Phenotypic and transcriptomic characterization of OsSWEET14-edited rice (cv. Samkwang) with enhanced bacterial blight resistance Authors: Yeeun Jun, Jiheon Han, Yurim Kim, Sieun Choi, Song Lim Kim, Jae-Hyeon Oh, Eun-Jung Suh, Sang-Ho Kang, Hye-Ran Park, Hyun-Cheol Jeong, Sang Ryeol Park, Youngjun Mo Published: 2025 ABSTRACT AbstractBackgroundBacterial blight (BB) caused by Xanthomonas oryzae pv. oryzae (Xoo) poses a serious threat to rice production. The pathogen promotes infection by targeting effector binding elements in the promoter regions of rice susceptibility genes such as SWEET (Sugars Will Eventually be Exported Transporters) genes. Previous studies have shown that natural or induced variations in SWEET genes can effectively enhance BB resistance in rice. However, the effects of variations in SWEET genes on disease resistance and agronomic performance vary depending on the Xoo strains and the genetic background of rice cultivars, highlighting the need for precise evaluations in breeding applications.ResultsIn this study, we developed a CRISPR/Cas9-edited OsSWEET14 knockout line (SK-s14) in the background of the elite Korean rice cultivar Samkwang. The SK-s14 line exhibited enhanced resistance against the Xoo strain KACC10859. Transcriptomic analysis revealed that defense-related genes, including pathogenesis-related genes and salicylic acid signaling genes, were downregulated in SK-s14 relative to Samkwang under Xoo inoculation. These results suggest that the enhanced resistance may result from pathogen starvation due to restricted sugar efflux to apoplast, rather than from classical defense activation. Field evaluations showed that SK-s14 plants headed 1–7 days earlier, had longer panicles (+ 8.0%), and exhibited modest reductions in grain weight (-3.0%) and grain fertility (-4.4%) compared to Samkwang.ConclusionsOur results show that knocking out OsSWEET14 in the Samkwang background enhances BB resistance against the KACC10859 strain, likely through a pathogen starvation mechanism. Despite a modest grain yield reduction (-7.2%) observed under field conditions, SK-s14 represents a promising breeding material for disease-prone rice growing regions.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12870-025-07899-4.Keywords: Bacterial blight, OsSWEET14, Samkwang, CRISPR/Cas9, RNA-seq Introduction Bacterial leaf blight (BB) caused by Xanthomonas oryzae pv. oryzae (Xoo) is a severe disease that threatens rice production. The pathogen infects rice through stomata or wounds and proliferates mainly through the xylem [1]. BB generally occurs at the maximum tillering stage around four to six weeks after transplanting, leading to reductions in grain quality as well as yield losses of 10% to 50% [2, 3]. Since its first identification in Japan in 1884, BB has become a major threat in rice production in various countries including China, Korea, India, Philippines, Indonesia, the United States, and West Africa [4]. Due to the limited effectiveness of chemical control methods, cultivating resistant varieties is considered the most efficient and economical strategy for managing BB [5–7]. To date, 47 resistance (R) genes for BB have been identified in rice and classified into five main types – receptor-like kinase genes (Xa3/Xa26, Xa4, Xa21), mutated alleles of SWEET (Sugars Will Eventually be Exported Transporters) genes (xa13, xa25, xa41), executor R genes (Xa7, Xa10, Xa23, Xa27), genes encoding nucleotide-binding leucine-rich repeat receptors (NB-LRR) (Xa1, Xa2, Xa14, Xa45(t), Xo1, Xa47(t)), and basal transcription factor genes (xa5) [4, 8, 9]. These R genes have been actively incorporated into rice breeding programs to enhance BB resistance [10]. However, the extensive use of single R genes in breeding programs has led to the emergence of new pathogen strains capable of overcoming resistance. For example, Xa3 and Xa4, which were widely used for over 20 years, exhibited resistance breakdown to different Xoo strains [10–12]. Xoo infects plants through the type III secretion system (T3SS), where the interaction of transcription activator like effectors (TALEs) from Xoo with effector binding element (EBEs) in the promoter region of plant susceptibility (S) genes transcriptionally activates target genes, leading to plant infection [4, 13]. SWEET genes encoding sugar transporters are the representative S genes for BB. In rice, a major strategy of Xoo is to exploit the host sugar transporters, particularly members of the OsSWEETs clade III family (OsSWEET11a, OsSWEET11b, OsSWEET12–OsSWEET15) [14, 15]. When activated by the TALEs of Xoo, these transporters facilitate the