The OsOXO2, OsOXO3 and OsOXO4 Positively Regulate Panicle Blast Resistance in Rice

Background Although panicle blast is more destructive to yield loss than leaf blast in rice, the cloned genes that function in panicle blast resistance are still very limited and the molecular mechanisms underlying panicle blast resistance remain largely unknown. Results In the present study, we have confirmed that the three Oxalate oxidase (OXO) genes, OsOXO2, OsOXO3 and OsOXO4 from a blast-resistant cultivar BC10 function in panicle blast resistance in rice. The expression of OsOXO2, OsOXO3 and OsOXO4 were induced by panicle blast inoculation. Subcellular localization analysis revealed that the three OXO proteins are all localized in the nucleus and cytoplasm. Simultaneous silencing of OsOXO2, OsOXO3 and OsOXO4 decreased rice resistance to panicle blast, whereas the OsOXO2, OsOXO3 and OsOXO4 overexpression rice plants individually showed enhanced panicle blast resistance. More H2O2 and higher expression levels of PR genes were observed in the overexpressing plants than in the control plants, while the silencing plants exhibited less H2O2 and lower expression levels of PR genes compared to the control plants. Moreover, phytohormone treatment and the phytohormone signaling related gene expression analysis showed that panicle blast resistance mediated by the three OXO genes was associated with the activation of JA and ABA signaling pathways but suppression of SA signaling pathway. Conclusion OsOXO2, OsOXO3 and OsOXO4 positively regulate panicle blast resistance in rice. The OXO genes could modulate the accumulation of H2O2 and expression levels of PR gene in plants. Moreover, the OXO genes mediated panicle blast resistance could be regulated by ABA, SA and JA, and may be associated with the activation of JA and ABA signaling pathways but suppression of the SA signaling pathway. Supplementary Information The online version contains supplementary material available at 10.1186/s12284-021-00494-9.


Background
Rice (Oryza sativa L.) is a major food crop for more than half of the world's population. The demand is increasing with growing population. However, rice production is facing many challenges. Rice blast disease, caused by Magnaporthe oryzae (M. oryzae) is one of the most destructive diseases in rice plant, causing 10% -30% of yield loss every year (Skamnioti and Gurr 2009). Use of resistant variety is considered to be the most economical and environment-friendly approach to solve this problem . Rice blast can be classified into leaf blast and panicle blast based on the infected parts, and generally panicle blast is more destructive in terms of yield loss (Sirithunya et al. 2002;Zhuang et al. 2002;Liu et al. 2016a;Liu et al. 2016b). Recently, more and more studies have shown that the correlations between leaf blast and panicle blast are not always positive, and there are different regulation mechanisms between leaf blast resistance and panicle blast resistance (Zhuang et al. 2002;Liu et al. 2016b;Fang et al. 2019). But, nowadays, almost all the work in rice blast is focused on leaf blast, and few genes for panicle blast resistance have been cloned (Liu et al. 2016a;Liu et al. 2016b;Inoue et al. 2017). Besides, our knowledge on the mechanism of panicle blast resistance is still very limited. Since blast disease may occur in different developmental stages and the blast pathogen may infect different parts (leaf or panicle) in rice, it is necessary for effective disease control to identify the genes associated with panicle blast resistance and understand their regulatory mechanisms.
Oxalate oxidases (OXOs) which belong to the germin protein family have been demonstrated to play important roles in various environmental stresses in plants (Hu et al. 2003;Dong et al. 2008;Wan et al. 2009;Partridge-Telenko et al. 2011;Karmakar et al. 2016). OXO catalyzes the degradation of oxalic acid to produce hydrogen peroxide (H 2 O 2 ) and carbon dioxide. H 2 O 2 generated from the reaction can function as a secondary messenger to activate the hypersensitive response, the phytoalexin biosynthetic pathways as well as the expression of pathogenesis-related (PR) genes in plants (Hammond-Kosack et al. 1994;Lamb and Dixon 1997;Carter et al. 1998). Owing to the function of H 2 O 2 , OXO genes were speculated to play a key role in plant disease resistance. Indeed, many studies have indicated the involvement of OXOs in plant basal host resistance. For example, overexpressing a wheat OXO gene (gf-2.8) resulted in the induction of defense proteins and increased resistance to Sclerotinia sclerotiorum in sunflower (Hu et al. 2003). Transgenic oilseed rape plants overexpressing a wheat OXO gene (pSBGer2) exhibited enhanced resistance to Sclerotinia sclerotiorum (Dong et al. 2008). Recently, Yang et al. (2019) reported that overexpressing a wheat OXO gene (GenBank No M21962.1) also showed enhanced resistance to Sclerotinia stem rot in Glycine max .
