Functional Analysis and Expressional Characterization of Rice Ankyrin Repeat-Containing Protein, OsPIANK1, in Basal Defense against Magnaporthe oryzae Attack

The ankyrin repeat-containing protein gene OsPIANK1 (AK068021) in rice (Oryza sativa L.) was previously shown to be upregulated following infection with the rice leaf blight pathogen Xanthomonas oryzae pv oryzae (Xoo). In this study, we further characterized the role of OsPIANK1 in basal defense against Magnaporthe oryzae (M.oryzae) by 5′ deletion analysis of its promoter and overexpression of the gene. The promoter of OsPIANK1 with 1,985 bps in length was sufficient to induce the OsPIANK1 response to inoculation with M.oryzae and to exogenous application of methyl jasmonate (MeJA) or salicylic acid (SA), but not to exogenous application of abscisic acid (ABA). A TCA-element present in the region between −563 bp and −249 bp may be responsible for the OsPIANK1 response to both M.oryzae infection and exogenous SA application. The JERE box, CGTCA-box, and two MYB binding sites locating in the region between −1985 bp and −907 bp may be responsible for the response of OsPIANK1 to exogenous MeJA. OsPIANK1 expression was upregulated after inoculation with M.oryzae and after treatment with exogenous SA and MeJA. Overexpression of OsPIANK1 enhanced resistance of rice to M.oryzae, although it did not confer complete resistance. The enhanced resistance to M.oryzae was accompanied by enhanced transcriptional expression of SA- and JA-dependent genes such as NH1, WKRY13, PAL, AOS2, PR1b, and PR5. This evidence suggests that OsPIANK1 acted as a positive regulator in rice basal defense mediated by SA- and JA-signaling pathways.


Introduction
Plants naturally co-exist with numerous and varied microbial pathogens and the diseases that can result represent a heavy loss in crop productivity worldwide. In consequence, plants have evolved effective immune systems to defend themselves against the attack of microbial pathogens. Two interconnected modes of the innate immune system are used by plants to defend against phytopathogens. The first mode, pattern-triggered immunity (PTI), results from the activation of plant cell-surface pattern recognition receptors (PRRs) following the exposure of these receptors to molecular patterns common to many types of microbes (pathogenassociated molecular patterns [PAMPs]). Infection by a particular pathogen can be blocked by PTI. Pathogens have evolved effector molecules that are secreted into plant cells and suppress PTI, and, in turn, plants have evolved resistance (R) proteins that recognize the effectors and activate the second mode of the innate immune system, effector-triggered immunity (ETI), in a gene-to-gene way [1][2][3][4]. In addition to the protein-protein interactions mediated by PRRs, R proteins, PAMPs, and effectors, other proteins such as the leucine-rich repeat (LRR) domain-containing proteins [5][6][7] and the ankyrin (ANK) domain-containing proteins [8,9] are important in plant defense.
Here, we report on a study of the ANK domain in the plant immune response, using rice as a model system. Rice is one of the most agriculturally important crops worldwide, but approximately a million tons are lost to disease each year. Better understanding of the genetic mechanisms of rice defense responses could lead to better agricultural management and reduced losses due to diseases.
The OsANK proteins in rice can be classified into 10 subfamilies. Expression profile analysis of OsANK proteins suggested that members of this gene family may play important roles in pollination, fertilization, and signaling [12]. Among all of the 175 known ANK-repeat genes in rice, only a few members have been functionally characterized. The ANK-repeat domaincontaining XA21 binding protein 3 was required for full Xa21mediated disease resistance [33], OsCBT, containing three ANK repeats, is a transcription activator modulated by CaM [43], OsBIANK1, encoding a plasma membrane-anchored ANK-repeat protein, was induced by treatment with either benzothiadiazole or by infection with Magnaporthe grisea [31]. OsPIANK1, an ankyrin protein in rice, was late up-regulated by infection of the rice leaf blight pathogen Xoo at both early and late infection stages, and had different expression patterns against the attack of Xoo [44]. Both PTI and ETI are accompanied by major transcriptional reprogramming, which is considered a key step of the plant defense program [45,46]. Up to 25% of all Arabidopsis genes respond to pathogen infection by altering their transcript levels [47,48]. Genetic evidence indicates that genes that are transcriptionally upregulated during plant immune responses are important in disease resistance [49][50][51]. This reasoning suggests that OsPIANK1 plays an important role in rice immunity, although there is no direct evidence for this role. In the present study, we performed functional identification of OsPIANK1 by its overexpression in rice and expressional characterization by a 59 deletion assay of its promoter. The results suggest that OsPIANK1 is a positive regulator in the rice basal defense mediated by the SA and JA signaling pathways.

