A Rice Gene of De Novo Origin Negatively Regulates Pathogen-Induced Defense Response

How defense genes originated with the evolution of their specific pathogen-responsive traits remains an important problem. It is generally known that a form of duplication can generate new genes, suggesting that a new gene usually evolves from an ancestral gene. However, we show that a new defense gene in plants may evolve by de novo origination, resulting in sophisticated disease-resistant functions in rice. Analyses of gene evolution showed that this new gene, OsDR10, had homologs only in the closest relative, Leersia genus, but not other subfamilies of the grass family; therefore, it is a rice tribe-specific gene that may have originated de novo in the tribe. We further show that this gene may evolve a highly conservative rice-specific function that contributes to the regulation difference between rice and other plant species in response to pathogen infections. Biologic analyses including gene silencing, pathologic analysis, and mutant characterization by transformation showed that the OsDR10-suppressed plants enhanced resistance to a broad spectrum of Xanthomonas oryzae pv. oryzae strains, which cause bacterial blight disease. This enhanced disease resistance was accompanied by increased accumulation of endogenous salicylic acid (SA) and suppressed accumulation of endogenous jasmonic acid (JA) as well as modified expression of a subset of defense-responsive genes functioning both upstream and downstream of SA and JA. These data and analyses provide fresh insights into the new biologic and evolutionary processes of a de novo gene recruited rapidly.


Introduction
To resist pathogen attacks, plants have evolved an efficient defense response system to elude or minimize the effects of diseases.Many genes are involved in this defense system.The major disease resistance (R) genes can directly or indirectly recognize the corresponding pathogen avirulence factors and trigger a highly specific resistance [1].The defense-responsive or defense-related genes also play an important role in the regulation of plant defense responses.Most of them encode the components of defense signal transduction pathways and provide nonspecific resistance to a broad spectrum of pathogens [2,3].An increasing number of R genes and defense-responsive genes encoding similar proteins have been identified as involved in plant-pathogen interactions in both dicots and monocots.The accumulated evidence, however, suggests that the defense responses to pathogen infection are not necessarily the same in dicots and monocots.For example, a rise in the level of endogenous salicylic acid (SA) is required for systemic acquired resistance in dicots, but rice (Oryza sativa L.), a model monocot, maintains a high level of SA without activating defense responses [4].Activation of indole-3-acetic acidamido synthetase can enhance disease resistance in both Arabidopsis (a model dicot) and rice, but the former is associated with an increase in SA and the latter is SA independent [5,6].Identification of species-specific genes will help to elucidate the differences in pathogen-induced defense responses among plant species.
A species-specific gene is the one that recently evolved or in rare cases the one that originated in the remote past but was lost in all other related species.A new R gene evolved a new and speciesspecific disease resistance trait.A number of mechanisms that can generate new genes are known [7].These mechanisms are often dependent on the duplication of ancestral genes or ancestral genomic sequences.Thus, the genes that originated from these mechanisms would have additional homologous gene copies in the same species or different species.However, recent studies in Drosophila melanogaster, D. simulans, D. yakuba, D. erecta, and Saccharomyces cerevisiae detected certain genes that have no homologs even in the species that diverged very recently [8][9][10][11][12].This raised the possibility of a de novo origination of new genes, to suggest the possibility that they may derive from ancestrally noncoding sequences or other unknown mechanisms [8,10], although there is much to do in understanding the biology of these de novo genes.It was speculated that many so-called ''orphan'' genes might actually be de novo genes [13]; http://blogs.nature.com/nature/journalclub/2007/10.The de novo evolution of genes is considered an important process in the genomes of some prokaryotes.Approximately 12% of genes in some archeal and proteobacterial species are thought to have de novo origination based on a computational analysis [14]; however, the origin of these de novo genes is unknown.Levine et al. [8] first reported de novo origination of whole genes from expressed noncoding regions in D. melanogaster and D. simulans.Other de novo genes from noncoding regions were then discovered in D. yakuba and D. erecta [9,10] as well as in S. cerevisiae [12].In comparison, the hydra gene in D. melanogaster was suggested to have evolved de novo from a DNA sequence that inserted into its present site; repetitive sequences in this region may have contributed to structural and expression-level evolution of this gene [11].Subsequently, we asked if plants have de novo genes.If they do, we can use the welldeveloped biologic analyses to further investigate their biologic processes to reveal how a function can be generated by a de novo gene.The defense system in rice is an excellent system to test this possibility because of the availability of rich analysis tools and knowledge about the disease resistance within the rice species.
Bacterial blight of rice, caused by Xanthomonas oryzae pv.oryzae (Xoo), is one of the most serious diseases of rice worldwide.Rice resistance to Xoo is governed by both R genes and resistance quantitative trait loci (QTLs).To date, six R genes for bacterial blight resistance have been isolated and numerous QTLs for bacterial blight resistance have been identified [15][16][17].However, only a few genes underlying the resistance QTLs are characterized [18].Our previous study showed that the cDNA clone BI71N2 (a fragment of Os08g05960, according to the rice genome annotation of The Institute for Genomic Research [TIGR], http://rice.tigr.org) showed suppressed expression in response to pathogen infection in different resistance-cultivated rice lines carrying an R gene for resistance to Xoo or Magnaporthe grisea, which causes fungal blast, another devastating rice disease worldwide [3].In addition, BI71N2 colocalizes with a previously identified resistance QTL in rice chromosome 8 [17].Those studies suggested that the gene represented by BI71N2 may be a negative regulator of disease resistance in a pathogen-nonspecific way, but the activation of R genes is the key to suppressing its expression.This finding also implies that suppression of the gene represented by BI71N2 in susceptible rice without the existence of an R gene may increase rice resistance to pathogen infections.
To test the above hypothesis, we monitored the expression of the gene represented by BI71N2, which we named OsDR10 (Oryza sativa defense-responsive gene 10).OsDR10 is an intronless gene and this gene and its homologs occur in the Oryza and Leersia genera of rice tribe, suggesting that it may have originated in the rice tribe.Suppressing OsDR10 mediated a broad-spectrum resistance to Xoo accompanying the activation of the SAdependent pathway and suppression of the jasmonic acid (JA)dependent pathway, suggesting that OsDR10 is a negative regulator of rice disease resistance.These data provided the first opportunity to examine how one of the most important plant traits -disease resistance -evolved with the origination of a young new gene.

