Novel Salinity-Tolerant Third-Generation Hybrid Rice Developed via CRISPR/Cas9-Mediated Gene Editing

Climate change has caused high salinity in many fields, particularly in the mud flats in coastal regions. The resulting salinity has become one of the most significant abiotic stresses affecting the world’s rice crop productivity. Developing elite cultivars with novel salinity-tolerance traits is regarded as the most cost-effective and environmentally friendly approach for utilizing saline-alkali land. To develop a highly efficient green strategy and create novel rice germplasms for salt-tolerant rice breeding, this study aimed to improve rice salinity tolerance by combining targeted CRISPR/Cas9-mediated editing of the OsRR22 gene with heterosis utilization. The novel alleles of the genic male-sterility (GMS) and elite restorer line (733Srr22-T1447-1 and HZrr22-T1349-3) produced 110 and 1 bp deletions at the third exon of OsRR22 and conferred a high level of salinity tolerance. Homozygous transgene-free progeny were identified via segregation in the T2 generation, with osrr22 showing similar agronomic performance to wild-type (733S and HZ). Furthermore, these two osrr22 lines were used to develop a new promising third-generation hybrid rice line with novel salinity tolerance. Overall, the results demonstrate that combining CRISPR/Cas9 targeted gene editing with the “third-generation hybrid rice system” approach allows for the efficient development of novel hybrid rice varieties that exhibit a high level of salinity tolerance, thereby ensuring improved cultivar stability and enhanced rice productivity.


Introduction
With the rapidly increasing population, changing diet, and increasing biofuel consumption, global crop productivity must double by 2050 [1]. However, abiotic stressors, including soil salinity, drought, cold, and heat can severely limit the feasibility of yield increase or even reduce crop productivity in large areas [2]. Salinity is a major limiting factor in crop productivity as most crops are not salt-tolerant [3]. The land affected by salt stress is also increasing for various reasons, including tsunamis, sea-level rise, and climate change [4]. For example, the 2011 earthquake and tsunami in Japan led to the flooding of more than 20,000 ha of rice paddy fields with seawater, resulting in increased soil salinity [5]. Over 800 Mha of arable land (approximately 6% of the world's total land used for cropping) is affected by salinity to varying degrees [3]. Therefore, developing salt-tolerant crop varieties is a promising strategy to mitigate the negative impact of salinity on crop yields.
Rice (Oryza sativa L.) is the dominant food crop in Asia and worldwide. However, rice is sensitive to salt stress [6,7], displaying reduced grain yields under salt stress. Many quantitative trait loci (QTLs) and genes associated with regulating salt stress tolerance in

Development of OsRR22 Mutagenized Plants
CRISPR/Cas9-based genome editing was employed to generate OsRR22 mutagenesis lines. The third exon of OsRR22 was targeted using three gRNAs designed for the 22KT1, 22KT2, and 22T3 target sites (depicted in Figure 1A). These sites were chosen because they are located near an SNP in the third exon of OsRR22, which was reported to be associated with increased tolerance to salinity [5]. Two different CRISPR/Cas9 vectors, including one containing two gRNAs (pC-22-KT12) and another containing a single gRNA (pC-22-T3), were constructed ( Figure 1B) [35]. These vectors were inserted into 733B and HZ varieties via Agrobacterium-mediated transformation. All T0 plants were tested for the presence of T-DNA using HPT-specific PCR primers HPT-F1/R1 (Supplementary Table S1). A target-specific PCR primer pair RR-CJ3F/RR-CJ3R (Supplementary Table S1) surrounding the target region was used to amplify the transgenic plants, and the resulting PCR products were Sanger-sequenced to determine the OsRR22 mutagenesis of the T0 generation plants. Sequencing revealed 21 and 10 edited plants in the 733B and HZ backgrounds, respectively, with an editing efficiency of 84% (21/25) and 58.82% (10/17). These results demonstrated that these pC-22-KT12 and pC-22-T3 constructs could rapidly generate OsRR22-mutagenized rice plants.
were constructed ( Figure 1B) [35]. These vectors were inserted into 733B and HZ varieties via Agrobacterium-mediated transformation. All T0 plants were tested for the presence of T-DNA using HPT-specific PCR primers HPT-F1/R1 (Supplementary Table S1). A targetspecific PCR primer pair RR-CJ3F/RR-CJ3R (Supplementary Table S1) surrounding the target region was used to amplify the transgenic plants, and the resulting PCR products were Sanger-sequenced to determine the OsRR22 mutagenesis of the T0 generation plants. Sequencing revealed 21 and 10 edited plants in the 733B and HZ backgrounds, respectively, with an editing efficiency of 84% (21/25) and 58.82% (10/17). These results demonstrated that these pC-22-KT12 and pC-22-T3 constructs could rapidly generate OsRR22mutagenized rice plants. Exons and introns are indicated with black rectangles and black lines, respectively. Both the translation initiation codon (ATG) and the termination codon (TGA) are shown. The target sit nucleotides are shown in lowercase, and the protospacer adjacent motif (PAM) site is in uppercase, italic and underline; (B) schematic presentation of the T-DNA structure in the CRISPR/Cas9-mediated genome editing construct. The expression of Cas9 is driven by the Maize ubiquitin promoter (Pubi); the expression of the gRNA scaffold is driven by the rice U3 or U6a small nuclear RNA promoter (OsU3 or OsU6a); the expression of hygromycin (HPT) is driven by two CaMV35S promoters (2 × 35S); Tnos, gene terminator; LB and RB, left border and right border, respectively.

