Rice (Oryza sativa L.) japonica variety “Kitaake” and transgenic Kitaake plants were grown at 28°C during the day and at 20°C at night with a light/dark cycle of 14/10 h (h) in a greenhouse facility at Sejong University in Korea. To grow rice plants in soil, rice seeds sterilized by heating at 65°C for 8 min were incubated in sterile water at 23°C for 7 days and then transferred to Baroker bed soil (Seoul Bio). Fully expanded leaves from 7-week-old rice plants were used for all experiments, unless otherwise indicated.

Arabidopsis thaliana ecotype Colombia-0 (Col-0) and transgenic Col-0 were grown on Sungro propagation mix soil (Sun Gro Horticulture) in a growth chamber under 16/8 h light/dark conditions at 23°C and 70% relative humidity. For stratification, Arabidopsis seeds sterilized with 70% ethanol and 20% sodium hypochlorite solution were incubated in sterile water at 4°C for 7 days.

Nicotiana benthamiana (tobacco) were maintained in Baroker potting soil (Seoul Bio)_in a growth chamber under 16/8 h light/dark conditions at 23°C and 70% relative humidity. Tobacco seeds sterilized with 70% ethanol and 40% sodium hypochlorite solution were germinated in Murashige and Skoog (MS) medium (Murashige and Skoog, 1962) at 23°C for 7 days and then transferred to soil.

Sequences of Xio1 homologous were downloaded from the entire set of available species in the Ensembl Plants database¹ (Zerbino et al., 2018). Of the 96 plant species occurring in the database, 11 belong to the genus Oryza. Xio1 protein sequence (LOC_Os09g13440.1) was used as a query against the available downloaded plant proteomes using BLASTp strategy (E value < 10^–6) and then against the available downloaded plant genomes using tBLASTn (E value < 10^–10). Afterward, multiple sequence alignments of protein sequences of Xio1 homologous were performed using the Mega7 software.

To generate a construct overexpressing Xio1 tagged with green fluorescent protein (Xio1-GFP) at its C-terminus in rice, full-length complementary DNA (cDNA) of Xio1 with a removed stop codon was amplified from the Kitaake cDNA library by the polymerase chain reaction (PCR) using Xio1-specific primers, CACC_09g13440_FL_F (5′- CACCATGTCGAAATGCACAGCAAATA-3′) and 09g13440_ FL_woSTOP_R (5′-TGCTGGGTTTCTTCTGCGCCC-3′). The PCR amplified product was cloned into pENTR/GW/TOPO^® vector (Invitrogen) to create an Xio1_woSTOP/pENTR. The overexpression construct, Ubi:Xio1-GFP/pC1300, was obtained by recombining the Xio1_woSTOP/pENTR construct into the gateway vectors Ubi:GFP/CAMBIA1300 (Rohila et al., 2006; Park et al., 2017) through a Gateway^® LR clonase^® II reaction (Invitrogen). The construct was also used for subcellular localization study in rice protoplasts.

As a positive control for accumulating ROS, full-length cDNA of rice salt intolerance 1 (OsSIT1) (Li et al., 2014), was amplified using OsSIT1-specific primers, CACC_OsSIT1_Short_FL_F (5′-CACCATGCTGACGAACGGCACGA-3′) and OsSIT1_FL_woS TOP_R (5′-TGCTCTCGCTCGAGGAATGTCA-3′). The final construct, Ubi:OsSIT1-GFP/pC1300, was obtained by recombining the OsSIT1_woSTOP/pENTR construct into the gateway vectors Ubi:GFP/CAMBIA1300 (Rohila et al., 2006; Park et al., 2017) through a Gateway^® LR clonase^® II reaction (Invitrogen).

To generate a construct overexpressing Xio1-GFP in Arabidopsis Col-0 and tobacco plants, the 35S:Xio1-GFP/pEG103 construct was generated by recombining the Xio1_woSTOP/pENTR construct into the gateway vector pEarleyGate 103 (pEG103) (Earley et al., 2006) through a Gateway^® LR clonase^® II reaction (Invitrogen).