In rice plants, four OXO genes (OsOXO1-OsOXO4) with > 90% nucleotide sequence identity are identified on chromosome 3. They form a tandemly duplicated cluster and co-localize with a blast disease resistance QTL (Ramalingam et al. 2003;Wu et al. 2004). Existing studies have shown that OsOXO4 was expressed during rice-M. oryzae infection in leaf blast, and the expression of OsOXO4 increased earlier in blast-resistant cultivar Moroberekan than blast-susceptible cultivar Vandana (Carrillo et al. 2009). Moreover, overexpressing OsOXO4 driven by the green tissue-specific promoter and coexpression of OsCHI11 and OsOXO4 all showed increased resistance to sheath blight pathogen in rice (Molla et al. 2013;Karmakar et al. 2016). Although the four OXO genes (OsOXO1-OsOXO4) have been demonstrated to co-localize with a blast resistant QTL, their specific regulation roles in rice blast resistance have not been reported so far.
In the present study, we identified that the transcription levels of OsOXO2, OsOXO3 and OsOXO4 were significantly induced by panicle blast inoculation. To confirm their functions in panicle blast resistance, we performed sub-cellular localization, spatio-temporal expression and transgenic analysis. Our results showed that OsOXO2, OsOXO3 and OsOXO4 cloned from a blast-resistant line BC10 function as positive regulators of panicle blast resistance in rice. The OXO genes could modulate the accumulation of H 2 O 2 and expression levels of PR gene in plants. Moreover, the OXO genes mediated panicle blast resistance could be regulated by ABA, SA and JA, and may be associated with the activation of JA and ABA signaling pathways but suppression of the SA signaling pathway.

Results
The OXO Genes Exhibit Different Temporal and Spatial Expression Patterns and are Localized in Nucleus and Cytoplasm in Rice In our previous study, we have performed a microarray analysis using a blast resistant line BC10 (Liu et al. 2004) and discovered that the expression levels of the three OXO genes, OsOXO2, OsOXO3 and OsOXO4, were strongly induced by panicle blast inoculation within a period of 48 h (Table S1). To further confirm these results, we analyzed the expression patterns of the three OXO genes in panicles in BC10 plants after blast infection using quantitative RT-PCR in this study. The results showed that the expression of OsOXO2 in infected panicles was induced dramatically at 6 h, 12 h and 24 h after inoculation. The expression levels of OsOXO3 also increased significantly at 6 h, 12 h, 24 h and 48 h after blast inoculation. Similarly, the transcription level of OsOXO4 was also induced at all time points after pathogen inoculation (Fig. 1A). These results suggest the possible roles of the three OXO genes in regulating panicle blast resistance in rice.
To analyze the sub-cellular localization of the three OXO genes, we fused the coding region of them with the red fluorescent protein (mcherry) fragment under the control of the cauliflower mosaic virus 35S promoter and expressed the fusion proteins in rice protoplasts, respectively. Laser confocal microscopy showed that the red fluorescent emitted by OsOXO2-mcherry, OsOXO3-mcherry and OsOXO4-mcherry fusion proteins were located in nucleus and cytoplasm (Fig. 1B).
