A new method for mutation inducing in rice by using DC electrophoresis bath and its mutagenic effects

Mutation breeding is a significant means of increasing breeding efficiency and accelerating breeding process. In present study, we explored a new method for mutations inducing in rice (Oryza sativa L.) by using direct current electrophoresis bath (DCEB). The results showed that 20 mM NaCl solution is the optimal buffer, and the mortality of rice seeds followed an upward trend with increasing voltage and processing time of DCEB. By exploring the mutagenic effects of γ-irradiation and DCEB on seed vigor and physiological damages, we found that the physiological damages induced by DCEB on seed vigor were significant compared with that by γ-irradiation. We screened two mutants with low filled grain percentage and one mutant with abnormal hull from the M2 generations. These three mutants were confirmed to be authentic mutants based on 48 SSR markers followed by the protocol NY/T 1433–2014. Whole-genome resequencing detected a total of 503 and 537 polymorphisms in the two mutants, respectively, and the DCEB mutagenesis induced mainly InDel variants, while the exon region of mutant genes occupied a large proportion, especially the SNP variants, which occupied about 20% of the mutation sites in the exon region.

www.nature.com/scientificreports/ era of electrophoresis technology 26 . In recent years, electrophoresis has been widely used in various fields, such as biomedicine 27,28 , analytical chemistry 29 and microbiology. The present study provides a new method for mutations inducing in rice by using DC (direct current) electrophoresis bath (DCEB). Aimed to (1) explore the optimal treatment for DCEB mutagenesis; (2) compare the mutagenic effects induced by DCEB and by γ-irradiation, and (3) predict the mutation frequencies of DC electrophoretic bath mutagenesis by WGR. The present studies are presented as successful examples of this method which is simple, convenient, safe, flexible, and low costly in rice on mutation breeding.