Isolation of the Full Length cDNA and Promoter of OsPIANK1
The full length cDNA of OsPIANK1 was amplified by reverse transcription-polymerase chain reaction (RT-PCR). Total RNA from leaves of Nipponbare (Oryza sativa L. japonica) was extracted using the TRIzol Reagent (Invitrogen, Carlsbad, CA, USA). Moloney Murine Leukemia Virus (M-MLV) Reverse Transcriptase (Invitrogen) was used for cDNA synthesis according to the manufacturer's protocol. The newly synthesized cDNA was used as the template for PCR amplification with the OsPIANK1-specific forward primer (59-TACGGCATgCATACTCCATCA-39) and reverse primer (59-TACATGCATGCAAGCTGTCA-39). Genomic DNA of Nipponbare was extracted from leaf tissues using the hexadecyltrimethylammonium bromide (CTAB) method. The OsPIANK1 promoter region was amplified by PCR using the specific primer pair (59-TTTGCACCTTTTACCCTg-3) and (59-CGCTTATTAGAGATAAAGAGGA-39). All of the PCR products was cloned into pMD18-T vector and sequenced. The OsPIANK1 promoter sequences were analyzed by the PLACE Web Signal Scan program (http://bioinformatics.psb.ugent.be/ webtools/plantcare/html/).

Vectors Construction and Rice Transformation
To construct the overexpression vector of OsPIANK1, the open reading frame of OsPIANK1 was amplified by PCR with Kpn 2 -Spel 2 (indicated by lowercase letters) linker primers (59-GGggatc-cATGCATACTCCATCAAT-39, 59-GGactagtTCAGGTCGT-CATCACAT-39) using pMD18T-OsPIANK1 as the template, which was further cloned into the Kpn I and Spel I sites of the modified binary expression vector pCAMBIA1390 under the control of the maize ubiquitin promoter.
After verification by sequencing, all of the vectors were transferred into Agrobacterium tumefaciens strain EHA105, and then were transformed into Nipponbare rice. The Agrobacterium-mediated transformation was performed according to the method of Toki S [53] using vigorously growing callus derived from mature embryos. More than 10 independent transgenic lines were obtained for each vector.

Pathogen and Exogenous Phytohormones Treatment
To analyze the response of OsPIANK1 promoter to inoculation with the pathogen and exogenous application of phytohormones,

GUS Activity Assay
Histochemical staining for beta-glucouronidase (GUS) activity was performed as reported previously [54]. Total protein was extracted from the leaf tissue of the OsPIANK1pro-GUS plants. GUS activity was measured using a spectrophotometer to quantify the rate of release of pnitrophenol (l = 415 nm) from p-nitrophenyl-b-D-glucuronide (PNPG) [54]. GUS activity assays were repeated with at least two independent lines, each with three replicates. When similar results were obtained in repeated experiments, only one experiment has been presented.

Subcellular Localization of OsPIANK1
For OsPIANK1 localization, the full-length coding sequence of OsPIANK1 was amplified and fused in frame to the N-terminus of the green fluorescent protein (GFP) gene to generate the CaMV35S: OsPIANK1-GFP construct by Gateway-mediated recombination into the vector pMDC83. The CaMV35S::GFP construct was used as the control. Both constructs were further transformed into Agrobacterium tumefaciens strain GV3101. The cells of GV3101 harboring CaMV35S::OsPIANK1-GFP constructs was co-infiltrated into N. benthamiana leaves. One day after agroinfiltration, the leaves were visualized under a fluorescence microscope (Olympus DP72), with an excitation wave length of 488 nm and a 505-530 nm band-pass emission filter. 49,6diamidino-2-phenylindole (DAPI) fluorescence was also imaged using an excitation wavelength of 405 nm and a 435-480 nm band-pass emission filter.

