An efficient CRISPR-Cas12a-mediated MicroRNA knockout strategy in plants

Abstract

In recent years, the CRISPR-Cas9 nuclease has been used to knock out MicroRNA (miRNA) genes in plants, greatly promoting the study of miRNA function. However, due to its propensity for generating small insertions and deletions, Cas9 is not well-suited for achieving a complete knockout of miRNA genes. By contrast, CRISPR-Cas12a nuclease generates larger deletions, which could significantly disrupt the secondary structure of pre-miRNA and prevent the production of mature miRNAs. Through the case study of OsMIR390 in rice, we confirmed that Cas12a is a more efficient tool than Cas9 in generating knockout mutants of a miRNA gene. To further demonstrate CRISPR-Cas12a-mediated knockout of miRNA genes in rice, we targeted nine OsMIRNA genes that have different spaciotemporal expression and have not been previously investigated via genetic knockout approaches. With CRISPR-Cas12a, up to 100% genome editing efficiency was observed at these miRNA loci. The resulting larger deletions suggest Cas12a robustly generated null alleles of miRNA genes. Transcriptome profiling of the miRNA mutants, as well as phenotypic analysis of the rice grains revealed the function of these miRNAs in controlling gene expression and regulating grain quality and seed development. This study established CRISPR-Cas12a as an efficient tool for genetic knockout of miRNA genes in plants.

Keywords: CRISPR‐Cas12a; MicroRNA; genome Editing; germplasm innovation; rice.

Figure 1

Cas12a is more efficient than Cas9 in generating null alleles of OsMIR390. (a) Comparison of the effects of different CRISPR‐Cas nucleases in gene editing of miRNAs. Cas12a is more likely to produce larger deletions, causing damage to secondary structure of pre‐miRNA, unable to produce mature miRNA and resulting in complete loss‐function of miRNA. The Cas12a nuclease outperforms the Cas9 nuclease for miRNA gene editing. (b) Secondary structure of primary OsMIR390. Matured osa‐miRNA 5P sequence is in red, matured osa‐miRNA 3P sequence is in blue, CRISPR‐Cas12a targeted site and CRISPR‐Cas9 targeted site are indicated by arrows. (c) Comparison of the effects of knockout OsMIR390 with Cas9 and Cas12a on the efficiency of regeneration and transformation in rice. (d) Knockout of OsMIR390 with Cas12a significantly inhibited the regeneration of rice calli. (e) Cas12a‐generated OsMIR390 T0 mutants were all heterozygous, with the genotypes of >3 nt deletion. (f) Cas9‐generated OsMIR390 T0 mutants were heterozygous and homozygous, with the genotypes of 1 nt insertion. (g) Comparison of pre‐miRNA prediction secondary structures of OsMIR390 mutants generated by Cas9 and Cas12a. (h) Comparison of seeds germination for 3 days of OsMIR390 T1 mutants generated by Cas9 and Cas12a. The T1 generation heterozygous seeds of the OsMIR390 mutants germinated normally like WT; Cas12a‐generated T1 homozygous seeds cannot germinate. Few of Cas9‐genetated T1 homozygous seeds can germinate, but only produce radicle without any bud. (i) Comparison of 14‐day seedlings of OsMIR390 T1 mutants generated by Cas9 and Cas12a. The T1 generation heterozygous seedlings of the OsMIR390 mutants grow normally as well as WT. Cas12a‐generated T1 seeds did not grow into seedlings, and few of Cas9‐generated T1 homozygous seeds had elongated lateral roots without any shoot.

Figure 2

Cas12a generated much larger deletions at miRNA loci in rice. (a) Expression patterns of nine primary miRNAs used in this study. All data come from PmiRExAt (http://pmirexat.nabi.res.in). (b) Secondary structure of nine primary miRNAs used in this study. All data comes from miRBase database (https://www.mirbase.org). The crRNA targeted site is underline. (c) List of efficiency in creating different miRNA mutants by CRISPR‐Cas12a nuclease. (d) Characteristic comparisons of the Cas12a and Cas9 nucleases in miRNA gene editing shows that Cas12a generates much larger deletions and outperforms the Cas9 for miRNA gene editing. Cas12a mutation data are from Cas12a‐generated T0 mutants of 9 OsMIRNAs shown in Figures S1–S3; All Cas9 mutation data are derived from Cas9‐generated T0 mutants published in 2017 (Zhou et al., 2017), including OsMIR408, OsMIR528, OsMIR815a, OsMIR815b, OsMIR815c, OsMIR820a, OsMIR820b and OsMIR820c. (e) The pre‐miRNA prediction secondary structures of different Cas12a‐genetated OsMIR394 mutants. (f) The pre‐miRNA prediction secondary structures of different Cas12a‐genetated OsMIR827 mutants. (g) The pre‐miRNA prediction secondary structures of different Cas12a‐genetated OsMIR3979 mutants. (h) The pre‐miRNA prediction secondary structures of different Cas12a‐genetated OsMIR5789 mutants. (i) The pre‐miRNA prediction secondary structures of different Cas12a‐genetated OsMIR5794 mutants.

