IR24 is a susceptible indica rice variety without the Xa7 gene, while IRBB7 is a near-isogenic line of IR24 containing the Xa7 gene. A japonica rice variety Zhonghua11 (ZH11) does not contain the Xa7 gene locus was selected as the wild-type control and transgenic recipient. Transgenic lines constitutively expressing the Xa7 gene (Xa7-OE-1 and Xa7-OE-2) were obtained in our previous research (Chen et al., 2021). The detailed information on rice varieties used in this study is listed in Supplementary Table S3 . All rice plants were grown in natural paddy fields during summer in southern China. Tobacco (Nicotiana benthamiana) plants were cultured in a growth chamber at 25°C (16-hour light/8-hour dark).

A synthetic DNA fragment containing three tandem repeats of the predicted EBE_AvrXa7 sequence was inserted into the pLacZ vector (digested with NcoI) via homologous recombination. Subsequently, the engineered pLacZ construct was linearized with NcoI and used for the transformation of the yeast strain YM4271. Positive colonies were able to grow on a nutrient-deficient medium (SD/-Ura) (Breton et al., 2016). A vector expressing the fusion protein of the GAL4 activation domain (AD) and the RVDs of AvrXa7 was constructed using the pGADT7 vector and then transformed into YM4271 harboring the corresponding pLacZ vector. Positive colonies were screened on a nutrient-deficient plate (SD/-Ura/-Leu). β-Galactosidase activity was assessed using a commercially available luminescent β-galactosidase substrate Beta-Glo, which was cleaved and then released d-luciferin for detection using firefly luciferase.

Constructs of 35S::avrXa7, Xa7_pro::Xa7, Xa7_pro::Xa10, Xa7_pro::Xa23, and Xa7_pro::Xa27 were transformed into Agrobacterium tumefaciens strain EHA105 using the heat-shock method. The bacteria were cultured in 20 mL of Lysogeny Broth (LB) liquid medium with 50 mg/L kanamycin and 25 mg/L rifampin until the density reached OD_600 = 0.6. The qualified bacterial cells were then collected by centrifugation at 4°C for 10 min at 3,000 rpm and re-suspended in an inoculation medium containing 10 mM MgCl₂, 10 mM MES, and 200 μM AS. Leaves of 4-week-old tobacco were inoculated using needleless syringes following the method described by Kay et al. (2007). The inoculated leaves were harvested after 2 days of inoculation and stained using 3,3′-diaminobenzidine (DAB), as described by Daudi and O'Brien (2012).

The PrimeSTAR HS DNA Polymerase Kit (TaKaRa, Mountain View, CA, USA) was used to amplify the fragments of interest with specific primer pairs listed in Supplementary Table S4 . The PCR products were then inserted into the frame vector pCAMBIA1300 using an In-Fusion HD Cloning Kit (TaKaRa). The verified vectors were transformed into the rice japonica variety Zhonghua11 using the Agrobacterium-mediated method, as described previously (Nishimura et al., 2006). DNA sequences of used constructs are shown in Supplementary 5 .

The Xoo strains PXO86 and PXO99 were obtained from the International Rice Research Institute (IRRI) in the Philippines. The engineered strain PXO99 ^avrXa7 was created by our group in the previous research, which contained an avrXa7 overexpression construct in the PXO99 strain (Chen et al., 2021). Among them, PXO86 and PXO99 ^avrXa7 are incompatible strains of the Xa7 gene in the rice variety IRBB7, while PXO99 is a compatible strain of the Xa7 gene in IRBB7. The strains were initially cultured on agar medium containing 20 g/L sucrose, 5 g/L peptone, 0.5 g/L Ca(NO₃)₂, 0.43 g/L Na₂HPO₄, and 0.05 g/L FeSO₄ at 28°C for 2 days. The bacterial colonies were then eluted with sterile water and diluted to a concentration of OD_600 = 1.0 for subsequent inoculation. The leaf-tip clipping method (Kauffman et al., 1973) was used for inoculation of the Xoo strains, and lesions on the inoculated leaves were measured 2 weeks after inoculation to evaluate BB resistance (Yin et al., 2000).

During the amplification of the Xa7 promoter, we obtained a mutated Xa7 promoter, which has two cytosine bases deleted; as a result, the Xa7 gene driven by the mutated Xa7 promoter could not induced by AvrXa7. Thus, we chose this promoter as a native control. We inoculated the leaves of Xa7_pro::Gus and Xa7_mpro::Gus transgenic plants with PXO86 at the early stage of tillering, and we detected Gus activity by submerging the infected leaves in GUS-staining solution (Zhang et al., 2018).

