AtCESA4, AtCESA7 and AtCESA8 were significantly higher than those in the wild type (Figures S4 and S5B-D).
The above results indicated that MdMYB46 positively regulated the transcriptional levels of secondary cell wall biosynthesisrelated genes and lignin deposition in apple and Arabidopsis.

MdMYB46 binds to M46RE and SMRE motifs in promoters of lignin biosynthesis-related genes
Lignin is the phenolic component of woody cell walls and plays the main role in providing compression strength to cell walls. In order to clarify the pattern of MdMYB46 in promoting the deposition of lignin in the secondary cell walls, we analysed the promoter region of lignin biosynthesis-related genes and specially validated the binding sites for genes involved in osmotic stress.
MYB58 and MYB63 are transcriptional activators that specifically promote lignin biosynthesis during Arabidopsis secondary wall formation and are also direct target genes for MYB46 (Demura and Ye, 2010;Wang and Dixon, 2012;Zhou et al., 2009). CAD, COMT and CCR are the key genes in the biosynthesis pathway of lignin monomers, which are significantly positively correlated with the drought/osmotic stress resistance of plants (Fan et al., 2006;Hu et al., 2009). So, the study mainly focused on the in-depth study of the regulation of MYB46 on the expression of these genes.
In order to detect the specificity of MYB46 binding to lignin regulatory genes and biosynthetic gene promoters, biotinylated probes of M46RE and SMRE elements were designed according to the elements in the red box of Figure 4a, and incubated with GST-MdMYB46 fusion protein. The interaction between the labelled probes and the GST-MdMYB46 fusion protein was tested by EMSA. The results showed that GST-MdMYB46 was able to interact with the active elements in the promoters of Meanwhile, the fragments containing the SMRE site and the M46RE site (about 200-500 bp) in the promoters of MdMYB58, MdMYB63, MdCAD, MdCOMT and MdCCR genes were inserted into the vector containing the luciferase reporter gene (these fragments shown by the black arrow in Figure 4c). p35S: MdMYB46 was used as an effector to inject with pRI-LUC, pMdMYB58-LUC, pMdMYB63-LUC, pMdCAD-LUC, pMdCOMT-LUC and pMdCCR-LUC, respectively. After 48 h, fluorescence was observed and the luciferase activity was detected in tobacco leaves. These results indicated that MdMYB46 could activate the transcription of the lignin-related genes (Figure 4d).
To further determine the effect of MdMYB46 on lignin-related genes, ChIP-qPCR primers were also designed in this experiment (shown by the black arrow in Figure 4a). The binding ability of MdMYB46 to the promoter of lignin biosynthesis genes was detected in vivo by transgenic pRI-MYC-MdMYB46 'Ourin' apple callus. The result of ChIP-qPCR is shown in Figure 4e and indicated that MYC-MdMYB46 could bind to the promoters of MdMYB58, MdMYB63, MdCAD, MdCOMT and MdCCR genes and directly participate in the regulation of lignin synthesis.
The expression of some stress-responsive genes involved in ABA-dependent or ABA-independent pathways is induced by MdMYB46 in apple In order to clarify the role of MdMYB46 in the regulation network of stress tolerance in apple, non-transgenic apple plants were treated with exogenous ABA, NaCl and mannitol, and the previously reported drought and salt stress related genes AREB1 (Yoshida et al., 2014(Yoshida et al., , 2015, RD22 (Abe et al., 2003) in ABAdependent pathway and DREB (Agarwal et al., 2006b;Lata and Prasad, 2011;Thomashow, 2010), RD29 (Yamaguchi-Shinozaki and Shinozaki, 1994) in the ABA-independent pathway were chosen for our study. The amino sequences in the Apple Genome (GDDH13 V1.1) protein database with highest similarity to AtAREB1A, AtAREB1B, AtDREB2A, AtDREB2B, AtRD29A, AtRD29B and AtRD22 in Arabidopsis were identified and named MdAREB1A, MdAREB1B, MdDREB2A, MdDREB2B, MdRD29A, MdRD29B and MdRD22 accordingly. In apple, some studies have also explored the involvement of these genes in stress signal transduction (Table S4). From Figure S6, it could be found that the expression level of these genes was significantly increased under the stress of exogenous ABA, NaCl and mannitol, except for MdDREB2B.