Quantitative Real-time PCR
For quantitative real-time PCR analysis, total RNA from the leaves of the wild-type (Nipponbare) and transgenic plants were extracted. First-strand cDNA was generated by converting 500 ng total RNA using Primescript RT reagent (perfect real time, TaKaRa), and the cDNA was diluted to 100 mL with water. Realtime PCR using Mastercycler ep realplex (Eppendorf, Hamburg, Germany) was performed with SYBRH Premix Ex Taq TM II (perfect real time, TaKaRa). Each reaction mix (25 mL) contained 12.5 mL SYBR Premix Ex Taq, 0.5 mL PCR forward/reverse gene specific primers (10 mM), 2.5 mL diluted cDNA, and 9.5 mL water. For each gene, three experimental replicates were obtained using different cDNAs synthesized from three biological replicates. The thermal cycle used was as follows: one cycle of 30 s at 95uC; 40 cycles of 5 s at 95uC, 34 s at 60uC; one cycle of 15 s at 95uC, 1 min at 60uC, 15 s at 95uC, 15 s at 60uC. Rice actin gene (X15865) was used for normalization. The gene-specific primer pairs are listed in Table 1. The relative expression levels were determined as described by Livak and Schmittgen [55].

Quantification of M.oryzae DNA in Rice Leaves
Rice and fungal DNA were extracted from leaves inoculated with M. oryzae and quantified by real-time PCR in accordance with the method of Qi and Yang [56] using two specific primer pairs, which were designed based on the 39 noncoding region of a MPG1 gene in M. oryzae (59-GGGATGATGGTGGTGGAGGAC-39; 59-GCCAGGTGCTTAGGACGAAAC-39). The data were normalized to the amount of DNA of a rice actin gene (AK060893), which was quantified using the forward primer (59-GAGTATGAT-GAGTCGGGTCCAG-39) and reverse primer (59-ACACCAA-CAATCCCAAACAGAG-39).

Expression Pattern of OsPIANK1 in Response to Pathogen Infection and Exogenous Application with Phytohormones
Expression of OsPIANK1 in rice has been reported to be transcriptionally up-regulated after infection of the rice leaf blight pathogen Xoo [44]. We wished to evaluate the possibility that OsPIANK1 is an important regulator for the rice response to other pathogens and to universal defense signaling molecules such as SA and JA. To this end, the transcriptional levels of OsPIANK1 following infection with the M.oryzae strain guy11 or exogenous application of SA and MeJA were examined by real-time PCR. Transcripts of OsPIANK1 increased approximately 10-fold 1 day after inoculation with M. oryzae, remained high for 2 to 4 days after inoculation, and decreased sharply 5 days after inoculation ( Figure 1A). OsPIANK1 transcripts increased markedly 3-12 h after treatment with SA and declined dramatically 24 h after treatment. SA treatment induced the rice acidic pathogenesis-related (PR) protein1 (PR1a) gene that was used as a comparable control ( Figure 1B). In response to MeJA, the mRNA accumulation of OsPIANK1was increased remarkably for 3 h, reached a maximal level after 6 h, and decreased sharply 12 h after treatment. MeJA induced the rice basic PR protein1 (PR1b) gene that was used as a comparable control ( Figure 1C). These results indicate that, in addition to responding to infection by Xoo, OsPIANK1 is involved in the response of rice to M.oryzae infection regulated by the signaling pathway mediated by SA and JA.

Deletion Analysis of the OsPIANK1 Promoter
Gene expression was largely controlled by its promoter. Therefore, deletion analysis of OsPIANK1 promoter was carried out to dissect the possible molecular mechanism of OsPIANK1 response to M.oryzae, SA, and JA. A 1985-bp upstream region of the OsPIANK1 open reading frame was isolated from Nipponbare genomic DNA by PCR amplification and cis-acting elements in this region were analyzed (http://bioinformatics.psb.ugent.be/ webtools/plantcare/html/). Several stress-related sequence motifs putatively acting as cis-elements were identified in the OsPIANK1 promoter ( Figure 2). These sequence motifs included two MYB binding sites (MBS) involved in response to drought; a cis-acting element involved in response to heat stress (HSE); a cis-acting element involved in abscisic acid responsiveness (ABRE); a cisacting element involved in dehydration, low-temperature, and salt stresses (DRE); two ethylene-responsive elements (GCC box and ERE); two cis-acting regulatory element involved in the MeJAresponsiveness (JERE and the TCACG motif); and a cis-acting element (TCA-element) involved in SA responsiveness.
To analyze the activity of the OsPIANK1 promoter in major organs of rice seedlings, the transgenic rice plants of OsPIANK1pro-  b-glucouronidase (GUS) and their corresponding T 1 lines were acquired. These lines were phenotypically normal and their growth was similar to wild-type plants. GUS expression was detected histochemically in leaves, stems, and roots of OsPIANK1pro-GUS plants ( Figure 3A). Next, the activation of the OsPIANK1 promoter by pathogen infection was assessed. OsPIANK1pro-GUS plants were inoculated with M.oryzae and the responses of OsPIANK1 promoter were monitored by determining GUS activity. GUS staining was more intense in M.oryzae-inoculated leaves of OsPIANK1pro-GUS rice plants 4 to 6 days after inoculation than in the uninoculated leaves ( Figure 3B).
To perform 59 deletion analysis of the OsPIANK1 promoter, serially 59-deleted OsPIANK1 promoters were created by PCR (Figure 4), cloned into the vector pMDC163 upstream of the GUS reporter gene, and transformed into rice via the Agrobacteriummediated method. The transgenic rice plants and their corresponding T 1 lines were acquired and detected by PCR. A 59 deletion assay of the OsPIANK1 promoter was performed with the transgenic plants. GUS expression in response to infection with M.oryzae or exogenous SA was similar for the following promoters: OsPIANK1pro-1985 bp, OsPIANK1pro-1101 bp, OsPIANK1pro-907 bp, and OsPIANK1pro-564 bp but different between these and OsPIANK1pro-249 bp ( Figure 5), suggesting that the cis-