To investigate the temporal and spatial expression patterns of the three OXO genes in rice plants, we analyzed the transcription of them in various tissues of Nipponbare by quantitative RT-PCR using EF1α as the internal Fig. 1 The temporal and spatial expression patterns of OsOXO2, OsOXO3 and OsOXO4 and their subcellular localizations. A. The expression of OsOXO2, OsOXO3 and OsOXO4 were assessed by quantitative RT-PCR at 6 h, 12 h, 24 h, 48 h after panicle blast inoculation. Error bars indicate the standard deviation (SD) from three biological replicates and ** indicates a statistically significant difference compared with mock (t test, P < 0.01). B. Sub-cellular localization of OsOXO2, OsOXO3 and OsOXO4 in rice protoplasts. Bar = 10 μm. C. Expression analysis of OsOXO2, OsOXO3 and OsOXO4 in different rice tissues by quantitative RT-PCR and GUS activity of OsOXO4 in the panicles of Nipponbare. Error bars indicate the standard deviation (SD) from three biological replicates. a, panicle at the early booting stage; b, panicle at the booting stage; c, panicle at the initial heading stage control. As shown in Fig. 1C, the expression of OsOXO2 was almost hard to be detected in all rice tissues examined. OsOXO3 expressed in callus, leaf and panicle, with the highest expression level in callus, but its background expression level is relatively low. Moreover, OsOXO4 expressed in all rice tissues examined, with a relatively higher expression levels in stem, leaf and panicle. To further confirm the high expression level of OsOXO4 in panicles, we generated transgenic Nipponbare plants in which the expression of a β-glucuronidase (GUS) was driven by the promoter of OsOXO4. Strong GUS activity was detected in panicles at the booting stage and heading stage (Fig. 1C), agreeing well with the result from quantitative RT-PCR.

Overexpression of OsOXO2, OsOXO3 and OsOXO4 Enhances Panicle Blast Resistance in Rice
To confirm the function of the three OXO genes in panicle blast resistance in rice, we generated the OsOXO2, OsOXO3 and OsOXO4 overexpressing plants in the japonica Nipponbare background, named OEOXO2, OEOXO3 and OEOXO4, respectively. The transgenic plants showed no obvious differences in non-target traits compared to Nipponbare plants and were fertile. Three independent homozygous lines of each OXO gene were used for panicle blast resistance evaluation. Quantitative RT-PCR analysis showed that the transcript levels of the three OXO genes were significantly increased in their corresponding transgenic lines (Fig. 2a). Cottonwrapping inoculation with M. oryzae isolate GD08-T13 showed that the infected main axis length was 82.1% for Nipponbare plants and 83.2% for the transformed empty vector control (PHQSN) plants, but 31.46% to 35.25% for the OXO overexpressing plants (Table 1, Fig. 2b). The enhanced panicle blast resistance was correlated with the increased expression of OXO genes in all transgenic plants (Fig. 2c). However, there were no significant difference in diseased leaf area, lesion size and fungal biomass between the OXO overexpressing plants and the control plants (PHQSN) after blast inoculation (Table  S2, Fig. S1).
OsOXO2, OsOXO3 and OsOXO4 Silencing Rice Plants are More Susceptible to Panicle Blast To further confirm the function of OXO genes in panicle blast resistance, we generated the OXO silencing plants in Nipponbare using an RNAi vector containing a 327 bp homologous coding sequence among OsOXO2, OsOXO3 and OsOXO4. We obtained 13 silenced transgenic lines, which showed no differences in the nontarget traits compared to the wild type Nipponbare and were fertile. Quantitative RT-PCR analysis showed that the expression levels of OsOXO2, OsOXO3 and OsOXO4 was remarkably decreased compared to the control plants (PHQSN) (Fig. 3a). The OXO silencing plants exhibited reduced resistance to panicle blast, with higher percentages of infected main axis length in both T 0 and T 2 generation plants when compared to the control plants (Table 2, Fig. 3b). The reduced expression levels of OsOXO3 and OsOXO4 were correlated with increased percentages of infected main axis length in the silencing plants (Fig. 3c). However, there were no significant difference in diseased leaf area, lesion size and fungal biomass between the OXO silencing plants and the control plants (PHQSN) after blast inoculation (Table S3, Fig.  S1). The results from gene overexpressing and silencing experiments suggest that OsOXO2, OsOXO3 and OsOXO4 positively regulate panicle blast resistance in rice.