Total RNA was extracted from rice leaves using the RNeasy Plant Mini Kit (Qiagen, Valencia, CA, USA) and was purified with the RNase-Free DNase Set (Qiagen) in accordance with the manufacturer’s instructions. Synthesis of first-strand cDNAs from the extracted RNA was performed using the M-MLV Reverse Transcriptase Kit (Promega, Madison, WI, USA) according to the manufacturer’s instructions. Real-time PCR (qRT-PCR) was carried out using the Fast SYBR Green Master Mix Reagent (Applied Biosystems, Foster City, CA, USA) on an ABI7500 Real-time PCR System. For qRT-PCR in rice, the housekeeping gene Actin was used as the internal control, while in N. benthamiana, the expression of the NbGAPDH was used as the internal control. The expression of target genes was quantified using the 2^−ΔΔCt method (Livak and Schmittgen, 2001). Standard errors were calculated based on a minimum of three biological replicates. The primers used are listed in Supplementary Table S4 .

The leaves of IRBB7 were inoculated with the positive control strain PXO86 (incompatible), the negative control strain PXO99 (compatible), and a mixed control strain (PXO86+PXO99) at the same time, while the remaining treatments were first inoculated with PXO86 strain simultaneously to induce the expression of the Xa7 gene and then inoculated with PXO99 strain at 1 day, 2 days, and 3 days after the initial inoculation, separately. One-month-old rice plants after transplantation were inoculated with Xoo strains at approximately 1 × 10⁹ cfu/mL using the leaf-clipping method described above.

Leaves of the transgenic plants of Xa7-OE-1 and Xa7-OE-2, as well as the wild-type control ZH11, were harvested with three biological replicates at the tillering stage (2 months old) and immediately frozen in liquid nitrogen. All samples were sent to the Beijing Genomics Institute (BGI) for RNA-seq using a HiSeq 2500 sequencing system (Illumina, San Diego, CA, USA). The raw data were processed to remove the adaptor sequences and low-quality reads by SOAPnuke (Kawahara et al., 2013). Then, the cleaning reads were mapped onto the reference genome sequence of Nipponbare (http://rice.plantbiology.msu.edu/) by Bowtie2 (Cock et al., 2010; Langmead and Salzberg, 2012). The transcriptional level of expression genes was calculated using RSEM (Li and Dewey, 2011). Differentially expressed genes (DEGs) were identified using DEGseq (Wang et al., 2010) with the Log₂FC ≥1 or ≤−1 and significance adjusted at p ≤ 0.05. Enrichment analysis of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways of the DEGs was performed on the BGI website (https://report.bgi.com). The raw sequencing data were deposited in the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/) with the BioProject accession number PRJNA954312.

The leaves of ZH11, Xa7-OE-1, and Xa7-OE-2 at the early stage of tillering were sampled and then ground into powder after fully drying for 96 hours. The powder was filtered using a 40 mesh sieve, and then 5 mg was weighed for the determination of lignin content. The measurement method was carried out according to the handbook of the Lignin Content Assay Kit (JC2203-S, China). The samples were mixed with acetic acid, and absorbance was measured at 280 nm. The content of lignin (%) was calculated using a computed formula with an extinction coefficient of 17.75 mL·mg⁻¹·cm⁻¹ for rice.

Data plotting and statistical analysis were performed using GraphPad Prism 8.0 software (https://www.graphpad.com/). Data are shown as the means ± SD, and significance analysis was conducted using a two-tailed Student’s t-test or a one-way analysis of variance (ANOVA) with least significant difference (LSD) multiple comparisons test. Asterisks represent statistical significance as *p < 0.05 and **p < 0.01.

In a previous study, a 26-bp putative EBE_AvrXa7 sequence (ATAACCCCCCCCCCCCCAGATAACCA) was predicted in the promoter of the Xa7 gene, which matched perfectly with the 25.5 RVDs of AvrXa7 based on the recognition principle between TALEs and EBEs (Chen et al., 2021). To confirm that AvrXa7 can indeed interact with EBE_AvrXa7, a yeast one-hybrid assay was carried out. In the reporter vector, three tandem repeats of both Xa7-EBE_AvrXa7 and SWEET14-EBE_AvrXa7 were inserted upstream of the LacZ gene, while the effector vector (AD-RVD) expressed the RVDs of AvrXa7 fused with the GAL4 activation domain ( Figure 1A ). The yeast strains co-transformed with Xa7-EBE_AvrXa7 and AD-RVD grew on a nutrient-deficient medium and showed significant β-gal activity, as detected by blue staining ( Figure 1B ). These results indicate that AvrXa7 can directly bind to the EBE_AvrXa7 sequence in the promoter of the Xa7 gene.

