Xanthomonas oryzae pv oryzae triggers immediate transcriptomic modulations in rice

Disease symptoms in IET8585 and IR24

In fifty five days old plants inoculated with Xanthomonas oryzae pv oryzae strain Bxo43, the symptoms first appear five days after inoculation (dai) as yellowish lesions around the site of inoculation in both the cultivars. In susceptible IR24 the whole leaf turned grayish yellow and dried up at 14 dai, lesions were visible on other leaves of the same plant as well. In resistant IET8585 at 14 dai the progression of lesion was limited to 16 +/- 2 cm from inoculation site whereas other leaves of the plant remained unaffected (Additional file 1). The symptoms in eighteen days old plants were also comparable with previously documented reports [12].

Microarray experiment and validation

Agilent Rice gene expression microarrays were used to examine differential transcript accumulation in resistant IET8585 and susceptible IR24 cultivars at 1 hai with Bxo43 or mock water treated control. The number of transcripts found to be differentially expressed in resistant plants compared to susceptible ones after pathogen inoculation were 378. Amongst them 104 were found to be differentially expressed after mock treatment as well and were not taken into consideration for further analysis (Figure 1A). A remaining subset of 274 transcripts were taken to be differentially expressed due to Bxo43 inoculation, of these 152 were found to be up-regulated and 122 down-regulated in IET8585 as compared to IR24. The microarray data has been submitted to ArrayExpress and is available as E-MEXP-3388.

Real-time qPCR was used to validate the microarray results. Fifteen genes were randomly selected from the differentially expressed genes. The differential expressions of selected genes were verified by quantitative PCR. Both the up- and down-regulated genes showed same trend of expression as obtained by microarray when further analyzed by qPCR (Additional file 2).

Identification and annotation of differentially expressed transcripts

The transcripts were annotated with the help of Rice Annotation Project Database http://rapdb.dna.affrc.go.jp. Annotations could be assigned to 112 up- and 73 down-regulated transcripts. On the basis of GO functional categories transcripts were assorted into several groups. The up-regulated transcripts were grouped into cell signaling (8%), transcriptional factors (10%), defense related (10%), general stress related (3%), transport (8%), metabolic (17%), structural (2%) and miscellaneous (16%) (Figure 1B, Additional file 3). Similarly down-regulated transcripts consisted of cell signaling (7%), transcription factors (2%), defense related (2%), stress related (7%), transport (7%), metabolic (23%), structural (3%) and miscellaneous (12%) (Figure 1C, Additional file 4). Interactions for 55 up-regulated (Additional file 5) and 31 down-regulated (Additional file 6) transcripts were mapped.

Expression of modulated transcripts at different time points

Cell signaling

It was interesting to note that the largest up-regulated cluster consisted of 15 transcription factors and 12 cell signaling related proteins, transcripts which or the products of which may act as mediators to usher defense response. In the corresponding down-regulated cluster 3 transcription factors and 9 signaling related proteins were found. Two receptor-like kinases (AK068504, AK111766) were found to be up-regulated (Table 1), one of them bearing a LysM domain (AK111766). The LysM domains are found in a variety of peptidoglycan or chitin binding proteins and have been implicated in perception of rhizobial lipichitooligosaccaride signals [17] and further elicitation of signals via its intercellular kinase domain [18]. A transcript for NB-ARC domain containing protein (AK069420) was found to be highly up-regulated (Table 1) as well. NB-ARC is an ancient highly conserved domain of a class of plant resistance proteins [19]. It has a functional ATPase domain and its nucleotide binding site is proposed to regulate activity of the R-protein in pathogen recognition.

