MdHIR proteins repress anthocyanin accumulation by interacting with the MdJAZ2 protein to inhibit its degradation in apples

In higher plants, jasmonate ZIM-domain (JAZ) proteins negatively regulate the biosynthesis of anthocyanins by interacting with bHLH transcription factors. However, it is largely unknown if and how other regulators are involved in this process. In this study, the apple MdJAZ2 protein was characterized in regards to its function in the negative regulation of anthocyanin accumulation and peel coloration. MdJAZ2 was used as a bait to screen a cDNA library using the yeast two-hybrid method. The hypersensitive induced reaction (HIR) proteins, MdHIR2 and MdHIR4, were obtained from this yeast two-hybrid. The ZIM domain of MdJAZ2 and the PHB domain of the MdHIR proteins are necessary for their interactions. The interactions were further verified using an in vitro pull-down assay. Subsequently, immunoblotting assays demonstrated that MdHIR4 enhanced the stability of the MdJAZ2-GUS protein. Finally, a viral vector-based transformation method showed that MdHIR4 inhibited anthocyanin accumulation and fruit coloration in apple by modulating the expression of genes associated with anthocyanin biosynthesis.


Results
MdJAZ2 inhibits anthocyanin accumulation in apple peel. To examine if MdJAZ2 influences anthocyanin accumulation and peel coloration in apple fruit, a viral vector-based transient transformation method was conducted to enhance the expression level of MdJAZ2 in the apple peel 20 . The viral overexpression vector pIR-MdJAZ2, plus a helper plasmid, IL-60-BS, were injected into the fruit peel of cultivar 'Red delicious, ' while the empty vector pIR, plus IL-60-BS, served as the control. The expression analysis demonstrated that pIR-MdJAZ2 injection noticeably enhanced the expression of MdJAZ2 gene in apple fruit peel than the pIR control (Fig. 1A). Subsequently, anthocyanin content was measured in the fruit peel around the injection sites. Anthocyanin content in the apple peel injected with pIR-MdJAZ2 was much lower than in the apple peel injected with the empty control (Fig. 1B). As a result, the injection of pIR-MdJAZ2 resulted in a loss of red colouration in the apple skin, compared with the empty control (Fig. 1C).
We previously reported that MdJAZ2 interacts with MdbHLH3 and inhibits the expression of MdMYB1, MdMYB9, MdMYB11 and their downstream anthocyanin structural target genes 15 . Therefore, the transcript levels of these genes were examined with real-time quantitative RT-PCR in the apple peel around the injection sites. The results indicated that MdJAZ2 transient overexpression slightly influenced the expression of MdbHLH3, but significantly repressed the transcript levels of MdMYB1, MdMYB9 and MdMYB11 (Fig. 1D). Consequently, the reduced expression of these genes repressed the expression levels of the anthocyanin structural genes to different degrees. Affected genes included MdCHS, MdCHI, MdANR, MdDFR, MdUFGT, MdF3H, MdANS and MdFLS (Fig. 1D).
Therefore, MdJAZ2 functions as a negative regulator of anthocyanin accumulation and fruit coloration by repressing the expression of the anthocyanin regulatory and structural genes in apple.
Yeast two-hybrid (Y2H) screening of an apple peel cDNA library reveals a putative MdJAZ2-interacting protein, MdHIR2. Y2H screening was conducted to screen MdJAZ2-interacting proteins. As a result, from a library with a primary library titer of 8.16 × 10 6 cfu/mL of which 95% of the clones had inserts, five positive colonies were obtained. Among the positive colonies a target cDNA was found that corresponds to the gene MDP0000653461.This gene encodes a putative protein that is similar to the hypersensitive induced reaction (HIR) proteins in Arabidopsis. There are a total of four HIR genes in Arabidopsis, AtHIR1, The anthocyanin content of the fruit peel around the injection sites. Anthocyanins were extracted from 8 pieces fruit skin, 1 cm 2 in size for each from one individual fruit. (C) The phenotype of the peel surrounding the injection sites. The apple fruit were injected with a viral based overexpression vector, pIR-MdJAZ2, while the empty vector, pIR, was used as the control. The injected apples were kept in an illumination incubator under white light at 17 °C for 4 days. (D) Quantitative real-time PCR analysis of the regulatory and structural genes associated with anthocyanin biosynthesis in the fruit peel around the injection sites. The data are shown as the mean ± SE, which were calculated based on 3 replicates. Mean differences in the bars are significant at P 0.05 level with different letters, not significant at P 0.05 level with the same letters. Repeat in following figure.