Rice protoplasts were harvested from 2-week-old rice seedlings using a previously described method (Bart et al., 2006) with minor modification. Fusion construct (NLS-RFP) of the nuclear localization signal (NLS) from simian virus 40 large T antigen and red fluorescent protein (RFP) was used for artificial localization of RFP into the nucleus (Lee et al., 2001). The vectors, 326-GFP and 326-RFP, were used as free GFP and RFP control, respectively (Lee et al., 2001). Ubi:Xio1-GFP/pC1300 and control plasmids were transformed into rice protoplasts through polyethylene glycol (PEG)-mediated transformation as previously described (Bart et al., 2006). For transient expression in tobacco leaves, Agrobacterium tumefaciens strain GV3101 was transformed with the plasmid 35S:Xio1-GFP/pEG103 or control plasmid 35S:YFP/pEG104 (Earley et al., 2006). A. tumefaciens cells were cultured, harvested, and resuspended in an infiltration buffer [10 mM MES (Biobasic), 10 mM MgCl₂ (Biobasic), and 100 μM acetosyringone (Sigma)] (pH 5.7) at a final OD600 = 0.4. Samples were infiltrated into 4 to 5-week-old tobacco leaves with 1-ml syringe. Fluorescences in protoplasts and leaves were photographed using an Eclipse Ti fluorescence microscope (Nikon) fitted with an objective (200× or 400×). The filter sets used were C-FL-C FITC (excitation 465 to 495 nm) and C-FL-C TRITC (excitation 537 to 552 nm) (Nikon) to detect GFP/YFP and RFP, respectively.

Rice transformations were carried out as described previously (Chern et al., 2005). For PCR selection of transformants carrying the Ubi:Xio1-GFP/pC1300 construct, hygromycin phosphotransferase (hpt)-specific primers, HygB_For (5′-CGC AAGGAATCGGTCAATACA-3′) and HygB_Rev (5′-CTACAC AGCCATCGGTCCAGA-3′) or Xio1-GFP-specific primers, 09g13440si_292-311_F (5′-GATGTTGTTCTTGGCCAGGG-3′) and smGFP2_308-327_R (5′-ACGTGTCTTGTAGTTCCCGT-3′) were used.

For the construction of transgenic Arabidopsis plants, the 35S:Xio1-GFP/pEG103 and 35S:YFP/pEG104 constructs were transformed into Arabidopsis Col-0 using the floral dip method (Clough and Bent, 1998) mediated by the Agrobacterium tumefaciens strain GV3101. Transgenic Arabidopsis Col-0 seeds were plated in selective MS medium containing 10 mg/L glufosinate-ammonium (Sigma) for 10 days and then transferred to soil and grown in growth chamber. For PCR selection, Xio1-GFP-specific primers, 09g13440_336-355_F (5′-GTTGGAGGAGATACACAAGG-3′) and smGFP2_308-327_R (5′-ACGTGTCTTGTAGTTCCCGT-3′) and YFP primers, EG FP_For (5′-CTACCCCGACCACATGAAG-3′) and smGFP2_ 308-327_R (5′-ACGTGTCTTGTAGTTCCCGT-3′) were used.

Xanthomonas oryzae pv. oryzae (Xoo) Philippine race 6 (PR6) strain PXO99 (referred to as Xoo throughout text) was used to inoculate Kitaake. Xoo was grown on a peptone sucrose agar (PSA) plate [10 g/L peptone (Duchefa), 10 g/L sucrose (Duchefa), 1 g/L glutamic acid (Duchefa), 16 g/L agar (Biobasic) and pH 7.0] containing 20 mg/L cephalexin (Duchefa) for 3 days. Two leaves of each tiller were inoculated by Xoo suspensions (OD600 = 0.5) by the scissors-dip method (Song et al., 1995; Chern et al., 2005). Lesion lengths were measured at the indicated days after inoculation. To harvest Xoo for bacterial colony counts, the entire leaves inoculated by the scissors-dip method were harvested and cut into approximately 1 mm pieces with sterile scissors and incubated in 10 mL sterile water for 1 h. After a series of gradient dilutions, the extract was plated on PSA plates containing 20 mg/L of cephalexin (Duchefa). The colonies of each plate were counted after incubation at 37°C for 2 days.

