CaHDZ27, a Homeodomain-Leucine Zipper I Protein, Positively Regulates the Resistance to Ralstonia solanacearum Infection in Pepper

Plants encounter a vast array of potentially phytopathogenic pathogens in their natural environments and have developed sophisticated defense strategies during coevolution with these pathogens. These strategies generally include constitutive chemical and physical barriers along with inducible innate immunity, which includes pathogen-associated molecular pattern–triggered immunity and effector-triggered immunity (ETI) (Hein et al. 2009; Jones and Dangl 2006). Inducible innate immunity is largely regulated at the transcriptional level by the actions of many transcriptional factors (TFs), which are generally encoded by large gene families. TFs are important players in regulated gene expression networks. Some studies have suggested that TFs are promising biotechnological targets for engineering-improved plant disease resistance (Goossens et al. 2017; Wang et al. 2015; Xu et al. 2011). To fully exploit these immunity-related TFs, a deeper understanding of their functional roles and mechanisms is required.

HD-Zip proteins contain a DNA-binding homeodomain (HD) and the adjacent leucine zipper (Zip) (Ariel et al. 2007). HD-Zip proteins constitute a large family of plant-specific TFs, with the identification of 55 HD-Zip TFs in corn (Mao et al. 2016), 27 in citrus (Ge et al. 2015), 51 in tomato (Zhang et al. 2014b), 33 in peach (Zhang et al. 2014a), 88 in soybean (Chen et al. 2014), 63 in Populus spp. (Hu et al. 2012), and 47 in Arabidopsis thaliana (Henriksson et al. 2005). HD-Zip protein family members are classified into four subfamilies (HD-Zip I to HD-Zip IV) based on their sequence similarities (Ariel et al. 2007; Henriksson et al. 2005; Hu et al. 2016; Schena and Davis 1994). The majority of HD-Zip genes in different classes are responsible for modulating developmental processes, such as those of vascular tissue, trichomes, cuticle (Müller et al. 2016; Yan et al. 2017), embryogenesis (Roodbarkelari and Groot 2017), leaf polarity (Merelo et al. 2016), and flowering (Kovalchuk et al. 2016). HD-Zip genes also are involved in plant responses to drought (Chen et al. 2014; Harris et al. 2016; Zhao et al. 2014), salinity (Chen et al. 2014; Zhao et al. 2014), and pathogen infection (AbuQamar et al. 2006; Oh et al. 2013).

HD-Zip class I functions by binding to a 9-bp pseudopalindromic DNA sequence (CAATNATTG) in the promoters of their target genes (Johannesson et al. 2001; Palena et al. 1999). HD-Zip I proteins form homodimers and heterodimers with HD-Zip proteins in the same class (Harris et al. 2016; Johannesson et al. 2001; Meijer et al. 2000). Most plant stress-responsive HD-Zip proteins belong to class I. Functional studies of HD-Zip I members indicate that they have important roles in abiotic stress responses to frost (Kovalchuk et al. 2016), flooding (Cabello et al. 2016), and drought (Harris et al. 2016; Romani et al. 2016; Zhao et al. 2014). Several reports suggested that HD-Zip I members are involved in plant responses to pathogen infection. For example, overexpression of the HD-Zip I gene ATHB13 in Arabidopsis leads to broad-spectrum disease resistance (Gao et al. 2014). The tomato HD-Zip I gene H52 is upregulated in response to Pseudomonas syringae pv. tomato infection, resulting in an oxidative burst that normally precedes the hypersensitive response (HR) (Mayda et al. 1999). The sunflower HD-Zip I gene HAHB4 is strongly induced by treatment with methyl jasmonate (MeJA) or ethylene (ET). Arabidopsis plants ectopically expressing HAHB4 are sensitive to bacterial infection and reduce levels of SA after infection with Pseudomonas syringae (Manavella et al. 2008). However, the involvement of HD-Zip I proteins in plant immunity is poorly understood.

Pepper (Capsicum annuum) is one of the most important vegetables worldwide. Pepper is a member of the Solanaceae family and is vulnerable to several soilborne diseases, such as bacterial wilt caused by Ralstonia solanacearum, which causes serious yield and economic losses (Flores-Cruz and Allen 2009). The most efficient and sustainable way to overcome pepper pathogenic diseases is to develop cultivars with disease resistance. A deeper understanding of the mechanisms involved in pepper immunity would facilitate the genetic improvement of pepper disease resistance. In the present study, we isolate and functionally characterize the Capsicum annuum HD-Zip I gene CaHDZ27. Our results indicate that CaHDZ27 functions as a positive regulator of pepper’s response to R. solanacearum infection.

RESULTS

Sequence analysis of CaHDZ27.