export of sugars from the cytosol to the apoplast, where the pathogen utilizes them as a nutrient source [16]. Beyond their key roles in plant-pathogen interactions, SWEET genes are also involved in various aspects of plant growth and development [17, 18]. For example, reduced expression of OsSWEET11 and OsSWEET14 in rice roots has been linked to lower carbohydrate levels in root exudate, potentially limiting nutrients for soil methanogenic bacteria and thereby reducing methane production [19]. These findings suggest that rice SWEET genes may also have important implications for climate change mitigation, as rice paddies represent a major source of agricultural methane [20]. To improve BB resistance in rice, many previous studies have focused on natural or induced variations in the promoter EBE regions of major susceptibility genes, such as OsSWEET11 (Xa13), OsSWEET13 (Xa25), and OsSWEET14 (Xa41) [21–26]. In addition to EBE editing, several studies have improved BB resistance by knocking out specific SWEET genes [27–29]. However, editing SWEET genes for enhancing BB resistance often deteriorates agronomic traits such as fertility and grain size [30, 31]. As the impact of editing SWEET genes on agronomic traits vary across editing sites, genetic backgrounds and environments [32, 33], precise assessment is required when applying these strategies in local breeding programs. In Korea, BB had been effectively controlled using Xa3 as the major R gene which is resistant to K1, K2, and K3 races, the major Xoo races in Korea [7]. However, the breakdown of Xa3 with the emergence of a newly virulent K3a race since 2003 prompted breeding efforts to diversify R gene sources against BB [34]. Samkwang is an elite japonica rice variety with multiple disease resistance, a mid-late maturation period, and high grain quality widely cultivated in Korea [35, 36]. Although Samkwang carries Xa3 and is known to be resistant to K1, K2, and K3 races of Xoo [37], a recent study reported that Samkwang is susceptible to a specific K1 Xoo strain, KACC10859 [38]. In this study, we developed an OsSWEET14 knock-out line (SK-s14) of Samkwang with enhanced BB resistance using a CRISPR/Cas9 mediated gene editing approach. An RNA-seq analysis was conducted to examine the transcriptomic changes induced by OsSWEET14 editing upon Xoo infection. Furthermore, major agronomic traits and methane emission of the OsSWEET14-edited rice were analyzed to assess its potential as a breeding material. Materials and methods Plant materials The CRISPR/Cas9 editing of OsSWEET14 (LOC_Os11g31190) in the rice cultivar Samkwang was conducted following the protocols of the Rural Development Administration (RDA, Korea), as reported previously [38]. Briefly, the sgRNA sequence (5’-GTTCCAGTCGGTACCCTACG-3’) (Fig. 1A) was designed using the CRISPR RGEN tools (https://www.rgenome.net), and the pRGEB32 vector was modified by inserting the sgRNA sequence through a ligation-independent cloning method [39]. The vector was then introduced into the Agrobacterium tumefaciens strain LBA4404 for rice transformation. Two previously developed gene-edited lines for OsWRKY102 (LOC_Os01g08710) and ONAC019 (LOC_Os06g01230) in the background of Samkwang [38] were also used in this study. CRISPR/Cas9-mediated gene editing of OsSWEET14 in the rice cultivar Samkwang. A Structure of OsSWEET14 (LOC_Os11g31190) and gRNA sequence; B Sanger sequencing chromatograms of the Cas9 target region from Samkwang and SK-s14 (T2). The expected Cas9 target site is marked with a red triangle, and the PAM sequence (TGG) is highlighted in green; C–D Screening for Cas9 null-segregants of SK-s14 at T1 and T2 generation, respectively. The T2 progenies in panel D were derived from SK-s14-2 highlighted in yellow in panel C (see Fig. S3 for the original gel images in panels C and D) Selection of Cas9 null SK-s14 plants and off-target analysis To confirm the sequence editing of SK-s14 (T2), the target region was amplified by PCR (Table S1) and sequenced using the CES (Capillary Electrophoresis Sequencing) service at Macrogen, Korea. To screen Cas9 null-segregants, PCR was conducted using Cas9-specific primers, along with OsPRR37 primers included as a positive control (Table S1), using a T100 Thermal Cycler (BIO-RAD, USA). The PCR products were separated on a 1% agarose gel (135 V, 30 min) and detected using a UV gel documentation system (Analytikjena, Germany). To assess potential off-target effects, Cas-OFFinder was used to predict 60 off-target sites (Table S2), allowing up to three mismatches and considering up to one bulge [40]. Among the predicted off-target sites, seven sites