Precision editing of GLR1 confers glufosinate resistance without yield penalty in rice.

Rice is one of the most important staple crops in the world and feeding over half of the global population. Weeds seriously affect the rice production and grain yield. Breeding and researching herbicide-resistant rice germplasm holds significant potential for effective weed management and the advancement of modern agriculture. Glufosinate has been widely used due to its non-selective broad-spectrum and high-efficiency weed control. However, there have been very few reports on rice germplasm that are resistant to the glufosinate herbicide. We obtained two glufosinate-resistant rice (glr1 and glr2) mutants from the japonica cultivar, Jingeng818 by treatment with a heavy ion beam and glufosinate screening (Figure S1). In this study, we selected the glr1 mutant for further research. To validate the glufosinate tolerance of glr1 mutant, seedlings were treated with 500 g ai ha−1 glufosinate. The results showed that glr1 mutant exhibited significant resistance to glufosinate (Figure 1a), and the resistance index (RI) of glr1 was 2.1 (Table S1), compared with its wild type. After treatment with glufosinate, the corresponding physiological indices of the glr1 were significantly better than those of the wild type, such as survival rate, chlorophyll content (Figure 1b), seedlings plant length (Figure S2a) and root length (Figure S2b), fresh weight (Figure S2c) and dry weight (Figure S2d). Glufosinate targets glutamine synthase (GS) and inhibits its activity, resulting in high levels of ammonia accumulation and a burst of reaction oxygen species (ROS), which in turn inhibits photosynthesis and causes lipid peroxidation, respectively, and eventually the plant death (Ren et al., 2023; Zhang et al., 2022). We quantified the levels of ammonia, H2O2, and malonaldehyde (MDA) in both wild-type and glr1 plants prior to and following exposure to glufosinate, the results showed that the contents of ammonia, H2O2, and MDA in glr1 were significantly lower than those of the wild type (Figure 1b). Our research findings indicate that glr1 plants demonstrate higher GS, SOD, and CAT activities compared to the wild type (Figure 1b). Additionally, we observed that the glr1 plants displayed initial accumulation of ammonia and ROS following glufosinate treatment, but this was swiftly cleared over a 7-day observation period (Figures S3 and S4). These results demonstrate that glr1 possesses an improved capacity to eliminate ammonia and ROS. The glufosinate-resistant trait exhibited by glr1 is the result of a recessive mutation that occurred in a single gene (Table S2). Map-based cloning revealed a deletion of G in the first exon of the LOC_Os06g47150 (Figure 1c), resulting in a garbled amino acid sequence and premature termination (Figure S5a). Genetic complementation and CRISPR-Cas9 mediated gene editing (Figures 1d,e and S6) also confirmed the LOC_Os06g47150 corresponds to GLR1, which encodes an Auxin response factor (ARF) family ARF18 (Figure S5b). The GLR1 gene is primarily expressed in the leaves (Figure S5c), and its expression is induced by glufosinate (Figure 1f). GLR1, as a transcription factor, exhibits nuclear localization characteristics (Figure S5d) and has the ability to inhibit transcription (Figure 1g). To investigate the molecular basis of GLR1 resistance to glufosinate, we performed RNA-seq analysis to identify differentaily expressed genes (DEGs) between glr1 and WT plants with or without glufosinate treatment (Figure S7). As glr1 plants showed lower levels of ammonia and H2O2 content compared to their wild-type counterparts when treated with glufosinate, we directed our attention towards the expression of genes involved in ammonia and ROS metabolism in the DEGs list (Table S3). According to our findings, glr1 plants exhibited upregulation of OsGS1, which encodes a glutamine synthase that converts glutamate and ammonia into glutamine, as well as OsCYP51G3 and OsCATA (Xia et al., 2015; Zhang et al., 2016), involved in eliminating ROS, compared to the wild type (Table S3). qRT-PCR assays further confirmed that OsGS1, OsCYP51G3 and