HD-Zip I proteins have been implicated in plant immunity, but no studies have been conducted on HD-Zip I proteins in pepper immunity. To isolate pepper genes associated with immunity, a cDNA library was constructed using RNA of pepper leaves inoculated with R. solanacearum. Random sequencing of this library identified a full-length cDNA containing an open reading frame (ORF) of 891 bp that harbored a homeodomain (HD) and an adjacent leucine zipper (Zip) domain without any other known domains in its deduced amino acid sequence (Fig. 1), suggesting that it was a pepper HD-Zip I protein (Ariel et al. 2007; Brandt et al. 2014). The deduced amino acid sequence of this cDNA was most similar to SlHZ27 (83.39%), an HD-Zip I protein family member in tomato (Solanum lycopersicum), we designated it as CaHDZ27. We performed a phylogenetic analysis using the full-length amino acid sequences of CaHDZ27, 22 HD-Zip I proteins from tomato (Zhang et al. 2014b), and 17 from Arabidopsis (Henriksson et al. 2005); the results indicated that CaHDZ27 was most closely related to SlHZ27 and belonged to a subgroup that included only three tomato proteins SlHZ27, SlHZ24, and SlHZ21(Supplementary Fig. S1). Other subgroups included members from both tomato and Arabidopsis. These results were consistent with a previous report (Zhang et al. 2014b). Alignment analysis of the deduced amino acid sequence of CaHDZ27 and HD-Zip I proteins in other plant species showed the high sequence similarity of CaHDZ27 with other HD-Zip I proteins. Interestingly, we also found a conserved motif in which serine (S) and aspartate (D) are rich (Fig. 1). Therefore, CaHDZ27 is likely an HD-Zip I protein.

Download as PowerPoint

CaHDZ27 is transcriptionally upregulated by R. solanacearum inoculation and exogenous application of salicylic acid (SA), MeJA, or ethephon (ETH).

To test whether CaHDZ27 is involved in pepper immunity, its transcriptional expression was examined by real-time quantitative polymerase chain reaction (qPCR) analysis of pepper plants inoculated with R. solanacearum, the causal agent of bacterial wilt of pepper. The results showed that CaHDZ27 transcript levels were upregulated significantly at 24 and 48 h postinoculation (Fig. 2A). SA, jasmonic acid (JA), and ET have been implicated as general defense signaling molecules in plant immunity; therefore, we tested the effects of exogenous application of SA, MeJA, or ETH on CaHDZ27 transcript levels in pepper seedlings. Different exogenous hormone concentrations were first investigated to determine suitable concentrations; 1 mM SA, 100 μM MeJA, and 100 μM are the optimal concentrations for spraying (Supplementary Fig. S2). qPCR analysis showed that MeJA application increased CaHDZ27 transcript levels at 6, 12, and 24 h posttreatment (hpt), with the peak at 24 hpt, compared with those in mock-treated control plants (Fig. 2B). SA application increased CaHDZ27 transcript levels at 1 and 6 hpt, compared with those in mock-treated control plants (Fig. 2C). ETH application markedly increased CaHDZ27 transcript levels at 6, 12, and 24 hpt, compared with those in mock-treated control plants (Fig. 2D). These combined results indicate that CaHDZ27 transcription is modulated by R. solanacearum inoculation and exogenous application of SA, MeJA, and ETH. Therefore, CaHDZ27 may have a functional role in pepper immunity against R. solanacearum.

Download as PowerPoint

CaHDZ27 protein is localized in the nucleus.

A nuclear localization sequence (NLS) at amino acid residue 78 to 86 (PEKKRRLLP, cNLS Mapper score 7) was determined using the open source software cNLS Mapper, suggesting that it is targeted to the nucleus. To determine the subcellular localization, CaHDZ27 was fused with the green fluorescent protein (GFP) gene under the control of the 35S promoter. This construct (35S::CaHDZ27-GFP) was transformed into Agrobacterium tumefaciens GV3101 and was infiltrated by injection into Nicotiana benthamiana leaves. GFP expression was examined after 48 h, using a Leica confocal microscopy. GFP fluorescence was detected exclusively in the nuclei of cells transiently overexpressing CaHDZ27-GFP, which also were stained with 4′,6-diamidino-2-phenylindole (DAPI) (Supplementary Fig. S3). By contrast, GFP fluorescence in control leaves, infiltrated with 35S::GFP, was observed throughout the cells. These results suggest that CaHDZ27 is localized in the nucleus.

CaHDZ27 has DNA-binding and transactivating activities.