Expression of OsDR10 in response to pathogens and phytohormones
Asian cultivated rice (Oryza sativa L.; AA genome) consists of two major groups, which are known by the subspecies names indica (Oryza sativa ssp.indica) and japonica (O.sativa ssp.japonica).To assess the effect of pathogens on OsDR10 expression, quantitative reverse transcription (qRT)-PCR was applied to examine the transcript level of OsDR10 in six different rice lines inoculated with Xoo strain PXO61.The indica cultivar Minghui 63, carrying two bacterial blight resistance genes, Xa3/Xa26 and Xa25(t), shows moderate resistance to PXO61 and indica cultivar Zhenshan 97, without any known R gene, is highly susceptible to PXO61 [19].IRBB4, carrying bacterial blight resistance gene Xa4, is a near-isogenic line with the genetic background of the susceptible indica variety IR24 and is highly resistant to PXO61 [20].MKbFZH2 is a transgenic line that overexpresses Xa3/Xa26 in the genetic background of susceptible japonica cultivar Zhonghua 11 and is highly resistant to PXO61 [21].The expression of OsDR10 showed a similar pattern in both resistant and susceptible rice lines on pathogen infection (Figure 1A).OsDR10 expression was first suppressed at 12 h in all rice lines except MKbFZH2 and then increased 1.8-to 8.1-fold at 24 to 72 h after PXO61 infection as compared to the corresponding uninfected plants.However, the expression level of OsDR10 in susceptible rice lines was obviously higher than that in resistant rice lines, with or without pathogen infection.Comparative analysis of the expression levels of OsDR10 in different rice lines showed that this gene had a similar expression level in the three resistant rice lines (Figure 1B), whereas the expression level in the three susceptible lines was approximately 3-to 10-fold higher than that in the resistant lines.These results suggest that a low expression level of OsDR10 is associated with disease resistance.
In comparison with untreated plants, wounding significantly suppressed (P,0.05)OsDR10 expression at 30 to 120 min after treatment (Figure S1).The phytohormones ethylene, JA, and SA, which are involved in pathogen-induced defense responses, also influenced the expression of OsDR10.Ethylene and SA first slightly induced and then suppressed OsDR10 expression, and JA suppressed OsDR10 expression as compared to the water (also as wounding) control.This result further suggests that OsDR10 is involved in rice-pathogen interactions.