Scientific RepoRts | 7:44484 | DOI: 10.1038/srep44484 AtHIR2, AtHIR3 and AtHIR4. In addition to MDP0000653461, four more MdHIR genes were found in the apple genome, MDP0000295316, MDP0000630084, MDP0000138908 and MDP0000122340. Reverse Transcription PCR (RT-PCR) and sequence analysis demonstrated that all five apple MdHIR genes transcribed cDNAs that encode putative MdHIR proteins (Fig. S1). The predicted MdHIR proteins were used for sequence alignment analysis with AtHIRs. The sequence alignment analysis showed that the five predicted MdHIR proteins are highly similar in amino acid sequence to the AtHIRs ( Fig. 2A).

MdHIR2 and MdHIR4 interact with MdJAZ proteins.
To verify the interaction between the MdHIRs and the MdJAZs, yeast two-hybrid assays were carried out. The full-length cDNA of each of the MdHIRs was inserted into the vector pGBT9 (BD-MdHIRs) as bait. The full-length CDSs of MdJAZ1, MdJAZ2, MdJAZ3, MdJAZ4, MdJAZ5, MdJAZ6 and MdJAZ8 were cloned with RT-PCR, and then inserted into the vector pGAD424 (AD-MdJAZs) as prey. Subsequently, each combination of the BD-MdHIRs and the AD-MdJAZs was co-expressed in yeast cells. The transformants were cultured on -Trp/-Leu/-His/-Ade screening medium and stained with X-α -gal. Yeast strains containing BD-MdHIR2 plus AD-MdJAZ1, AD-MdJAZ2 or AD-MdJAZ4, and BD-MdHIR4 plus all seven of the AD-MdJAZs were positive for X-α -gal activity when grown on -Trp/-Leu/-His/-Ade screening medium. Yeast strains containing BD-MdHIRs or BD-MdHIRs plus the empty pGAD424 vector were negative for X-α -gal activity. Therefore, MdHIR2 interacted with MdJAZ1, MdJAZ2 and MdJAZ4, while MdHIR4 interacted with all seven of the MdJAZs tested (Fig. 3A). Interestingly, it was also found that MdHIR4 interacted with itself and four other MdHIR proteins (Fig. 3B).
To determine which region of the MdHIR proteins is necessary for interaction with the MdJAZ proteins, MdHIR2 and MdHIR4 were divided into the N-terminal PHB domain (MdHIR2N, MdHIR4N)    To further verify the interactions between MdHIR2, MdHIR4 and MdJAZ2 an in vitro pull-down assay was performed with MdHIR2-GST, MdHIR4-GST and MdJAZ2-HIS proteins that were expressed in and purified from the Escherichia coli strain BL21. The results showed that MdHIR2-GST and MdHIR4-GST, but not GST alone, interacted with the MdJAZ2-HIS protein (Fig. 3G,H).
Additionally, yeast two-hybrid and pull-down assays were also performed to examine the interactions between the four AtHIRs and the twelve AtJAZs in Arabidopsis. The yeast two-hybrid assays indicated that either AtHIR1 or AtHIR4 interacted with AtJAZ3, AtJAZ4 or AtJAZ9, while either AtHIR2 or AtHIR3 interacted with only AtJAZ3 (Fig. S2A). Subsequently, pull down assays confirmed the interaction between AtHIR1 and AtJAZ3 or AtJAZ9 (Fig. S2B,C).