Characterization of OsRR22 T0 Mutant Plants
Ten (pC-22-T3, HZ) T0 osrr22 plants were selected to analyze the target site mutations. These genotypes were divided into two categories ( Figure 2A): (1) Homozygotes or bi-allelic (editing efficiency: 90%, 9/10, e.g., T1347 or T1349 at T3); (2) chimeras (editing efficiency: 10%, 1/10, e.g., T1352 at T3). Furthermore, only one type of mutation was identified: Deletions. Single-nucleotide deletions were the most common forms of mutations, accounting for 55.56% of all mutations ( Figure 2B). Exons and introns are indicated with black rectangles and black lines, respectively. Both the translation initiation codon (ATG) and the termination codon (TGA) are shown. The target sit nucleotides are shown in lowercase, and the protospacer adjacent motif (PAM) site is in uppercase, italic and underline; (B) schematic presentation of the T-DNA structure in the CRISPR/Cas9-mediated genome editing construct. The expression of Cas9 is driven by the Maize ubiquitin promoter (Pubi); the expression of the gRNA scaffold is driven by the rice U3 or U6a small nuclear RNA promoter (OsU3 or OsU6a); the expression of hygromycin (HPT) is driven by two CaMV35S promoters (2 × 35S); Tnos, gene terminator; LB and RB, left border and right border, respectively.
While powerful, CRISPR/Cas9 technology has been reported to be susceptible to introducing off-target mutations [36,37]. We, therefore, assessed the off-target efficiency of 22KT1, 22KT2, and 22T3 at high-sequence-similarity sites (<5 mismatch bp), with PAM, and in the coding sequence (CDS) or intron, to our target sites. Ten different potential off-