To investigate whether the promoter of the Xa7 gene can activate other executor R genes by binding with AvrXa7, six constructs were created ( Figure 2A ) and transiently transferred into tobacco for HR analysis—a plant-specific basic immunity that causes infected cell death via the bursts of reactive oxygen species (ROS; Lorrain et al., 2003). Among them, four constructs containing the executor R gene Xa7, Xa10, Xa23, and Xa27 were driven by the native promoter of Xa7 (Xa7_pro ). The remaining two constructs were driven by the CaMV35S promoter that constitutively expressed Xa7 or GFP ( Figure 2A ). The empty vector (EV) was used as a negative control. The results showed that significant cell death and excessive ROS accumulation occurred only when transforming the 35S::Xa7 or co-transforming the Xa7_pro::executor (Xa7, Xa10, Xa23, or Xa27) with 35S::avrXa7 in the tobacco leaves. No or only slight (Xa7_pro::Xa10) HR phenotype was observed in other situations ( Figures 2B–E ; Supplementary Figure S1 ). The above results suggested that the promoter of the Xa7 gene can activate other executor R genes to trigger HR in plants by inoculation with AvrXa7.

To further verify whether the executor R gene driven by the Xa7 promoter can enhance the disease resistance of rice plants, the mentioned construct Xa7_pro::Xa23 was transformed into a BB-susceptible rice cultivar Zhonghua11 (ZH11). Twenty-five stable transgenic T₂ lines were obtained and used for the analysis of the BB resistance ( Figure 3A ). The transgenic recipient ZH11 was highly susceptible to BB, while the transgenic plants with Xa7_pro::Xa23 exhibited a similar disease resistance spectrum of Xa7, instead of Xa23. The Xa7_pro::Xa23 transgenic plants were resistant to the Xa7-incompatible strain PXO86 and susceptible to the Xa7-compatible strain PXO99, while rice cultivar CBB23 containing the native Xa23 gene was resistant to both stains ( Figures 3B, C ). Additionally, the expression of Xa23 gene in those transgenic lines was significantly induced at different time points after PXO86 inoculation ( Figure 3D ). The above results showed that the executor R gene driven by the promoter of Xa7 gene enhanced host resistance to BB and indicated that the disease resistance characteristics of the executor R genes mainly depend on their promoters.

To explore the mechanism of Xa7 gene-mediated disease resistance, the expression pattern of the Xa7 gene after inoculation with Xoo strains was analyzed. A native promoter of the Xa7 gene (abbreviated as Xa7_pro ) and a mutated promoter of the Xa7 gene (abbreviated as Xa7_mpro ) with two base deletions in the EBE_AvrXa7 sequence were constructed upstream of the Gus gene and obtained corresponding transgenic rice plants designated as Xa7_pro::Gus and Xa7_mpro::Gus. After inoculation with the Xa7 incompatible Xoo strain PXO86, which contains an AvrXa7, GUS activity was significantly detected in leaves of the Xa7_pro::Gus transgenic plants, and it was localized around the infection site, while no induced GUS activity was detected in the leaves of the Xa7_mpro::Gus transgenic plants ( Figure 4A ).

Additionally, leaves from rice cultivar IRBB7 that contain the Xa7 gene were inoculated with PXO86 by clipping the leaf tips. After 24 hours, the infected leaves were collected and divided into five different regions based on distance from the clipping site (0–1 cm, 1–2 cm, 2–3 cm, 3–5 cm, and 5–7 cm) ( Figure 4B ). The expression level of the Xa7 gene was then analyzed via RT-qPCR. The results showed that expression of the Xa7 gene was strongly induced near the infection site of the leaves (<1 cm) and gradually decreased as the distance increased ( Figure 4C ). These results suggested that the Xa7 gene was locally induced in the host cells by Xoo infection.

As the induced expression of the Xa7 gene is localized, it is likely that the pathogen would be controlled locally in the host. To investigate this, the leaves of IR24 and its near-isogenic line IRBB7 were inoculated with PXO86 via leaf-tip clipping, and the growth of Xoo strains was analyzed in the five different regions of the inoculated leaves ( Figure 4B ) at 1 day, 3 days, 5 days, 9 days, and 14 days after inoculation (DAI) ( Figures 5A, B ). At 1 DAI, the growth of Xoo was only detected in the 0–1-cm region of the IRBB7 leaves, whereas it was detected in the 0–3-cm region of the IR24 leaves. Subsequently, the growth of Xoo spread on the leaves from the tips of both IRBB7 and IR24. However, the pathogen was limited to the 0–5-cm region of IRBB7 from 5 DAI, while it had spread to areas beyond 7 cm in IR24. Moreover, the amount of Xoo strains in IRBB7 was at least 10-fold lower than that in IR24 over 0–1-cm regions. These results indicated that the induced expression of the Xa7 gene could retard and restrict the growth of Xoo strains.