HD-Zip I proteins act as TFs by binding to the pseudopalindromic sequence CAAT(A/T)ATTG in their target genes (Ariel et al. 2007). To test if CaHDZ27 can bind this cis element, we performed chromatin immunoprecipitation (ChIP) analyses. We constructed two β-glucuronidase (GUS) reporter vectors (2×cis-p35Score::GUS and 2×mcis-p35Score::GUS), which were driven by a synthetic promoter containing two copies of the cis element (cis, CAATAATTG) or its mutant (mcis, CAGGGGTTG) and an adjacent Cauliflower mosaic virus (CaMV) 35S core promoter sequence. The control vector construct (p35Score::GUS) contained only the CaMV 35S core promoter sequence (Fig. 3A). The effect vector (35S::CaHDZ27-HA) and reporter vectors were transformed into GV3101. Cells containing both effect and reporter vectors were mixed at a 1:1 ratio and were infiltrated into pepper leaves, which were harvested after 48 h and were used for ChIP analyses. The results revealed that CaHDZ27 bound to the synthetic promoter containing the CAATAATTG cis element but did not bind to the promoter containing the mcis element (CAGGGGTTG) or only the CaMV 35S core promoter sequence (Fig. 3B). We analyzed GUS activities in pepper leaves at 48 or 72 h after infiltration with GV3101 cells carrying effect and reporter vectors. The results showed that GUS activities were significantly higher after infiltration with 2×cis-p35Score::GUS compared with 2×mcis-p35Score::GUS or p35Score::GUS (Fig. 3C). These results indicate that CaHDZ27 can activate GUS expression by binding to the cis element (CAATAATTG).

Download as PowerPoint

CaHDZ27 silencing significantly impairs pepper resistance against R. solanacearum infection.

The obtained results suggest that CaHDZ27 might function as a TF and is upregulated by R. solanacearum infection. Therefore, we performed a gene silencing analysis using virus-induced gene silencing (VIGS) and a Tobacco rattle virus (TRV) vector containing a 185-bp fragment in the 3′ untranslated region (UTR) of CaHDZ27. Three independent experiments were performed, and each experiment included approximately 100 TRV::00 and 100 TRV::CaHDZ27 plants. The control plants expressed a pepper phytoene desaturase gene (TRV::CaPDS), which enabled visual inspection of gene silencing based on photobleaching (Supplementary Fig. S4). Reverse transcription (RT)-PCR demonstrated that the transcript level of CaHDZ27 is reduced in most of the TRV::CaHDZ27-infiltrated plants, as compared with TRV::00-infiltrated plants (Supplementary Fig. S5). Subsequently, the CaHDZ27 silenced plants and TRV::00-infiltrated plants were inoculated by infiltrating leaves with an injection of 20 μl of R. solanacearum FJC100301 cells (10⁸ CFU/ml). The result of qPCR analysis showed that CaHDZ27 transcript levels during R. solanacearum infection in TRV::CaHDZ27-infiltrated plants were significantly down-regulated compared with TRV::00 control plants (Fig. 4A).

Download as PowerPoint

Wounded root inoculations were performed by pouring 10 ml of the 10⁸ CFU/ml suspensions into each pot. TRV::CaHDZ27 plants displayed leaf wilting at 10 days after root inoculation, and this progressed until complete leaf wilting was observed approximately 12 days after root inoculation. By contrast, plants inoculated with TRV::00 displayed delayed symptom development at 10 days after root inoculation (Fig. 4B; Supplementary Table S1) and symptom progression (leaf wilting) significantly differed in TRV::CaHDZ27 and control TRV::00 plants (Fig. 4C).

To investigate whether CaHDZ27 silencing affects HR-mediated cell death in response to R. solanacearum infection, we stained leaves of TRV:00 and TRV::CaHDZ27 plants with diaminobenzidine (DAB) (indicator of H₂O₂ accumulation) and trypan blue (indicator of cell death or necrosis) 48 h after R. solanacearum inoculation. The results showed that DAB and trypan blue staining intensities were higher in TRV::00 control leaves than in TRV::CaHDZ27 leaves (Fig. 4D). Consistently, electrolyte leakage levels were significantly higher in TRV::00 leaves than in TRV::CaHDZ27 leaves at 48 or 72 h after R. solanacearum infection (Fig. 4E). Notably, H₂O₂ accumulation and cell death during R. solanacearum infection were significantly attenuated in CaHDZ27-silenced pepper leaves.

To confirm these phenotypic results and to investigate the mechanism of CaHDZ27 activity in pepper immunity, we performed qPCR analysis of the following pepper defense-related gene expression levels in TRV::CaHDZ27 and TRV::00 control plants challenged with R. solanacearum: the HR marker gene CaHIR1 (Choi et al. 2011), the ET biosynthesis–associated gene CaACO1 (Cai et al. 2015), ET/JA-responsive gene CaPR4 (Cai et al. 2015), the reactive oxygen species (ROS) detoxification–associated gene CaPO2 (Choi et al. 2007), the SA-responsive gene CaPR1 (Jung and Hwang 2000), and the basal defense–related gene CaBPR1 (An et al. 2008). The results showed that transcript levels of CaHIR1, CaACO1, CaPR4, CaPO2, CaPR1, and CaBPR1 were significantly reduced in CaHDZ27-silenced leaves compared with the levels in control plants with or without R. solanacearum infection (Fig. 4F). These combined results suggest that CaHDZ27 silencing directly or indirectly downregulates the expression of these defense-related genes during R. solanacearum infection, and enhances the susceptibility of pepper plants to R. solanacearum infection.