To inoculate Arabidopsis, Pseudomonas syringae pv. tomato (Pst) strain DC3000 was cultured in King’s B (KB) broth [20 g/L peptone (Duchefa), 10 ml/L glycerol (Biobasic), 1 g/L K₂HPO₄ (Biobasic), 1.5 g/L agar (Biobasic), 1.5 g/L MgSO₄ (Biobasic), and pH 7.0 containing 30 mg/ml rifampicin (Duchefa) and 50 mg/L kanamycin (Duchefa)] at 28°C. Bacterial cells harvested by centrifugation were re-suspended in 10 mM MgCl₂ (Biobasic) solution to OD600 = 0.002. The inoculation was performed by a syringe infiltration method (Liu et al., 2015). Bacterial cell suspension was infiltrated into rosette leaves using 1-ml syringes without needle, and the inoculated plants were maintained in sealed containers to facilitate disease development under high humidity. For quantification of bacterial growth in Arabidopsis, five leaf discs from inoculated leaves were collected at indicated time points and homogenized in sterile water. After a series of gradient dilutions, the homogenates were plated on KB plates containing 15 mg/L agar (Biobasic), 30 mg/L rifampicin (Duchefa) and 50 mg/L kanamycin (Duchefa), and bacterial colonies were counted at 1 day after incubation at 28°c.

For Western blot analyses, total proteins were extracted from rice or Arabidopsis plants using a protein extraction buffer [6.8 mM Na₂HPO₄ (Biobasic), 3.2 mM NaH₂PO₄ (Biobasic), 150 mM NaCl (Biobasic), 2 mM EDTA (Biobasic), 200 mM NaF (Biobasic), 1% Triton X-100 (Biobasic), and 1 mM PMSF (Biobasic)]. Protein concentration was determined using a previously described method (Bradford, 1976). Proteins (100 μg) were loaded per well, separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and subsequently blotted onto nitrocellulose membrane (Bio-Rad) for immunodetection. For Xio1-GFP detection, anti-GFP antibody (B-2) SC-9996 (Santa Cruz Biotechnology) and horseradish peroxidase (HRP)-conjugated anti-mouse IgG antibody SC-2031, (Santa Cruz Biotechnology) were used as a primary and a secondary antibody at a final dilution of 1:1,000 and 1:10,000 for 2 h, respectively. GFP bands were visualized using the SuperSignal West Pico Chemiluminescent Substrate (Thermo).

For hormone and peptide treatments, detached leaf bioassays were conducted. Rice leaf sections (1.5 to 2.0 cm) were prepared from fully expanded leaves of Kitaake plants. The leaf tip and basal portion were discarded. Leaf sections were equilibrated overnight in 6-well cell culture plates (SPL Life Sciences) filled with sterile water under constitutive light. Then, 200 μM salicylic acid, 1 mM ethephon, or 100 μM methyl jasmonate were substituted with sterile water. A synthetic 21-amino acid derivative of tyrosine-sulfated peptide called RaxX-sY and inactive non-sulfated RaxX (RaxX) were synthesized (Anygene) and treated as described previously (Pruitt et al., 2015; Thomas et al., 2016). For Xoo treatment, smear inoculation method was carried out as described previously (Nozue et al., 2011). Three layers of gauze were soaked with Xoo suspension (OD600 = 0.8) and smeared onto leaves five times. To extract RNA, leaf sections were collected immediately in the indicated time points, frozen in liquid nitrogen, and stored in a deep freezer at 80°C.

Total RNA was extracted by TRIzol reagent (Invitrogen) and two micrograms of total RNA were reverse-transcribed using M-MLV reverse transcriptase (Promega) according to the manufacturers instructions. Expression of Xio1 and Xio1-GFP were confirmed by RT-PCR using 09g13440_F (5′-GCAAGA ACTACCTCGCCGTGGAC-3′)/09g13440_R (5′-CCGTGGGGA GCTGTTCTTGATAG-3′) and 09g13440_F/smGFP2_308- 327_R (5′-ACGTGTCTTGTAGTTCCCGT-3′), respectively. Expression of each gene was determined by PCR using a specific set of primers: pathogenesis-related protein 3 (PR3), PR3-1 (5′-CTTGGACTGCTACAACCAGA-3′) and PR3-2 (5′-CATTGTGGGCATTACTGATG-3′); PR10, PR10_RT_F (5′-CGCAGCTCACATTATCAAGTCAGA-3′) and PR10_RT_R (5′-GAAGCAGCAATACGGAGATGGATG-3′); lipoxygenases (LOX), LOX_F (5′-TGTACGTGCCGAGGGACGAG-3′) and LOX_R (5′-GCGAGCGTCTCCCTCGCGAACTC-3′); ascorbate peroxidase 8 (APX8), APX8-5′ (5′-TGGTCTGATGACCTC CTCTGA-3′) and APX8-3′ (5′-CATGAGCCATGACAACTAGA-3′); rice translation elongation factor 1 alpha (rEF1α), rEF1α1048F (5′-ACTGCCACACCTCCCACATTG-3′) and rEF1α1552R (5′-CAACAGTCGAAGGGCAATAATAAGTC-3′). Quantification of RT-PCR bands on raw gel images in the expression analysis was performed in ImageJ by drawing a rectangular region-of-interest within each band and calculating the gray value of each gene relative to one of rEF1α.