Serine/theonine protein kinases (STK) have long been implicated to play a role in signaling processes concerned with self verses non-self recognition and disease resistance [20]. A up-regulated STK was found to be similar to MAPKKK17 (AK071585) (Additional file 5), which is known to be induced by pathogens [21]. Another MAPK cascade initiating protein MAPKKK3 (AK068725) [22] as well as Protein Phosphatase 2C (AK100561) or PP2C (Additional file 5) a regulator of MAPK pathway, known to be activated in stress [23], was found to be up-regulated in the present case study, suggesting that there may be a remarkable fine tuning of MAPK cascade at such early a time point. Calcium ion is the most important signal entity in cell, its importance is reflected in present work by differential regulation of transcripts of several associated proteins, up-regulation of a calmodulin-like protein (AK070889), Os12g0556500 (AK110372) a calmodulin-binding protein-like protein and Os06g0683400 (AK111852) an EF-hand domain containing protein, similar to CCD1 (Table 1). Studies in wheat cultured cells have revealed that ccd-1 mRNA are strongly responsive to elicitors of snow mold and gene product CCD1 plays a role in elicitor provoked Ca²⁺ mediated signal transduction [24]. Some Ca²⁺ binding proteins Os05g0583500 (AK100131) a calcium binding protein and an EF-hand domain containing protein (AK100302) were also found to be down-regulated indicating tight regulation of signaling pathways. Besides Ca²⁺ flux transporters for several other ions across membrane were also stimulated. These ion fluxes affect membrane potential which in turn affect uptake by other channels and activation of defense response [25]. A net K⁺ eflux in elicitor treated cells is a common early response [26]. In present data set a transcript similar to high affinity K⁺ uptake pump HAK5 (AK066544) (Additional file 6) [27] and K⁺ channel protein (AK069303) (Table 2) were found to be down-regulated whereas a low to high affinity K⁺ pump KUP3 (AK066194) [28] is up-regulated implying a regulation of K⁺ uptake. An additional complexity is introduced by the fact that K⁺ uptake is sensitive to ammonium transport and a transcript for AMT2-like protein (AK100411) an ammonium transporter has also been found to be up-regulated [29, 30]. Proton influx is another important defense response and three H⁺ transporters were found to be up-regulated in present study (Additional file 5), one of them was a proton-peptide symporter PTR2 (AK064899) and other two were sugar-proton symporters STP1 (AK099079) and PLT5 (AK073216) like proteins. The sugar-H⁺ symporters probably serve dual purpose, they relocate H⁺ and sugars into infected cells, which behaves as a sink and draws carbon resources for energy consumption to put up defense on one hand [31, 32] while on the other hand modulate sugar signals (Figure 2) which are known to influence SAR pathway [33, 34]. Monosaccharide-H⁺ symportes in Arabidopsis are rapidly induced by pathogenic elicitors [31].

Lipid signaling has emerged as an important component of stress signaling and Phospatidic acid (PA) is a key lipid second messenger. PA is directly formed via activation of Phospholipase D (PLD) and indirectly by phosphorylation of Diacylglycerol (DAG) by DAG kinase (AK100906) or DGK (Additional file 5) [35]. The DGK pathway has been known to be activated in defense response against pathogens within few hours of infection [36], a transcript similar to PA generating DAG kinase is upregulated in the present data set as well whereas a phospholipase A2 family protein (AK105828) was down-regulated (Additional file 6). PA can be deacylated by phospholipase A2 (PLA2) to produce LPA and free fatty acids (Figure 2), which are signaling compounds in plant responses to auxin [37]. Further a transcription factor Auxin responsive factor 5 or ARF5 (AK099793) that binds to AUX/IAA family proteins (Additional file 6), a critical component in auxin signaling pathway, as well as an AUX/IAA family protein (AK066518) were found to be down-regulated [38], incidentally another AUX/IAA family protein (AK068232) was found to be up-regulated (Additional file 5). This two-way modulation of AUX/IAA family proteins is in agreement with previous findings. AUX/IAA genes were discovered based on their induction by auxin and some can promote auxin response, however many AUX/IAA proteins actually inhibit auxin response [39]

The down-regulated cell signaling associated transcripts include two interesting genes a Leucine-rich repeat 2 containing protein (AK060253) and a Remorin domain containing protein (AK060392) (Table 2). LRR repeats occur in various proteins to provide structural framework for protein-protein interaction and when associated to kinase domain are known to function as receptors. Remorins are unique to plants and are ubiquitously expressed. Differential expression of remorins have been well documented in Arabidopsis, interestingly most of the remorins were found to be down-regulated in case of bacterial-plant incompatible interactions [40].

