CaCRK5 encodes a pathogen induced receptor-like kinase of the arginine aspartate (RD) family
In cDNA-AFLP experiments designed to isolate genes involved in the pepper resistance to the infection of R. solanacearum, a partial cDNA fragment of CaCRK5 was obtained. As its expression was significantly upregulated after inoculation of R. solanacearum, we decided to study the function of this gene further. CaCRK5 cDNA clones were isolated from a cDNA library made from R. solanacearum inoculated leaves of pepper inbred line CM334. The proteins deduced from the cDNA clones of CaCRK5 contained 669 residues. SMART (http://smart.embl-heidelberg.de/) analysis of the domain architecture showed that their primary structures were composed of two cysteine-rich DUF26 domains, a transmembrane region and a serine/threonine kinase domain , and therefore CaCRK5 belongs to the family of cysteine-rich kinases . In addition, CaCRK5 contains a conserved arginine-aspartic (RD) motif in the protein kinase domain (Fig.1). For most RD kinases, the phosphorylation in the activation loop is crucial for triggering kinase activity which usually displays phosphorylation/autophosphorylation ability. Non-RD kinases usually exhibit lower kinase activities because of the lack of activation loop autophosphorylation. RD and non-RD kinases often cooperate to control innate immune signaling in plants [16-19].
Genome-wide identification of CRK family in pepper
In the start of this work, we evaluate the CRK gene family in the pepper genome, Hidden Markov Model (HMM) profile of the Stress-antifung domain (PF01657) was used to search the pepper genome database PGP (http://peppergenome.snu.ac.kr/). The Arabidopsis CRK gene family sequences  were also used as query sequences to search against the PGP and the NCBI database. All of them were further subjected to domain analysis using NCBI-CDD (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) and SMART, to ensure the presence of three important domains for typical CRK proteins including stress-antifung domains, transmembrane domain, and kinase domain. A total of 27 CRK genes (CaCRKs) were identified in the pepper cultivar CM334 genome, which were numbered from CaCRK1 to CaCRK27 according to their localization on chromosomes. Except for CaCRK13/24/25 with only one DUF26 domain, most CRKs contained two DUF26 domains. The length of CaCRK proteins varied from 332 to 1120aa, and molecular weights from 36.76 kDa to 128.25 kDa. The predicted isoelectric points of CaCRK ranged from 5.53 to 9.37(Additional file 1).
Chromosome physical localization analysis revealed that 27 CaCRKs were distributed on 7 of the 12 chromosomes in pepper. Chromosomes 2 contained the largest number of CRKs with 9 genes (34.6%), chromosomes 12 had only 1 gene. Gene clustering was the most common feature of the CRK genes distribution (Fig. 2a). Phylogenetic analysis was performed using the full length protein sequences of CRK from pepper, and an unrooted phylogenetic tree constructed by the neighbor-joining (NJ) method. Multiple CRKs were present in tandem repeat on the same chromosome (e.g. CaCRK1-7, CaCRK8/9, CaCRK10/11, CaCRK13-15 and CaCRK24/25) (Fig. 2b). These results suggested that the expansion of CRKs in pepper may be due to tandem duplication. This distribution and physical clustering pattern is consistent with CRKs in Arabidopsis and soybean.
Expression of CaCRK5 response to infection of R. solanacearum and treatment with exogenous SA, MeJA and ETH
To characterize the expression pattern of CaCRK5 in detail, quantitative real-time PCR (qRT-PCR) analyses were performed. 6-week-old pepper plants were leaf-inoculated with R. solanacearum. CaCRK5 was rapidly up-regulated over time after inoculation of R. solanacearum, and the highest level of expression was observed at 12 h post inoculation (hpi) with about 12.4-fold of that in control plant (Fig. 3a). Next, we investigated the expression of CaCRK5 treated with the defense-related signaling molecules salicylic acid (SA), jasmonic acid (JA) and ethylene (ET). The analysis showed the expression of CaCRK5 was increased after SA treatment and reached a peak at 48h, and no significant change was observed after methy jasmonate (MeJA) or ethephon (ETH) treatment (Fig. 3b). These results supported that CaCRK5 was involved in pepper defense against R. solanacearum invasion.