OsOXO2, OsOXO3 and OsOXO4 Modulate the Expression of Defense-Related Genes
It is well-known that PR genes play important role in plant defense responses (Kaur et al. 2017). Previous study reported that the expression levels of PR genes were significantly increased in OsERF83 overexpressing plants which showed enhanced rice blast resistance (Tezuka et al. 2019). In this study, the expression levels of several PR genes were analyzed in the OXO transgenic plants and control plants (PHQSN). The results showed that the expression levels of all the chosen PR genes were up-regulated in OEOXO3, and all the others except PR10 were up-regulated in OEOXO4 plants, while only the transcriptions of PR1b, PR8, PR9 and PR12 were upregulated in OEOXO2 compared to the control plants. Moreover, all the nine PR genes were down-regulated in OXO silencing plants compared to the control plants, suggesting that the three OXO genes may mediate different PR-involved disease resistance regulatory pathway in rice (Fig. 4).

The OsOXO2, OsOXO3 and OsOXO4 Influence the Endogenous Levels of H 2 O 2 in the Transgenic Plants
Previous studies have shown that higher levels of H 2 O 2 contribute to OXO genes mediated fungal resistance in plants (Lamb and Dixon 1997;Carter et al. 1998;Wan et al. 2009). To confirm if the three OXO genes mediated panicle blast resistance is associated with the altered endogenous levels of H 2 O 2 , we firstly measured the H 2 O 2 concentrations in the panicles of control (PHQSN) and transgenic plants using the xylenol orange method (Kim and Hwang 2014). The results showed that the H 2 O 2 concentrations were significantly higher in the OXO overexpressing plants, but were lower in the OXO silencing plants when compared with the control plants ( Fig. 5a). Similar results were also observed in leaves using the DAB staining method (Thordal-Christensen et al. 1997). Brown staining was observed in the leaves of both control and OXO overexpressing plants, but the staining was stronger in the OXO overexpressing plants than in control plants. No visible staining was observed in the leaves of OXO silencing plants (Fig. 5b).
The Three OXO Genes Mediated Panicle Blast Resistance is Associated with Activation of the JA and ABA Signaling Pathways but Suppression of the SA Signaling Pathway To dissect the potential mechanisms of OXO genesmediated panicle blast resistance in rice, we firstly analyzed the cis-elements in the promoters of OsOXO2, OsOXO3 and OsOXO4, respectively. One ethylene (ET)-responsive element (ERE), two TC-rich repeats involved in defense and stress responses and one salicylic acid (SA)-responsive element (TCA-element) were found in the promoter of OsOXO2. Three abscisic acid (ABA)-responsive elements (ABREs), two methyl jasmonic acid (MeJA)-responsive elements (CGTCA-motif and TGACG-motif) and one TC-rich repeats were identified in the promoters of both OsOXO3 and OsOXO4 (Table  S4). The existence of TC-rich repeats further supported the regulatory roles of OsOXO2, OsOXO3 and OsOXO4 in disease resistance. Meanwhile, the presence of the hormone response elements in the promoter region Error bars indicate the SD from three biological replicates. ** indicates a statistically significant difference compared with PHQSN (t test, P < 0.01). b Disease phenotypes of OEOXO2, OEOXO3, OEOXO4, and the empty vector control (PHQSN) plants at the heading stage after inoculation with M. oryzae using the cottonwrapping method. c Correlation analysis between enhanced resistance and the expression levels of OsOXO2, OsOXO3 and OsOXO4 in the overexpressing plants after panicle blast inoculation. The correlation coefficient is calculated by linear regression implies that the expression of the three OXO genes could be regulated by these hormones. To confirm this inference, we treated the wild-type Nipponbare plants with exogenous ABA, SA, ET and JA at the three leaf stage, respectively. The expression levels of the three OXO genes in leaf were analyzed both before and after hormone treatments using quantitative RT-PCR. Unfortunately, the expression of OsOXO2 could not been detected. The expression of OsOXO3 and OsOXO4 were significantly induced by ABA, SA and JA treatment (Fig. 6). Transcription of OsOXO3 was significantly induced while the transcription of OsOXO4 was remarkably reduced by ET treatment (Fig. 6). These observations indicated that ABA, SA and JA may act upstream of OsOXO3 and OsOXO4 to induce their expression.