located in exon or intron regions were selected for validation. PCR amplification followed by Sanger sequencing (Table S1) was performed to compare the sequences between Samkwang and SK-s14. Bacterial blight resistance assay For BB inoculation assay, Samkwang and SK-s14 (T2) plants were sown in GS150 pots (diameter 14.5 cm, height 13.5 cm) in a greenhouse at the National Institute of Agricultural Sciences (NIAS), on July 3, 2023. Additionally, we included previously developed knockout mutants of OsWRKY102 (SK-w102) and ONAC019 (SK-n019) in the Samkwang background [38], which were selected based on their potential roles in disease resistance from previous studies [41–44]. On August 9, 2023 (five weeks after sowing), 10 plants with uniform growth were selected per genotype and arranged in a randomized complete block design (RCBD) for two treatments (Xoo and mock inoculations) with five replications (Fig. S1). Cultures of the Xoo strain KACC10859 (K1) were grown for two days at 28℃ on PSA agar medium. The bacterial culture was diluted in 10 mM MgCl༒ to an OD600 of 0.5 for Xoo infection, while 10 mM MgCl༒ was used for mock infection. Six-week-old plants were inoculated on August 14, 2023, using the leaf clipping method by cutting at 2 cm from the leaf tip. After two weeks of infection, lesion lengths were measured on the inoculated leaves. RNA-seq analysis For RNA extraction, leaf samples from two genotypes, Samkwang and SK-s14 (T2), subjected to two treatments (Xoo and mock inoculation) were collected at 24 h post-inoculation (hpi). From the leaf clipping site, 1 cm of leaf tissue was sampled from three leaves of a single plant and pooled to form one biological replicate. Three plants were sampled per condition, resulting in a total of 12 RNA-seq libraries (2 genotypes × 2 treatments × 3 biological replicates). The samples were ground in liquid nitrogen, and RNA was extracted using the RNeasy Plant Mini Kit (Qiagen, Germany). RNA concentration was analyzed with the Qubit 3.0 Fluorometer (Thermo Fisher Scientific, USA), and only samples with a RIN 6.7 were selected based on analysis with the Agilent 2100 Bioanalyzer (Agilent Technologies, Germany). VAHTS mRNA Capture Beads 2.0 (Vazyme, China) was used for mRNA isolation from the extracted RNA. RNA-seq libraries were then constructed using the MGIEasy RNA Directional Library Prep Set (MGI, China), followed by PE-150 sequencing on the DNBSEQ-T7RS platform (MGI, China). The raw reads in fastq format were processed using Fastp ver 0.23.4 [45] for adapter trimming and quality filtering (-length required 25, --quality phred 20). Fastqc ver. 0.12.0 (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) was used to perform quality control. The reads were then mapped to the Nipponbare IRGSP-1.0 reference genome from RAP-DB using HISAT2 ver.2.2.1 [46], converted to BAM files, and sorted. After sorting, gene expression levels were quantified using HTSseq-count ver.2.0.5 [47] with the IRGSP-1.0 GTF annotation file from RAP-DB. The flagstat option of BAM sorting was used to determine the number of mapped reads and the mapping rate. The raw counts were normalized using DESeq2 ver.1.44.0 [48] and differentially expressed genes (DEGs) were filtered with the fold-change (FC) threshold of |log2FC| > 1 and adjusted P < 0.05. Gene ontology (GO) analysis was performed using RGI (Rice Gene Index, https://riceome.hzau.edu.cn/) by a threshold FDR (false discovery rate) < 0.05 [49]. To analyze genes with significant GO terms, gene expression levels were normalized using the TPM (transcripts per kilobase million) method. Validation of RNA-seq data To verify the RNA-seq results, five DEGs were selected based on their absolute fold change values. Total RNA was extracted from the same samples used for transcriptome analysis, treated with TURBO DNA-free kit (Invitrogen, USA) to remove genomic DNA, and cDNA was synthesized using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, USA). qRT-PCR was performed with gene-specific primers (Table S1) and TOPreal SYBR Green qPCR PreMIX (Enzynomics, Korea) in a reaction volume of 20 µL, using CFX Opus 96 (BIO-RAD, USA). The cycling conditions were as follows: 40 cycles of denaturation at 95℃ for 10 s, annealing at 58℃ for 15 s, and extension at 72℃ for 30s. Relative gene expression levels were calculated using the 2-ΔCt method. Evaluation of agronomic traits Samkwang and SK-s14 (T3) were sown on May 7, 2024 and transplanted on June 3, 2024 at the LMO experimental field of NIAS, with a planting density of 30 cm between rows and 15 cm between plants in a row. For each genotype (Samkwang and SK-s14), six