Subcellular localization of CRK5
Since CaCRK5 encoded a potential transmembrane domain, it was predicted to localize to plasma membrane. To test the hypothesis, CaCRK5 was fused with green fluorescent protein (GFP) under the control of the CaMV 35S promoter. CBL1n protein, which is known to be localized to the plasma membrane , was fused with red fluorescent protein (RFP). 35S:GFP, 35S:CaCRK5-GFP were transiently co-expressed with 35S:CBL1n-RFP in leaf epidermal cells of N. benthamiana. As shown in Fig.4, 35S:GFP construct served as a negative control, and green fluorescence was ubiquitously distributed throughout the cell. CaCRK5-GFP green fluorescence overlapped very closely with the red fluorescence in the plasma membrane suggesting that they are localized on the plasma membrane in the plant cell.
Silencing of CaCRK5 in pepper plants increases susceptibility to R. solanacearum infection
To assess the role of CaCRK5 in the interaction between pepper and R. solanacearum, loss-of-function experiments in pepper seedlings by virus-induced gene silencing (VIGS) were performed . As shown in Additional file 2, photobleaching was observed in newly emerged true leaves of plants infiltrated with Agrobacterium carrying CaPDS, indicating that the VIGS system worked efficiently. qRT-PCR analysis showed that expression of CaCRK5 in pepper leaves was significantly down-regulated during R. solanacearum infection in VIGS plants (Fig. 5a), indicating that CaCRK5 were efficiently silenced.
CaCRK5 silenced and control (TRV:00) pepper plants were subjected to R. solanacearum challenge. Phenotypic analysis indicated that CaCRK5 silenced pepper plants showed more severe disease symptoms than control plants after R. solanacearum infection (Fig. 5b). From 6 days post inoculation with R. solanacearum, CRK5 silenced pepper plants exhibited significant delay of wilt symptom. The disease index of CaCRK5 silenced plants was significantly increased, compared to the control plants (Fig. 5c). To address whether silencing of CaCRK5 affect growth of R. solanacearum, the bacterial population was determined. As shown in Fig. 5d, growth of R. solanacearum was significantly enhanced in CaCRK5 silenced plants 3 day post leaf-inocualtion, compared with control plants. Based on these observations, we believed that CaCRK5 participate defense responses in pepper.
Next, we investigated the cell death and oxidative burst in CaCRK5 silenced and the control leaves. Trypan blue and DAB staining confirmed that hypersensitive cell death and H2O2 accumulation were significantly reduced in CaCRK5 silenced leaves 48 h after inoculation with R. solanacearum (Fig. 5e), indicating that CaCRK5 play pivotal roles in early defense response associated with cell death during R. solanacearum infection. We further determined the effects of CaCRK5 silencing on the expression of defense related genes in pepper during R. solanacearum infection. qRT-PCR analyses showed that CaCRK5 silencing in pepper leaves significantly attenuated expression of defense related genes, including CaNPR1, CaSAR8.2, CaDEF1 and CaACO1, during R. solanacearum infection (Fig. 5f).
Overexpression of CaCRK5 in tobacco reduces susceptibility to R. solanacearum infection
A gain-of-function approach was also employed to study the function of CaCRK5 in the defense response. Due to technical difficulties, we used the tobacco plant (Nicotiana benthamiana), which is also a host for R. solanacearum. At least 10 transgenic tobacco lines were obtained and confirmed by kanamycin resistance analysis. Two T3CaCRK5 overexpressed lines exhibited constitutive expression (Fig. 6a), and were used in subsequent experiments.
Wild-type and transgenic plants (L3 and L7) at 4 weeks old were root-inoculated with R. solanacearum. The effect of CaCRK5 overexpression on development of bacterial wilt disease was observed. As shown in Fig. 6b, the CaCRK5 overexpressed plants showed much weaker symptoms at 12 d post-inoculation, compared with the wild-type. The wild-type plants showed wilt symptom from 6 day after inoculation and completely died 14 d after inoculation with R. solanacearum. The CaCRK5 overexpressing plants were observed significant delay of wilt symptom compared to wild-type plants (Fig. 6c), showing that overexpression of CaCRK5 conferred increased disease tolerance to R. solanacearum in tobacco. In addition, we also checked the expression of defense related genes, including NtPR2, NtPR3, NtHSR201 and NtHSR505. qRT-PCR analysis indicated that expressions of these tested defense related genes were increased in CaCRK5 overexpressed plants compared with wild-type plants, during the inoculation of R. solanacearum (Fig. 6d). Collectively, these data suggested that CaCRK5 overexpression enhanced defense responses against R. solanacearum infection in tobacco.