Transcription factors

The host cell gears up its arsenal on receiving signals of pathogen attack through elaborate signal transduction network by modulating expression of several transcription factors. These act as hubs that further modulate defense, hormone and growth related genes. The up-regulated cluster consists of several transcription factors some of them are R2R3 type MYB transcription factor (AK105817) related to stress response [41], MYC7E (AK099506) related to ABA response, OsNAC protein (AK073848), DREB1B(AK062422), Ethylene responsive factor (ERF1) family protein (AK073812), WRKY69 (AK111606) (Table 1). WRKY69 is a member of the well documented WRKY family related to regulation of many cellular processes including defense and is known in turn to be modulated by MAPK cascade [42]. OsNAC is induced by abiotic and biotic stress as well as ABA, Jasmonic acid and hydrogen peroxide [43, 44] and plays a role in activation of defensin [45]. DREB1B responds to variety of stress thus inducing expression of PR genes [46]. ERF1 is induced by both ethylene and JA [47, 48]. It integrates JA and ethylene signaling [49] and both signaling pathways are required simultaneously [50] for ERF1 induction. The tight co-regulation of ERF1 by both stress hormones is quite understandable considering the fact that of all the total number of defense related genes induced by ethylene and JA about 80% are ERF1 mediated. There are also reports of ERF1 being under MAPK cascade control mediated by EIN2 and EIN3 [51]. The most important genes up-regulated by ERF1 are basic chitinase, defensins and glutathione synthases (AK099489, AK063796) which have been found to be up-regulated in present study as well (Table 1). Ethylene and JA play a pivotal role in plant defense, plants with impaired ethylene signal transmission exhibit susceptibility to necrotrophic bacterium [52]. Ethylene induces cell-wall strengthening, xylem occlusion response, phenylpropanoid derived phytoalexin production, PR-genes including Beta-glucanase and chitinase [53] both of which have been found to be up-regulated in the present study. While Ethylene and JA have synergistic functions ABA is antagonistic to both, it interferes at different levels with ethylene and JA signaling [54]. The transcription factors found to be down-regulated were Auxin response factor 5, AUX/IAA and B3 transcription factor (Additional file 6). All three are growth related and are expressed in all organs, Auxin response factor 5 is required for normal growth [38] and B3 belongs to a family of plant transcription factors with various roles in plant development.

Metabolism

Secondary metabolites play an important role in plant defense [55]. The present study also documents similar findings; there was up-regulation of seven genes related to flavanoid biosynthesis (Table 1), three cytochrome P450 monoxygenases (AK071599, AK064764, AK107349), a flavanoid 3-monoxygenase (AK104472), A flavanone 3-hydroxylase-like protein (AK061337) [56] and a Glutathione S-transferase-GST 23 (AK099489) (Figure 2). GST proteins are known to act as escort proteins in flavanoid transport. WD-40 repeat like domain containing protein (AK109657) was down regulated (Figure 2). WD-40 repeat containing protein have been found to be necessary for anthocyanin biosynthesis at the DFR (Dihydroflavonol reductase) step in Arabidopsis leaves but does not seem to affect upstream genes involved in flavanoid bio-synthesis. It appears that flavanoid biosynthesis pathway is modulated to produce excess flavanoides rather than anthocyanins. The flavanoides may produce lignins to strengthen cell wall or phytoalexins the classical anti-microbial plant compound [57]. Transcripts for several well studied pathogen induced genes were found to be up-regulated these include: a chitinase (AK099355), a chitinase precursor Oschib (AK100973), an aspartic protease Os11g0183900 (AK061884) as well as two Harpin-induced 1 (Hin1) domain containing proteins (AK068115, AK108457) (Table 1). Chitinases are induced by environmental stress and considered to play a role in active or passive defense [58, 59]. Aspartic protease is a major family of protease enzymes, expression of an aspartic protease was found to be up-regulated in case of incompatible interaction between potato and the fungus Phytophthora infestans [60]. Hin1 is induced by bacterial effector, harpin through MAPK activity [61].