Transient CaHDZ27 expression triggers cell death and upregulates immunity-associated marker genes in pepper leaves.

We performed gain-of-function analysis to further confirm the role of CaHDZ27 in pepper immunity. For these experiments, we tested the transient expression of CaHDZ27, using Agrobacterium infiltration into pepper leaves, and monitored CaHDZ27 expression by qPCR analysis. The results showed that CaHDZ27 transcript levels were substantially higher in leaves transiently overexpressing CaHDZ27 (35S::CaHDZ27) compared with those in control leaves (35S::00) (Fig. 5A).

Download as PowerPoint

HR-mediated cell death is a hallmark indicator of ETI. To determine whether HR-mediated cell death is triggered by transient CaHDZ27 overexpression, the infiltrated leaves were stained with trypan blue and DAB. The results showed that transient CaHDZ27 overexpression resulted in more intense trypan blue and DAB staining, indicating that infiltrated leaves transiently overexpressing CaHDZ27 have higher levels of cell death and H₂O₂ accumulation than control leaves (35S::00) (Fig. 5B). Higher levels of electrolyte leakage also occurred in leaves expressing 35S::CaHDZ27 than in control leaves (35S::00) at 48 and 72 h after infiltration (Fig. 5C). We also performed a qPCR analysis of defense-marker gene expression, and the results showed higher expression levels of CaHIR1, CaACO1, CaPR1, CaPR4, CaPO2, and CaBPR1 in leaves transiently expressing CaHDZ27 than in control leaves (Fig. 5D). These combined results indicate that CaHDZ27 has a role in regulating HR- and defense-associated gene expression during pepper defense responses.

Ectopic CaHDZ27 overexpression confers R. solanacearum resistance in transgenic tobacco.

Stable transformation of pepper plants is still technically challenging. Therefore, we generated transformed tobacco (Nicotiana tabacum) plants overexpressing CaHDZ27, using A. tumefaciens–mediated leaf disc transformation. We acquired a number of homozygous T₃ lines that did not exhibit significant morphological or developmental differences from wild-type plants. Three lines (#3, #5, and #8) were randomly selected for further functional characterization of CaHDZ27 (Supplementary Fig. S6). By inoculating roots with R. solanacearum, leaf-wilting symptoms were observed in wild-type plants at 12 days after inoculation, whereas CaHDZ27-overexpressing transgenic plants (lines #3, #5, and #8) exhibited only minimal leaf-wilting symptoms (Fig. 6A). The temporal progression of symptoms significantly differed between wild-type and CaHDZ27-overexpressing transgenic plants (Fig. 6B; Supplementary Table S2).

Download as PowerPoint

Next, we investigated H₂O₂ accumulation and HR-mediated cell death in response to R. solanacearum infiltration into leaf veins. After 48 h, the infiltrated leaves were stained with DAB and trypan blue. The results showed that H₂O₂ accumulation and cell death were higher in the CaHDZ27-overexpressing line #3 challenged with R. solanacearum (Fig. 6C). Consistently, bacterial growth was significantly reduced in leaves of line #3 compared with that in wild-type plants (Fig. 6D). These results indicate that CaHDZ27 overexpression enhances the resistance of tobacco plants against R. solanacearum infection. The qPCR analysis indicated that transcript levels of HR marker genes (NtHSR201 [Sohn et al. 2007], ET biosynthesis-associated NtACC oxidase [Chen et al. 2003], and SA- or JA-associated NtPRa/c, NtPR2, NtPR3, and NtPR4 [Brogue et al. 1991; Ward et al. 1991; Dang et al. 2014]) were significantly higher in CaHDZ27-overexpressing plants with or without R. solanacearum inoculation than in wild-type tobacco plants (Fig. 6E).

CaHDZ27 can form homodimers.