CaCRK5 is directly regulated by transcription factor CaHDZ27
To better understand the regulatory mechanism of CaCRK5, a 2000-bp promoter region upstream of the CaCRK5 coding sequence was identified. Sequence analysis using the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) suggested that the cis-elements in the promoter of CaCRK5 included two TCA-element involved in salicylic acid responsiveness, five binding sites for MYB transcription factor, and one binding sites for MYC. In addition, a known binding site (CAATTATTG) for the HD-Zip subfamily I member CaHDZ27 was located between positions -625 and -617. As CaHDZ27 also regulate the defense in pepper against R. solanacearum infecction, we speculated that CaCRK5 might be targeted by CaHDZ27.
An electrophoretic mobility shift assay (EMSA) was conducted to access the interaction of CaHDZ27 with the CaCKR5 promoter, and CaHDZ27 was found to bind to the Cy5-labeled CaCKR5 promoter probe. Moreover, binding was gradually attenuated by increasing concentrations of unlabeled probe, indicating that CaHDZ27 binds specifically to the CAATTATTG in the CaCKR5 promoter in vivo (Fig. 7a). Chromatin immunoprecipitation (ChIP)-qPCR was conducted to confirm interaction of CaHDZ27 and the CaCKR5 promoter in vitro. The pepper leaves infiltrated with Agrobacterium strain GV3101 containing 35S:HA-CaHDZ27 or 35S:HA were harvested for ChIP assay. Chromatin from these pepper leaves was immunoprecipitated using anti-HA antibodies and enrichment of DNA sample was determined by qRT-PCR. The result showed that CaHDZ27 was significantly enriched in the CaCKR5 promoter, and the enrichment was significantly enhanced by the inoculation of R. solanacearum (Fig. 7b and c), suggesting that CaHDZ27 could bind to the CAATTATTG in the CaCKR5 promoter in vitro, and binding was increased by the R. solanacearum infection.
To further investigate the CaCKR5 regulated by CaHDZ27 in the transcriptional level, we transiently expressed 35S:CaHDZ27-GFP in pepper leaves, and 35S:GFP infiltrated leaves were used as a negative control. qRT-PCR analysis demonstrated that the transcript level of CaCKR5 was increased in 35S:CaHDZ27-GFP infiltrated leaves compared with the control(Fig. 8a). Using anti-GFP immune-blotting, CaHDZ27 was confirmed to be expressed in pepper leaves (Fig. 8b). We used a VIGS approach to silence CaHDZ27. The transcript level of CaCKR5 is reduced significantly in CaHDZ27-silenced plants compared with TRV:00 infiltrated plants, 48 h post the inoculation of R. solanacearum (Fig. 8c and d). These data suggested that the transcriptional expression of CaCKR5 was positively regulated by CaHDZ27.
CaCRK5 interacts with CaCRK6
Previous studies indicated that plants employ immune receptor complex for sensing microbe-derived molecular patterns and effectors to trigger inducible immune defenses[28-31], and a few CRKs have been known to function in association with each other, such as AtCRK28/AtCRK29 and AtCRK39/AtCRK40. Thus, we investigated the ability of CaCRK5 to heterodimerize with the closely related CaCRK6, which has 77.4% amino acid sequence identity to CaCRK5.
The cDNAs of CaCRK5 and CaCRK6 were cloned into the pGBKT7 and pGADT7 vectors separately, and generated DNA-binding domain (BD) and activation domain (AD) fusions. BD-CaCRK5/AD-CaCRK6 and BD-CaCRK6/AD-CaCRK5 transformant yeast cells can grow on SD/-Trp-Leu-His-Ade medium, as did positve control (pGBKT7-53/pGADT7-T). In contrast, no growth was observed in the negative controls. The result indicated that CaCRK5 interacts with CaCRK6 each other in yeast cells (Fig. 9a). To provide more evidence for the interaction between CaCRK5 and CaCRK6, we performed bimolecular fluorescence complementation (BiFC) assays. CaCRK5 and CaCRK6 were fused to the N- and C-terminal ends of yellow fluorescent protein (YFP) to generate CaCRK5-nYFP/CaCRK6-cYFP and CaCRK5-cYFP/CaCRK6-nYFP, respectively. Then, these constructs were transiently co-expressed in Nicotiana benthamiana leaves. At 48h post infiltration, the YFP fluorescence signals were checked by by confocal microscopy. As showed in Fig. 9b, YFP fluorescence was oboserved at the plasma membrane in N. benthamiana leaves. These results indicate that CaCRK5 heterodimerizes with CaCRK6 at the plasma membrane of plant cells.