Additionally, immunoblotting assays were conducted with an anti-GUS antibody to examine the abundance of MdJAZ2-GUS in the tested transgenic calli. The response of MdJAZ2-GUS to jasmonate signal was measured and indicated that MdJAZ2-GUS was degraded upon Me-JA treatment (Fig. 4C). The influence of MdHIR4 on the stability of MdJAZ2-GUS was examined. The results indicated that compared with the GUS protein in the 35S::GUS control calli, both the 35S::MdJAZ2-GUS and 35S::MdJAZ2-GUS+ 35S::MdHIR4 transgenic calli produced MdJAZ2-GUS, a fusion protein that has a higher molecular weight than GUS alone (Fig. 4D). In addition, there is a smear of proteins in the 35S::MdJAZ2-GUS calli, which suggests a degradation of the larger molecular weight MdJAZ2-GUS fusion protein (Fig. 4D). However, there was almost no degradation product obviously found in the 35S::MdJAZ2-GUS+ 35S::MdHIR4 transgenic calli (Fig. 4C,D). This indicates that the 35S::MdJAZ2-GUS+ 35S::MdHIR4 transgenic calli accumulated more MdJAZ2-GUS protein than the 35S::JAZ2-GUS transgenic calli, suggesting that MdHIR4 enhanced the stability of the MdJAZ2-GUS protein.

MdHIR4 represses the accumulation of anthocyanins in the apple peel. To examine if and how
MdHIR4 influences the accumulation of anthocyanins in the apple fruit peel, a TRV(Tobacco Rattle Virus)-based VIGS technique was used to transiently suppress the expression of MdHIR4 and MdJAZ2 in the peel of bagged fruit immediately following their detachment from the apple tree. The plasmids TRV-MdJAZ2, TRV-MdHIR4 or TRV-MdJAZ2+ TRV-MdHIR4 were injected into the fruit peel and treated the apple leaf. Quantitative real-time PCR analysis showed that the injection or treatment correspondingly suppressed the expression levels of the MdJAZ2 and MdHIR4 genes in the peel and leaf (Fig. 5A,B; Fig. S3A,B). However, the TRV-induced MdJAZ2 suppression did not influence the expression of MdHIR4, while MdHIR4 did not influence MdJAZ2 (Fig. 5A and B;  Fig. S3C).
As a result, the suppression of MdJAZ2 or MdHIR4 alone noticeably promoted anthocyanin accumulation and fruit/leaf coloration in the peel and leaf. A co-injection with TRV-MdJAZ2+ TRV-MdHIR4 resulted in the production of more anthocyanins and the redder color in the fruit peel and leaf than a single injection of either TRV-MdJAZ2 or TRV-MdHIR4 (Fig. 5C,D; Fig. S3C,D). Furthermore, the relative expression of the regulatory genes, MdMYB1, MdMYB9 and MdMYB11, as well as the structural genes, MdCHS, MdCHI, MdF3H, MdDFR, MdANS, MdUFGT, MdANR and MdFLS, were determined through quantitative real-time PCR of the fruit peel around the injection sites. The results indicated that the suppression of MdHIR4 and MdJAZ2 remarkably enhanced the expression levels of these ten regulatory and structural genes (Fig. 5E).
Additionally, a viral vector-based transformation method was carried out to enhance the expression of the MdHIR4 and MdJAZ2 genes. The plasmids pIR, pIR-MdHIR4 and pIR-MdJAZ2+ pIR-MdHIR4 were injected into the fruit peel. The results showed that MdHIR4 negatively regulated anthocyanin accumulation and peel coloration in the apple fruit (Fig. 6).

Discussion
Plants produce anthocyanins in almost all organs, such as fruit, flowers and leaves, in response to the ripening and maturation processes, and in response to biotic and abiotic stresses due to its antioxidant activity 37,38 . As is well known, jasmonic acid (JA) is an important phytohormone that induces anthocyanin biosynthesis via the JA signaling pathway. On the contrary, JAZ proteins interact with the WD40/bHLH/MYB complex and act as negative regulators for anthocyanin accumulation in various plant species 15,34 . In this study, we identified the MdJAZ-interacting proteins, the MdHIRs, through a yeast two-hybrid screen and then functionally characterized and identified them as negative regulators of anthocyanin biosynthesis in apple.