After isolation of rice protoplasts, Ubi:Xio1-GFP/pC1300 was transformed into the protoplasts. Ubi:OsSIT1-GFP/pC1300 and Ubi:GFP/pC1300 were used as a positive and a negative control, respectively. ROS production was measured in the transformed rice protoplasts using the luminol-based chemiluminescence assay described previously with minor modifications (Murphy and Huerta, 1990; Tiew et al., 2015). After transformation with each construct, 1 × 10⁵ protoplasts were suspended in each well of a 96-well plate and incubated at RT for approximately 16 h in the dark. 2 × luminol solution (5 mM luminol L-012, 0.25 mg/mL HRP) was added to each well, and luminol-dependent chemiluminescence was measured for 3 h with a signal integration time of 10 s on a luminometer (Berthold Mithras LB940).

Each sequence in this study can be found in the GenBank/EMBL database under following accession numbers: Xio1 (LOC_Os09g13440), Xio1 homologous [O. sativa ssp. indica (B8BE71), O. rufipogon (A0A0E0QPF5), O. meridionalis (A0A0E0EQK1), O. glumipatula (A0A0E0B139), O. barthii (A0A0D3H4Y5)], OsUDT1 (LOC_Os07g36460), PR3 (LOC_Os06g51060), PR10 (LOC_Os12g36380), LOX (LOC_Os08g39850), APX8 (LOC_Os02g34810), rEF1α (LOC_Os03g08020), OsSIT1 (LOC_Os02g42780), AtPR1 (AT2G14610), AtPR4 (AT3G04720), AtEIF4A1 (AT3G13920).

All assays described above were repeated at least three times. All data are presented as means ± standard error. Data was analyzed by a one-way analysis of variance (ANOVA) using Sigma Plot 11.0 (Systat Software Inc). Means with different letters in the figures indicate significantly different values at p < 0.05 or 0.001 according to the Tukey test.

Previously, we identified XA21-mediated defense-related genes induced by Xoo. One of them was predicted to encode an uncharacterized protein without any known domain² (Figure 1A) and cloned for further characterization. To identify its potential homologous using sequence similarity, we ran BLASTp and tBLASTn against the whole plant genome database from Ensembl Plants (see text footnote 1) (Supplementary Table 1). This strategy revealed that its potential homologs are present only in the species within genus Oryza (Figure 1B and Supplementary Figure 1). They were restricted to exclusively AA genomes species in the genus Oryza, and all hits were mapped to chromosome 9. No variants were identified outside of the genus in any other sequenced monocotyledon and dicotyledon genomes. Therefore, the putative uncharacterized protein was concluded to be a taxonomically restricted Oryza-specific orphan protein and designated as Oryza-specific Xoo-induced orphan 1 (Xio1).

To confirm XA21-dependent induction of Xio1 upon Xoo inoculation, its expression pattern was monitored in Kitaake lacking XA21 and transgenic Kitaake carrying XA21 driven by the maize ubiquitin (Ubi) promoter in the Kitaake background (XA21/Kit) (Park et al., 2010; Figure 2A). In RT-PCR analysis, Xio1 in Kitaake was barely detectable at all time points after Xoo inoculation. In contrast to the result in Kitaake, Xio1 in XA21/Kit was significantly induced at 3 days after Xoo inoculation. As a positive control for Xoo inoculation, the expression of the rice pathogenesis-related protein 10 gene (PR10) (McGee et al., 2001) was examined. Strong induction of the PR10 was observed at 1 day after Xoo inoculation and maintained up to 3 days in XA21/Kit only.