Results
The optimal conditions for DC electrophoresis bath mutagenesis. Four buffer solutions, NaCl solution, NaOH solution, TAE buffer, and TBE buffer, were used to explore their effects on the germination of rice seeds, sterile water was used as the negative control. The results showed that the 20 mM NaCl solution was the most effective and accelerated the germination of rice seeds, followed by the NaOH solution, the germination rate (GR) showed a sharp downward trend with increasing concentrations of TAE and TBE buffers. It is worth noting that the effect of 20 mM NaCl solution on the germination rate of rice seeds was more pronounced than that of sterile water (Fig. 1). Therefore, 20 mM NaCl solution was selected as the buffer for DCEB.
Seven voltages (20 V, 50 V, 80 V, 110 V, 140 V, 170 V and 200 V) and three treatment times (12 h, 24 h and 48 h) were used for a total of 21 treatment combinations, and 20 mM NaCl solution was used as the  Effect of γ-irradiation and DC electrophoresis bath mutagenesis on rice seed vigor. Finally, we selected six DCEB treatments with the three highest voltages (140 V, 170 V and 200 V), and six doses of γ-irradiation (50 Gy, 100 Gy, 150 Gy, 200 Gy, 250 Gy, and 300 Gy), as well as CK, consisting of thirteen treatments, were obtained to analyze seed vigor (Table 2). Compared with the control, γ-irradiation had a positive effect on rice seed vigor at 50 Gy and 100 Gy and a negative effect at other doses. The rice seed vigor index (VI) showed a trend of increasing and then decreasing with different doses of γ-irradiation (Fig. 3, Table 2). There were extremely significant differences between the positive and negative effects of the 50 Gy and 300 Gy doses, respectively (Table 3). This result indicates that when the dose of γ-irradiation reached 100 Gy, it had a positive effect on the seed vigor, and beyond this threshold, the seed vigor decreased. In contrast, DCEB mutagenesis had a negative effect on rice seed vigor under all six-level voltages treatments, the GR, GI and VI decreasing with increasing electrophoresis time and voltage (Fig. 2, Table 1). This result indicates that the both methods have effects on seed viability, and the DCEB mutagenesis has more negative significant effects on rice seed vigor than that γ-irradiation.
Effect of γ-irradiation and DCEB mutagenesis on agronomic traits. We investigated the agronomic traits of M 1 generations induced by DCEBs and γ-irradiation, respective. In Table 4, the number of productive tillers of plants treated with γ-irradiation were significantly higher than that of control plants, while plant height and filled grain percentage were lower in treated plants than the control; no significant differences were found for other agronomic traits, such as thousand-grain weight and flag leaf length and width. In particular, the 250 Gy and 300 Gy doses produced more productive tillers and lower filled grain percentages than the control. There was almost no significant effect of DCEB treatment on the contemporary agronomic traits compared to the control (Table 4).   www.nature.com/scientificreports/ 97.6%, respectively, and the filled grain percentage was 6.3%, 13.4%, 26.8% and 93.2%, respectively (Fig. 4). The genetic analysis of the mutants according to the 48 SSR markers showed that all three mutants shared the wild type bands (Fig. 5). These results clearly indicated that all three mutants are true mutants.
Evaluation of resequencing results. To further investigate the effects of DCEB mutagenesis at the genetic level, three experimental materials (WT, M1 and M2) were subjected to WGR in this study. Approximately 37.43 Gbp of clean data was produced, with a Q20 of 97.03%, Q30 of 92.25% and GC of 42.43% (averaged from WT, M1 and M2, respectively, Table 5). According to the results of WGR, the average sample-to-reference genome match was 98.70%, the average depth of coverage was 29X, and the genome coverage was 97.15%. The above results indicate that the sequencing data are reliable and can be used for SNP and InDel analysis.
Mutations that were identical to those of the wild type were filtered out. After filtering, the M1 mutant had a total of 101 SNPs and 402 InDels, and the M2 mutant had a total of 113 SNPs and 424 InDels. The mutation frequencies of M1 and M2 were calculated to be 2.39 × 10 -6 and 2.56 × 10 -6 (Calculated according to reference 30 ), respectively. The chromosome with the highest number of mutations was Chr7, followed by Chr9 and Chr12, and the chromosomes with the lowest number of mutations were Chr8 and Chr6 (the distribution of the SNPs and InDels on each chromosome is shown in Supplementary Figure S1). SNP variants was present at the highest levels in downstream regions, followed by upstream regions, exon regions and intergenic regions. Meanwhile, a large proportion of the InDel variants were found in downstream and upstream regions and only about 7% was present in exon regions (Fig. 6A,B). SNP variants occupied 20% of the mutation sites in the exon region, which indicated that the effect of DCEB mutagenesis on the function of rice genes is relatively large. SNP and InDel variants of the two mutants. In contrast to the wild type, among the two mutants M1 and M2, SNP mutations included 78 A to G transition (36.4%), 66 C to T transition (30.8%); and 23 A > C/T transversion (10.7%), 15 G > C/T transversion (7.0%), 13 C > A/G transversion (6.1%), and 19 T > A/G transversion (8.9%). In both mutants, A: G transitions were the most abundant type of mutations detected by DC bath, followed by the C:T transitions. The most frequent types of transversions in M1 mutants were A > C/T and T > A/G, and in M2 mutants, the most frequent type of transversions was A > C/T (Fig. 6C).
In both mutants, InDels mutations in the coding region were analyzed, and frame-shift were the most common type of mutation, accounting for approximately 37% of the entire coding region (Fig. 6D). The longest insertion mutation in the coding region was a frame-shift mutation at 97 bp, and the longest deletion mutation in the coding region was a stop-gained mutation at 99 bp. The most common mutations are 1-2 bp insertion and deletion mutations (Fig. 6E).

Discussion
Definition of the optimal median lethal dose for DCEB. According to the definition of median lethal dose (LD50), the result should be 50 V and 80 V for 24 h and 140 V for 12 h for DCEB, under which the seed germinations were close to 50%, however we didn't screen the significant phenotypically mutants in the M 2 generation of these treatments; while M1, M2, M3 and other types of suspected mutants appeared in 140 V, 200 V and 140 V for 48 h. The mutagenic effect on rice was further increased along with the increasing voltage and processing time in treatments above the LD50, so that we easily obtained significant phenotypically mutants. Therefore, LD50 is not necessarily suitable for DCEB mutagenesis, and a higher lethal dose (higher voltage or/ and longer processing time) can be taken appropriately to be able to increase the probability of mutagenesis. Table 3. Effects of γ-irradiation and DC electrophoresis bath mutagenesis on seed viability across the different treatments. Data are expressed as mean ± S.E. *p < 0.05, **p < 0.01, and ***p < 0.001.