To further validate the relationship between the three OXO genes and hormone signaling pathways, we analyzed the expression patterns of several well-known stress-related genes that were involved in ABA, JA or SA signaling pathways in panicle in all the OXO transgenic lines, including ICS1 and NH1 which were related to the SA pathway (Deng et al. 2012), LOX2 and AOS2 which were related to the JA signaling pathway (Deng et al. 2012;Liu et al. 2016) and ABA signaling pathway related genes LEA3, NCED3, NCED4 and Rab16A Chen et al. 2015). Compared with the control plants, the transcription levels of ICS1 and NH1 were significantly decreased in OEOXO2, OEOXO3 and OEOXO4 plants. In contrast, the transcription level of LOX2 was increased in OEOXO3 and OEOXO4 plants, while the transcription of AOS2 was induced in all OEOXO2, OEOXO3 and OEOXO4 plants. The expression levels of LEA3 and NCED3 were significantly higher in OEOXO3 and OEOXO4 plants than in control plants, while the transcription levels of another two ABA-dependent pathway related genes NCED4 and Rab16A were significantly higher in all OEOXO2, OEOXO3 and OEOXO4 plants than in control plants (Fig. 7). Just contrary to the results of overexpressing plants, the expression of ICS1 and NH1 were significantly up-regulated in the OXO silencing plants compared to the control plants, and the expression levels of LOX2, AOS2, LEA3, NCED3, NCED4 and Rab16A were remarkably lower in the OXO silencing plants than in the control plants (Fig.  7). These results together suggest that the disease resistance conferred by OsOXO2, OsOXO3 and OsOXO4 could be associated with activation of JA and ABA signaling pathways while suppression of SA signaling pathway.

Discussion
OsOXO2, OsOXO3 and OsOXO4 Positively Regulate Panicle Blast Resistance in Rice Since oxalate oxidases were first isolated and characterized in barley and wheat, they have been reported to play important roles in disease resistance in many plant species (Hu et al. 2003;Livingstone et al. 2005;Welch et al. 2007;Dong et al. 2008;Walz et al. 2008;Barman and Banerjee 2015;Yang et al. 2019). The four OXO genes in cluster on chromosome 3 have been reported to be co-localized with a QTL for rice blast resistance (Carrillo et al. 2009). However, the other study showed that the transgenic plants overexpressing OsOXO1 or OsOXO4 of Zhonghua11 did not show improved resistance to rice blast disease . To further confirm the actual functions of the three OXO genes on blast resistance, we used BC10, a strong blast-resistant line (Liu et al. 2004) instead of Zhonghua11, a blast susceptible variety in the study conducted by Zhang et al. (2013) for cloning of the OXO genes in the present study. We identified that the expression of OsOXO2, OsOXO3 and OsOXO4 were induced by    (Table S2, Table S3 and Fig. S1). From these results, we can conclude that the three OXO genes (OsOXO2, OsOXO3 and OsOXO4) function as positive regulators to modulate panicle blast Error bars indicate the SD from three biological replicates and asterisks indicate statistically significant differences compared to the control plants (t test, **P < 0.01; *P < 0.05) resistance in rice. These results are different from the previous study . We believed that the conflict could be attributed to the use of different materials for cloning of the three OXO genes. To confirm this inference, we compared the sequences of the protein coding regions of the three OXO genes between BC-10 and Zhonghua11 (Fig. S2). The results revealed that there were several SNPs which lead to the changes of translated proteins. The changes of translated proteins could result in difference in blast resistance. The fact that the three OXO genes have positive functions in regulating panicle blast resistance but no effect in leaf blast resistance provides a new evidence for the difference between leaf blast resistance and panicle blast resistance in rice.