Development of Transgene-Free Mutant and Homozygous Osrr22 Lines
To develop homozygous osrr22 lines free of T-DNA, ten edited lines from T0 progeny (five lines from 733B and five lines from the HZ background) were bred to the T1 generation. All T1 plants from the selected, edited lines were further tested for T-DNA and mutation using HPTand Cas9-specific PCR primers and RR-CJ3F/RR-CJ3R target-specific PCR primers (Supplementary Table S1). Transgene-free plants were those in which neither Cas9 nor HPT was detected. Such plants were detected at frequencies of 0.00 to 33.33% among the analyzed plants (Supplementary Table S2 and Figure 3). Furthermore, three (733B background) and four (HZ background) transgene-free plants bearing homozygous osrr22 were selected. Seeds of these verified, transgene-free, homozygous mutant plants (designated as 733B rr22 -T1430-7, 733B rr22 -T1431-5, 733B rr22 -T1447-1, HZ rr22 -T1347-6, HZ rr22 -T1349-3, HZ rr22 -T1350-2, and HZ rr22 -T1351-8) were harvested and used to generate the T2 progeny. In addition, the genetic segregation analysis of self-pollination of three mutagenesis maintainer lines (733B rr22 -T1430-7, 733B rr22 -T1431-5, and 733B rr22 -T1447-1) demonstrated fluorescent and nonfluorescent seeds ( Figure 4). The nonfluorescent seeds were selected to generate a T2 population as the GMS lines (named 733S rr22 -T1430-7-T2, 733S rr22 -T1431-5-T2, and 733S rr22 -T1447-1-T2), transgene-free and homozygous for osrr22. target sites within the rice genome were examined. However, no evidence of off-target events among the 180 T1 plants was detected (Supplementary Table S3).   target sites within the rice genome were examined. However, no evidence of off-target events among the 180 T1 plants was detected (Supplementary Table S3).   While powerful, CRISPR/Cas9 technology has been reported to be susceptible to introducing off-target mutations [36,37]. We, therefore, assessed the off-target efficiency of 22KT1, 22KT2, and 22T3 at high-sequence-similarity sites (<5 mismatch bp), with PAM, and in the coding sequence (CDS) or intron, to our target sites. Ten different potential off-target sites within the rice genome were examined. However, no evidence of off-target events among the 180 T1 plants was detected (Supplementary Table S3).
2.6. Agronomic Trait Analysis of 733S rr22 -T1447-1, HZ rr22 -T1349-3 The T4 population carrying the homozygous 733S rr22 -T1447-1 and HZ rr22 -T1349-3 alleles and wild-type (733S and HZ) were cultivated under natural growth conditions to evaluate the effects of the 733S rr22 -T1447-1 and HZ rr22 -T1349-3 alleles on the major agronomic traits. No significant differences were detected in the plant height, the number of tillers per plant, the number of grains per panicle, spikelet fertility, or 1000-seed weight at the maturation stage between osrr22 and wild-type (Table 2 and Figure 7). Therefore, these findings suggest that the 733S rr22 -T1447-1 and HZ rr22 -T1349-3 alleles did not affect the major agronomic traits when the plants were cultivated under natural growth conditions. wild-type (733S and HZ) plants were aligned using DNAMAN5.2.2. The results demonstrated that the two mutagenesis genotypes encoded OsRR22, 187 or 116 amino acids in length, respectively. In contrast, the wild-type (733S and HZ) OsRR22 was 695 amino acids in length ( Figure 6). The amino acid sequences in these novel OsRR22 alleles were altered, emphasizing that these mutant alleles expressed truncated, disrupted, and altered OsRR22 proteins. evaluate the effects of the 733S rr22 -T1447-1 and HZ rr22 -T1349-3 alleles on the major agronomic traits. No significant differences were detected in the plant height, the number of tillers per plant, the number of grains per panicle, spikelet fertility, or 1000-seed weight at the maturation stage between osrr22 and wild-type (Table 2 and Figure 7). Therefore, these findings suggest that the 733S rr22 -T1447-1 and HZ rr22 -T1349-3 alleles did not affect the major agronomic traits when the plants were cultivated under natural growth conditions.

Development of the New Promising Third-Generation Hybrid Rice Line with Novel Salt Tolerance
In order to develop the novel salt-tolerant third-generation hybrid rice line, the 733S rr22 -T1447-1-T4 (as the female parent) and HZ rr22 -T1349-3-T4 (as the male parent) were crossed, and the promising hybrid line was named ST1. The salt-tolerant hybrid rice cultivar Xiang Liang You 900 (XLY900) was used as a control (CK). Whole growth stage trials were conducted in a salt-treated pool during the 2022 growing season to assess the practical applicability of ST1 for salt-tolerant rice breeding. Following the local farmers' cultivation practice, four-week-old ST1 and CK seedlings were transplanted and cultivated. Plants in the high-salinity plot were treated by periodically supplying water containing seawater (an ultimate density of approximately 0.6% NaCl concentration) until the maturation stage. The soil salinity level in the salt-treated plot was monitored by measuring the EC values in the soil. Before the end of February, the EC value was <1 dS/m, but the EC value gradually increased over time. In late March, the EC value exceeded 6 dS/m (equivalent to approximately 0.5% salinity) ( Figure 8C). During the whole growth stage trial, the grain yield per ST1 plant in salt-treated plots was significantly increased to 16.22 ± 0.61 g (equivalent to 3.65 tons/hectare, calculated at approximately 225,000 plants per hectare), demonstrating a production increase of more than 45% when compared to the CK plants ( Figure 8B). All ST1 plants in the salt-treated plots survived until the end of the trial ( Figure 8A). These findings provide further evidence that the CRISPR/Cas9-mediated mutagenesis of the OsRR22 gene combined with the "third-generation hybrid rice system" can be used to produce promising third-generation hybrid rice varieties with novel salt tolerance. This strategy could increase grain yield by enabling the cultivation of salt-tolerant varieties in salt-contaminated land. hectare), demonstrating a production increase of more than 45% when compared to the CK plants ( Figure 8B). All ST1 plants in the salt-treated plots survived until the end of the trial ( Figure 8A). These findings provide further evidence that the CRISPR/Cas9-mediated mutagenesis of the OsRR22 gene combined with the "third-generation hybrid rice system" can be used to produce promising third-generation hybrid rice varieties with novel salt tolerance. This strategy could increase grain yield by enabling the cultivation of salt-tolerant varieties in salt-contaminated land.