To test whether CaHDZ27 can form homodimers, we performed bimolecular fluorescence complementation (BiFC) and coimmunoprecipitation (CoIP) analyses. For the BiFC assay, CaHDZ27 was fused to the N- and C-terminal ends of yellow fluorescent protein (YFP) to generate CaHDZ27-nYFP and CaHDZ27-cYFP, respectively. Each construct was introduced into A. tumefaciens GV3101, and cells carrying different constructs were simultaneously infiltrated into N. benthamiana leaves. The infiltrated leaves were examined by confocal microscopy after 48 h. YFP fluorescence was observed exclusively in nuclei in leaves infiltrated with both CaHDZ27-nYFP and CaHDZ27-cYFP, whereas no YFP fluorescence was observed in negative controls (coinfiltration of nYFP and CaHDZ27-cYFP, or cYFP and CaHDZ27-nYFP) (Fig. 7A). We also examined CaHDZ27 homodimerization using CoIP. FLAG-tagged CaHDZ27 and hemagglutinin (HA)-tagged CaHDZ27 were transiently coexpressed in N. benthamiana leaves by infiltration. After 48 h, total protein was extracted and immunoprecipitated with FLAG-antibody–conjugated agarose beads, and the isolated proteins were then analyzed with HA antibody. The results show that CaHDZ27 can bind to itself to form homodimers (Fig. 7B).

Download as PowerPoint

CaHDZ27 dimerization is enhanced by R. solanacearum infection.

Next, we tested whether CaHDZ27 dimerization was affected by R. solanacearum infection by performing floating-leaf luciferase (LUC) complementation imaging using Gateway-compatible split-LUC destination vectors (Gehl et al. 2011). We generated 35S::CaHDZ27-NLUC, 35S::CaHDZ27-CLUC, and pEXP-GUS (control) transient overexpression constructs and transformed them into A. tumefaciens GV3101 cells. Then, pepper leaves were coinfiltrated with the transient overexpression constructs. GUS activity was measured and used to normalize total protein levels, and dynamic protein-protein interactions were detected by luminescence activity measurement assays. The negative controls (coinfiltration of CaHDZ27-NLUC and CLUC, or CaHDZ27-CLUC and NLUC) had lower luminescence activity than leaves coinfiltrated with CaHDZ27-NLUC and CaHZ27-CLUC. These results suggest that CaHDZ27 can form homodimers. The luminescence activity in pepper leaves coinfiltrated with CaHDZ27-NLUC and CaHDZ27-CLUC was significantly increased at 1 day after R. solanacearum inoculation (Fig. 8), suggesting that CaHDZ27 homodimerization was enhanced by R. solanacearum infection.

Download as PowerPoint

DISCUSSION

HD-Zip proteins are unique TFs in plants, and HD-Zip I proteins have been implicated in abiotic stress responses (Harris et al. 2011; Zhang et al. 2014b; Zhao et al. 2014). Several HD-Zip I members have important roles in biotic stress responses, such as ATHB13 in Arabidopsis (Gao et al. 2014), H52 in tomato (Mayda et al. 1999), and HAHB4 in sunflower (Manavella et al. 2008). Information is lacking about pepper HD-Zip I members, and their biological functions are unknown. This study performed a functional genetic analysis of the CaHDZ27 HD-Zip I protein of Capsicum annuum. The results indicate that CaHDZ27 functions as a positive regulator of R. solanacearum infection in pepper.

The deduced amino acid sequence of CaHDZ27 contains a HD and an adjacent Zip domain and exhibits the highest sequence similarity with SlHZ27, among all HD-Zip I proteins in tomato and Arabidopsis. CaHDZ27 also has high sequence similarities with homologs from other species, including potato, tomato, and tobacco. We also show that CaHDZ27 binds to the 9-bp pseudopalindromic DNA sequence (CAATNATTG), which is specifically bound by HD-Zip class I proteins (Johannesson et al. 2001; Palena et al. 1999), and activates GUS expression in a CAATAATTG-dependent manner. These results suggest that CaHDZ27 is an HD-Zip class I protein in pepper. This conclusion was further supported by the observation that CaHDZ27-GFP expression was observed exclusively in the nuclei, which is frequently observed in other HD-Zip I proteins (Cao et al. 2016; Zhao et al. 2014).