rows (20 plants per row) were planted per plot as an experimental unit, in an RCBD with four replications (plots) for each genotype. Days to heading were determined by counting the days from sowing until at least 40% of the plants in each plot had headed. Seven random plants from each plot were used to measure culm length, panicle length, panicle number per plant, and yield components. Culm length was measured from the ground to the panicle node of the main culm, while panicle length was measured from the panicle node to the tip of the main panicle. Using main panicles harvested from seven random plants per plot, the number of grains per panicle, grain fertility, and thousand-grain weight were measured. For plot yield, forty plants were harvested from each plot, threshed, and the rough rice weight and moisture content was measured. The rough rice yield was adjusted to a moisture content at 15%. Grain shape analysis was conducted using Image J [50] to measure grain area, length, and width of 100 rough rice grains. Evaluation of methane (CH4) emission Paddy soil collected from the research field at Jeonbuk National University was used to grow Samkwang and SK-s14 (T3) in Wagner pots (1/2000a) for a methane emission experiment. The plants were sown on May 30, 2024, and seedlings with uniform growth were transplanted into the Wagner pots (three plants per plot), in a greenhouse at NIAS on July 4, 2024. As controls, the experiment included pots without plants: two pots containing only paddy soil, and two pots containing paddy soil mixed with 90 g of finely ground rice straw. The closed chamber for gas collection consisted of an acrylic upper chamber (34 cm diameter with 130 cm height) and a lower chamber (34 cm diameter, 54 cm height), with a water reservoir in the middle to maintain complete sealing (Fig. S2). The upper chamber was placed only during gas sampling to ensure optimal conditions for plant growth. Gas sampling was conducted once a week between 10:00 AM and 11:00 AM, after three flushing cycles. The collected gas samples were analyzed for CH4 concentration using a gas chromatograph (7890 A, Agilent Technologies, USA) at NIAS. CH4 flux was calculated using the following equation. F = CH4 flux (mg m−2 h−1). ρ = gas density. V = volume of chamber (m3). A = surface area of chamber (m2). = rate of increase of gas concentration (mg m−3 h−1). T = absolute temperature (273 + mean temperature in chamber, ℃). Statistical analysis All statistical analyses and visualizations were performed using R (The R Foundation for Statistical Computing Platform) ver.4.2.1 and Microsoft Excel 2016. One-way and two-way ANOVAs were conducted using the ‘DescTools’ package. For analyzing the RNA-seq results, principal component analysis (PCA) was performed using the ‘ggplot2’ package, volcano plots for DEG selection were generated using the ‘EnhancedVolcano’ package, and DEG comparisons were conducted using the ‘VennDiagram’ package in R. Results CRISPR/Cas9 mediated editing of OsSWEET14 in the rice cultivar Samkwang To enhance bacterial blight (BB) resistance of Samkwang, a Korean elite rice cultivar, we used CRISPR/Cas9 to edit OsSWEET14, a major susceptibility gene for BB. OsSWEET14 possesses five exons and four introns with the CDS region of 912 bp encoding 303 amino acids (Fig. 1A). By transforming Samkwang with a CRISPR/Cas9 construct targeting the third exon, we obtained a T0 transformant with a homozygous 1-bp (T) deletion in exon 3 (named SK-s14), resulting in frame-shift mutation that altered the downstream codons with a premature stop codon at the 223rd amino acid (Fig. 1A, B). OsSWEET14, like other plant SWEET proteins, consists of seven transmembrane domains arranged in a 3-1-3 topology – two parallel triple-helix bundles (THB domains, also referred to as MtN3 domains) connected by a central linker helix, forming the essential structure of the sugar transport route [51]. Therefore, the frameshift mutation in exon 3 is expected to severely disrupt the second THB domain, likely abolishing the protein’s sugar transport function. By conducting a screening for Cas9 null-segregants on four T1 progeny plants, we obtained one Cas9 null-segregant plant, which was further confirmed in the next generation (T2) (Fig. 1C, D, Fig. S3). Additionally, potential off-target effects were assessed by sequencing the top seven candidates located in genic regions among the 60 predicted off-target sites (Table S1, S2), revealing no mutations in any of them (Fig. S4). Bacterial blight resistance of the OsSWEET14 knockout line To test whether OsSWEET14 knockout enhances BB resistance of Samkwang, we