CRISPR/Cas9‐Mediated Base Editing of SiGS1 Confers Glufosinate Resistance in Foxtail Millet ( Setaria italica )

Foxtail Millet [Setaria italica (L.) P. Beauv.] is a drought-tolerant, soil-poor-resistant crop whose yield is significantly limited by field weeds. Weed control remains a major constraint on its production and industrial development (Darmency et al. 2017). Developing non-transgenic herbicide-resistant germplasm offers a sustainable solution to mitigate production losses (Jin et al. 2022). Recent gene-editing technologies represented by the CRISPR/Cas9 have enabled the specific enhancement of crop tolerance to herbicides through precision editing of crop herbicide target genes, with breakthrough progress already achieved in rice, maize, and other crops (Zhang et al. 2023). Therefore, it has become an urgent task to create herbicide-resistant germplasm using this molecular method. Glufosinate (phosphinothricin, PPT) is a widely used non-selective, broad-spectrum herbicide that is used globally to indirectly inhibit photosynthesis and culminate in the death of the plant by irreversibly inhibiting glutamine synthetase (GS), which leads to the high accumulation of ammonia (Takano et al. 2020). In this study, the identification and phylogenetic analysis of the GS gene in foxtail millet revealed the presence of five SiGS genes (Figure 1a). Complementary bioinformatic analyses—encompassing gene structure architecture, conserved motif identification, chromosomal localization, synteny assessment, and promoter cis-element profiling—demonstrated that these five SiGS members reside across three chromosomes and exhibit lineage-specific collinearity (Figure S1). Tissue-specific expression profiling via RT-qPCR further delineated functional divergence: Si1G311400 displayed constitutive transcription across all examined organs (Figure 1b). This spatiotemporal regulation suggests subfunctionalization within the SiGS family, positioning Si1G311400 (hereafter SiGS1) as the primary cytosolic GS isoform for subsequent functional investigation. Building upon established evidence that concurrent overexpression of OsGS1;1 and OsGS2 confers glufosinate resistance in rice (James et al. 2018). To further verify the bioinformatic predictions, we transiently expressed SiGS1 and tagged it with a GFP in Nicotiana benthamiana leaves. Confocal microscopy revealed exclusive cytoplasmic localization (Figure S2). Subsequently, we constructed overexpressing SiGS1 (SiGS1-OE) lines in Arabidopsis (Figure S3a). Then, we carried out a quantitative evaluation of glufosinate. The results after 2 weeks showed that SiGS1-OE plants had limited tolerance to glufosinate at concentrations of 0.96 mg/L and 0.32 mg/L (Figure S3b). To definitively establish SiGS1 as the functional glutamine synthetase isoform governing glufosinate sensitivity, we engineered knockout mutants in the elite foxtail millet cultivar Ci846 using CRISPR/Cas9-mediated targeted mutagenesis. Molecular characterisation of T0 transformants confirmed five independent frameshift alleles (Figure S4a). Homozygous T1 generation (SiGS1-ko2) was sensitive to glufosinate after 2 weeks of treatment (0.5 g/L, 1 g/L, and 2 g/L; Figure S4b). Beyond the above two strategies, structural modification of glutamine synthetase represents an alternative approach to confer glufosinate resistance by reducing herbicide binding affinity (Zhang et al. 2022). Sequence alignment of GS1 orthologs revealed absolute conservation of Ser59 across angiosperms (Figure 1c), positioning this residue as an optimal target for precision editing in foxtail millet. We employed base editing-mediated gene evolution (BEMGE) using the high-efficiency adenine base editor ABE8e (Wang et al. 2022), which catalyzes A•T → G•C conversions coupled with optimized sgRNAs targeting SiGS1 exon 3 (Figure 1d,e). Following Agrobacterium-mediated transformation of foxtail millet cultivar Ci846, genomic DNA from 25 independently transgenic lines was analyzed by PCR amplification and Sanger sequencing, and 11 (44%) homozygous editors, 7 (28%) heterozygous editors, and 1 (4%) biallelic edited individual were identified (Figure 1f). Line #1 and line #3 carried the A−17–G, A−14–G conversions (S59G-T60A, designated SGTA thereafter) in the targeting region. Line #13 carried the A−17–G, A−14–G and C−13–G/A conversions (S59G-T60A/G, designated SGTA/SGTG thereafter) in the targeting region (Figure 1g). Subsequently, potential sgRNA-dependent off-target sites for SiGS1-SGTA were predicted using the web tool CRISPR-GE (http://skl.scau.edu.cn/). Two potential off-target sites containing 2-nt and 3-nt mismatches relative to SiGS1-SGTA were evaluated, and no off-target events were detected across the 19 lines tested (Table S1). To validate the herbicide resistance phenotype conferred by the SGTA allele, we exposed SiGS1-SGTA and wild-type plants to glufosinate challenge (1 g/L). Phenotypic assessment 14 days after application revealed a significant difference in response: the SiGS1-SGTA plants remained viable, in contrast to the necrosis observed in wild-type plants. Sixty days following glufosinate application, unlike the death of the wild type, the SiGS1-SGTA plants grew normally (Figure 1h). The agronomic characteristics of glufosinate-treated SiGS1-SGTA plants were similar to those of wild-type plants (Figure S7). Prior to treatment, SiGS1-SGTA plants exhibited substantially elevated basal GS activity compared to wild-type counterparts. Following herbicide application, SiGS1-SGTA plants maintained significantly higher enzymatic activity than wild-type plants despite marked inhibition, indicating partial preservation of ammonium assimilation capacity. Concurrently, analysis of oxidative defence enzymes revealed comparable pre-treatment levels of peroxidase (POD), superoxide dismutase (SOD), and catalase (CAT) activities across genotypes. Post-application profiling demonstrated pronounced induction of these antioxidant systems exclusively in SiGS1-SGTA plants, consistent with enhanced mitigation of herbicide-induced oxidative stress (Figure S5a). Photosynthetic pigment quantification further corroborated this physiological resilience: chlorophyll a, chlorophyll b, carotenoids, and total chlorophyll displayed no genotypic differences before spraying. After glufosinate exposure, SiGS1-SGTA plants retained significantly greater pigment levels than severely depleted wild-type plants, confirming protection of photosynthetic apparatus integrity (Figure S5b). This aligns with base-edited GLR1 rice (Ren, Liu, et al. 2023). Collectively, our findings establish that CRISPR-mediated S59G-T60A substitution results in glufosinate-resistant foxtail millet. Research has shown that the growth rate of rice mutants lacking OsGS1;1 is severely reduced (Tabuchi et al. 2005). Fortunately, SiGS1-SGTA plants did not show significant growth inhibition under normal growth conditions (Figures S6 and 1i–l) we will use cytosine base editors (CBEs) and adenine base editors (ABEs) to generate additional glufosinate-resistant alleles (e.g., AVPS, +AF, D171N, H249Y) as previously reported (Ren, Kuang, et al. 2023), and design a multi-target base-editing library covering the full-length SiGS gene to screen for novel resistance loci. Ultimately, we will pursue multiple resistance alleles to create foxtail millet materials with enhanced glufosinate tolerance. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Appendix S1: pbi70440-sup-0001-AppendixS1.doc. Appendix S2: pbi70440-sup-0002-AppendixS2.docx. 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.