The HIR proteins ubiquitously exist in plant species. In Arabidopsis, there are four HIR genes, HIR1, HIR2, HIR3 and HIR4, in the genome. Similarly, in the apple genome there are five MdHIR genes, MdHIR1-1, MdHIR1-2, MdHIR2, MdHIR3 and MdHIR4 (Fig. 2). Additionally, the HIR genes have been cloned and identified in various plants, such as tobacco, maize, barley, rice, wheat, pepper and legumes. In these plant species, the HIR proteins contain highly conserved amino acid sequences, and possess the SPFH domain (also named the PHB domain) 39 . They belong to a protein superfamily that controls cell proliferation, ion channel regulation and cell death 40,41 .
In this study, it was found that the PHB domain of MdHIR2 and MdHIR4 interacted with the N-terminal ZIM domain of the MdJAZ proteins ( Fig. 3E and F). The ZIM (also named the TIFY) domain primarily mediates the homo-and heteromeric interactions between most of the JAZs in Arabidopsis and is involved in JA signaling output 42 . In addition, it also binds to Novel Interactor of JAZ (NINJA), which contains an EAR motif, and mediates the interaction between the JAZ proteins and TOPLESS to negatively regulate JA signaling 43 . Jasmonate does not affect the stability of NINJA, and NINJA overexpression does not affect JAZ3 stability 43 . The MdHIR proteins interacted with the ZIM domain of the MdJAZ proteins and inhibited their degradation in apple (Figs 3C-F and 4D).
In addition to the ZIM domain, the JAZ proteins also contain a C-terminal Jas motif that is required for the interactions between the JAZ proteins and COI1 and MYC2. The degradation of the JAZ proteins is induced by JA, and depends on the interaction between JAZ and the SCF COI1 protein complex 32 . Jasmonate induces anthocyanin accumulation by promoting the degradation of JAZ proteins in plants 15,34 . The MdHIR proteins inhibited the degradation of the MdJAZ proteins (Fig. 4C,D), thereby negatively regulating the biosynthesis of anthocyanins in apple. A model summarizing our findings regarding the regulatory pathway through which the MdHIR proteins inhibit the biosynthesis of anthocyanins is presented in Fig. 7. Additionally, the AtHIRs interacted with the AtJAZ proteins in Arabidopsis (Fig. S2), suggesting that this mechanism could work in other plant species.
In plants, the HIR proteins are involved in the hypersensitive response (HR). The HR is one of the various defense mechanisms mounted by plants in response to pathogen attack. This response is now almost universally accepted as a form of programmed cell death (PCD) characterized by the rapid death of plant cells at the site of pathogen infection. It generally causes localized cell death and results in necrotic lesions around infection sites in different plant organs. In barley and wheat, the Hv-HIR1, Hv-HIR2, Hv-HIR3 and Hv-HIR4 genes are induced by pathogens and are involved in the induction of the HR 44,45 . In Arabidopsis, the HIR proteins physically associate with the immune receptor RPS2 in plant immune responses 46 . In rice, a novel simple extracellular leucine-rich  repeat (eLRR) domain protein, OsLRR1, enters the endosomal pathway and interacts with OsHIR1 to participate in PCD 47 . In wheat, the hypersensitive-induced reaction genes, TaHIR1 and TaHIR3, play positive roles in resistance to the stripe rust fungus 48 . In pepper, CaLRR1 specifically binds to the plasma membrane (PM)-localized CaHIR1 to regulate PCD in leaves in response to an infection by Xanthomonas campestris pv. Vesicatoria 49 .
The HR is one of the most characteristic plant defenses against biotrophic pathogens. Salicylic acid (SA) plays a primary role in the activation of disease resistance mechanisms frequently associated with the HR 50 . The resistance (R) protein, which is a pathogen-encoded avirulence protein, can trigger the HR 51 . The R protein-mediated HR and SA-mediated basal resistance are generally considered effective against biotrophic pathogens, but ineffective against necrotrophic pathogens. Instead, plant resistance to necrotrophic pathogens is often mediated by jasmonic acid (JA) signaling, a process also involved in plant responses to wounding 52 .