As the induced expression of Xa7 gene inhibited the growth of incompatible strain PXO86, we hypothesized that the activated expression of the Xa7 gene may also enhance the host’s defense against compatible Xoo strains. To test this hypothesis, a frequently used Xoo strain PXO99, which does not contain an AvrXa7 and cannot activate the expression of the Xa7 gene, was selected as a compatible strain of the Xa7 gene (Chen et al., 2021). First, the leaves of IRBB7 were infected with the incompatible strain PXO86 to activate the expression of the Xa7 gene and then further inoculated with the compatible strain PXO99 by clipping closely to the PXO86-infected site at 1 day, 2 days, and 3 days after PXO86 inoculation (<0.5 cm). The lesion lengths on leaves were measured 2 weeks later after Xoo inoculation. The results showed that the lesions of leaves inoculated with combined PXO86 and PXO99 strains were significantly shorter than those only inoculated with PXO99 ( Figures 6A, B ). To further verify that the observed effect was solely due to AvrXa7-mediated induction of Xa7, the leaves of IRBB7 were infected with PXO99 ^avrXa7 , PXO99, and combined strains (PXO99 ^avrXa7 and PXO99 mixed). The lesion lengths were measured 2 weeks later after Xoo inoculation ( Supplementary Figure S2 ), and the results were consistent with those shown in Figure 6 . Thus, the induced expression of the Xa7 gene could significantly increase the resistance of host plants to the compatible strains.

To further explore the molecular basis of cell death and defense response mediated by the Xa7 gene, transgenic rice plants constitutively expressing the Xa7 gene that was constructed in our previous study were applied (Chen et al., 2021). Two different lines, named Xa7-OE-1 and Xa7-OE-2, were selected. Compared to the wild-type control ZH11, Xa7-OE-1 plants display normal phenotypes and have low expression levels of the Xa7 gene, while the growth and development of the Xa7-OE-2 plants are obviously inhibited and have significantly higher expression of the Xa7 gene ( Figures 7A, B ). RNA-seq analysis was performed on both transgenic lines and the wild-type ZH11. Compared to ZH11, 897 and 2,561 DEGs were detected in Xa7-OE-1 and Xa7-OE-2, respectively, and among them, 617 DEGs were shared by the two transgenic lines ( Figure 7C ), in which 491 genes were collectively upregulated and 87 genes were downregulated ( Supplementary Table S1 , S2 ).

The Gene Ontology (GO) enrichment of the upregulated DEGs showed that genes involved in biological processes and defense response were prevalent ( Figure 7D ), such as the transcription factor OsWRKY19, which can activate the transcription of PR gene and play an important role in the defense response against pathogens (Du et al., 2021). To further explore Xa7-mediated pathways, KEGG enrichment analysis of the DEGs was performed, and the most enriched pathway was phenylpropanoid biosynthesis (ko00940) ( Figure 7E ). It has been reported that this pathway is involved in lignin biosynthesis and plays a key role in plant protection under biotic stress (Hamberger et al., 2007; Vogt, 2010). Enzymes such as phenylalanine ammonia-lyase (PAL), 4-coumarate–CoA ligase (4CL), cinnamoyl-CoA reductase (CCR), and cinnamyl alcohol dehydrogenase (CAD) are involved in the biosynthesis of monolignol from phenylalanine (Vanholme et al., 2010; Tianpei et al., 2015), and then laccases (Lac) or peroxidases (POD) oxidize monolignols to monolignol radicals in the cell wall, which subsequently combine in a combinatorial fashion to form lignin (Miedes et al., 2014). In the Xa7-OE lines, genes encoding the enzyme involved in lignin biosynthesis, such as PAL (OsPAL4 and OsPAL6), 4CL (Os4CL5), CAD (OsCAD6), and POD (OsPRX30, OsPRX62, OsPRX115, and OsPRX117), were found to be significantly upregulated ( Figure 7F ).

The transcription levels of selected genes, including OsPAL4, Os4CL5, OsPRX62, and OsCAD6, were verified via qRT-PCR, and the results were consistent with the RNA-seq data ( Figure 7G ). Furthermore, the expression of these genes was significantly induced in the leaves of IRBB7 but not in IR24 after inoculation with the PXO86 strain ( Figures 7H–K ). Moreover, qRT-PCR was selected to detect the expression level of two lignin-related genes in the different regions of IRBB7 after inoculation with PXO86 24 hours, and the results showed that the expression pattern was similar to that of the Xa7 gene. Subsequently, quantitative determination of lignin content in the leaves from ZH11, Xa7-OE-1, and Xa7-OE-2 proved that with the increased expression of Xa7, the lignin content was significantly enriched ( Figure 7M ). The above results suggested that the executor R gene Xa7 may enhance the resistance of rice plants to Xoo pathogens by increasing lignin synthesis.