A Transiently Expressed OsPIANK1-GFP Fusion Protein is Localized in the Nucleus
The full-length cDNA of OsPIANK1 was amplified with a specific primer pair of OsPIANK1 by PCR with cDNA synthesized from the total RNA of rice leaves inoculated with M. oryzae. To determine the subcellular localization of OsPIANK1, the construct of a Cauliflower mosaic virus 35S-controlled OsPIANK1-green fluorescent protein chimaera (CaMV35S::OsPIANK1-GFP) and the control vector (CaMV35S::GFP) were constructed, transformed into Agrobacterium strain GV3101, and infiltrated into N. benthamiana leaves. Localization of the OsPIANK1-GFP fusion protein was visualized exclusively in the nuclei ( Figure 6).

OsPIANK1 Overexpression Enhanced the Resistance of Rice to Infection by M.oryzae
We wished to characterize the function of OsPIANK1 in rice against M. oryzae attack. First OsPIANK1-overexpression (OE) lines were generated. From these, three lines (lines 3#, 4#, and 6#) were chosen based on their high transcript levels of OsPIANK1. The function of OsPIANK1 was studied in T 2 generations of those transgenic lines compared to WT rice plants. The expression of OsPIANK1 in the three OsPIANK1-OE lines was higher than in WT plants ( Figure 7A).
Three-week-old plants of OsPIANK1-OE transgenic and WT plants were inoculated with M.oryzae strain guy11 and evaluated for symptoms of rice blast 6 days later. The three OsPIANK1-OE lines had fewer and smaller expanding lesions ( Figure 7B) with less fungus in the leaves ( Figure 7C) than WT plants. OsPIANK1-OE and wild-type rice plants, both at the four-leaf stage, were inoculated with spores of M.oryzae strain guy11 and grown in a greenhouse. The OsPIANK1-OE and WT rice plants began to wilt at 15 days after inoculation and some of the WT plants were dead at 30 days after inoculation. In contrast, all of OsPIANK1-OE rice plants survived with slight disease symptoms ( Figure 7D). These results suggest that OsPIANK1-OE enhanced resistance to M.oryzae.

Overexpression of OsPIANK1 Upregulated the Expression of Defense Marker Genes
The mode of action of OsPIANK1 in rice was studied by comparing the transcriptional expression of a set of pathogeninduced genes involved in defense in OsPIANK1-OE rice seedlings inoculated with M.oryzae strain guy11 and in un-inoculated controls. The genes included phenylalanine ammonia lyase (PAL; X87946), Arabidopsis NPR1 homolog 1 (NH1; AY9123983), thaumatin-like protein (PR5; X68197), basic PR protein 1 (PR1b; U89895), OsWRKY13 (EF143611), and allene oxide synthase 2 (AOS2; AY062258). The transcriptional expression of PAL, NH1, PR1b, and AOS2 was enhanced by overexpression of OsPIANK1 in both inoculated and control seedlings. Expression of PR5 was transcriptionally enhanced only in the control seedlings while WRKY13 was enhanced only in the inoculated seedlings ( Figure 8).