To investigate the effects of MdMYB46 on the expression of MdRD22, MdRD29A, MdRD29B, MdAREB1A, MdAREB1B, MdDREB2A and MdDREB2B, the transcriptional levels of these genes were studied in the MdMYB46-overexpressing and MdMYB46-RNAi plants without stress treatment (the control in Figure 5). The results showed that the expression level of MdRD22, MdRD29A, MdAREB1A, MdDREB2A and MdDREB2B genes was positively correlated with the expression level of MYB46 in apple plants, while the expression levels of MdRD29B and MdAREB1B changed insignificantly in the overexpressing and RNAi plants, respectively.
The transcriptional levels of stress-responsive genes in transgenic apple plants under salt and osmotic stress were also studied. After short-time stress treatment, the transcription levels of MdRD22, MdRD29A, MdAREB1A and MdDREB2A showed similar trends with MdMYB46 ( Figure 5). Compared with the control, the expression of the four genes in MdMYB46-overexpressing and MdMYB46-RNAi plants increased after stress treatment, but the expression levels of these genes in RNAi plants were lower than those in the non-transgenic and MdMYB46overexpressing plants, indicating that MdMYB46 can regulate the expression of these genes. After stress treatment, the expression levels of MdRD29B and MdAREB1B increased, but there was no significant difference in the expression level of these genes in MdMYB46-overexpressing and MdMYB46-RNAi plants, indicating that there are still other regulatory factors in apple involved in the ABA-dependent and ABA-independent pathway for stress responding.
MdMYB46 activates the expression of stress signalling genes by binding to SMRE,M46RE and MYBCORE motifs present in their promoters With the exclusion of genes that were not significantly associated with stress induction or MdMYB46 expression, the promoter sequences (2000 bp) of MdRD22, MdRD29A, MdAREB1A and MdDREB2A were analysed. As shown in Figure 6a, there were five SMRE sites, one M46RE site and one MYBCORE site in the promoter of MdAREB1A; one SMRE site and one M46RE site in the promoter of MdDREB2A; three SMRE sites, one M46RE site and one MYBCORE site in the promoter of MdRD22; and one SMRE site and one M46RE site in the promoter of MdRD29A, which indicated that they all had the potential to be activated by MdMYB46. In addition, different from the promoter sequences of AtRD22 and AtRD29A Shinozaki, 1993, 1994), two DRE sites were found in the promoter of MdRD22, but not in MdRD29A, and two ABRE sites were found both in the promoter of MdRD22 and MdRD29A which indicated that a different regulatory pathway of these genes may be existed in apple.
To explore the interaction of MdMYB46 with the MYB binding site in the promoters of MdRD22, MdRD29A, MdAREB1A and MdDREB2A, ChIP-qPCR primers were designed in the black arrow of Figure 6a. The binding signal of SMRE, M46RE and MYBCORE sequence in each gene promoter was detected in transgenic callus containing pRI101-MYC empty vector or pRI-MYC-MdMYB46 vector and showed that MdMYB46 bound to these sites with varying degrees (Figure 6b).
Based on the results of ChIP-qPCR, the SMRE and M46RE sequences (Figure 6c) in the MdAREB1A-P2, MdAREB2A-P1, MdRD22-P4 and MdRD29A-P1 fragments with the strongest binding ability were selected to design biotin labelling probes. These probes were then incubated with the GST-MdMYB46 protein. A 200-fold unlabelled probe was used as a competition assay, and a protein gel migration assay was used to detect the binding between MdMYB46 and these promoters. Figure 6d indicates that the MdMYB46 protein could be directly bound to the promoters of MdRD22, MdRD29A, MdAREB1A and MdDREB2A genes.
To further demonstrate the interaction between MdMYB46 and the promoters of these genes, the fragments of MdAREB1A-P2, MdAREB2A-P1, MdRD22-P4, MdRD29A-P1 and the corresponding mutation-binding site sequences were inserted into the luciferase-containing reporter gene vector (Figure 6e), which was then injected into the tobacco leaves together with the effector p35S:MdMYB46 and pRI101AN empty vectors, respectively. The fluorescence distribution and the luciferase activity were detected 48 h later (Figure 6f-g). The results showed that MdMYB46 could directly bind to the promoters of MdRD22, MdRD29A, MdAREB1A and MdDREB2A.