Genes related to lipid metabolism were also found to differentially regulated, a lipase (AK066720) (Additional file 5) was up-regulated whereas a gene related to lipid biosynthesis, an Acyl carrier protein (AK060566) was down-regulated (Table 2). The cell lipid metabolism perhaps is so diverted to provide ingredients and energy for mounting defense respose. A lipoxygenases (AK066825), chloroplast precursor of lipoxygenase LOX2 (Additional file 5), was also found to be significantly up-regulated. Lipoxygenases are key enzymes of lipid metabolism and JA biosynthesis [62]. LOX2 is required for wound induced JA accumulation and is involved in early defense response to pathogens [63]. The expression of LOX2 in turn is enhanced by JA through a positive feedback loop. The up-regulation of JA producing enzyme indicates the important role played by JA signaling in the Xanthomonas-rice incompatible interaction. Along with chloroplastid lipoxygenase other plastidial proteins including Tic32 (AK071972) and a protein similar to 1-deoxy-D-xylulose 5-phosphate synthase 2 precursor or DXPS (AK100909) were up-regulated (Additional file 5) as well. Tic32 is a NADPH-dependent dehydrogenase and its dehydrogenase activity is affected by Calmodulin. It is associated with Tic translocon on the stomatal side of the plastidial inner envelope. It may serve as a switch to differentially integrate redox signals from inside of chloroplast with calcium signals outside and influence the activity and/or specificity of Tic translocon [64, 65].

Defense involves induction as well as repression of several proteins. In order to meet the demand the cell modulates several components of transcriptional, translational and post-translational modification machinery. Present work documents up-regulation of Nuclease1(AK059608), a Ribosomal T2 family protein (AK061438), Ribosomal protein S8e domain containing protein (AK103678), Ribosomal protein S14 domain containing protein (AK107269) and a protein related to ARP6 (AK068395) (Table 1). ARP6 is a component of chromatin modifying complex implicated in maintaining state of gene activation [66]. While several others were down regulated (Table 2), these included TFIIH domain containing protein (AK062946), methyltransferase type 12 domain containing protein (AK111609), RNA-binding protein 8A-like protein (AK103277), pre-mRNA slicing factor SRP31 (AK071503), ribosomal protein S31 (AK060420) and an adenine salvage related protein APRT2 (AK069606). Transcripts for F-box/LRR-repeat MAX2 homolog (AK107102) and RING-H2 finger protein ATL8 (AK070038) were found to be up-regulated. F-box/LRR-repeat proteins function as substrate recruiting subunit of SCF-type Ubiquitin E3 ligases [67]. ATL is a multigenic family of putative RING-type E3 ubiquitin ligases [68], the specificity determinants that mediate the transfer of ubiquitin to the ε-amino group of target proteins resulting in mono-ubiquitination, additional ubiquitin moieties are transferred to the target protein by E4, a multiubiquitin chain assembly factor [69]. While multi-ubiquitination generally tag proteins for degradation, mono-ubiquitination of a target results in non-proteolytic events such as changes in protein activity, histone modification, localization or protein-protein interactions [70]. Incidentally beta 5 subunit of 20S proteosome (AK071652), the core complex of the 26S proteasome, was also found to be up-regulated implying modulation of ubiquitin mediated protein degradation.

Protein transport apparently has also been affected, the relocation of proteins to new sites for defense was evident by the down-regulation of B14 protein (AK061365) (Table 2) involved in peptide transport from ER to golgi and up-regulation of a DnaJ like protein (AK069823), two peptide disulphide isomerases (PDI) like proteins (AK068268, AK069583) and a peptide-proton symporter PTR2 (Table 1). DnaJ and PDI are molecular chaperones and rapid induction of PDI in wheat after fungal inoculation during early response has been previously documented [71].