A tyrosine-sulfated peptide called RaxX-sY (required for activation of XA21-mediated immunity X, tyrosine-sulfated) secreted from Xoo is sufficient to activate XA21-mediated immune responses (Pruitt et al., 2015). Therefore, to assess whether Xio1 is specifically induced after XA21 recognizes the sulfated peptides, leaf discs prepared from XA21/Kit were treated with a synthetic 21-amino acid derivative of RaxX-sY or inactive non-sulfated RaxX (Figure 2B). Strong induction of PR10 gene was observed at 3 and 9 h after RaxX-sY treatment but not after RaxX treatment, indicating that sulfated peptide successfully triggered the XA21-mediated immune response in XA21/Kit. Transcripts corresponding to the Xio1 gene accumulated preferentially at 9 h after treatment with RaxX-sY but not RaxX. Taken together, our results suggest that Xio1 is a novel Oryza-specific orphan gene having unknown function during the XA21-mediated immune response.

In O. sativa ssp. japonica and O. glumipatula, possible alternative splicing forms were predicted and annotated in the databases of the Rice Genome Annotation Project³ and Ensemble Plants (see text footnote 1) (Supplementary Table 1). Five exons from Xio1 are spliced together, resulting in Os09g13440.1, which encodes the putative full-length Xio1 protein of 24.9 kDa (Supplementary Figure 2A). Based on a database prediction, the Os09g13440.2 transcript results from the alternative splicing of a 95-bp alternative exon within intron II of the Xio1 gene. The inclusion of the alternative exon in Os09g13440.2 provides the translation termination at the end of the exon, leading to a shorter Os09g13440.2 transcript encoding a protein of 21.8 kDa. To investigate whether both predicted alternatively spliced transcripts, Os09g13440.1 and Os09g13440.2, are expressed in japonica variety Kitaake, we performed RT-PCR analysis after treatment of RaxX or RaxX-sY in XA21/Kit plants (Supplementary Figure 2B). Primer sets specific to Os09g13440.1 and Os09g13440.2 were used (Supplementary Figure 2A). PR10 was significantly induced in RaxX-sY-treated XA21/Kit plants, indicating that RaxX-sY peptide treatment successfully induced the XA21-mediated immune response. Os09g13440.1-specific product of 490 bp was preferentially amplified from RaxX-sY-treated XA21 plants. However, although we tried several times, our attempt to amplify Os09g13440.2 failed. These results suggest that Os09g13440.1 is a major transcript of Xio1 in O. sativa ssp. japonica.

Xio1 is a basic protein with pI 10.08 of 230 amino acids not carrying a predicted N-terminal signal peptide. However, the prediction in IUPred3⁴ (Erdos et al., 2021) showed that C-terminal region of Xio1 with score above 0.5 could be regarded as disordered regions (red line) and disordered binding sites (blue line) (Supplementary Figure 3). In addition, it contains two putative conserved motifs, nuclear localization signals (NLS) and nuclear export signals (NES) predicted by the NLStradamus⁵ (Nguyen Ba et al., 2009) and NetNES 1.1⁶ (la Cour et al., 2004), respectively (Figure 1A). The putative motifs suggest that Xio1 can translocalize into the nucleus for its novel function. To determine its subcellular location experimentally, Xio1 was tagged with green fluorescent protein (GFP) at its C-terminus and transiently expressed in rice protoplasts using PEG-mediated transformation (Figure 3A). In rice protoplasts co-transformed with Xio1-GFP and control RFP constructs, all transformed cells displayed identical subcellular localization patterns, characterized by GFP fluorescence mainly located in the nucleus-like compartment but also in the cytosol at a lower level (Figure 3A, Xio1-GFP + RFP). When NLS-RFP construct encoding a nuclear localization signal (NLS)-tagged RFP was co-delivered as a nucleus marker, Xio1-GFP was co-localized with NLS-RFP, indicating the nuclear localization of the Xio1 protein (Figure 3A, Xio1-GFP + NLS-RFP). To confirm its location, Xio1-GFP was transiently expressed in tobacco leaf epidermal cells using Agrobacterium-mediated infiltration (Figure 3B). In infiltrated leaves, fluorescence microscopy detected Xio1-GFP in the nucleus and the cytosol, which is consistent with its localization observed in rice protoplasts.

To investigate the unknown function of Xio1, we analyzed the expression pattern of Xio1 in various tissues of 10- and 3-week-old Kitaake plants using RT-PCR (Figure 4A and Supplementary Figure 4). The transcripts of Xio1 were detected at relatively high levels in fully expanded leaves (4th leaf) from main-tiller, but at significantly low levels in stem. The expression of Xio1 in leaf tissue was also observed in 3-week-old plants. Xio1 was barely expressed in both roots of 10-week-old and 3-week-old rice. Overall, Xio1 displayed relatively high expression in mature leaf tissue.