conducted a disease resistance assay using the Cas9 null-segregant T2 plants of SK-s14. The Xoo strain KACC10859 was used for inoculation by leaf clipping, as this strain was previously shown to cause infection in Samkwang [38]. At 14 days post inoculation, the average lesion length of SK-s14 (3.1 cm) was significantly shorter than that of Samkwang (11.1 cm), demonstrating enhanced resistance (Fig. 2). In our assay, two additional lines developed in our previous study – OsWRKY102 and ONAC019 knockout mutants (SK-w102 and SK-n019) [38] – were included in the BB resistance evaluation based on their potential roles in disease resistance [41–44]. However, the average lesion lengths of SK-w102 and SK-n019 were 12.8 cm and 10.8 cm, respectively, showing no significant difference compared to Samkwang (Fig. 2). Bacterial blight resistance of SK-s14. A Lesions induced by Xoo (KACC10859) at 14 days post-inoculation (dpi). Scale bar = 1 cm; B Lesion lengths at 14 dpi. Each bar represents mean ± standard error (***P < 0.001, ns not significant) Transcriptomic characterization of SK-s14 under bacterial blight inoculation To investigate transcriptomic changes associated with the OsSWEET14 knockout under Xoo infection, RNA-seq analysis was performed on the leaf samples of Samkwang and SK-s14 plants at 24 h post-inoculation (hpi) with Xoo or a mock treatment. We conducted RNA-seq at 24 hpi, as previous transcriptome studies of Xoo infection in rice reported substantial transcriptional differences at this time point, either between resistant and susceptible varieties or between mock and inoculated treatments [26, 52–54]. Three biological replicates were sampled and analyzed for each genotype and treatment, resulting in 12 RNA-seq libraries. These were sequenced, with an average of 104.1 M reads (pair-ended 150-bp) per sample and an average of 102.2 M reads (98.17%) mapped to the rice reference genome (IRGSP-1.0) (Table S3). PCA analysis of the 12 RNA-seq samples revealed that the difference caused by gene editing was more pronounced than the difference caused by Xoo infection (Fig. 3A). A total of 122 (90 up-regulated and 32 down-regulated) differentially expressed genes (DEGs) were identified between SK-s14 and Samkwang under mock condition, while 515 DEGs (up: 186, down: 329) were identified under Xoo infection (Fig. 3B, C). To focus on DEGs specifically expressed in SK-s14 under Xoo infection, 80 DEGs (up: 62, down: 18) detected under both mock and Xoo infection were excluded. After this filtering, 435 DEGs (up: 124, down: 311) remained (Fig. 3D, E). Additionally, one gene that was upregulated under Xoo infection but downregulated under mock infection was excluded, resulting in a final set of 434 DEGs for the subsequent analysis (Table S4). Transcriptome analysis of SK-s14 in comparison with Samkwang under Xoo inoculation. A Principal component analysis (PCA) of the 12 RNA-seq samples; B–C Volcano plots of differentially expressed genes (DEGs); D–E Venn diagrams of DEGs in response to Xoo inoculation GO analysis on the 434 DEGs revealed that the upregulated genes in SK-s14 relative to Samkwang under Xoo infection were enriched by two GO terms (GO:0048572, GO:0048575), with Hd3a (Os06g0157700) and OsFTL12 (Os06g0552900), involved in heading date regulation [55, 56] (Table 1). On the other hand, for the downregulated genes, a total of 26 enriched GO terms were identified. Many of these GO terms were associated with plant defense mechanisms such as “defense response”, “defense response to fungus”, and “regulation of defense response”, including 54 genes. This was followed by 18 genes associated with the GO terms involved in salicylic acid (SA) pathways, including “response to salicylic acid”, “regulation of salicylic acid biosynthetic process”, and “cellular response to salicylic acid stimulus”. The downregulated defense-related genes included PR (pathogenesis-related) genes such as OsPR1a (Os07g0129200) and OsPR1b (Os01g0382000), OsWRKY67 (Os05g01383100), and XinN1 (Os02g0282000), all of which are typically upregulated in response to pathogen attack [57–59]. Similarly, the downregulated SA-related genes included OsWAK71 (Os07g0493800) and OsWRKY47 (OS07g0680400), which are also known to be induced upon pathogen infection [60, 61] (Table 1). The downregulation of defense-related genes in SK-s14 under Xoo infection in our results suggest interesting insights into the mechanism of enhanced BB resistance conferred by OsSWEET14 knock-out, which likely starves Xoo by limiting sugar release into the apoplast, thereby precluding the expression of typical defense response genes (see Discussion). qRT-PCR