Plant and bacterial culture

Seeds of bacterial blight resistant cultivar IET8585 [72] were obtained from Central Rice Research Institute Orissa, India and those of susceptible IR24 from Chinsurah Rice Research Station, West Bengal, India. Both varieties were grown on synthetic soil (Kaltech Energies, Karnataka, India) in a green house under 16 hrs light/8 hrs dark photoperiod at 28+/-2°C temperature.

The Xanthomonas oryzae pv oryzae culture Bxo43 [73] was obtained from Centre for Cellular and Molecular Biology, Andhra Pradesh, India.

Inoculation

Eighteen and fifty-five days old plants were inoculated with Bxo43 by clipping method [74]. Sterile water treated plants served as mock control. Leaf samples from eighteen days old seedlings were collected one hour after inoculation (hai), flash frozen in liquid nitrogen and subjected to molecular analyses. Disease progression was studied in both adult plants and seedlings.

RNA extraction and microarray hybridization

RNA from leaves of eighteen days old seedlings of both inoculated and mock-inoculated samples was extracted using tri-reagent (Sigma-Aldrich, St.Louis, MO, U.S.A.) and purified by Qiagen RNeasy Maxi Kit (Qiagen, Valencia, CA, U.S.A.) following manufacturers' instructions. The quality and purity of RNA was analyzed using spectrophotometer (Nano-Drop, Wilmington, DE, U.S.A.) and Agilent 2100 Bioanalyzer (Agilent Tecnologies, Santa Clara, CA, U.S.A.). Total RNA (200ng) was labeled with Cy5 or Cy3 using an Agilent Quick Amp Kit (Agilent Technologies).The amplified products were purified using Qiagen RNeasy Mini Kit (Qiagen), the recommended amount, 825 ng of each of the labeled products were used for array hybridization. Labeled targets of resistant and susceptible genotypes similarly treated (inoculated or mock inoculated) were hybridized to the same Agilent 44K custom oligo DNA microarray G2519F (Agilent). Dye-swap procedure was followed for two independent biological replicates (Additional file 8). Hybridization and wash processes were performed according to the instructions of the manufacturer. Microarrays were scanned using an Agilent Microarray Scanner (G256CA) at recommended settings (Additional file 9).

Data analysis

Data from each of the four arrays was extracted using Agilent Feature Extraction 10.5.1.1 software following protocol recommended by the manufacturer. Raw data was exported to Genespring GX11 (Agilent Technologies). Signals were background corrected and baseline transformed to the median of all spots. The data was log2 transformed and normalized to 75^th percentile using Loess normalization. The log2 ratios were averaged for replicate spots. Saturated spots and oligonucleotides with more than fifty percent replicate spots flagged as absent were excluded from analysis. Differentially expressed genes were identified using Students unpaired t-test with a corrected p-value of < = 0.05 and fold change of two or above. Gene interaction pathways were generated with the help of the software Pathway Studio 7.1 (Ariadne Genomics,Rockville, MD, U.S.A.).

Real-time qRT-PCR

RNA from independent biological replicate was used to synthesize cDNA employing Fermentas Revert Aid H minus first strand kit (Fermentas Life Sciences, Glen Burnie, MD, U.S.A.). Fifteen genes were randomly selected from among those that showed a significant up- or down-regulation in response to treatments. Specific primers (Additional file 10) were designed from the selected genes employing Primer3 software [75] and by comparison and alignment with available rice gene sequences from NCBI and Rice Annotation Project Database (RAP-DB). Actin [76] and Ubiquitin-conjugating enzyme E2 [77] were used as internal controls. PCRs were carried out in Bio-rad iQ5 Multicolor Real-Time PCR Detection System (Bio-rad Laboratories, Hercules, CA, U.S.A.) using iQ Syber Green Supermix (Bio-rad Laboratories) (Additional file 9). Quantification was based on cycle threshold (Ct value) and PCR efficiency determined by iQ5 Optical System Software 2.0 (Bio-rad Laboratories). The expression of each gene was normalized with internal controls and relative fold change was calculated using 2^-ΔΔCt method [78].