Real-time qPCR analysis indicated that CaHDZ27 transcription was significantly induced by R. solanacearum inoculation, implying an involvement in pepper response to R. solanacearum. Plant genes that are upregulated in response to pathogen attack frequently have important roles in plant immunity (Cai et al. 2015; Dang et al. 2013, 2014; Shen et al. 2016a and b). Loss-of-function VIGS silencing experiments in pepper plants confirmed this hypothesis, because CaHDZ27 silencing increased susceptibility to R. solanacearum infection and downregulated the defense-related marker genes CaHIR1 (Choi et al. 2011, 2013; Jung and Hwang 2007; Jung et al. 2008), CaACO1 (Dong et al. 2014; Cheng et al. 2017), CaPR1 (Yi et al. 2013; Kim et al. 2014), CaPR4 (Dong et al. 2014), CaPO2 (Choi and Hwang 2012; Choi et al. 2007), and CaBPR1 (Hong et al. 2005; Sarowar et al. 2005). By contrast, transient CaHDZ27 overexpression in pepper plants increased HR-mimicking cell death and upregulated defense-related marker genes. Consistently, ectopic CaHDZ27 overexpression in transgenic tobacco enhanced resistance against R. solanacearum infection and induced H₂O₂ accumulation and defense-related gene expression of NtHSR201 (Czernic et al. 1996; Qiu et al. 2016; Takahashi et al. 2004), NtACC oxidase (Avni et al. 1994; Fammartino et al. 2010; Lasserre et al. 1997), NtPR1a/c (Jamir et al. 2004; Matsuoka et al. 1987; Zhang et al. 2012b), NtPR2 (Ghanta et al. 2011; Naoumkina et al. 2008; Wang et al. 2013), NtPR3 (Zhang et al. 2012a), and NtPR4 (Li et al. 2015; Maimbo et al. 2007). These combined results indicate that R. solanacearum infection enhances CaHDZ27 expression and enhanced CaHDZ27 expression in turn activates HR- or immunity-associated gene expression, leading to enhanced disease resistance.

The plant hormones SA, JA, and ET have key roles in regulating defense signaling networks that are activated in response to pathogen invasion (Pieterse et al. 2012). SA is primarily involved in biotrophic pathogen defense signaling, whereas JA and ET are primarily involved in necrotrophic pathogen defense signaling (Glazebrook 2005; Spoel et al. 2007). The production of SA, JA, and ET is frequently coupled with ETI or pathogen-triggered immunity (PTI) (Tsuda and Katagiri 2010), and the defense phytohormone levels depend on the recognized pathogen (De Vos et al. 2005) and the invading population density (Pétriacq et al. 2016). The defense phytohormones can act synergistically or antagonistically, depending on their concentration levels (Kunkel and Brooks 2002; Mur et al. 2006). Recent work reported that SA and JA/ET are synergistically related in PTI but compensatorily related in ETI (Tsuda et al. 2009). The present study found that the CaHDZ27 transcriptional response to exogenous application of MeJA, SA, and ETH was consistent with SA-, JA-, and ET-responsive pathogenesis-related marker gene expression in pepper and tobacco plants overexpressing CaHDZ27. The expression of CaPO2, a marker gene for ROS-dependent resistance, also was upregulated by transient CaHDZ27 overexpression in pepper. These combined results indicate that R. solanacearum infection in pepper upregulates CaHDZ27 expression, which in turn synergistically boosts SA-, JA-, and ET-mediated defense responses.

HD-Zip proteins form homodimers and heterodimers with other HD-Zip proteins in the same family (Johannesson et al. 2001; Meijer et al. 2000). Our results showed that CaHZ27 forms homodimers in pepper and this homodimeriztion was found to be enhanced by R. solanacearum infection, indicating that some unidentified signaling components might be activated by R. solanacearum, which potentiate homodimeriztion of CaHZ27. We speculate that the homodimeriztion of CaHZ27 might play a role in resistance of pepper to R. solanacearum infection; to confirm this speculation, further investigation and more evidence is required.

In summary, our data indicate that CaHDZ27 is a member of the HD-Zip I protein family in pepper. CaHDZ27 is transcriptionally upregulated by R. solanacearum infection, thereby integrating SA-, JA-, and ET-dependent defense signaling. CaHDZ27 may act as a positive regulator of defense responses against R. solanacearum infection.

MATERIALS AND METHODS

Plant materials and growth conditions.

Seeds of pepper (Capsicum annuum L. cv. Fj8) were obtained from the pepper breeding group in Fujian Agriculture and Forestry University (Fuzhou, Fujian, China). Tobacco (Nicotiana tabacum L. cv. Honghuadajinyuan) was supplied by the tobacco breeding group in Fujian Agriculture and Forestry University. Pepper and tobacco plants were grown in steam-sterilized soil mix in plastic pots at 25°C under a 16-h day and 8-h night photoperiod.

Pathogens and inoculation procedures.

Ralstonia solanacearum FJC100301 was cultured in potato sucrose agar (medium overnight at 28°C, according to the method of Dang et al. (2013). The bacterial culture density was measured as 10⁸ CFU/ml (optical density at 600 nm [OD₆₀₀] = 0.8) from serial dilutions in 10 mM MgCl₂. Wounded root inoculation was performed by pouring 10 ml of the 10⁸-CFU/ml bacterial suspension into each pot. Before inoculation, roots were slightly damaged by making three holes in the soil. Then, the pots were kept in a growth room at 28°C with soil moisture >90%.

For RNA extraction, trypan blue and DAB staining, and ion conductivity experiments, plants were inoculated by infiltrating bacterial suspension (OD₆₀₀ = 0.8) into the third leaf from the apical meristem, using a syringe without a needle, and mock control plants were inoculated with sterile 10 mM MgCl₂.