validations of the top five DEGs with the highest fold-change values confirmed the expression patterns observed in the RNA-seq analysis (Fig. S5). Table 1Gene ontology (GO) enrichment analysis of DEGsGO termDescriptionNo. of genesFDRUpregulatedGO:0048572short-day photoperiodism20.0315GO:0048575short-day photoperiodism, flowering20.0315DownregulatedGO:0007166cell surface receptor signaling pathway124.03 × 10− 6GO:0009626plant-type hypersensitive response135.74 × 10− 6GO:0050832defense response to fungus176.07 × 10− 6GO:0080142regulation of salicylic acid biosynthetic process61.02 × 10− 6GO:0006952defense response311.22 × 10− 6GO:0016102diterpenoid biosynthetic process60.0002GO:0009751response to salicylic acid100.0002GO:0006032chitin catabolic process50.0014GO:0031347regulation of defense response60.0014GO:0006076(1->3)-beta-D-glucan catabolic process30.0014GO:0009992cellular water homeostasis50.0026GO:0010200response to chitin80.0026GO:0000272polysaccharide catabolic process50.0026GO:0009311oligosaccharide metabolic process50.0026GO:0006468protein phosphorylation170.0026GO:2,000,022regulation of jasmonic acid mediated signaling pathway50.0028GO:1,900,056negative regulation of leaf senescence30.0066GO:0009809lignin biosynthetic process50.0121GO:0010150leaf senescence60.0129GO:0009826unidimensional cell growth70.0152GO:0006816calcium ion transport30.0232GO:0071230cellular response to amino acid stimulus30.0232GO:0009607response to biotic stimulus40.0255GO:0006979response to oxidative stress80.0408GO:0071446cellular response to salicylic acid stimulus20.0448GO:0042744hydrogen peroxide catabolic process60.0448 Gene ontology (GO) enrichment analysis of DEGs Agronomic traits of SK-s14 To evaluate the impact of OsSWEET14 knockout on major agronomic traits, we evaluated days to heading, culm length, panicle length, and other yield-related traits of SK-s14 in field condition (Fig. 4). Compared to Samkwang, SK-s14 (T3 generation) exhibited one day earlier days to heading, 1.7 cm longer panicle, 2.8%p lower grain fertility, and 1.1 g lighter thousand-grain weight (Fig. 4C, E, H, L). No significant difference was observed in culm length, panicle number per plant, grain number per panicle, and grain size (area, length, and width) (Fig. 5D, F, G, I, J, K). The plot yield (estimated from the rough rice weight of 40 harvested plants per plot, with four RCBD replications) of SK-s14 was 1,218 g/plot, which was 7.2% lower than that of Samkwang (1,312 g/plot) (Fig. 5M). The earlier heading and longer panicle of SK-s14 compared to Samkwang were consistent with preliminary observations from the previous T2 generation, in which four SK-s14 T2 sister lines exhibited seven days earlier heading and 1.2 cm longer panicles than Samkwang (Fig. S6). Agronomic traits of SK-s14 in comparison with Samkwang. A Plants at maturity; B Rough rice grains. Bar = 1 cm; C Days to heading; D Culm length; E Panicle length; F Panicle number per plant; G Grain number per panicle; H Grain fertility; I Grain area; J Grain length; K Grain width; L 1000-grain weight; M Grain yield harvested from 40 plants, adjusted to 15% moisture content. Each bar represents mean ± standard error (** P < 0.01, * P < 0.05, ns not significant) CH4 emission of SK-s14 in comparison with Samkwang. A Cumulative CH4 emissions during the growth period; B Weekly CH4 emissions (shown as days after sowing). Green and gray arrows indicate heading date (days to heading from sowing) of SK-s14 and Samkwang, respectively. Error bar represents mean ± standard error Methane emission analysis in SK-s14 OsSWEET14 is a sugar transporter gene also expressed in rice roots, and its downregulation has been linked to reduced methane emissions from paddy fields [19, 62]. To test whether SK-s14 shows altered methane emission compared to Samkwang, we measured methane flux and cumulative emissions using the closed-chamber method in a greenhouse (Fig. S2). The experiment included three pots (three plants per pot) each for SK-s14 and Samkwang, along with four control pots without plants (two containing paddy soil only, and two containing paddy soil mixed with 90 g of ground rice straw). All pots were maintained under constantly flooded conditions from transplanting until maturity (two weeks before harvest). The paddy soil mixed with rice straw exhibited the highest cumulative methane emission (33.3 g m− 2), highlighting the strong influence of soil organic matter on methane production (Fig. 5A). While the difference was not statistically significant, the cumulative methane emissions of SK-s14 (5.5 g m− 2) were 29% lower than that of Samkwang (7.7 g m− 