Suppression of OsDR10 enhances resistance to bacterial blight
Comparative analysis of the genomic and cDNA sequences of OsDR10 (GenBank accession no.FJ194952) from indica cultivar Minghui 63 showed that OsDR10 was an intronless gene consisting of 617 nucleotides and encoding an unknown protein of 100 amino acids (Figure S2A).Approximately half (49%) of the amino acid residues that compose the OsDR10 protein are charged (Figure S3).Structural analysis of OsDR10 provided no clue as to the mode of action of the protein.
To determine whether OsDR10 had a phenotypic impact on rice disease resistance, the RNA interference (RNAi) strategy was used to suppress the expression of OsDR10 in Minghui 63, which is moderately resistant to Xoo strain PXO61 and susceptible to Xoo strain PXO99 [19].Twenty-three independent transgenic plants were obtained.Each T 0 transgenic plant was divided into two parts by separating the tillers at the tillering stage.The two parts were inoculated with PXO61 and PXO99, respectively, at the booting stage.Thirteen of the 23 plants showed significantly enhanced resistance to the two bacterial stains (P,0.01;Table S1).The lesion area of the 13 plants ranged from 10% to 24%, compared with 34% for the wild type after PXO61 infection, and from 16% to 45%, compared with 60% for the wild type after PXO99 infection (Table S1).The reduced lesion area of transgenic plants was highly correlated (r = 0.918, a = 0.01, n = 13 for PXO61 infection; r = 0.892, a = 0.01, n = 13 for PXO99 infection) with suppressed accumulation of OsDR10 transcripts; transgenic plants showing enhanced resistance had 3.1-to 12.5-fold lower OsDR10 transcripts than the wild-type plants (Figure 2A and Table S1).
To verify that the enhanced resistance of the transgenic plants was due to suppression of OsDR10, three T 1 families generated from independent resistant T 0 plants carrying a single copy of RNAi construct, D27RMH3, D27RMH15, and D27RMH17, were examined individually at the booting stage for resistance to PXO61 and the presence of the RNAi construct.All the T 1 plants showing significantly enhanced resistance (P,0.05) to PXO61 carried the RNAi construct, whereas other plants showing no significant difference (P.0.05) from wild type in response to PXO61 infection were free of the RNAi construct (Table S2).A T 2 family from a resistant T 1 plant D27RMH3-5 and a T 1 family from a resistant T 0 plant D27RMH17 were further analyzed at the booting stage for resistance to PXO61 and expression of OsDR10.
The enhanced resistance cosegregated with the reduced OsDR10 transcripts in the T 1 and T 2 families (Figure 2B).The reduced lesion area of the transgenic plants was highly correlated with reduced OsDR10 transcripts (r = 0.918, a = 0.01, n = 16).These results suggest that the enhanced bacterial resistance of the transgenic plants is due to suppressed expression of OsDR10.
The growth of Xoo strains PXO61 and PXO99 on resistant transgenic plants was 5-to 167-fold and 3-to 42-fold lower than that on corresponding wild-type control at 4 to 14 d after inoculation, respectively (Figure 3A).OsDR10-suppressed plants showed enhanced disease resistance not only at the adult stage but also at the seedling stage.Two T 2 lines (D27RMH3-5 and D27RMH17-23) from two independent T 0 plants were inoculated with PXO61 at the 5-or 6-leaf stage.They also showed enhanced resistance to Xoo as evaluated by lesion area (Figure 3B).The two transgenic lines were further examined for their resistant spectrum to different Xoo strains representing different races.Pathogen inoculation analysis demonstrated that these OsDR10-suppressed plants were also resistant to different Xoo strains (Figure 3C).Compared with the resistant rice line IRBB4, which carries the bacterial blight resistance gene Xa4, OsDR10-suppressed plants showed better resistance to Xoo strain PXO99 and similar resistance levels to strains PXO61, PXO86, and PXO341.

Suppression of OsDR10 affects transcript levels of a set of defense-responsive genes
The transcript levels of 12 defense-responsive genes, including those known to function in JA-and SA-dependent pathways, were examined in OsDR10-suppressed lines (D27RMH3-5 and Ten of the 12 genes showed differential expression either with or without pathogen infection (Figure 4).The 10 genes showed five types of expression patterns, considering the repetition of the two OsDR10-suppressed lines.PAL1 (phenylalanine ammonia-lyase 1, GenBank accession no.X16099), Cht1 (chitinase 1, D16221), and PR1a (acidic pathogenesis-related protein, AJ278436) showed 7-to 98-fold and 2-to 67-fold higher transcript levels in OsDR10suppressed plants than in wild-type plants before and after bacterial infection, respectively.Compared to the wild-type plants, the transcript level of PAD4 (phytoalexins deficient 4, CX118864) was 2-fold higher without pathogen infection and approximately 2-fold higher at some time points and 2-to 6-fold lower at other points after pathogen infection in OsDR10-suppressed plants.The expression of ICS1 (isochorismate synthase 1, AK120689), AOS2 (allene oxide synthase 2, AY062258), PR10, and OsWRKY13 (EF143611) was approximately 2-to 15-fold higher in OsDR10suppressed than in wild-type plants without bacterial infection, but there was a less than 2-fold difference in the two types of plants after bacterial infection.The transcript level of CHS (chalcone synthase, X89859) showed no significant difference (P.0.05) in OsDR10-suppressed and wild-type plants before bacterial infection, but it was 2-to 8-fold lower in OsDR10-suppressed plants than in wild-type plants after bacterial infection.The transcript level of LOX (lipoxygenase, D14000) was also not influenced (P.0.05) before bacterial infection, but was 2-fold higher at some time points and 2-to 3-fold lower at other points after pathogen infection in OsDR10-suppressed plants than in wild-type plants.The expression of NH1 (Arabidopsis NPR1 homologue 1, AY923983) and OsMPK6 (EF174189) showed no marked difference in OsDR10-suppressed and wild-type plants.