The expression pattern of Xio1 was also examined after treatments with phytohormones of salicylic acid (SA), ethylene-releaser ethephon (ET), and methyl jasmonic acid (MeJA/JA), which play key roles in coordinating plant immune responses (Figure 4B). PR10 and PR3 were used as marker genes for SA/ET- and MeJA-dependent signaling pathways, respectively (Xu et al., 1996; Jwa et al., 2001; Sathe et al., 2019), and demonstrated that phytohormone treatments were successful in the detached leaf bioassay. Xio1 expression was induced to varying degrees by SA or ET (Figure 4B, left panel). Particularly, strong and rapid induction was observed by SA treatment. However, after MeJA treatment, Xio1 expression was repressed to barely detectable levels repeatedly (Figure 4B, right panel), suggesting that there are antagonistic impacts of SA/ET and MeJA in its expression.

In order to examine the molecular function of Xio1 in planta, Agrobacterium-mediated transformation was performed with the overexpression construct, Ubi:Xio1-GFP/pC1300. Numbers of putative transgenic Kitaake plants (Xio1-GFPox) were generated, and they were phenotypically normal in root-inducing MS medium. However, all putative transgenic plants displayed growth retardation and eventually died approximately 2 months after transfer to potting soils in greenhouse. Despite many attempts, we were able to neither maintain the transgenic plants nor harvest seeds from any of them. Alternatively, we decided to analyze the putative T₀ transgenic lines (Figure 5). Three putative independent transgenic Kitaake (Xio1-GFPox T₀, line 1, 2, and 3) were generated and moved to soils in the greenhouse. After 6 weeks, all of them displayed growth retardation compared to the wild type Kitaake plant (Figure 5A). The putative transgenic plants were subjected to PCR-based selection using the hpt gene-specific primer set, HygB_For and HygB_Rev. All of them generated a hpt-specific 555 bp amplicon (Figure 5B), indicating that transgenic lines carrying the overexpression construct were generated.

We next examined the transcript level of Xio1-GFP in the transgenic plants (Figure 5C). In RT-PCR analysis, Xio1-GFP-specific primers (Figure 5C, Xio1-GFP) for transgene Xio1-GFP were used to determine its overexpression. As expected, significant amounts of Xio1-GFP transcripts were detected in the transgenic lines (T₀), Xio1-GFPox 1 and 2, indicating that the Xio1-GFP transgene was constitutively overexpressed. In addition, we also investigated the expression levels of two defense-related marker genes, PR3 and PR10 (Xu et al., 1996; Jwa et al., 2001; McGee et al., 2001; Sathe et al., 2019), in the Xio1-GFPox lines (Figure 5C). Both PR3 and PR10 genes were significantly induced in the transgenic lines compared with the wild type Kitaake, suggesting that Xio1-GFP overexpression triggered defense response, regardless of Xoo inoculation. The transgenic lines (Xio1-GFPox T₀, line 1 and 2) were assayed for the accumulation of the fusion protein, Xio1-GFP, by Western blot analysis using the anti-GFP antibody (Figure 5D). In the two lines, a significant accumulation in Xio1-GFP was observed, compared to the wild type Kitaake. These results indicated that transgenic lines overexpressing Xio1-GFP were successfully generated and that the phenotype of growth retardation and induction of defense-related marker genes should be caused by Xio1-GFP accumulation.

Another three independent Xio1-GFPox lines (T₀, line 4, 5 and 6) were generated and all of them generated a Xio1-GFP-specific 792 bp amplicon (Supplementary Figure 5A). The transgenic lines displayed significantly increased Xio1 transcript and protein levels (Supplementary Figures 5B,C). Then, 6-week after transferring to soil in our greenhouse, the three Xio1-GFPox lines were inoculated with Xoo (Figure 6). They exhibited significantly enhanced resistance to Xoo with very short lesions of approximately 0.47–0.96 cm compared to Kitaake with lesion lengths of approximately 11.10–20.10 cm (Figures 6A,B). The bacterial population also correlated well with lesion length measurements (Figure 6C). Xoo populations in transgenic lines reached approximately 1.66 × 10⁵ – 1.75 × 10⁷ colony-forming units per leaf (cfu/leaf), whereas the population in Kitaake (Kit) plants reached to more than 1.29 × 10⁹ cfu/leaf. The inoculation experiment was repeated using different batch of transgenic lines, 7 and 8, with similar results (Supplementary Figure 6). These results demonstrated that Xio1 overexpression conferred enhanced resistance to Xoo under our greenhouse condition.