Discussion
Salinity is a major abiotic stress that negatively affects crop growth and production. Crop productivity in rice is severely limited due to the high salt concentrations in the soil. Generally, genetic improvement is deemed a cost-effective and environmentally friendly means to enhance salinity tolerance in rice. In recent years, many genes involved in salinity tolerance have been identified in rice, including OsNHX1, SKC1, OsbZIP71, and OsNAC2 [8,13,15,38]. Nonetheless, most of these positively regulate the salinity tolerance of rice, and over-expressing these genes shows improved tolerance to salt stresses. However, transgenic crops generated by gene addition are subjected to rigorous genetically modified management. The latest breeding strategy to knock down or knock out a negatively regulated gene is an alternative approach for the genetic improvement of rice and avoiding the transgenic issue using CRISPR/Cas9 technology. Nowadays only a few genes acting as negative regulators of salt tolerance in rice have been identified (e.g., the OsRR22 gene). A previous study suggested that the OsRR22 loss-of-function mutation showed significantly enhancing salinity tolerance but does not affect other major agronomic traits [5]. Therefore, this gene has great potential for genetic improvements via CRISPR/Cas9-mediated editing.
In this study, the pC-22-KT12 and pC-22-T3 Cas9 vectors were used to edit OsRR22, achieving 84.00% and 58.82% editing efficiency, respectively, in T0 transgenic plants. By transgene self-pollination segregation, seven independent osrr22-mutagenesis lines were generated bearing T-DNA-free and homozygous OsRR22 mutations in the T2 progeny. The transgene-free and rr22 GMS lines were obtained via fluorescent sorting. The salinity tolerance evaluations at the seedling stage demonstrated that the salinity tolerance was markedly enhanced in two novel OsRR22 alleles (733S rr22 -T1447-1-T3 and HZ rr22 -T1349-3-T3), suggesting that these alleles are valuable resources for developing elite rice varieties that can be cultivated on salt-contaminated soil. Further analysis of the OsRR22 translation, OsRR22 amino acid sequences of 733S rr22 -T1447-1, HZ rr22 -T1349-3, and wild-type lines demonstrated that OsRR22 proteins were truncated, altered, and disrupted in these particular alleles. The previous study indicated that an SNP representing a nonsense mutation of TGG to TAG resulting in a loss-of-function mutation in OsRR22 is responsible for the salinity-tolerance phenotype [5]. Therefore, the 110 bp and 1 bp deletions at 22KT12 and 22T3 caused a frameshift and resulted in premature translation termination and may lead to the enhanced salinity tolerance of 733S rr22 -T1447-1 and HZ rr22 -T1349-3, respectively. In summary, a loss of function of the OsRR22 protein was the cause of the destruction of the two-component cytokinin signaling in these rr22 plants, resulting in the enhanced salinitytolerance phenotype. Furthermore, a new promising third-generation hybrid rice line (ST1) with a novel salt-tolerant phenotype was developed by crossing the 733S rr22 -T1447-1-T4 and the HZ rr22 -T1349-3-T4 lines. The ST1 line exhibited favorable agronomic traits in salt stress, was transgene-free, and could be readily used for farming production if permitted under government policies. These results highlight the potential to use CRISPR/Cas9 technology to develop promising hybrid rice lines and parental varieties bearing targeted mutations capable of improving the salinity-tolerance trait, with OsRR22 being a putatively ideal target for the accelerated development of novel salt-tolerant rice varieties.
Technological advances in the agricultural sector, particularly breeding techniques, have increased agricultural productivity [39][40][41][42]. CRISPR/Cas9 technology has recently become a useful tool for editing genes owing to its simple methodology, mature technology, high mutation efficiency, and ability to avoid the transgenic issue [43]. It has been used in many plant species to improve major agronomic traits, including yield traits, fertility, disease resistance, and others [44][45][46][47][48]. Meanwhile, the development and adoption of hybrids for several key crops have significantly contributed to increased global agricultural productivity since the 1960s. As the latest rice heterosis utilization technology, the "thirdgeneration hybrid rice system" holds great promise to boost hybrid rice production further. The previous report suggests that a promising third-generation hybrid rice combination exhibits more than a 13% yield advantage over the control combinations [49]. Therefore, the CRISPR/Cas9 approach was combined with a "third-generation hybrid rice system" to rapidly produce one new promising hybrid rice line with a novel salt-tolerant phenotype with desirable agronomic traits. The positive results suggest that integrating the CRISPR/Cas9 approach and heterosis utilization may represent a powerful, highly efficient, and green approach to rice and other hybrid crop breeding genetic improvement. Since environmental deterioration is extensive throughout salt-contaminated paddy fields, the resulting crop productivity in rice is seriously limited. Therefore, enhanced rice salinity tolerance is one of the best strategies to expand the rice planting area, ensuring better stability and improved rice yield. As a result, this study offers a new strategy and novel OsRR22 alleles for breeding new rice varieties with novel salt tolerance and favorable agronomic traits under the conditions of severe salt pollution.