The Panicle Blast Resistance Conferred by OsOXO2, OsOXO3 and OsOXO4 is Associated with the Accumulation of H 2 O 2 and the Expression of PR Genes The enzymes encoded by OXO genes degrade oxalate acid to generate CO 2 and H 2 O 2 , the latter has been proved to play a key role in plant disease resistance through cell wall modification by cross linking of plant cell wall proteins in papillae at the infection sites (Olson and Varner 1993;Wei et al. 1998;Liu et al. 2016;Xu et al. 2018;Li et al. 2019;Pei et al. 2019). In the present study, our results showed that the H 2 O 2 concentrations in panicles were significantly higher in the OXO overexpressing plants, but were lower in the OXO silencing plants when compared with that in the control plants (Fig. 5a), suggesting that the panicle blast resistance conferred by OsOXO2, OsOXO3 and OsOXO4 was associated with the accumulation of H 2 O 2 . Interestingly, higher H 2 O 2 concentrations were also observed in the leaves of the OXO overexpressing plants whereas lower H 2 O 2 contents were also observed in the leaves of the OXO silencing plants compared to that in the control plants. However, there were no significantly differences in disease leaf area, lesion size and fungal biomass among the control, OE and RNAi plants (Table S2,  Table S3 and Fig. S1). Different disease reactions of the  (Hammond-Kosack et al. 1994;Carter et al. 1998). Here, we also identified that the expression levels of several PR genes were significantly up-regulated in OXO overexpressing plants (which harbors more H 2 O 2 ) when compared to the control plants. This is consistent with the previous report that the sunflower plants overexpressing a wheat OXO gene could regulate a number of defense-related genes after pathogen infection (Hu et al. 2003). However, it should be noted that the expression patterns of the PR genes were different among different OXO gene overexpressing plants. For instance, the induction of PR2, PR3, PR5 and PR5-1 were identified in the OEOXO3 and OEOXO4 plants but not in the OEOXO2 plants. The induction of PR10 was only observed in the OEOXO3 plants, and the induction of PR1b, PR8, PR9 and PR12 were identified in all the OEOXO2, OEOXO3 and OEOXO4 plants. These results implied that though OsOXO2, OsOXO3 and OsOXO4 all positively regulate panicle blast resistance in rice, their regulatory mechanisms may be different at least to some extent.
The Important Roles of Hormone Signaling Pathways in OsOXO2, OsOXO3 and OsOXO4 Mediated Panicle Blast Resistance in Rice Phytohormones are the well-known endogenous signal molecules that function in diverse biological processes including plant defense responses. Phytohormones generate and transmit distinct defense signals and the crosstalks between different hormones have been considered as universal defense responses employed by many plant species (Robert-Seilaniantz et al. 2011;Yang et al. 2013;Huot et al. 2014;Takatsuji and Jiang 2014;Yang et al. 2019). In this study, the crosstalks between JA, SA and ABA were identified for the panicle blast resistance conferred by OsOXO2, OsOXO3 and OsOXO4. We Error bars indicate the SD from three biological replicates and asterisks indicate statistically significant differences compared with water treatment (t test, **P < 0.01 and *P < 0.05) discovered that there were JA, SA and ABA response elements in the promoter region of the three OXO genes, and the expression levels of OsOXO3 and OsOXO4 were significantly induced by exogenous JA, SA and ABA. Furthermore, our results showed that the transcription of LOX2 and AOS2 which are involved in JA biosynthesis, LEA3, NCED3, NCED4 and Rab16A which are involved in the ABA signaling pathway were remarkably up-regulated in OXO gene overexpressing plants compared to the control plants. Reversely, the expression of ICS1 and NH1 which are involved in the SA signaling pathway were remarkably reduced in the overexpressing plants compared to the control plants. The expression patterns of these stress-related genes in the OXO Fig. 7 The expression of plant hormone related genes in panicles of the OXO transgenic plants and control plants by quantitative RT-PCR. Error bars indicate the SD from three biological replicates and asterisks indicate statistically significant differences compared to the control plants (t test, **P < 0.01 and *P < 0.05) silencing plants were just opposite to the results of the overexpressing plants. These results together suggest that the OXO genes mediated panicle blast resistance may be regulated by ABA, SA and JA, and associated with the activation of JA and ABA signaling pathways but suppression of SA signaling pathway.