Germination rate (%)
Relative www.nature.com/scientificreports/  www.nature.com/scientificreports/ Effects of γ-irradiation and DCEB on physiological damage. This study investigated the effect of two methods (γ-irradiation and DCEB) on the physiological damages to rice seeds. The results showed that at low doses, γ-irradiation had a positive effect on the germination potential and germination index (GI), while high doses of γ-irradiation significantly reduced the shoot length and vigor index, which is consistent with the results of previous studies [31][32][33] . However, rice seeds treated with γ-irradiation in this study all had significantly higher panicle numbers and lower filled grain percentage than controls, especially after a dose of 250 Gy. These results may be due to the high doses of γ-irradiation causing pollen sterility 34,35 . In contrast, the rice seeds treated with the DCEB method all had extremely significantly lower values for rice seed viability, and the performance of the contemporary agronomic trait was essentially indistinguishable. The result suggests that γ-irradiation mainly affects contemporary agronomic traits, whereas the DCEB mainly affects rice seed vigor.
The effect of DCEB mutagenesis on gene function. In present study, SNPs were found to be the most frequent mutation type, with C: T > A: G being the most frequent type of transitions, which is consistent with the results of other common mutagens. Transitions occurred more frequently than transversions, with an observed transition/transversion ratio of 2.0. For InDels, 1-2 bps insertions or deletions were detected most frequently, and the longest insertion in the coding region was a 97 bps frameshift mutation, while the longest deletion was a 100 bps deletion with a stop codon gain mutation. The DCEB method produced a large proportion of mutations detected in the exon regions of genes, particularly for SNP variants, with 20 (19.8%) mutations on M1 and 23 (20.4%) on M2 were screened.
The advantages of DCEB. Compared with traditional irradiation mutagenesis methods 36 , DCEB only needs electrophoresis instrument to mutagenize, which has the advantages of low requirements for instruments and equipment, simple, convenient and flexible operation, safe experimental environment, low experimental cost, etc. According to the reported studies, the number of mutations in exon regions is less than 10% of the genome mutations caused by γ-radiation 37,38 , DCEB can cause about 20% exon mutation, this result suggests that the effect of DCEB mutagenesis on gene function in rice is relatively large, especially the mutagenic effect is apparent at a treatment time of 48 h. It provides a new method for future mutagenesis breeding in rice and even in other crops.

Conclusions
The present study provides a new mutagenesis method by using DCEB. We identified three mutants (M1, M2 and M3) from M 2 generation and did the WGR for M1 and M2. M1 had a total of 101 SNPs and 402 InDels, and the M2 had a total of 113 SNPs and 424 InDels, indicating that insertional or deletion mutations were the main type  www.nature.com/scientificreports/ of mutations as induced by DCEB. The mutation frequencies were calculated at 2.39 × 10 -6 and 2.56 × 10 -6 for M1 and M2, respectively. Meanwhile, the exon region of mutant genes occupied a large proportion, especially for SNP variants, which occupied about 20% of the mutation sites in the exon region. The DCEB may provide a simple, convenient, safe, flexible, and low-cost method for future mutagenesis breeding in rice and even in other crops.  DCEB treatment of seeds. Plump seeds of XiangGeng 365 were soaked in water for 24 h and disinfected with 3-6% sodium hypochlorite for 30 min. For each DCEB treatment, 400 seeds were selected and placed in an electrophoresis tank lined with filter paper and buffered for various processing time, after which the buffer residue was removed by washing with water. Then, the seeds were placed on petri dishes lined with filter paper for germination, and physiological indexes such as the GR and GI were calculated. Plants, named as M 1 generation that germinated normally were transplanted to the field (Fig. 7), harvested when mature and examined for agronomic traits including plant height, filled grain percentage and thousand grain weight; and around one thousand plants of each M 2 generation (obtained from the self-crossed M 1 plants) were transplanted in the field for screening of mutations.
γ-irradiation of seeds. In February 2021, rice seeds were irradiated in a γ-irradiator provided by Guangzhou Huada Biological Company. Each well of a 450-well plate filled with 1000 seeds were exposed to different doses of γ-irradiation ranging from 50 to 300 Gy with an interval of 50 Gy.
The mutagenic effects induced by γ-irradiation and DCEB on M 1 generation. The VI including GR, GI, bud length, and agronomic traits of M 1 generations including plant height, flag leaf length and width, panicle weight, panicle length, number of productive tillers, filled grain percentage and thousand-grain weight were compared between the two methods. The mutagenic effects of DCEB and γ-irradiation on the physiological damage to seeds were estimated by quantifying four parameters on M 1 plants: GR, bud length, panicle weight and filled grain percentage. The formula for calculating each physiological index is as follows 39 : (1) GR(%) = means of germinated seeds means of total seeds × 100 (2) GI = (Gt/Dt) Figure 7. Protocol for the DC electrophoresis bath. (A) Soak the rice seeds in clean water for 24 h and disinfected with 3-6% sodium hypochlorite for 30 min; (B) the soaked and disinfected seeds were placed in an electrophoresis tank lined with filter paper and buffer, and then start the DC electrophoresis for a certain period of time; (C) After electrophoresis, the buffer residue was removed by washing with water; (D) Seeds were placed on petri dishes lined with filter paper for germination, and (E) Seedlings from these germinated seeds were transplanted to the field.