Plant Materials and Wild-Type Growth Conditions
Oryza sativa L. Indica lines 733-3B and HZ that had stable inheritance and were easily distinguishable were used in this study. 733-3B is a maintainer of the GMS 733-3S. HZ is an elite restorer line of hybrid rice. Plants were cultured at 28-35 • C in the greenhouse in Changsha or were grown in the screen house of the Hunan Hybrid Rice Research Centre in Changsha or Sanya during the typical rice-growing season. Management followed standard agricultural practices.

Vector Construction and Transformation
The CRISPR/Cas9 system was used to generate knockout mutants. For high-efficiency and targeted mutagenesis of 733-3B, two candidate sites for OsRR22 (Os06g0183100) were selected as target single guide RNAs (gRNAs): 22KT1 (5 -tgatcagaagaaccacaggtAGG-3 ) and 22KT2 (5 -ctgggcttctttgcagctgaGGG-3 ) with the anti-sense strand sequence containing the PAM motif 20 bp downstream. Another target gRNA 22T3 (5 -agacaaagactgtgatgaagGGG-3 ) was selected for targeted mutagenesis in HZ. The specificity of the target gRNAs was evaluated using biology-website CRISPR-GE (http://skl.scau.edu.cn/) and NCBI (https: //blast.ncbi.nlm.nih.gov) [50]. The target sequence had at least two bases different from similar off-target sequences within the PAM or PAM-proximal region. The two gRNAs under the U3 and U6a promoters were introduced into the pYLCRISPR/Cas9 vector to generate the knockout construct pC-22-KT12. All the constructs were confirmed by sequencing. The primers used for polymerase chain reactions (PCRs) during pC-22-KT12 construction are presented in Supplementary Table S1.
The correct pC-22-KT12 vector was introduced into Agrobacterium tumefaciens EHA105 strains. Bacteria containing the pC-22-KT12 construct were transformed into the 733-3B line as previously described [51] or by Wuhan Biorun Biological Technology Co., Ltd., to generate T0 plants for salt-tolerant mutant screening. After four weeks of rooting, regenerated plants were transferred to plastic buckets and grown in a greenhouse at 28 • C for 14 h day and 26 • C for a 10 h night photoperiod.