OsCATA were upregulated in glr1 plants after being treated with glufosinate (Figure 1h). ARF transcription factors recognize and bind to TGTCTC motifs through their highly conserved B3-type DNA-binding domain. Accordingly, chromatin-immunoprecipitation qPCR (ChIP-qPCR) (Figure 1i) and electrophoretic mobility shift assays (EMSAs) (Figure 1j) confirmed that GLR1/OsARF18 binds directly to the TGTCTC-containing segments within the promoter regions of OsGS1, OsCYP51G3 and OsCATA. Consequently, GLR1/OsARF18 suppressed transcription from the promoters of OsGS1, OsCYP51G3 and OsCATA in transient transactivation assays performed in rice protoplasts (Figure 1k). Therefore, in the wild type, treatment with glufosinate induced the expression of the GLR1 gene (Figure 1f), which subsequently suppressed the expression of downstream genes (OsGS1, OsCYP51G3 and OsCATA) involved in the clearance of ammonia and ROS. This resulted in the failure of timely clearance of accumulated ammonia and ROS in the body, ultimately leading to plant death. However, when the GLR1 gene is mutated, it cannot inhibit the expression of related genes. Consequently, the proteins encoded by these genes remain active and facilitate the clearance of accumulated ammonia and ROS, thus preventing plant damage and death (Figure 1l). Although the glr1 mutant exhibits resistance to glufosinate, it exerts some effects on yield-related traits (Figure S8). To investigate whether there is an ideal allelic variation of GLR1 that confers glufosinate resistance and without yield penalty in rice. We utilized editing tool CRISPR-Cas9 to modify GLR1 at various locations corresponding to different domains (Figure S9). We found that retention of the DNA binding domain (DBD) and middle or repressor domain (MD/RD) of GLR1 can trade-off between resistance to glufosinate and grain yield (Figures 1m, S10 and S11 and Table S4). After utilizing the GLR1 knockout line 4 (glr1-ko4) allele, which has been demonstrated to confer glufosinate resistance without negative impact on yield (Figures 1m and S11), we were able to breed several strains of glufosinate-resistant rice that exhibit exceptional resistance due to the presence of this specific mutation site (Figure S12). In addition, we can also consider introducing this specific mutation into both sterile lines and restorer lines for the breeding of glufosinate-resistant hybrid rice. The GLR1 mutation eliminates the suppression of gene expression involved in ROS clearance, which can be beneficial for plants facing various abiotic stressors since many such stresses can cause a burst of ROS (Mittler et al., 2022). Therefore, we hypothesize that the GLR1 mutation also provides resistance to various abiotic stress. To further investigate our hypothesis, we conducted experiments where we treated both WT and glr1-ko4 plants with NaCl and glyphosate herbicide. The results showed that the glr1-ko4 plants demonstrated significant tolerance to salt stress (Figure S13) and to glyphosate herbicide (Figure S14). Moreover, due to the evolutionary and functional conservation of GLR1 in plants, engineering of GLR1 has the potential to improve tolerance to abiotic stress in other important crop species and without any yield penalty. This work was supported by the National Natural Science Foundation of China (32171962), the Youth Innovation Promotion Association CAS (2022454), the Natural Science Foundation of Anhui Province (Grant 2108085MC99), the State Key Laboratory of Plant Cell and Chromosome Engineering (PCCE-KF-2021-01), the HFIPS Director's Fund (No. YZJJKX202201), the Excellent scientific research and innovation team of the Education Department of Anhui Province (2022AH010087), Hainan Yazhou Bay Seed Lab (B23CQ15IP). The authors have declared no conflict of interest. Y.Y., X.F., Y.W. and C.Z. together designed the experiments. Y.R., B.L., H.J., W.C., L.T., H.W., G.S., K.W., Y.F. and C.Z. carried out the experiments and analysed the data. Y.Y. wrote the manuscript. All authors read and approved of the manuscript. Figure S1-S14 Supplementary Figures. Table S1-S5 Supplementary Tables. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.