During the HR, ROS production activates PCD 53 . Interestingly, anthocyanin works as an active scavenger of ROS in plant cells 54 . During PCD in lace plant (Aponogeton madagascariensis) leaves, the first visible change observed is the reduction of visible anthocyanin 55 , suggesting that anthocyanin may be involved in PCD. Therefore, the MdHIR-mediated inhibition of anthocyanin accumulation may be conducive to the occurrence of HIR and PCD.
Both isoflavonoid phytoalexin concomitants and anthocyanins are biosynthesized through the phenylpropanoid metabolic pathway. They belong to two different branches of this pathway, thereby being theoretically competitive with each other. For example, there in soybean is a strong bias towards decreasing the synthesis of anthocyanins and proanthocyanins, but increasing the synthesis of isoflavonoid phytoalexin concomitants during the resistance response 56 . It seems reasonable to suppose that the MdHIR-mediated inhibition of anthocyanin and proanthocyanin accumulation promotes the production of isoflavonoid phytoalexin concomitants, which may be conducive to PCD and HR. In addition, MdHIR proteins stabilize MdJAZ2 proteins (Fig. 4C,D), while JA promotes the degradation of the JAZ proteins. Therefore, JA should inhibit the function of the MdHIR proteins. As a result, JA seems to restrict cell death processes associated with hypersensitive induced reaction (HIR) in response to SA, ethylene, and ROS 57 , and thus is part of the machinery that prevents excessive damage to host tissues.  Determination of total anthocyanins. Anthocyanins were extracted from 8 pieces of apple skin, each of 1 cm 2 in size from one individual fruit, in 1 mL 1% (v/v) HCl-methanol for 24 h at room temperature in the dark. The upper aqueous phase was separated by centrifugation for 5 min at 13000 g and was subjected to spectrophotometric quantification at 650, 620 and 530 nm using a UV-Vis spectrophotometer (Shimadzu UV-2450, Kyoto, Japan). The content of the anthocyanins was determined using the following formula: OD = (A530-A620)-0.1 (A650-A620) 58 .

Material and Methods
Gene cloning and expression analysis. Total RNA was extracted from plant material for gene cloning and expression analyses using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). For quantitative real-time PCR analyses, we used the clone enzyme Kit (Transgene, Beijing, China) and SYBR Green MasterMix (SYBR Premix EX Taq TM, Dalian, China), according to the manufacturer's instructions. The primer sequences used for quantitative real-time PCR analyses are listed in Supplemental Table S1.
DNA constructs and genetic transformation. The full length coding regions of MdJAZ2 and MdHIR4 were amplified from the cDNA of Malus domestica using standard molecular biology protocols and enzymatic digestion technology (Invitrogen, Carlsbad, CA, USA). The MdHIR4 PCR product was recombined with vector pCXSN-HA 59 to create the pCXCN-MdHIR4 plasmid. The MdJAZ2 PCR product was cloned into pMD18-T (Takara Bio, Atsu, Japan) to create a fusion protein of the GUS gene and the coding region of MdJAZ2, for convenient detection of MdJAZ2. The MdJAZ2-GUS fusion was recombined into the vector pRI101-AN (Takara Bio, Atsu, Japan), to form the pRI-MdJAZ2-GUS plasmid. In both expression vectors the promoter was the cauliflower mosaic virus (CaMV) 35S promoter. All the specific primers and restriction enzymes used are listed in Supplemental Table 1. The two plasmids were introduced into the Agrobacterium tumefaciens strain LBA4404, and the resultant Agrobacterium tumefaciens transformants were used to transform the apple callus as described by An et al. 18 .