To minimize the possibility that unknown environmental stresses in our greenhouse, instead of Xio1 overexpression, induced stress-related responses in Xio1-GFPox lines, the next batch of five independent Xio1-GFPox lines (T₀, line 9 to 13) were investigated in an environmentally controlled growth chamber. The transgenics and Kitaake control were grown in sterilized glass container containing MS medium and maintained under growth chamber condition. Two weeks (line 9, 10, and 11) and 3 weeks (line 12 and 13) after transferring to MS medium, the expression patterns of defense-related genes were examined (Figure 7 and Supplementary Figure 7). In both stages, two PR genes, PR3 and PR10, were significantly induced in the transgenic lines. Interestingly, the lipoxygenase (LOX) gene encoding a lipid peroxidation-related enzyme also displayed high level expression in all Xio1-GFPox lines. In contrast, ascorbate peroxidase 8 (APX8) encoding an antioxidative enzyme was reduced. These results suggest that Xio1 overexpression elicited defense responses to Xoo by expressing various defense- and antioxidant-related genes directly or indirectly.

Because of the premature death of all Xio1-GFPox lines, it was impossible to further analyze the transgenic plants for investigating antioxidant enzymes, in conjunction with ROS accumulation. ROS homeostasis has been successfully analyzed in various protoplast systems including Arabidopsis, Cucumis sativus (cucumber), and Solanum lycopersicum (tomato) (Ondrej et al., 2010; Tiew et al., 2015; Gandhi et al., 2020). Therefore, to measure extracellular H₂O₂ production, we performed the luminol-based chemiluminescence assay in rice protoplasts. Rice salt intolerance 1 (OsSIT1) was used as a positive control in the assay because its overexpression prompted the accumulation of ROS in rice (Li et al., 2014). Xio1-GFP, OsSIT1-GFP, and GFP were successfully expressed in rice protoplasts transformed with each construct (Figure 8A). In Figures 8B,C, protoplasts expressing GFP control displayed slightly elevated H₂O₂, which is commonly observed in freshly isolated protoplasts (Tiew et al., 2015). In contrast, dramatically increased H₂O₂ was observed after transformation with OsSIT1-GFP, indicating that the luminol-based chemiluminescence assay in protoplasts worked successfully. Xio1-GFP expression was also significantly increased H₂O₂ production at a higher level than that caused by OsSIT1-GFP.

To investigate whether Xio1 plays distinct role in disease resistance in non-Oryza plant species, Xio1 was overexpressed in Arabidopsis Col-0. We generated 12 independently transformed transgenic Arabidopsis lines (Xio1-GFPox/Col-0, T₁) carrying 35S:Xio1-GFP/pEG103 construct (Supplementary Figure 8A). All of them displayed significant expression of Xio1-GFP (Supplementary Figure 8B). Of these, two lines (6 and 11) with relatively higher expression of Xio1-GFP were selected for further analysis. Western blot analysis performed with their progenies (6-1 and 11-13-1) exhibited high accumulation of fusion protein, Xio1-GFP (Supplementary Figure 8C). In the transgenic Arabidopsis, fluorescence from Xio1-GFP was detected in the nucleus and the cytosol (Supplementary Figure 9), consistent with previous observations in rice protoplasts and tobacco leaves (Figure 3). However, unlike in rice, Xio1-GFP overexpression in Arabidopsis did not cause any obvious morphological changes, including growth retardation (Supplementary Figures 10A,B) and increased expression of PR genes, AtPR1 and AtPR4 (Supplementary Figure 10C), when compared to non-transgenic Col-0 plants. In addition, after challenging with a virulent strain of Pseudomonas syringae pv. tomato DC3000 (Pst DC3000), they did not display any significant difference in disease symptom and bacterial population compared with non-transgenic Col-0 (Figure 9). This indicated that Oryza-specific orphan gene, Xio1, was not able to induce disease resistance in non-Oryza plant species, Arabidopsis.