Figure 3

Characterization of Cas12a‐generated knockout mutants of select miRNAs by small RNA‐seq and mRNA‐seq. (a) Small RNA seq results of matured OsMIR394, OsMIR3979 and OsMIR5794 expression amount in WT and three Cas12a‐generated mutants. (b) Comparison of changes in expression of pri‐miRNA and mature‐miRNA. According to the asymmetry of pri‐miRNA and mature‐miRNA regulation (e.g., precursors are up‐regulated but mature miRNAs do not change), miRNAs are divided into nine categories, and the number is counted by category. (c) Differential expression miRNA of Cas12a‐generated mutants of OsMIR394, OsMIR3979 and OsMIR5794 by sRNA‐seq. (d) Differential expression genes of Cas12a‐generated mutants of OsMIR394, OsMIR3979 and OsMIR5794 by mRNA‐seq. (e) Wayne diagram of up‐regulated genes in Cas12a‐generated mutants of OsMIR394, OsMIR3979 and OsMIR5794 by mRNA‐seq. (f, g) Wayne diagram of up‐ and down‐regulated genes in Cas12a‐generated mutants of OsMIR394, OsMIR3979 and OsMIR5794 by mRNA‐seq.

Figure 4

Characterization of grain phenotypes in knockout mutants of five miRNAs in rice. Comparison of grain length (a), grain width (b) and grain thickness (c) of five miRNA knockout mutants. n = 10 grains × 3 times. (d) The grain length and grain width of OsMIR394 knockout mutants increased significantly; Bar = 8 mm. (e) The seed of the OsMIR827 knockout mutants became elongated, significantly increased in grain length and decreased in width; Bar = 8 mm. (f) The grain length of OsMIR3979 knockout mutants increased significantly, and the grain width did not change; Bar = 8 mm. (g) 1000‐grain weight comparison of five miRNA knockout mutants. n = 3 times. (h) The epidermal cell length, width and numbers of OsMIR394, OsMIR827 and OsMIR3979 knockout mutant seeds by SEM. Illustration of SEM at 15×, 150× and 400× magnification. (i) Comparison of the average length and the cell numbers of epidermal cells in the longitudinal direction of five miRNA knockout mutant seeds, n = 10 capsules × 3 times. (j) Comparison of the average length and the cell numbers of epidermal cells in the transverse direction of five miRNA knockout mutant seeds, n = 10 capsules × 3 times. Data are means ± SD. Statistical significance is indicated by asterisks. *P < 0.05; **P < 0.01; ***P < 0.001, two‐way ANOVA, Student's t‐test.

Figure 5

Characterization of starch composition and structure in the mutants of five miRNAs in rice. (a) Comparison of seed transparency of 5 miRNA knockout mutants. Bar = 5 mm. (b) Cross‐section of five miRNA knockout mutant seeds, showing chalky white (heart white). Bar = 1 mm. (c) Comparison of chalkiness of five miRNA knockout mutant seeds. n = 3 times, 50 grains each time. Total starch content (d), amylose content (e) and Amylopectin content (f) of five miRNA knockout mutant seeds. (g) The SEM of five miRNA knockout mutant seeds cross‐section, showing the starch grain structure at 30×, 1000× and 2000× magnification. (h) KEGG‐enrichment of starch and sucrose metabolism pathway genes affected by the three miRNA mutants. Expression of ADP‐glucose pyrophosphorylase small subunit 1 (APS1) and soluble starch synthase I (SSI) in the OsMIR5794 knockout mutant increased significantly, which may have caused an increase in amylopectin content in the OsMIR5794 mutant. Data are means ± SD. Statistical significance is indicated by asterisks. *P < 0.05; **P < 0.01; ***P < 0.001, two‐way ANOVA, Student's t‐test.

Figure 6

Phenotypic analysis seed germination and seedling growth of miRNA mutants. (a) Comparison of germination ratio and phenotypes of five miRNA knockout mutant seeds at 10 days; n = 30 seeds × 3 times. (b) Comparison of germination rate of five miRNA knockout mutant seeds, showing the germination ratio at 1‐, 4‐, 7‐ and 10‐day. n = 30 seeds × 3 times. Comparison of seedling height (c) and root length (d) of five miRNA knockout mutant seedlings at 10 days. n = 10 seedlings × 3 times. (e) Phenotype of five miRNA knockout mutant seedlings at 10 days. (f) KEGG‐enrichment of the auxin biosynthesis and signalling transduction pathway genes affected by the three miRNA mutants. The expression of TAA1, a key gene in the tryptophan‐dependent IAA synthesis pathway in OsMIR394 and OsMIR5794 knockout mutants, was up‐regulated, which may promote endogenous IAA synthesis and affect the seedling growth of mutants. In the auxin signalling pathway, GH3.8 expression of OsMIR394 knockout mutants is up‐regulated, which may promote seedling growth. (g) KEGG‐enrichment of the diterpenoid biosynthesis pathway genes affected by the three miRNA mutants. The CYP714B gene of OsMIR3979 mutant is up‐regulated and GA20ox1 is down‐regulated, which may affect the synthesis of GA4 and GA1, thereby inhibiting seed germination and affecting the germination rate of mutant seeds. Data are means ± SD. Statistical significance is indicated by asterisks. *P < 0.05; **P < 0.01; ***P < 0.001, two‐way ANOVA, Student's t‐test.