Suppression of OsDR10 enhances the SA level and suppresses JA level
To examine whether the modified expression of defenseresponsive genes caused by OsDR10 influences the endogenous levels of JA and SA, we quantified the concentrations of the two signal molecules in the leaves of OsDR10-suppressed T 2 lines.Pathogen infection significantly induced (P,0.05)JA accumulation in both OsDR10-suppressed plants (maximum 1.8-fold) and wild-type plants (maximum 5.7-fold; Figure 5).However, JA concentration was significantly lower (P,0.05) in OsDR10suppressed plants both before (approximately 1.5-fold) and at 72 h (5.8-to 6.5-fold) after bacterial infection than that in wildtype plants.Pathogen infection only slightly influenced SA accumulation in OsDR10-suppressed and wild-type plants (Figure 5), but SA concentration was significantly higher (P,0.05) in OsDR10-suppressed plants both before (1.4-to 1.9-fold) and at 24 to 48 h (1.2-to 1.5-fold) after bacterial infection than that in wild-type plants.The concentration of conjugated SA, SA bglucoside, was not influenced by the suppressed expression of OsDR10 (data not shown).

OsDR10 and its homologs are only detected in the species of rice tribe
There is one homolog (named OsDR10-NipponbareA and OsDR10-9311A in this paper) of OsDR10 in the genomes of japonica rice cultivar Nipponbare and indica 93-11 based on BLAST analysis against the whole genome sequences of the two cultivars [22], suggesting that OsDR10 is member of a small gene family in O. sativa (Table 1).A cDNA harboring the complete coding region of OsDR10-NipponbareA was identified to support its expression in Nipponbare (Table 1).In addition, another OsDR10-homologous cDNA (OsDR10-NipponbareB) from Nipponbare and another OsDR10-homologous genomic sequence (OsDR10-9311B) from 93-11, which putatively encode a same protein of 105 amino acids, were identified from GenBank (Table 1 and Figure S3).Several OsDR10 homologous cDNA or genomic sequences were also identified from japonica rice cultivar Nackdong, different accessions of common wild rice O. rufipogon (AA genome), and wild rice O. australiensis (EE genome) by BLAST analysis (Table 1).These OsDR10 homologous genes putatively encode proteins consisting of 99 to 106 amino acids (Figure S3).
OsDR10 and its product have no discernible sequence similarity to the genes and proteins from organisms other than rice, according to BLAST analysis of the nucleotide database GenBank and the protein database of the National Center for Biotechnology Information.However, BLAST analyses showed that sequences flanking OsDR10 were homologous to the sequences of other species.A 3490-nt sequence, located upstream of OsDR10 and 9912 nt away from OsDR10 and a 620-nt gene (Os08g05970), located immediately downstream of OsDR10 and 998 nt away from OsDR10 based on TIGR annotation, were homologous to the sequences of monocots maize and dicots Medicago truncatula (barrel medic) (Table 2).These results suggest that OsDR10 may be a rice tribe-specific gene.
To test this hypothesis, DNA gel blot analysis was performed to examine homologous sequences of OsDR10 in different plant species.The analysis showed that OsDR10 occurred in cultivated rice, common wild rice (O.rufipogon), and other wild rice, including O. punctata (BB genome), and O. latifolia (CCDD genome; Figure 6 and Figure S4).One to three hybridization bands were detected for the DNA from various rice accessions digested with different restriction enzymes under high-stringency hybridization.No homologous sequence was detected in other cereal species (maize, wheat, and foxtail millet) or dicot crops (cotton, potato, tomato, and rapeseed) under low-stringency DNA hybridization (Figure 6 and Figure S4).These results suggest that OsDR10 does not occur in the species outside of rice tribe Oryzeae and this gene has sequence diversity or a copy number difference in various rice species.
We also used PCR sequencing to detect OsDR10 homologs from the species that were tested by DNA gel blot analyses and their closer relatives.The beginning and end sequences of the coding regions of the sequences listed above are conserved (Figure S5).The conserved regions were used to design PCR primers to amplify OsDR10 homologs from different species of Poaceae (grass) family.Clear PCR product bands were detected after electrophoresis in other two wild rice species, O. punctate and O. latifolia, of Oryza genus, two Leersia species, L. tisserantii and L. JX , of Leersia genus, and the Bambusoideae species, bamboo, but not in wheat, barley, foxtail millet, maize, and sorghum.The Oryza and Leersia genera belonging to rice tribe Oryzeae and Leersia is the closest relatives of Oryza genus [23].Sequencing analysis of these PCR products showed that O. punctuate, O. latifolia, L. tisseranti, and L. JX but not bamboo contained OsDR10 homologs (Table 1 and Figure S5).The sequences from bamboo matched with only the primer sequences but not with OsDR10 (data not shown).These analyses further support the conclusion from the DNA gel blot and BLAST analyses: OsDR10 is a rice tribe-specific gene that may be generated by de novo evolution.