MdMYB46 regulates the biosynthesis of secondary cell wall and deposition of lignin in apple
With the evolution of plants and lignin biosynthesis pathways, complex and sophisticated lignin synthesis regulatory networks have been formed in plants. NST1, SND1, NST2, VND6 and VND7 are the main upstream regulators of lignin biosynthesis pathway, while MYB transcription factors (MYB46, MYB83 and MYB55) and NAC transcription factors (SND3 and XND1) (Yamaguchi et al., 2011;Zhong et al., 2010) belong to the Tier 2 regulators in the regulatory network (Kumar et al., 2015). In Arabidopsis, the transcription factor MYB46 showed to be a key transcriptional switch controlling Arabidopsis secondary cell wall biosynthesis (Ko et al., 2009;Zhong et al., 2007), which can act directly on the M46RE cis-regulatory element (

[A/G][G/T] T[A/T]GGT[A/G]) or SMRE element (secondary wall MYB-responsive element, ACC[A/T]A[A/C][T/C])
, to activate the expression of related genes and regulate lignin synthesis (Kim et al., 2012;Ko et al., 2009;Zhong and Ye, 2015). In this study, multiple MYB46 binding sites were present in the promoters of the regulatory (MdMYB58 and MdMYB63) and biosynthetic (MdC4H, MdC3H, MdCAD, MdF5H, MdHCT, Md4CL, MdCOMT and MdCCR) genes for apple lignin metabolism (Figure 3), which was consistent with previous findings in Arabidopsis (Kim et al., 2012(Kim et al., , 2014Zhong et al., 2007). The EMSA, luciferase reporter assays and ChIP further indicated that MdMYB46 could directly bind to the promoters of MdMYB58, MdMYB63, MdCAD, MdCOMT and MdCCR to activate the expression of these genes.
In our transgenic Arabidopsis, the overexpression of MdMYB46 caused a dwarf plant phenotype ( Figure S3A  MdMYB46 enhances the salt and osmotic stress tolerance of apple plants by directly activating ABAdependent and ABA-independent stress signals Extreme drought and high salt stress usually produce high osmotic stress causing dehydration and even death of plant cells, which significantly affect plant growth and development, leading to severe crop yield loss (Agarwal et al., 2012;Chaves et al., 2009). The salt and osmotic stress tolerance of plants are the results of interaction between genotype and environment. Lignin deposition is related to osmotic stress and salt tolerance of plants and many studies have shown that lignin biosynthesis is enhanced under osmotic stress. Lignin can reduce plant cell wall water leaching and transpiration, which helps maintain cell osmotic balance and protects membrane integrity (Monties and Fukushima, 2005). In addition to the accumulation of lignin, salt and osmotic stress also trigger a series of stress-responsive pathways in plants. AREB/ABFs are a family of ABA-dependent bZIP TFs that interact with the ABRE (PyACGTGG/TC) motif in promoter of many ABA-responsive genes in many plant species (Kang et al., 2002;Yoshida et al., 2010Yoshida et al., , 2014. Among the nine AREB/ABF genes identified in Arabidopsis, AREB1/ABF2, AREB2/ABF4 and ABF3 can be induced by osmotic stress. DREB proteins, belonging to AP2/ERF TF family, are indeed implicated in plant responses to drought/osmotic stress tolerance (Lata and Prasad, 2011;Thomashow, 2010), of which DREB1A and DREB2A specifically interact with cis-acting dehydration-responsive element (C-repeat) involved in drought stress-responsive gene expression in A. thaliana (Sakuma et al., 2006). Several stress-responsive genes, such as RD29A induced through ABA-independent pathway and RD22 mediated by ABA, can be activated by dehydration, high salinity and cold stress Shinozaki, 1993, 1994). In apple, some studies have explored the involvement of these genes in stress signal transduction (Table S4). Sixty-eight MdDREB genes have been classified into six subgroups against the entire genome of apple, of which MdDREB2A and MdDREB2B (called MDP0000147009 and MDP0000153866 in their study) were identified and found they could be up-regulated under various abiotic stress treatments (Zhao et al., 2012). MdAREB1A and MdAREB1B (called MdAREB3.2 and MdAREB2 in their study) were identified and studied for their sensitivity to ABA in apple (Ma et al., 2017). MdRD22, MdRD29A and MdRD29B were also found to be responsive for drought or osmotic stress Shao et al., 2019). These results indicated that these genes in apple were also active in stress signal transduction. The MdMYB46-overexpressing plants had higher salt and osmotic stress tolerance with lower levels of reactive oxygen species and membrane lipid peroxidation products such as MDA, and higher relative water content and proline content in leaves compared with the non-transgenic apple plants, while the MdMYB46-RNAi plants had lower resistance with higher level of active oxygen, lipid peroxidation and water loss. These results indicated that in addition to promoting lignin deposition, MdMYB46 might also increase the strength of other stressresponsive pathways to enhance the stress tolerance of apple plants.