In general, JA and SA are in most cases antagonized in regulating plant disease resistance (Robert-Seilaniantz et al. 2011). Our results here also indicated the antagonized roles between JA and SA in rice blast resistance. ABA has been well known for its regulatory roles in abiotic stress response and plant development. However, recently, more and more studies have discovered the important roles of ABA in regulating plant biotic stresses (Nambara and Marion-Poll 2005;Ton et al. 2009;Jiang et al. 2010;Jiang et al. 2013;Liu et al. 2018). For instance, ABA has been identified to positively regulate plant resistance to Alternaria brassicicola and Plectospharella cucumerina in Arabidopsis (Nambara and Marion-Poll 2005;Ton et al. 2009). Furthermore, our previous study also demonstrated that the transcription factor ONAC066 mediated leaf blast resistance is involved in suppression of ABA signaling pathway, indicating the negative role of ABA in leaf blast disease (Liu et al. 2018). Nevertheless, in this study, we found that the panicle blast resistance conferred by OXO genes is associated with the activation of ABA signaling pathway, suggesting the positive role of ABA in panicle blast disease. The opposite regulatory roles of ABA may also partially explain the differential mechanisms between leaf blast resistance and panicle blast resistance.

Conclusion
In conclusion, we have confirmed that OsOXO2, OsOXO3 and OsOXO4 from blast-resistant line BC10 positively regulate panicle blast resistance in rice in the present study. The OXO genes could modulate the accumulation of H 2 O 2 and expression levels of PR genes in plants. Moreover, the OXO genes mediated panicle blast resistance may be regulated by ABA, SA and JA, and associated with the activation of JA and ABA signaling pathways but suppression of SA signaling pathway. However, there are still some issues to be elucidated in the future study. For example, are there any additive effects among the three OXO genes (OsOXO2, OsOXO3 and OsOXO4). What are the common and different mechanisms of the three OXO genes in panicle blast resistance? Further studies are needed to address these issues.

Vector Construction and Rice Transformation
For overexpression vectors construction, the coding region sequences of OsOXO2, OsOXO3 and OsOXO4 were amplified from the blast-resistant line BC10 using the primers in Supplemental Table 1. The RNAi vector was generated by cloning a homologous sequence among OsOXO2, OsOXO3 and OsOXO4 using the primers in Supplemental Table 1. The resulting products were cloned into pEASY-T1 (TransGen) vector and verified by sequencing. The entry clones for overexpressing plants were then inserted into PHQSN (modified from pCAMBIA1390) which harbors a CaMV35S promoter. The clone for RNAi plants was constructed into pRNAi-Ubi, which was suitable for generation of hairpin-RNA constructs. For P OsOXO4 -GUS construction,~2.0-kb fragment was amplified from the upstream of OsOXO4 in BC10 genomic DNA using specific primers (Table  S1). Then the fragments were sub-cloned into pCAM-BIA1381Z. All the positive plasmids and control vectors were electroporated into Agrobacterium tumefaciens EHA105 and then introduced into calli of the cultivar Nipponbare via Agrobacterium-mediated genetic transformation.

Total RNA Extraction and Real-Time Quantification of mRNAs
The samples for total RNA extraction in the study of hormone treatments were collected from leaves, the other samples panicles for total RNA extraction were collected from panicles. Total RNA was extracted with Trizol reagent (Invitrogen) and purified with NucleoSpin RNA Clean-up (MACHEREYNAGEL) according to the manufacturers' instructions. RNA quality and quantity were assessed by formaldehyde denaturing agarose gel electrophoresis and spectrophotometry (Nanodrop-1000), respectively. The purified total RNA was reversetranscribed using the Primescript™ RT reagent kit (Takara) to generate cDNA and quantitative RT-PCR was carried out using SYBR ExTaq™ (Takara). EF1a gene was chosen as a reference gene. Gene expression was quantified by the comparative CT method. Experiments were performed in triplicate, and the results were presented by their means ± standard derivation (SD). T test was used for statistical analysis. Gene-specific primers used were listed in Table S5.