The evolution of OsDR10 was conservative
The OsDR10 gene and its homologs in different species of Oryza and Leersia genera share 67% to 100% sequence identity each other in the coding region (Table S3).These genes putatively encode nine types of proteins (Figure S3) that share 55% to 99% sequence identity and 57% to 99% sequence similarity with each other (Table S4).Phylogenetic analysis of the coding regions of these genes showed the similar evolution relationship as reported previously, in which the Leersia genus is more closely related to O. australiensis than to other species of the Oryza genus examined (Figure 7) [23].Thus OsDR10 from rice line Minghui 63 has the most distant phylogenetic relationship with Leersia in the coding region.The ratio of nonsynonymous (Ka) to synonymous (Ks) substitution is a measure of natural selection acting on the protein.
If there is no selective pressure (neutral evolution) on a gene, the ratio will be approximately 1.A Ka/Ks ratio ,1 indicates functional constraint with purifying selection on the gene, and a Ka/Ks ratio .1 indicates accelerated evolution with positive selection.The Ka and Ks of the coding sequences of OsDR10 and OsDR10-L.JX are 0.1331 and 0.3361, respectively, for a ratio of 0.3960 (P = 0.0126,0.05).Within the Oryza genus, OsDR10 has the highest sequence diversity from OsDR10-O .punctata.The Ka