MYB proteins are involved in ABA and stress responses as well as many other cellular processes (Chen et al., 2006). In Arabidopsis, AtMYB52 was regulated by AtMYB46 directly during secondary wall synthesis, and overexpression of AtMYB52 could improve plant drought tolerance by regulating ABA signalling pathway in which AtRD29B and AtNCED3 may be involved (Ko et al., 2009;Park et al., 2011). AtMYB2 protein also functioned as a transcriptional activator in ABA-regulated gene expression under drought and salt stress, and ABA-induced gene expression of AtRD22 and AtADH1 was enhanced in AtMYB2 transgenic plants (Abe et al., 2003). AtMYB14, AtMYB15 and AtMYBS3 participated in freezing tolerance in Arabidopsis by repressing the expression of CBF (C-REPEAT BINDING FACTOR) genes (Agarwal et al., 2006a;Chen et al., 2013;Su et al., 2010). Overexpressing of TaMYB33 (Triticum aestivum) in Arabidopsis led to the downregulation of AtABF3 and AtABI1, and the up-regulation of AtAAO3 (Qin et al., 2012). In apple, MdMYB88 and MdMYB124 positively regulated the expressions of cold hardiness and coldresponsive genes under cold stress by CBF-dependent and CBFindependent pathways (Xie et al., 2017). In this study, the transcription levels of MdRD22, MdRD29A, MdAREB1A and MdDREB2A were positively correlated with that of MdMYB46 in the wild and transgenic apple under salt and osmotic stress. EMSA, luciferase reporter assays and ChIP further confirmed that MdMYB46 could bind to the SMRE, M46RE and MYBCORE motifs in the promoter of the four genes and activate their expression.
ABA-dependent and ABA-independent signalling pathways play signal transduction roles during plant adaptation to osmotic stress (Vishwakarma et al., 2017;Yoshida et al., 2014;Zhu, 2016), which are closely related to the active oxygen scavenging system and the production of antioxidants in plants (Krasensky and Jonak, 2012;Zhao et al., 2018). So, the activation of MdABRE1A, MdDREB2A, MdRD22 and MdRD29A might be beneficial to stimulate the anti-stress system in the whole plant, improve the level of stress-resistant components (such as proline) and enhance the stress tolerance in apple. In addition to regulating the formation of secondary cell wall, MdMYB46 could also enhance the salt and osmotic stress tolerance of apple plants by directly activating MdABRE1A in the ABA-dependent pathway, MdDREB2A in the ABA-independent pathway and dehydrationresponsive genes MdRD22 and MdRD29A (Figure 7). Jeong et al. (2018) found that the expression level of AtSND1 in Arabidopsis was significantly increased under stress treatment conditions, such as mannitol and NaCl. SND1 could coordinate the relationship between plant growth and ABA metabolism by regulating the expression of ABI4 and MYB46. Under the stress simulated by mannitol and NaCl, the increase of MdMYB46 expression in apple may also be activated by MdSND1. MdSND1, the key upstream transcription factor of MdMYB46 involved in lignin biosynthesis, was identified in apple and its expression was also induced by ABA, salt and mannitol. The stress treatment experiment of transgenic apple lines indicated that MdSND1 could also enhance the salt and osmotic stress tolerance in apple (unpublished date). In further studies, the transcriptional regulation mechanism of MdSND1 on MdMYB46 and stress response pathways under salt and osmotic stress conditions will be a potential direction to analyse the stress response mechanism in apple. In addition, drought stress could induce the expression of MdMYB88 and MdMYB124 in apple, which can activate the transcription of MdMYB46 in direct or indirect ways (Geng et al., 2018). A similar regulation mechanism may happen under our mannitol and NaCl treatment.
In summary, MdMYB46 can enhance the salt and osmotic stress tolerance of apple not only by activating secondary cell wall biosynthesis pathways, but also by directly activating stressresponsive signals.

