Development of new mutant alleles and markers for KTI1 and KTI3 via CRISPR/Cas9-mediated mutagenesis to reduce trypsin inhibitor content and activity in soybean seeds

Abstract

The digestibility of soybean meal can be severely impacted by trypsin inhibitor (TI), one of the most abundant anti-nutritional factors present in soybean seeds. TI can restrain the function of trypsin, a critical enzyme that breaks down proteins in the digestive tract. Soybean accessions with low TI content have been identified. However, it is challenging to breed the low TI trait into elite cultivars due to a lack of molecular markers associated with low TI traits. We identified Kunitz trypsin inhibitor 1 (KTI1, Gm01g095000) and KTI3 (Gm08g341500) as two seed-specific TI genes. Mutant kti1 and kti3 alleles carrying small deletions or insertions within the gene open reading frames were created in the soybean cultivar Glycine max cv. Williams 82 (WM82) using the CRISPR/Cas9-mediated genome editing approach. The KTI content and TI activity both remarkably reduced in kti1/3 mutants compared to the WM82 seeds. There was no significant difference in terms of plant growth or maturity days of kti1/3 transgenic and WM82 plants in greenhouse condition. We further identified a T1 line, #5-26, that carried double homozygous kti1/3 mutant alleles, but not the Cas9 transgene. Based on the sequences of kti1/3 mutant alleles in #5-26, we developed markers to co-select for these mutant alleles by using a gel-electrophoresis-free method. The kti1/3 mutant soybean line and associated selection markers will assist in accelerating the introduction of low TI trait into elite soybean cultivars in the future.

Keywords: CRISPR/Cas9; KTI1; KTI3; Kunitz trypsin inhibitor (KTI); allele-based selection marker; anti-nutritional factor; soybean.

Figure 1

Expression levels of KTI genes in WM82. (A) RNA sequencing data of 38 KTI genes in 30 different tissue types of cv. Williams 82 acquired from Phytozome soybean database was used to construct the heatmap to visualize their expression patterns. (B) The expressions of four soybean KTI genes were monitored by real-time PCR. Samples of leaf, flower, pod, and seed tissues from 2 breeding lines, V98-9005 (normal-TI line) and V03-5903 (low-TI line), were collected for RNA extraction. After reverse transcription, real-time PCR was used to evaluate the expressions of 4 genes including Gm01g095000, Gm08g341000, Gm08g342300, and Gm08g341500 in different tissues with the ELF1B (Gm02g276600) as the reference gene. The expression data was normalized as ΔCT and shown as mean ± s.e. Experiments were repeated three times and obtained similar results.

Figure 2

The scheme of binary vector used for CRISPR/Cas9 mediated gene editing on KTI1/KTI3, and transgenes and gene editing have been detected in the leaves of four T0 soybean plants. (A) The CRISPR/Cas9 construct harbors three necessary elements exhibited as below: the selection cassette consists of MAS promoter, Bar gene (soybean transformation selection marker), and MAS terminator; the Cas9 cassette consists of U10 promoter, Cas9 gene, and OCS terminator; three guide RNA cassettes and each of them consists of a U6 promoter, and one sgRNA. (B) The sequences of three sgRNAs is shown here. Two sgRNAs were designed, synthesized, and assembled to the plasmid to target on KTI1, while one sgRNA was designed, synthesized, assembled to the plasmid to target on KTI3. The fragments of two transgenes, (C) Cas9 and (D) Bar, have both been detected in lines #2, #5, #11, and #17 by PCR, but not lines #4 and #7. The WM82 gDNA serves as the template for negative control, while the plasmid DNA serves as the template for positive control. (E) The gene editing on KTI1 and KTI3 has also been observed in the leaf tissues of plants at T0 generation. The double peak sequence around the sgRNA region indicates the gene editing was ongoing but not completed.

Figure 3

Gene editing on KTI1 has been completed for all seeds of T0 generation while it has been completed on KTI3 for some seeds of T0 generation. From each transgenic line (#2, #5, #11 and #17), four seeds of T0 generation were selected randomly for genotyping. (A) The gr of mutant kti1 in T0 seeds and T1 plant (#5-26) leaf, where the wild type KTI1 in WM82 was the control. (B) Gel electrophoresis of kti1 PCR products showed 16 seeds from line #2, #5, #11, and #17 had the same mutant on kti1, in which 66 nucleotides are lost between two sgRNAs. (C) Sanger sequencing result displayed the identical mutant kti1. (D) The alignment of mutant kti3 in T0 seeds (#2-3, #5-4, #5-26, #11-2 and #11-4) and T1 plant (#5-26) leaf, where the wild type KTI1 in WM82 was the control. (E–G) showed the sanger sequencing results of kti3 mutant in #2-3, #5-4, #11-2, #11-4, and #5-26.

Figure 4

KTI content declined dramatically in gene-edited seeds. KTI content was measured in 4 double mutated seeds (#2-3, #5-4, #11-2, and #11-4) and 4 seeds with a single mutation on KTI1 (#2-1, #5-1, #11-1, and #17-1), where the KTI content in WM82 seed served as the control. Experiments were conducted with three technical replicates and showed comparable results, shown as mean ± s.e. Different letters indicate significant differences.

Figure 5

TIA declined dramatically in the gene-edited seeds. Bovine trypsin enzyme activities were measured using crude extracts of 4 double mutated seeds (#2-3, #5-4, #11-2, and #11-4), 4 seeds with a single mutation on KTI1 (#2-1, #5-1, #11-1, and #17-1), and WM82. Experiments were conducted with three technical replicates and showed comparable results, shown as mean ± s.e. Different letters indicate significant differences.

Figure 6

The development of selection markers for breeding low KTI soybean varieties based on the kti1 and kti3 mutants generated by CRISPR/Cas9-mediated gene editing. (A) Schematic development of primers for amplification of wild type KTI1 (1), mutant KTI1 (2), wild type KTI3 (3), and mutant KTI3 (4). The red lines indicate the lost fragment in KTI1 or KTI3 during gene editing. The green lines indicate new DNA regions in kti1 or kti3 generated by splicing two fragments. (B) The 4 pairs of primers in (A) were utilized to amplify the alleles of KTI1, kti1, KTI3, and kti3 with gDNA of four different soybean genotypes, including WM82, three transgenic lines #5-26, #5-9, and #2-30. Based on our genotyping data, #5-26 has homozygous mutations of kti1 and kti3; #5-9 only has a homozygous mutation of kti1 but carries the heterozygous mutation of kti3; #2-30 only has a homozygous mutation of kti3 but carries the heterozygous mutation of kti1. Thus, it was clear that the pair of ZW1/ZW2 can amplify wild type KTI1 from WM82 and #2-30 gDNA in PCR tests, while the pair of ZW1/ZW3 can amplify mutant kti1 from #5-9 and #5-26 gDNA. Also, the pair of ZW4/ZW5 can amplify wild type KTI3 from WM82 and #5-9 gDNA, while ZW4/ZW6 can amplify mutant kti3 from #2-30 and #5-26 gDNA. As shown in the bottom panel, only the positive PCR products incubated with the dye of sybrgreen at 75°C can display the fluorescent signals, suggesting the reliability of the developed gel-electrophoresis-free method for screening mutant alleles of kti1 and kti3.