Discussion
OsDR10 is a rice tribe-specific gene OsDR10 is a new gene and may have originated de novo.This hypothesis is supported by the following evidence.First, OsDR10 or its homologs were detected only in Oryza species and Leersia species, the closest relative of Oryza genus, based on sequencing, BLAST and DNA gel blot analyses, whereas the sequences homologous to the flanking sequences of OsDR10 were identified in other plant species.Furthermore, the close relative of rice tribe, the Bambusoideae subfamily [24], did not carry OsDR10 homologs.These results suggest that OsDR10 may originate in the rice tribe and is a new rice-specific gene.Second, the predicted OsDR10 protein and its homologs do not have sequence homology with any known, expressed, or hypothetical proteins.Third, OsDR10 is a single-exon gene, and it has been reported that the coding region of recently evolved de novo genes in D. melanogaster are not interrupted by intron [8].
An alternative interpretation with the de novo origination would be that the rapid sequence evolution of this gene could have lost identity with the homologous genes, if they exist, in non-Oryza species.We tested this hypothesis by considering the age of Oryza and the age of maize.The age of the rice tribe is around ,14 million years [25,26].Given the average substitution rate for grasses (5.9610 29 substitutions per synonymous site per year) [26][27][28] under the assumption of the neutral constant substitution process, the divergence at synonymous sites in the evolutionary time of 2614 million years would be 16.5% between the rice genes and the hypothetical homologs in the grass branches from the most recent ancestors of the rice genus.This study showed that the replacement site evolved 2.5 to 3.9 times (3.2 of the median) lower than synonymous sites; thus, the divergence at the replacement would be 16.5/3.2= 5.2%.The divergence of the entire coding region would be expected to be 8% (the weighted average of the divergence at the two types of sites).Very conservatively, we can take maize as the most recent outgroup that does not have any homologs of OsDR10, which diverged from rice 50 million years ago [29].A similar estimation led to an expected divergence of 28% between maize and rice in the entire gene region.Even using the fastest known rate of substitution for grasses (7.0610 29 substitutions per synonymous site per year) [28] to test this hypothesis, the expected divergence between maize and rice in the entire gene region is 67%.Thus, we should have been able to detect the homologous sequences in maize with this level of similarity.But we did not observe this homologous sequence using the all three methods: the genomic Southern analysis, the BLAST analysis of the sequenced genomes of maize, and the PCR amplification, in contrast to an expectation that can be made given the highly constraint detected from this gene.Therefore, we rejected this alternative hypothesis of identity loss in evolution, in strong support of the conclusion of the de novo origination.
Our Southern DNA gel blot analysis suggests that there are one to three OsDR10 homologous genes in different rice varieties and wild rice accessions.Rice variety Minghui 63 and some common wild rice accessions contain only one copy of OsDR10, whereas other rice varieties, including Nipponbare, and wild rice accessions may contain two or three copies of OsDR10 homologs.This suggestion can be supported by DNA sequence analysis.The homologous rice genes of OsDR10 have been identified based on BLAST analysis.The sequences of OsDR10 and its homologous genes in different rice accessions do not contain the digestion sites of restriction enzymes used in our DNA gel blot analysis.OsDR10 is a suppressor of the NH1-uninvolved SAdependent pathway and an activator of the JAdependent pathway Although several whole genes of de novo origination have been identified outside plants, the roles of these genes in molecular regulation are largely unknown.It has been reported that spermatogenesis appears to be the functional target for newly evolved genes in the lineages of Drosophila.Several de novo genes have been reported to show male-specific expression in these species [8][9][10][11].The de novo gene BSC4 may be involved in the DNA repair pathway during the stationary phase of S. cerevisiae [12].
Our results suggest that OsDR10 is a negative regulator of rice defense responses against bacterial infection.Low expression level of OsDR10 is associated with bacterial resistance.Both R genes and Xoo influence OsDR10 expression.In the same genetic background, rice lines carrying an R gene showed a significantly lower level of OsDR10 transcripts than those without an R gene, suggesting that OsDR10 may be negatively involved in some R gene-mediated defense signaling transduction.In contrast, OsDR10 expression was induced by Xoo, indicating that OsDR10 is a potential target of pathogen infection.
JA and SA are important signaling molecules in induced disease resistance of rice [30,31].Our results showed that the enhanced disease resistance of OsDR10-suppressed plants was accompanied by increased accumulation of SA and suppressed accumulation of JA, suggesting that OsDR10 is a negative regulator of SA synthesis and a positive regulator of JA synthesis.SA can be synthesized via both the isochorismate pathway and phenylpropanoid pathway in Arabidopsis and putatively in rice as well [31][32][33].ICS1 and PAD4 are involved in SA synthesis in the isochorismate pathway [34].PALs belong to a large family and are key enzymes for SA synthesis in the phenylpropanoid pathway.OsWRKY13 is an activator of ICS1 and PAD4 in rice disease resistance [31].OsDR10 appears to inhibit the expression of OsWRKY13, ICS1, PAD4, and PAL1 when free of pathogen infection and only inhibits the expression of PAL1 after pathogen infection.However, OsDR10 is an activator of CHS, which functions on a branch of the phenylpropanoid pathway and catalyzes synthesis of defensive secondary metabolites using the intermediate of SA synthesis as a matrix [35].Our results suggest that OsDR10 inhibits SA synthesis, mainly via the phenylpropanoid pathway during defense responses.PR1a functions in an SA-dependent pathway in rice bacterial resistance [31].The increased Cht1 expression in OsDR10-suppressed plants was associated with accumulation of SA, suggesting that it may also function in an SA-dependent pathway.PR1a and Cht1 showed a similar expression pattern in both OsDR10-suppressed and wild-type plants, suggesting that the two genes function in the same SA-dependent pathway, which is suppressed by OsDR10.NH1 is both the sequence and functional ortholog of Arabidopsis NPR1, which is a key regulator of SA-dependent systemic acquired resistance [36] and functions downstream of OsMPK6 in bacterial blight resistance [37].The expression of NH1 and OsMPK6 was not influenced by suppressed expression of OsDR10.Thus, the SA-dependent pathway suppressed by OsDR10 is an NH1-independent pathway.LOX and AOS are two important enzymes in the biosynthesis of JA [30].The reduced accumulation of JA in OsDR10-suppressed plants was associated with suppressed expression of LOX and AOS2 after bacterial infection, suggesting that OsDR10 is an activator of the JA-dependent pathway and it functions upstream of JA.