Plant materials
The tissue culture plants of apple (Malus9domestica Borkh.) used for transformation and resistance treatment were named GL-3 with high efficiency for transformation (Dai et al., 2013), which were cultured in the subculture medium (MS medium supplemented with 0.3 mg/L 6-BA, 0.2 mg/L IAA, 0.1 mg/L GA 3 ) and managed under long-day conditions (14 h : 10 h, light : dark) at 25°C. The callus of 'Ourin' apple for transformation was kept in our laboratory at 25°C in the dark, and tobacco used for subcellular localization and transient expression of genes was Nicotiana benthamiana. The Arabidopsis thaliana was managed under long-day photoperiod (16 h : 8 h, light : dark) at 23°C.

Genetic transformation
The MdMYB46 full-length coding sequence (CDS) region [the same to the sequence of MD03G1176000 in the Apple Genome (GDDH13 V1.1) database] amplified with MdMYB46-pRI primer pair (Table S1) was inserted into the pRI101-AN vector to construct the overexpression vector of pRI-MdMYB46 (Figure S2A). Forward and reverse fragments of MdMYB46 gene were amplified with the RNAi-MdMYB46 ZF/ZR and FF/FR primer pairs (Table S1), respectively, and inserted into the pRNAi-E vector (Song et al., 2017) to construct the MdMYB46-RNAi vector ( Figure S2A). The overexpression and RNAi vectors of MdMYB46 were introduced into the Agrobacterium tumefaciens strain EHA105. The genetic transformation of GL-3 was according to the method of Dai et al. (2013).
For callus transformation, the MYC tag was added before the MdMYB46 full-length CDS and inserted into the pRI101-AN vector, labelled as pRI-MYC-MdMYB46. The fusion vector was introduced into EHA105. The 'Ourin' callus, subcultured in MS medium supplemented with 1.5 mg/L 2,4-D and 0.4 mg/L 6-BA, was transferred into the bacterial solution, oscillating cultivated for 15 minutes, and transferred to fresh callus medium for preculture. After 2-4 days of dark culture at room temperature, the pre-cultured callus was transferred to callus screening medium (MS medium supplemented with 1.5 mg/L 2,4-D and 0.4 mg/L 6-BA, 250 mg/L cefotaxime and 30 mg/L kanamycin, pH = 5.8). After about 30 days of dark culture, the newly grown callus was transferred to a new screening medium and positive callus was screened for subsequent chromatin immunoprecipitation (ChIP).
For A. thaliana transformation, the overexpression vector of MdMYB46 was introduced into Agrobacterium strain GV3101. A. thaliana (Col-0) was transformed using the floral dip method. The seeds collected were screened in 1/2 MS medium supplemented with 30 mg/L kanamycin. Then, the seedlings were transplanted into the matrix for phenotypic observation and those with the similar changes were chosen for further qRT-PCR detection. The screened T 1 transgenic lines were used for further analysis.