Sub-Cellular Localization Analysis
We amplified the protein coding region of OsOXO2, OsOXO3 and OsOXO4 from BC10 using the primers in Table S5 and cloned them into the pGY1-mcherry vector to generate the OXO-mcherry fusion proteins, respectively. The plasmids of OsOXO2-mcherry, OsOXO3-mcherry and OsOXO4-mcherry were extracted using UNlQ-50 Column Plasmid Max-Preps Kit (Sango). For the transient expression assay, 1 μg of every plasmid DNA was introduced into rice protoplasts. After 24 h incubation at 28°C without light, the rice protoplasts were observed and photographed under a laser confocal microscopy (Zeiss LSM710, Germany).

Phytohormone Treatments
Hormone treatments were performed using the same method as our previous study (Liu et al. 2016b). Mature seeds of Nipponbare were soaked in distilled water for 2 days and pre-germinated for 2 days at 30°C without light. Germinated seeds were placed in a salver for incubation in a growth chamber at 28°C, 70% relative humidity and 12 h photoperiod. When the seedlings grown to three-leaf stage, they were sprinkled with different plant hormone solution and distilled water (control), respectively. The concentration of each hormone solution was 100 μM. Sampling for RNA extraction was conducted at 0 h before treatment and 3 h, 6 h, 12 h, 24 h after treatment. The experiments were repeated twice.

Evaluation of Disease Resistance
We got T 0 transgenic plants by vector construction and rice transformation. T 1 and T 2 segregating progeny germinated from T 0 transgenic seeds were grown in soil in greenhouse. M. oryzae isolate GD08-T13 inoculum was used for blast resistance evaluation and was prepared as described by Beltenev et al. (2007). For leaf blast inoculation, 2 week seedings were inoculated by spraying with spore suspension of M. oryzae isolate GD08-T13 (1 × 10 6 spores/ml). Inoculated plants were maintained in a growth chamber (25°C, 16000 Lux, and 100% relative humidity) in the dark for 24 h; then, the growth chamber was set to a photoperiod of 16 h of light and 8 h of darkness at 25°C and 100% relative humidity. Disease was assessed 5 days after inoculation by measuring the disease leaf area percent. Each treatment was repeated three times. We also used the punch method as described by Ding et al. (2012) for leaf blast evalution. For panicle blast inoculation, cotton-wrapping inoculation method was used as described by Liu et al. (2016b). Briefly, we wrapped the upper-middle part of a panicle by cotton in 1-2 days after heading, and injected 2 ml spore suspension of 1 × 10 6 spores/ml of GD08-T13 into the cotton and then wrapped the cotton with foil. Each inoculated panicle was sprayed with water for 2-3 min every 2 h to maintain the humidity. Disease was assessed at 3 weeks after inoculation by calculating the percentage of infected main axis length (infected main axis length/main axis length of the inoculated panicle). Each treatment was repeated twice. T-test was used for test the significant level of difference.

The 3, 3′-Diaminobenzidine (DAB) Staining and Measurement of H 2 O 2
The DAB staining of H 2 O 2 was conducted according to the previously reported method (Thordal-Christensen et al. 1997;Kim and Hwang 2014). DAB was dissolved by sterile ddH 2 O and reduced PH to 3.8 with HCl to get the 1 mg/ml DAB staining solution. Similar leaves of the transgenic and PHQSN plants were selected and immersed into the DAB staining solution immediately. After 3 h incubation in a growth chamber at 28°C with relative humidity of 60% and light intensity of 50 μmol m − 2 s − 1 , the leaves were decolorized in a water-bath at 80-90°C for 1 h. Then, the decolorized leaves were photographed.
The measurement of H 2 O 2 was conducted using the hydrogen peroxide assay kit (Beyotime Institute of Biotechnology, China) as described in our previous study (Liu et al. 2016). The similar panicles of the transgenic and PHQSN plants were collected and grinded with liquid nitrogen. The 200 mL lysis buffer solution was added to the 10 mg panicle dry powder and blended fully. The supernatant was collected by centrifuging at 12,000 g for 5 min. Then, the 50 μL of the supernatants and 100 μL of test solutions were transferred to the test-tubes immediately and the mixed solution was kept for 30 min at room temperature. H 2 O 2 concentration was monitored by measuring the absorbance at 560 nm using a Thermo Scientific Multiskan Spectrum (Thermo, USA).