Treatment of plants with exogenous hormones.

Pepper plants at the six-leaf stage were used to investigate CaHDZ27 expression patterns in response to exogenous application of hormones or growth regulators. Plants were sprayed with 0.1, 1, or 10 mM SA (dissolved in 10% ethanol), 10, 100, or 1,000 μM MeJA (dissolved in 10% ethanol), or 10, 100, or 1,000 μM ETH (dissolved in sterile distilled and deionized H₂O). Mock control plants were sprayed with 10% ethanol or sterile distilled and deionized H₂O.

Vector construction.

Vectors were constructed using a Gateway cloning technique (Invitrogen, Carlsbad, CA, U.S.A.), and a series of Gateway-compatible destination vectors were employed. The ORF of CaHDZ27, with or without the stop codon, was initially amplified by PCR with its specific primer pair (5′-ATGGAAGTTGGAGGAGTTATT-3′ and 5′-CTATTGCCAGAACCAACAAGG-3′) flanked with attB for Gateway cloning and GXL DNA polymerase (Takara, Osaka, Japan). The amplified product was confirmed by sequencing. The CaHDZ27 ORF was cloned into the entry vector pDONR207, using the BP reaction and was then cloned into destination vectors such as pMDC83, pK7WG2, and pEarleyGate201, using the LR reaction for subcellularr localization, transient overexpression, or ChIP analysis, respectively.

DNA fragments, containing two copies of conserved HD-Zip I protein-binding cis element (CAATAATTG), or two copies of mutant element (CAGGGGTTG) and the 54-bp minimal CaMV 35S promoter flanked with attB, were synthesized by Convenience Corporation (Suzhou, China). These were cloned into the pMDC163 destination vector by sequential BP and LR reactions for promoter expression assays of the GUS reporter gene in pepper plants.

To construct the gene-silencing vectors for VIGS analysis, a fragment (approximately 185 bp) of the 3′ UTR of CaHDZ27 was amplified by PCR with a specific primer pair (5′-TTCCACAAGAGAATAGTG-3′ and 5′-GGAACAAAGCTAATAAA-3′). The fragment was cloned into the VIGS vector TRV2 (PYL279), using the Gateway cloning technique, as described above.

For construction of the BiFC vector, the CaHDZ27 ORF without the stop codon was cloned into pDONR207, using the BP reaction. Then, the LR reaction was used to recombine the entry plasmids into the Gateway binary destination vectors puc-SPYNE^GW and puc-SPYCE^GW, which contain the N-terminal or C-terminal fragment of YFP (nYFP or cYFP, respectively). Thus, the CaHDZ27 protein was independently tagged with nYFP or cYFP at either the N or C terminus.

Split-LUC vectors were constructed by amplifying the attB-flanked CaHDZ27 ORF without stop codon, using gene-specific primers. The PCR products were introduced into the satellite vector pDONR207. The CaHDZ27 entry vector was transformed into the destination vectors (pDEST-^GWNLUC and pDEST-^GWCLUC) to generate 35S::CaHDZ27-NLUC and 35S::CaHDZ27-CLUC, respectively.

Transient expression of CaHDZ27 in pepper leaves.

The 35S::CaHDZ27 vector was transformed into A. tumefaciens GV3101. GV3101 cells carrying the 35S::CaHDZ27 vector were grown overnight in YEP medium with appropriate antibiotics. Cells were suspended in infiltration buffer (10 mM MES, pH 5.7, 10 mM MgCl₂, and 200 μM acetosyringone). The cell suspension (OD₆₀₀ = 0.8) was infiltrated into leaves of pepper plants at the eight-leaf stage, using a syringe without a needle. The plants were kept in a growth room at 25°C, 60 to 70 mmol photons/m/s, 70% relative humidity, and a 16-h light and 8-h dark photoperiod, for 2 days before injected leaves were used for assays.

Tobacco transformation.

Transgenic tobacco plants overexpressing CaHDZ27 were generated using Agrobacterium-mediated tobacco leaf disc transformation (Oh et al. 2005). Eight kanamycin-resistant lines of transgenic tobacco plants harboring the 35S::CaHDZ27 construct were selected and were confirmed with RT-PCR analysis. Seeds of T₁ plants were collected from regenerated T₀ plants, and seedlings of T₂ and T₃ lines were further selected on Murashige-Skoog agar plates containing 50 μg of kanamycin per milliliter. T₃ plants were used for analyses.

Subcellular localization of CaHDZ27.