Stress treatment
For the short-time stress treatment, the top young part of the GL-3 apple plants that has been subcultured for 20 days was cut and then transferred to a liquid subculture containing 50 lM ABA, 200 mM NaCl or 300 mM mannitol. After 0, 2, 4 and 6 h, young tissues were taken for RNA extraction.
For the long-time stress treatment, the young shoot tip of the 20-day-old apple tissue-cultured plants was cut and transferred to the subculture medium containing 200 mM NaCl or 300 mM mannitol. After 10 days of the treatment, the samples were photographed and their leaves were collected for further tests.

Measurement of lignin, proline, H 2 O 2 , MDA and relative water content
The stem tissue of 1-month-old apple tissue culture plants was used to detect lignin content according to the method of Wang et al. (2016). The leaf relative water content was calculated according to the method of Farooqi et al. (2000). The proline content was measured spectrophotometrically by the method of Bates et al. (1973). The H 2 O 2 level was measured as described by Velikova et al. (2000). Thiobarbituric acid (TBA) test was used to determine the malondialdehyde (MDA) content in leaf with the method described by Heath and Packer (1968).

NBT and DAB staining
The 3,3 0 -diaminobenzidine (DAB) and nitrotetrazolium blue chloride (NBT) staining were performed according to the method described by Kumar et al. (2014).

ChIP Assay
The positive callus transformed with pRI-MYC-MdMYB46 of 'Ourin' was screened. After the positive callus was grown for 20 days, the ChIP experiment was performed. The main steps were based on the instructions of EPIGENTEC's Plant ChIP kit (Farmingdale). The formaldehyde was firstly used to cross-link with the chromatin in the 'Ourin' callus overexpressing MYC-MdMYB46, and then, MYC immuno-antibody was used to precipitate and enrich the DNA fragment bound to MYC-MdMYB46. The immunoprecipitated DNA fragments were then used as templates in qRT-PCR for detection of MdMYB58, MdMYB63, MdCAD, MdCOMT, MdCCR, MdAREB1A, MdDREB2A, MdRD22 and MdRD29A promoter sequences by ABI 7500 real-time PCR instrument. We used the MdACTIN as the internal reference gene, and the 2 ÀMMCt method (Livak and Schmittgen, 2001) was used to calculate the fold enrichment of each promoter fragment with ABI 7500 software v2.0.6. All primers are listed in Table S1.  Table S2. The number of each gene is listed in Table S3.

EMSA
The full-length CDS region of MdMYB46 was inserted into the prokaryotic expression vector pGEX-5X-1 and labelled as pGEX-MdMYB46. Then, the constructed vector was separately introduced into the competent cells of Escherichia coli BL21 to induce protein production. The GST protein was purified according to the instructions of the Thermo GST Fusion Protein Purification Kit (Massachusetts). The EMSA probe was prepared according to the Biyuntian Biotin Probe Labeling Kit (Shanghai, China), and the fusion between protein and labelled probes was made with reference to the EMSA kit (Thermo, Rockford, IL, USA). The nondenaturing polyacrylamide gel was prepared, and the test was completed by spotting, electrophoresis detection, film transfer, film washing and development.