Discussion
One hundred seventy-five OsANK proteins have been identified in rice, which have been classified into 10 subfamilies based on their domain compositions. Seventy-three (41.7%) belong to subfamily ANK-M, a classification based on the presence of the ANK domain in the absence of other known functional domains. Members of the other subfamilies contain not only the ANK domain but several other known functional domains [12]. The only distinguishing feature of the OsPIANK1 protein is its three ANK domains; therefore, it may belong to the ANK-M subfamily. Little is known about the ANK-M rice proteins, and this report may be the first characterization of a protein in this subfamily.
Plant immune responses are accompanied by major transcriptional reprogramming and the genes involved appear to have important roles in the immune response [45][46][47][48][49][50]57]. OsPIANK1 was upregulated following infection by the rice leaf blight pathogen Xoo [44], making it a likely player in the immune response of rice to pathogens. In the present study, OsPIANK1 also responded to M. oryzae, exogenous SA, and exogenous MeJA, suggesting that OsPIANK1 is a regulator in basal defense or PTI of rice against different pathogens. Transcriptional expression of genes is largely regulated by cis-elements in their promoters. Some stress-related cis-elements: the CGTCA-box and JERE, involved in the MeJA-responsiveness [58,59]; the ethylene-responsive element [60]; the GCC-box [61]; the TCA-element, involved in SA responsiveness [62]; MBS, involved in drought responsiveness [63]; and HSE, involved in heat stress responsiveness [64] were found in the promoter of OsPIANK1. GUS expression controlled by the full-length promoter was enhanced by inoculation of M.oryzae, a finding that confirmed our finding that overexpression of OsPIANK1 enhanced the resistance of rice to M.oryzae. The region responsible for M.oryzae infection and response to exogenous SA was between 2563 bp and 2249 bp, where only one ERE and one TCA-element was found, suggesting that this TCAelement was responsible for the response of OsPIANK1 to M.oryzae infection and that this TCA-element mediated SA-signaling pathway is sufficient for the response of OsPIANK1 to M.oryzae infection. This result is consistent with previous studies that SA is important in modulation of the redox balance and protection of rice plants from the oxidative stress caused by infection of Magnaporthe grisea [65] or Xanthomonas oryzae [66]. The regions responsible for the MeJA response in the promoter of OsPIANK1 located between 21985 bp and 2907 bp, where the CGTCAbox, JERE, the GCC-box, and MBS were located. Among these cis-elements, JERE and CGTCA-box cis-element may be responsible for the response of OsPIANK1 to exogenous application of MeJA. In addition, MBS may play a role in the response of OsPIANK1 to exogenous application of MeJA. In support of this, a myb gene that may bind to MBS was induced by jasmonic acid [67].
The data of the present study also suggested that the OsPIANK1 protein targeted to the nuclei and acted as a positive regulator in basal defense in rice. The latter conclusion was supported by the findings that inoculation of the WT, Nipponbare, with spores of M.oryzae strain guy11 caused a compatible reaction. Resistance to M.oryzae attack was greater in OsPIANK-OE rice plants than in WT, as manifested by fewer, smaller lesions, less M. oryzae, and greater vigor at 30 days after inoculation. These results indicated that OsPIANK1 was a regulator of basal defense or PTI, although overexpression did not confer complete resistance.
The transcriptional expression of NH1, WKRY13, PAL, AOS, PR1b, and PR5 were consistently enhanced by the overexpression of OsPIANK1. Of these genes, NH1, WKRY13, and PR5 are either SA-dependent or involved in the regulation of SA [68][69][70], and PAL is associated with SA biosynthesis and induced by SA application [71]. AOS2 is involved in JA synthesis [72][73][74] and PR1b has been reported to be associated with both JA and SA signaling pathways [69]. These results strongly suggest that OsPIANK1 primes the expression of SA-and JA-responsive endogenous genes. Although the activated immune responses in ETI are more prolonged and robust than those in PTI [75], differences observed between compatible and incompatible interactions are temporal and quantitative, not qualitative [47,48]. Common defense signaling molecules such as SA and JA are often common to both PTI and ETI [75], but the way they use the common network is very different: synergistic relationships among the signaling sectors are evident in PTI, compensatory relationships among the sectors are dominate in ETI [2]. Thus, the co-upregulation of JA-and SA-dependent genes by the overexpression of OsPIANK1 is consistent with the role of OsPIANK1 as a regulator in PTI or basal defense.
OsPIANK1 was late up-regulated by infection of rice plants with the leaf blight pathogen (Xoo) and had different expression patterns against the attack of Xoo in the leaf-blight resistant cultivar IET8585 and the leaf-blight susceptible cultivar IR24 [44], as would be predicted. In the present study, OsPIANK1 was cloned from Nipponbare, a japonica rice cultivar susceptible to M.oryzae and Xoo. Therefore, its expression patterns in response to pathogen attack and its function in disease resistance might be different from its analogues in resistant cultivars. Characterization of analogues of OsPIANK1 should supply new insight into its function and underlying mechanism in rice immunity.