The 35S::CaHDZ27-GFP and 35S::GFP constructs were transformed into A. tumefaciens GV3101 and were grown overnight. Then, GV3101 cells were suspended in infiltration buffer (10 mM MES, pH 5.7, 10 mM MgCl₂, and 200 μM acetosyringone) and were infiltrated into N. benthamiana leaves, using a syringe without a needle. For DAPI staining, N. benthamiana leaves were immersed in DAPI solution (0.1% DAPI and 5% dimethyl sulfoxide [DMSO]) and were incubated at 37°C for 1 h, before observation. GFP fluorescence was imaged using a laser scanning confocal microscope (TCS SP8; Leica, Solms, Germany) with an excitation wavelength of 488 nm and a 505 to 530 nm band-pass emission filter.

VIGS of CaHDZ27 in pepper plants.

VIGS of CaHDZ27 was performed according to a previously described method (Dang et al. 2013). GV3101 cells harboring TRV1 (PYL192) with TRV2 (PYL279) and TRV2::CaHDZ27 (OD₆₀₀ = 0.8) were mixed at a 1:1 ratio, and the mixtures were coinfiltrated into cotyledons of 2-week-old pepper seedlings. The seedlings were incubated at 16°C for 56 h and were then grown at 25°C.

Histochemical staining and ion conductivity measurement assay.

Histochemical staining and ion conductivity measurements were conducted using previously published methods (Choi et al. 2012). Agrobacterium-infiltrated leaves were stained with DAB for H₂O₂ detection and with trypan blue for cell-death detection. To quantify cell death in pepper leaves, ion conductivity was measured, using a Mettler Toledo 326 (Mettler, Zurich, Switzerland), as described previously (Cai et al. 2015).

Real-time qPCR.

The qPCR analysis was performed, using a Bio-Rad real-time PCR system (Foster City, CA, U.S.A.) and the SYBR Premix Ex Taq II system (Takara), as described previously (Cai et al. 2015; Dang et al. 2013). The expression of CaHDZ27 and defense-related genes in pepper or tobacco plants were examined, using the primers listed in Supplementary Tables S3 and S4. Relative transcript levels were normalized with respect to the levels of CaActin and 18srRNA in pepper or NtActin and NtEE1α in tobacco. Data were calculated by the by the 2^−ΔΔCT method, as described previously (Livak and Schmittgen 2001). In each case, three technical replicates were performed for at least three independent biological replicates.

ChIP analysis.

ChIP assays were performed as described previously (Cai et al. 2015). The leaves of pepper plants at the eight-leaf stage were infiltrated with A. tumefaciens GV3101 cells carrying the effect vector (35S::CaHDZ27-HA) and reporter vectors (2×cis-p35Score::GUS, 2×mcis-p35Score::GUS, and p35Score::GUS). T-DNA insertion into the pepper genome was confirmed. After 48 h, approximately 2 g of leaves were harvested and treated with either 10 mM bithionol sulfoxide or DMSO (solvent control) for 16 h and were then fixed with 1.0% formaldehyde for 5 min. The chromatin was isolated and sheared to an average length of 500 bp. Chromatin fragments that bound to CaHDZ27 were immunoprecipitated using the HA antibody (Santa Cruz Biotechnology, Santa Cruz, CA, U.S.A.). The resulting DNA fragments were isolated and were used as templates for PCR analysis with attB-specific primer pairs.

BiFC analysis.

BiFC was performed as described previously (Liu et al. 2016). CaHDZ27-nYFP and CaHDZ27-cYFP were introduced into A. tumefaciens GV3101, and cells carrying the BiFC constructs were infiltrated into N. benthamiana leaves. After 48 h, leaf fluorescence was analyzed using a confocal microscope.

CoIP assay.

The 35S::CaHDZ27-FLAG and 35S::CaHDZ27-HA constructs were transformed into A. tumefaciens GV3101, and cells carrying 35S::CaHDZ27-FLAG or 35S::CaHDZ27-HA were coinfiltrated into N. benthamiana leaves at a ratio of 1:1. The Co-IP assay was performed as described previously (Liu et al. 2016).

LUC activity determination.

GV3101 cells carrying the appropriate split-LUC constructs were to infiltrate pepper leaves. For coexpression analyses, cells carrying the different vectors were mixed before infiltration. GV3101 cells carrying the appropriate split-LUC constructs were infiltrated by injecting 5-week-old pepper leaves. After 1 day, the infiltrated leaves were inoculated with 20 μl of R. solanacearum cells, and leaves were harvested for LUC activity assay after 1 day. LUC activity reconstituted from the 35S::CaHDZ27-NLUC and 35S::CaHDZ27-CLUC vectors was determined as described previously (Liu et al. 2016).

ACKNOWLEDGMENTS

We thank M. D. Curtis for kindly providing the Gateway destination vectors and S. P. Dinesh-Kumar (Yale University) for the pTRV1 and pTRV2 vectors. This work was supported by grants from the National Natural Science Foundation of China (31301254, 31372061, 31401890, 31401312, 31260482, and 31060263).