OsDR10 has been functionally constrained during evolution
During the evolution of a genome, those genes required for or that facilitate the survival of a species are created and maintained, whereas unwanted genes are discarded.As a negative regulator of rice disease resistance, OsDR10 appears to be important for rice.This hypothesis is supported by the following evidence.First, OsDR10 originated after cereal crops diverged and the evolution of OsDR10 was conservative.Plant disease resistance signaling pathways consist of both active and negative factors.Elimination of the functions of negative regulators of disease resistance frequently results in abnormal plants [37][38][39], which suggests that functional balance of the activities of the two types of factors in disease resistance is essential for the normal activity of plants.OsDR10 contributes to this balance.Second, the open reading frame of OsDR10 has been maintained during evolution.At least six different sizes (95, 99, 100, 101, 105, and 106 amino acid residues) of OsDR10 protein and its homologs were identified.Comparative alignment of the coding regions of the OsDR10 gene and its homologs showed that all the deletions and insertions of the genes during evolution occurred in triplet or multiple triplets in one site, which did not shift the original reading frames (Figure S5).For example, 75 deletions/insertions in 12 sites were observed in the coding region of OsDR10-O.australiensis compared to that of OsDR10, but all the deletions/insertions showed reading frame maintenance.In comparison, 17 deletion/insertion sites were detected in the 39-and 59-untranslated regions of OsDR10 homologs compared to those of OsDR10, but 15 of the 17 sites showed deletion/insertion with one, two, four, five, seven, or eight nucleotides (Figure S6).
In conclusion, our study identified a novel gene of de novo origination in rice tribe.The encoding product of this gene functions as a negative regulator in the SA-dependent pathway to balance rice defense response induced by pathogen infection.These results provide fresh insights into the new biologic and evolutionary processes of a de novo gene recruited rapidly in plant.The present data also put OsDR10 as a tentative ''orphan'' gene for it is unknown how this gene appeared in the rice tribe.Further studies on the identification of low homologous sequences of OsDR10 from the other species of rice tribe may help us to disclose how this gene was generated, when the genomic sequence of all the species in rice tribe are available.Further studies of tissuespecific expression pattern of OsDR10 in the entire life cycle of rice will help us target the putative function of this gene in other physiologic or developmental processes in addition to disease resistance.
PCR amplification of OsDR10 homologous from different grass species was performed using PCR primers 87G8-R1 and 87G8-R2 (59-AAGGTACCATGGCGTTCTACAAGTACGG-39; underlined, KpnI digestion site) with different annealing temperatures to obtain PCR products.The PCR products were excided from agarose gel after electrophoresis and purified for sequencing.Each PCR product was sequenced at least three times.The consensus sequences were used for further analysis.

Pathogen inoculation
To examine the resistance of plants to bacterial blight disease, plants were inoculated with Philippine Xoo strain PXO61 (race 1) at the seedling stage and four Philippine Xoo strains, PXO61, PXO86 (race 2), PXO99 (race 6), and PXO341 (race 10), and two Chinese Xoo strains, Z173 and K-S-21, at the booting stage by the leaf clipping method [40].Disease was scored (3-5 leaves for each plant) as the percent lesion area (lesion length/leaf length) at 2 to 3 weeks after inoculation.The bacterial population in rice leaves was determined by counting colony-forming units [19].

Hormone treatments
Phytohormone treatments were applied according to the procedures described previously [31].Leaf segments approximately 2-cm long were cut from the fully expanded leaves of 21-d Minghui 63 and floated on 30 mL of solution containing 100 mM of SA, JA, or ethephon (an ethylene generator) in covered sterile Petri dishes.Leaf segment wounding by cutting and floating on deionized water served as a control.

RNA expression analyses
Aliquots (15 mg) of total RNA were used for RNA gel blot analysis [31].A 617-bp probe of OsDR10 digested from cDNA clone EI87G08 with restriction enzymes KpnI and BamHI was used for the hybridization.qRT-PCR was conducted using genespecific primers (Table S5) as described previously [31].The assays were repeated at least twice, with each repetition having three replicates; similar results were obtained in repeated experiments.Standard deviation was calculated for technical replicates.

Rice transformation
The RNAi vector of OsDR10 was constructed by inserting a fragment of cDNA digested from cDNA clone BI71N2 using restriction enzymes KpnI and BamHI into the pDS1301 vector (Figure S2B) [37].The recombinant plasmid was introduced into Agrobacterium tumefaciens strain EHA105 by electroporation.Agrobacterium-mediated transformation was performed using calli derived from mature embryos of Minghui 63 according to a published protocol [19].

Quantification of SA and JA
The leaves of rice plants at the booting stage were used to quantify endogenous SA and JA.The SA and conjugated SA samples were prepared and quantified using high-performance liquid chromatography as described previously [31].The JA samples were prepared as described previously [6], with some modifications.The leaf samples were purified and diluted with 300 mL methanol using the method described by Ding et al [6].
Derivatization of JA using ethereal diazomethane was accessional.Ethereal diazomethane was synthesized from N-nitrosomethyl urea following the method of Mu ¨ller et al [41].Ethereal diazomethane (150 mL) was added to the methanol redissolved samples for derivatization at room temperature for 15 min, followed by the addition of 40 mL 0.2 M acetic acid/methanol to counteract the excessive diazomethane.The samples were dried by evaporation with nitrogen gas at 35uC.Samples were dissolved in 30 mL hexane and a 1-mL aliquot of the sample was injected into the gas chromatograph/mass spectrometer system (GCMS-QP2010S, Shimadzu, Japan) under the conditions of 77uC for 1.2 min, 10uC/min to 180uC, and 20uC/min to 250uC.The quantitative data of JA and the internal standard dihydro-JA were obtained using the peak of the product ion 151 with retention times 12.613 min and 12.680 min, respectively.