2) (Fig. 5A). This difference was mainly attributed to two emission peaks observed in Samkwang at 114 and 147 days after sowing during weekly gas sampling (Fig. 5B). In the pot experiment, SK-s14 plants headed six days earlier than Samkwang, consistent with the results observed under field conditions (Fig. 4C, Fig. S6). Discussion Enhanced bacterial blight resistance through editing OsSWEET14 in rice In this study, we successfully improved the BB resistance of Samkwang, an elite Korean japonica rice cultivar, by generating an OsSWEET14 knock-out line through CRISPR/Cas9. Samkwang was originally released as a BB-resistant cultivar against the three major Xoo races (K1, K2, and K3) in Korea [35]. However, recent reports have shown that Samkwang is susceptible to a specific K1 strain, KACC10859 [38]. To address this, we edited OsSWEET14, a major S gene for BB in rice, and achieved significantly enhanced resistance against the KACC10859 strain (Fig. 2). While previous studies have demonstrated that editing the EBE or exon regions of OsSWEET11, 13, and 14 confers resistance to Xoo, the effectiveness of resistance can vary depending on the edited site, Xoo strain, and cultivar background [24, 25, 28, 33, 63, 64]. Additionally, editing the promoter regions has limitations in providing broad-spectrum resistance, as different pathogens often target distinct EBE sites. Therefore, we aimed to abolish the function of OsSWEET14 by targeting its exon, and to evaluate the resulting effects on BB resistance and agronomic performance. However, in the current study, the resistance of SK-s14 was evaluated against only a single strain. Further testing with diverse Xoo strains will be necessary to fully characterize the resistance spectrum conferred by this mutation and to inform targeted breeding strategies for broader BB resistance. The RNA-seq analysis exploring gene expressions in SK-s14 relative to Samkwang under Xoo inoculation suggested interesting molecular insights of the enhanced BB resistance in SK-s14 (Fig. 3; Table 1, Fig. S7). Among 123 upregulated DEGs, the GO analysis revealed that only two GO terms represented by two genes were enriched – Hd3a and OsFTL12, both associated with flowering regulation [55, 56]. Interestingly, previous studies showed that Hd3a knockout, not overexpression, enhances bacterial blight resistance in rice [56, 65]. In addition, we also detected the upregulation of genes typically associated with enhanced resistance when downregulated (Table S4), such as Pi-21 (Os04g0401000), known to enhance resistance to blast and sheath blight when knocked out [66, 67], and OsMATE1 (Os03g0188100), which negatively affects disease resistance in Arabidopsis [68]. Other upregulated DEGs included OsPRX125 (Os10g0109300), reported to participate in Xa23-mediated bacterial blight resistance [69] and a putative NB-LRR gene (Os11g0249000). Downregulated DEGs in SK-s14 also included many genes that are typically upregulated under pathogen attack. These included PR (pathogenesis related) genes (OsPR1a, OsPR1b), widely used as markers for bacterial blight resistance [57, 58, 70], OsWRKY67, a transcription factor that enhances resistance via PR gene activation [58], and XinN1, a Xa21-responsive gene that promotes PTI (pattern-triggered immunity) signaling [59]. Additionally, SA-related genes included OsWAK71, which mediates DAMP (damage-associated molecular patterns)-triggered immunity, and OsWRKY47, which is known to enhance blast resistance when overexpressed [60, 61]. Previous transcriptomic studies on BB resistance in rice have consistently reported the upregulation of PR genes, SA-signaling genes, WAK (wall-associated kinase) genes and NB-LRR genes in resistant cultivars following pathogen inoculation [71–73]. However, in this study, many of these genes were downregulated in SK-s14 relative to Samkwang under Xoo inoculation, suggesting that enhanced BB resistance of SK-s14 may be mediated by a distinct mechanism. The two major models explaining how sugar affects pathogen resistance are pathogen starvation and sugar signaling via activation of the salicylic acid (SA) pathway [74, 75]. Given the downregulation of SA-related genes, it is likely that the enhanced resistance observed in SK-s14 is mediated by pathogen starvation, rather than SA-related immune signaling. This model proposes that reduced sugar efflux into the apoplast limits the nutrient availability to the pathogen, thereby suppressing its proliferation. As a result, the plant may not need to activate typical defense-related genes, because the pathogen is effectively starved and unable to continue the infection. This