Analysis of transcriptional activity
The constructed pRI-MdMYB46 vector was used as an effector. The clipped region (about 200-500 bp fragment containing the corresponding binding site) of MdMYB58, MdMYB63, MdC4H, MdC3H, MdCAD, MdF5H, MdAREB1A, MdDREB2A, MdRD22 or MdRD29A was inserted into the vector of pRI-mini35S-LUC and used as a reporter. The conserved bases of the binding sites in the MdAREB1A, MdDREB2A, MdRD22 and MdRD29A promoter fragments were mutated by PCR and ligated into the vector of pRI-mini35S-LUC as a control.
The above-constructed vector was transferred into the leaves of 1-month-old N. benthamiana for transient expression. The infected tobacco was cultured for 24 h in the dark and then transferred to a light incubator (25°C, 16 h : 8 h, light : dark). After 48 h, the tobacco was photographed by a living fluorescence imager (Lb985, Berthold, Germany). The luciferase activity of these leaves was measured using a Transgen Biotech firefly luciferase assay kit (Beijing, China).

RNA extraction and qRT-PCR
Total RNA was extracted from leaves using the modified CTAB method (Chang et al., 2007), and cDNA was obtained with reference to the Takara Bio PrimeScript TM RT Reagent Kit (Dalian, China). The expression level of these genes was detected by ABI 7500 real-time PCR instrument. The reaction system was in a total volume of 10 lL, containing 0.5 lL of cDNA, 5.0 lL of TaKaRa SYBR Premix Ex Taq, 0.5 lL of upstream and downstream primers for the relevant detection gene (with each primer at 0.2 lM), 0.2 lL 50 9 ROX Reference Dye II, and 3.3 lL ddH2O. The amplification programme was as follows: one cycle of 10 min at 95°C followed by 40 cycles of 10 s at 95°C and 30 s at 60°C; and the dissolution curve was plotted at 95°C for 15 s, 60°C for 1 min, 95°C for 30 s and 60°C for 15 s. All reactions were run in triplicate, and average values were calculated. MdActin and At18S were used as the internal reference for apple and Arabidopsis, respectively, and the 2 ÀMMCt method (Livak and Schmittgen, 2001) was used to calculate the relative expression level of each gene with ABI 7500 software v2.0.6.

Subcellular localization
The MdMYB46 full-length CDS without the stop codon was amplified with MdMYB46-eGFP primer pair (Table S1) and inserted into the pRI-eGFP vector to constitute MdMYB46-eGFP fusion expression vector. The fusion vector was introduced into EHA105 and transiently expressed in leaves of N. benthamiana (1-month-old) by infiltration method (Yang et al., 2000). The infected tobacco was cultured for 24 h in the dark and then transferred to a light incubator (25°C, 16 h : 8 h, light : dark). After 48 h, the tobacco leaves were photographed by a confocal laser-scanning microscope (TCS SP8-SE; Leica, Wetzlar, Germany).

Analysis of transcriptional activity of MdMYB46
The full-length CDS of MdMYB46 and their N-terminal DNAbinding region and C-terminus were inserted into pGBT9 vector containing the DNA-binding region of GAL4, and were labelled as pGBT9-MdMYB46, pGBT9-MdMYB46-N and pGBT9-MdMYB46- C, respectively. The vector plasmid was transferred to the Y2H strain and cultured according to the manufacturer's instructions (Clontech). The yeast on the selective medium lacking Trp (-T) solid medium plate was activated onto the medium lacking Trp, His and adenine (-T/-H/-A) + X-alpha-gal (5-Bromo-4-chloro-3indolyl-a-D-galactoside) solid medium plate. If the yeast grew and turned blue, it was indicated that the gene had transcriptional activation activity.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article.

Figure S1
Multiple sequence alignments and phylogenetic analysis of MdMYB46 orthologs. Figure S2 Transcriptional levels of MdMYB46 in transgenic apple. Figure S3 Phenotypes of MdMYB46-overexpressing Arabidopsis and apple.  Figure S4 Differences in transcriptional levels of lignin biosynthesis-related genes between MdMYB46-overexpressing Arabidopsis plants and wild type. Figure S5 Changes in expression of genes related to cellulose biosynthesis in MdMYB46-overexpressing Arabidopsis and apple. Figure S6 Transcriptional levels of stress signaling genes under ABA, salt and drought stress and in MdMYB46 transgenic apples.