DNA gel blot analysis
Rice total DNA was digested with BamHI, EcoRI, and HindIII, separated by electrophoresis on 0.8% agarose gels, and blotted onto nylon membranes.The cDNA of OsDR10 was used as probe for hybridization.DNA filters were hybridized at 55uC overnight and subjected to two low-stringency washes (10 min at room temperature and 5 min at 55uC) with a solution composed of 26SSC and 0.1% SDS.The same filters were also hybridized at 65uC overnight and subjected to two high-stringency washes (10 min at room temperature and 5 min at 65uC) with a solution composed of 26SSC and 0.1% SDS.The hybridization signals were detected using a fluorescent image analyzing system.

Analysis of gene evolution
Multiple-sequence alignment of the coding sequences of OsDR10 and its homologs was performed using the ClustalX program [42].The Ka and Ks were determined from the alignments with a modified Nei-Gojobori model in MEGA4 and the difference between Ka and Ks was tested using the Z-test embedded in MEGA4 [43].Table S2 Cosegregation analysis of enhanced resistance to Xoo strain PXO61 and existence of the RNAi construct in OsDR10suppressed T1 families Found at: doi:10.1371/journal.pone.0004603.s008(0.13 MB PDF)

Figure 3 .
Figure 3. OsDR10-suppressed plants showed enhanced resistance to Xoo at both adult and seedling stages.PXO61, PXO86, PXO99, PXO341, Zhe173, and KS-1-21 were different Xoo strains that had different levels of virulence to the wild-type plant Minghui 63.D27RMH3-5 and D27RMH17-23 are transgenic lines of T 2 generation.Each data point (A) or bar (C) represents mean (three technical replicates)6standard deviation.(A) Growth of Xoo strains PXO61 and PXO99 in leaves of OsDR10-suppressed and wild-type plants at booting stage.Bacterial populations were determined from three leaves at each time point by counting colony-forming units (CFUs).(B) OsDR10-suppressed plants showed enhanced bacterial resistance at the seedling stage.(C) OsDR10-suppressed plants showed broad-spectrum resistance to different Xoo strains at the booting stage.IRBB4, with bacterial blight resistance gene Xa4, is a near-isogenic line with the genetic background of the susceptible indica variety IR24.doi:10.1371/journal.pone.0004603.g003

Figure 4 .
Figure 4. Expression analysis of defense-responsive genes in OsDR10-suppressed plants by qRT-PCR.Samples were collected before (0 h) and after inoculation of Xoo strain PXO61.Each sample was from 5 to 8 plants.Bars represent mean (three technical replicates)6standard deviation.Circles indicate a significant difference (P,0.05) between noninoculated and inoculated plants and asterisks indicate a significant difference (P,0.05) between the OsDR10-suppressed plants and corresponding wild-type plants within the same treatment.doi:10.1371/journal.pone.0004603.g004

Figure 7 .
Figure 7. Phylogenetic analysis of the coding regions of OsDR10 and its homologs from different species.The tree was constructed by the neighbor-joining method.The numbers for interior branches indicate the bootstrap values (%) for 1000 replications.The scale at the bottom is in units of number of nucleotide substitutions per site.doi:10.1371/journal.pone.0004603.g007

Figure S1 Figure S6
Figure S1The effects of defense signal molecules on the expression of OsDR10.OsDR10 expression was analyzed by qRT-PCR in the leaves of rice cultivar Minghui 63.Each sample was from 10 to 20 plants.Bars represent mean (three technical replicates)6standard deviation.Asterisks indicate a significant difference (P,0.05) between the hormone treatment and the water (also as wounding) control, which also served as wounding treatment for the leaves were cut from plants and treated in vitro, at the same time point.Found at: doi:10.1371/journal.pone.0004603.s001(1.96 MB TIF) Figure S2 Schematic diagrams of the OsDR10 gene and its transformation construct.(A) OsDR10 gene.The coding region (black box), 59-and 39-untranslated regions (hatched boxes), translation start codon (ATG), and translation stop codon (TAG) are also indicated.The numbers indicate the base pairs of each substructure.(B) RNAi construct of OsDR10.RB and LB, right and left T-DNA border; GUS, b-glucuronidase gene; Hpt, hygromycin phosphotransferase gene; 35S, cauliflower mosaic virus 35S promoter; OCS, octopine synthase polyadenylation signal.Found at: doi:10.1371/journal.pone.0004603.s002(1.88 MB TIF) Figure S3 Alignment of OsDR10 protein sequence with its homologs.Dash indicates a gap.Found at: doi:10.1371/journal.pone.0004603.s003(2.03 MB TIF)

Table 1 .
OsDR10 gene and its homologs in different plant species.
a Identity to OsDR10 gene.The two numbers represent the identity of whole homologous region/identity to the coding region of OsDR10.b Length of sequence homologous to OsDR10 gene.doi:10.1371/journal.pone.0004603.t001