was supported by our qRT-PCR validation (Fig. S5), which showed that the key defense-related genes including PR genes and SA-signaling genes were upregulated in Samkwang upon Xoo infection, but not in SK-s14. These findings suggest that reduced sugar efflux caused by OsSWEET14 knockout may limit pathogen proliferation to an extent that downstream defense responses are not activated. Effects of OsSWEET14 editing on major agronomic traits Agronomic evaluation of SK-s14 showed earlier heading of one to seven days, along with an increase in panicle length by 1.2–1.7 cm compared to the wild-type, Samkwang (Fig. 4, Fig. S6). However, SK-s14 showed a reduction in thousand grain-weight and fertility rate by 3.0% and 4.4%, respectively, leading to a 7.2% decrease in overall grain yield. Interestingly, the earlier heading of SK-s14 is consistent with a recent finding that phloem sucrose promotes flowering during the floral transition stage [76]. The knockout of OsSWEET14 may have restricted sucrose efflux into the apoplast, thereby maintaining higher intracellular sucrose level that could contribute to accelerated heading. Additionally, methane emission analysis indicated that SK-s14 mutants emitted approximately 29% less cumulative methane compared to Samkwang (Fig. 5). However, since methane emissions are affected by various environmental factors such as soil characteristics, microbial communities, and water and temperature managements, more comprehensive evaluations will be required to draw consistent and reliable conclusions [77–79]. Previous studies have reported contrasting effects of OsSWEET14 disruption on agronomic traits depending on the genetic background or the editing site. For instance, no significant agronomic changes were observed in Kitaake and Nipponbare backgrounds [14, 30], while seed size was significantly reduced in Zhonghua11 background [32]. Despite a modest yield reduction in SK-s14 (cv. Samkwang), its enhanced disease resistance and generally stable growth traits support its potential for use in BB-endemic regions in Korea. Nevertheless, caution is warranted when interpreting the effects of OsSWEET14 editing on agronomic traits, as only a single OsSWEET14-edited line was analyzed in this study. Although no mutations were detected among the top seven off-target candidate sites in SK-s14 (Fig. S4), the potential influence of somaclonal variation cannot be excluded. Further evaluation using additional independent OsSWEET14-edited lines or backcrossed progenies will be necessary to confirm the stability of the observed agronomic traits. Conclusion This study demonstrated that CRISPR/Cas9-mediated knockout of OsSWEET14 in the Samkwang background enhanced resistance to Xoo (KACC10859), with minimal impact on agronomic traits. The enhanced BB resistance is likely attributed to the pathogen starvation mechanism, in which limited sugar efflux suppresses pathogen proliferation and diminishes the need for activating typical host immune responses. Further research should involve multi-strain disease assays, long-term and diverse field evaluations, and investigation of Korean Xoo strains and their interaction with host resistance mechanisms to support sustainable and climate-resilient rice breeding. Supplementary Information Additional file 1. [80] Additional file 2. Abbreviations Bacterial Blight Differentially Expressed Genes Effector Binding Elements Gene Ontology Principal Component Analysis Type III Secretion System Transcription Activator Like Effectors Xanthomonas oryzae pv. oryzae Authors’ contributions Conceptualization: SRP, YM; Formal analysis: YJ, JH; Investigation: YJ, JH, YK, SC; Methodology: SLK, JHO, HRP, HCJ; Resources: SLK, JHO, EJS, SHK; Project administration: SRP, YM; Visualization: YJ, JH; Writing – original draft: YJ, YM; Writing – review & editing: YJ, YM. Funding This research was funded by the Rural Development Administration (RDA) of South Korea (grant # RS-2022-RD010034, and RS-2024-00322166). Data availability The RNA-seq datasets generated in this study are available in the NCBI’s Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra), Bioproject PRJNA1209593. Declarations Ethics approval and consent to participate Not applicable. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Sang Ryeol Park, Email: srpark@korea.kr. Youngjun Mo, Email: yjmo@jbnu.edu. References Associated Data Supplementary Materials Additional file 1. [80] Additional file 2. Data Availability Statement The RNA-seq datasets generated in this study are available in the NCBI’s Sequence Read Archive (https://www.ncbi.nlm.nih.gov/sra), Bioproject PRJNA1209593.