Yeast two-hybrid (Y2H) screening and assays. To screen for proteins that interact with MdJAZ2, the full-length CDS of the MdJAZ2 gene was inserted into pGBT9 vector which contain the DNA binding domain of GAL4. The resulting construct, MdJAZ2-BD, was used as bait vector to screen an apple cDNA library. The cDNA library was made with total RNAs isolated from apple fruit peel and constructed by Oebiotech Company (Shanghai, China).
Y2H assays were carried out as described by Xie et al. (2012). All of the coding regions of the JAZ genes used in this study were amplified from the cDNA of Arabidopsis and Malus domestica. The JAZ genes were recombined with the vector pGAD424 to generate constructs containing the JAZ genes fused to the GAL4 activation domain for Y2H analysis. The coding regions of the HIR genes used in this study were amplified from the cDNA of Arabidopsis and Malus domestica. They were then recombined with the vector pGBT9 to generate constructs of the HIR gene fused to the GAL4 DNA binding domain for Y2H analysis. The specific primers and restriction enzymes that were used are listed in Supplemental Table 1.
Yeast transformants were exhaustively screened on synthetic defined (SD) media (-Leu/-Trp/-His/-Ade) according to the manufacturer's instructions (Clontech, Palo Alto, CA, USA). The JAZ-AD and HIR-BD plasmids were co-transformed into the Y2HGOLD yeast cell strain using the lithium acetate method and were cultured at 30 °C. The resulting yeast transformants were filtered on medium lacking Trp and Leu (-Trp/-Leu), and putative transformants were subsequently transferred to medium lacking Trp, Leu, His and Adenine (-Leu/-Trp/-His/-Ade) with and without X-alpha-gal.
Pull-down assays. The full-length coding sequences of MdJAZ2, AtJAZ3, AtJAZ9 and MdHIR2, MdHIR4, AtHIR1 were cloned into the pGEX-4T-1 and pET32a vectors, respectively, to generate the JAZ-HIS and HIR-GST constructs used for the pull-down assays. All the specific primers and restriction enzymes that were used are listed in Supplemental Table 1.
For the in vitro pull-down experiments, expression of the proteins MdJAZ2/AtJAZ3/AtJAZ9-HIS and MdHIR2/MdHIR4/AtHIR1-GST was induced and the proteins were purified from E. coli BL21 cells. The purified protein mixtures of the JAZ and HIR genes were incubated for 1 h at 4 °C, 80 rpm, and purified using a His purification kit (Cwbio, Beijing, China), according to the manufacturer's instructions. Finally, the proteins were detected through Western blot analysis using anti-His and anti-GST antibodies.

Construction of the viral vectors and the Agrobacterium infiltration of the apple fruit. Viral
vectors were used as described by Li et al. 20 . to observe the effects of MdJAZ2 and the MdHIR genes overexpression or suppression in apple fruit. The full-length sequences of MdJAZ2 and MdHIR4 were cloned into the pIR vector under the control of the 35S promoter. The overexpression constructs were named pIR-MdJAZ2 and pIR-MdHIR4. To generate antisense expression vectors for MdJAZ2 and MdHIR4, the fragment of MdJAZ2 and MdHIR4 used in the above-mentioned genetic transformation was cloned into the tobacco rattle virus (TRV) vector in an antisense orientation under the control of the 35S promoter. All the specific primers and restriction enzymes that were used are listed in Supplemental Table 1. The resultant vectors were named TRV-MdJAZ2, and TRV-MdHIR4. The resultant viral vectors were used for injection into the apple fruit as described by Xie et al. 12 .
The degradation of the protein. The callus of the 'Orin' cultivar, which was transformed with the pRI-MdJAZ2-GUS fusion gene, was grown on the culture medium for 15 days and treated with 50 mM MG132 (proteinase inhibitor, Sigma-Aldrich) overnight in the dark. Then, the calli were washed clean and treated with 50 μ M Me-JA. Untreated calli was used as a control. The JAZ protein was detected through Western blotting using an anti-GUS antibody (BGI, Beijing, China). The 'Orin' callus co-transformed with 35S::MdJAZ2-GUS and 35S::MdHIR4 was obtained and used to testing for the function of the HIR protein in regards to the stability of