Mutant Identification and Characterization in Edited Plants
T0 plants were collected from plates containing 50 mg/L G418 or 50 mg/L hygromycin. The T0 plants were then transferred to grow in a greenhouse, and the T0 seeds were harvested for further analysis. To confirm the targeted mutagenesis in the plants, genomic DNA (gDNA) was extracted using the cetyltrimethylammonium bromide (CTAB) method from a minimum of five leaves from different tillers during the mature period of growth [52]. The gDNA was used as the temple for target-PCR. PCR amplification with KOD FX (TOYOBO, Osaka, Japan) was performed using specific primers RR-CJ3F/RR-CJ3R (Supplementary Table S1) containing the CRISPR/Cas9 target sites. Following PCR, the products were purified and subjected to sequencing. Mutations were identified by comparing the T0 plant sequences to those of the wild-type. Plants were considered homozygous when mutations were associated with a normal sequencing chromatogram. Alternatively, mutations that exhibited superimposed sequencing chromatograms were considered heterozygous or biallelic. Samples with superimposed sequence chromatograms were cloned into the pEASY-Blut vector (TransGen Biotech, Beijing, China), and ten positive clones were sequenced to determine the mutation genotype. DNAMAN5.0 and MAGE4.0 were used for sequence alignment analysis.

Transgene-Free Detection
The identification of the absence of the transfer DNA (T-DNA) containing Cas9 and HPT in the T1 generation was determined based on Cas9and HPT-specific PCRs. The primer pairs Cas9-F/Cas9-R and HPT-F/HPT-R were used to amplify the Cas9 and HPT genes, respectively (Supplementary Table S1). Plants were considered transgene-free when amplicons lacked both of these genes, with the pYLCRISPR/Cas9 plasmid and wild-type plants serving as positive and negative controls, respectively.

Measurement of Salinity Tolerance
A salt stress test was performed using the Takagi et al. [5] method to evaluate the salinity tolerance of osrr22 lines at the seedling stage. Briefly, three-week-old osrr22 and wild-type plants were selected and treated with Yoshida nutrient solution and 0.8% NaCl nutrient solution, respectively. The pH value of the nutrient solution was adjusted to 5.6-5.8, and the solution was renewed every 3 days. The plants were grown in a greenhouse at 28 • C for 12 h light and 22 • C for 12 h dark photoperiod. After 10 days of treatment with the 0.8% NaCl nutrient solution and 7 days of recovering with the nutrient solution, the different genotypes were photographed and compared. The salinity tolerance was determined by measuring the fresh weight, seedling length, root length, dry weight, and survival ratios of five plants per line. The measurements for each line were repeated in triplicate.
Whole growth stage trials were carried out in a salt-treated pool during the 2022 growing season in Sanya (a coastal city in southern China) to assess the salt tolerance of the promising third-generation hybrid rice line. Four-week-old seedlings of the new hybrid rice line and the control were transplanted. These were periodically irrigated with a mix of seawater (an approximately 1.0-1.5% NaCl concentration) and fresh water. The yield of these two hybrids was further analyzed. The soil salinity level in the seawater-treated plot was monitored by measuring the mixed soil's electrical conductivity (EC) value, which was selected from five plots of the salt-treated pool using a Soli-EC tester (ZD-EC).

Amino Acid Sequence Analysis
The amino acid sequence of OsRR22 was deduced and aligned with the OsRR22 mutants and wild-type using MAGE4.0 and DNAMAN5.0 software.

Agronomic Trait Characterization under Natural Growth Conditions
In order to assess the agronomic traits under normal conditions, both osrr22 and wild-type plants were grown under standard conditions in a screenhouse in Changsha or Sanya. Four-week-old seedlings were transplanted, and management was conducted according to local conventional methods. Five plants in the middle rows of each line were sampled for the following traits: Plant height, the number of tillers per plant, the number of grains per panicle, spikelet fertility, and 1000-seed weight. These traits were all measured in five biological replicates per line.

Statistical Analysis
Three independent samples were taken from osrr22 lines and wild-type to measure fresh weight, seedling length, root length, dry weight, and survival ratios. Five plants from each T4 population and wild-type were used for agronomic trait measurement. SPSS20.0 (IBM, Armonk, NY, USA) was used for the variance least significant difference test.  Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data generated or analyzed during this study are included in this published article and its supplementary information files.