Efficient deletion of multiple circle RNA loci by CRISPR-Cas9 reveals Os06circ02797 as a putative sponge for OsMIR408 in rice

Abstract

CRISPR-Cas9 is an emerging genome editing tool for reverse genetics in plants. However, its application for functional study of non-coding RNAs in plants is still at its infancy. Despite being a major class of non-coding RNAs, the biological roles of circle RNAs (circRNAs) remain largely unknown in plants. Previous plant circRNA studies have focused on identification and annotation of putative circRNAs, with their functions largely uninvestigated by genetic approaches. Here, we applied a multiplexed CRISPR-Cas9 strategy to efficiently acquire individual null mutants for four circRNAs in rice. We showed each of these rice circRNA loci (Os02circ25329, Os06circ02797, Os03circ00204 and Os05circ02465) can be deleted at 10% or higher efficiency in both protoplasts and stable transgenic T0 lines. Such high efficiency deletion enabled the generation of circRNA null allele plants without the CRISPR-Cas9 transgene in the T1 generation. Characterization of the mutants reveals these circRNAs' participation in salt stress response during seed germination and in particular the Os05circ02465 null mutant showed high salt tolerance. Notably, the seedlings of the Os06circ02797 mutant showed rapid growth phenotype after seed germination with the seedlings containing higher chlorophyll A/B content. Further molecular and computational analyses suggested a circRNA-miRNA-mRNA regulatory network where Os06circ02797 functions to bind and sequester OsMIR408, an important and conserved microRNA in plants. This study not only presents genetic evidence for the first time in plants that certain circRNAs may serve as sponges to negatively regulate miRNAs, a phenomenon previously demonstrated in mammalian cells, but also provides important insights for improving agronomic traits through gene editing of circRNA loci in crops.

Keywords: CRISPR-Cas9; circle RNA; large deletion; microRNA sponge; rice.

Figure 1

CRISPR‐Cas9‐mediated deletion of four circRNA loci in rice. (a) Genomic locations of four circRNA genes in this study. Exons are indicated as black boxes. UTRs are indicated as grey boxes. The sgRNA region of each circRNA is indicated by two red triangles. Note the circRNA genes (Os02circ25329 and Os06circ02797) are located in intragenic introns, while the other two circRNA genes (Os03circ00294 and Os05circ02465) are located in intergenic regions. (b) Schematics for four multiplexed CRISPR‐Cas9 T‐DNA vectors (pZJP053, pZJP054, pZJP055 and pZJP057) with each expressing two sgRNAs for targeted deletion of each circRNA gene. The Cas9 gene is expressed under a maize ubiquitin 1 promoter (pZmUbi1), and the sgRNAs are expressed under a rice U6 promoter (pU6) and U3 promoter (pU3), respectively. (c) Targeted chromosomal deletion of four circRNA loci in rice protoplasts. Compared to the untransformed protoplasts, chromosomal deletions of circRNA genes were detected by PCR in transformed protoplasts.

Figure 2

Genotypes of transgene‐free homozygous deletion lines in the T1 generation. Genotypes of homozygous lines containing chromosomal deletions of rice circRNAs were validated by PCR followed with Sanger sequencing. For each circRNA, two independent homozygous genotypes were identified. The CRISPR‐Cas9 transgenes were segregated out from these lines as confirmed by PCR.

Figure 3

Transcriptional characterization of the circRNAs and their flanking genes in the rice circRNA mutants. (a) Detection of circRNAs in the wild‐type (WT) and circRNA mutants. Note two pairs of PCR primers were used for detecting each circRNA. The convergent PCR primers (indicated by two black face‐to‐face inward arrows) were used to detect genomic DNA (gDNA) and complementary DNA (cDNA) of each circRNA gene. The divergent PCR primers (indicated by two black back‐to‐back outward arrows) were used to specifically detect circRNAs. (b) (First) Relative gene expression of the Os02circ25329‐containing gene, Os02g50174, in the WT and circRNA mutant backgrounds. Relative gene expression of the Os06circ02797‐containing gene, Os06g04610, in the WT and circRNA mutant backgrounds. (Second) Relative expression of the two genomic genes (Os03g01350 and Os03g1360) flanking the Os03circ00204 gene in the WT and circRNA mutant backgrounds. (Third) Relative expression of the two genomic genes (Os05g04950 and Os05g04960) flanking the Os05circ02465 gene in the WT and circRNA mutant backgrounds. (Fourth) Gene expression was analysed by quantitative reverse transcription PCR (qRT‐PCR). The effort bars represent standard deviations of three biological replicates. Statistical significance is indicated by asterisks (** indicative of a P‐value < 0.01 by the student’s t‐test).

Figure 4

Differential responses of rice circRNA mutants to salt stress during seed germination. (a) A representative picture of seed germination under salt stress treatment for the WT plants and rice circRNA mutants. The seeds were germinated under different NaCl concentrations as indicated, and the picture was taken 7 days after sowing the seeds. (b) Quantification of germination rates for different genotypes under different NaCl concentrations. The error bars represent standard deviations (n = 3). Statistical significance is indicated by asterisks (* indicative of a P‐value < 0.05 by the Student t‐test).

Figure 5

Transcriptome analysis of rice circRNA mutants. (a) Quantification for the numbers of differentially expressed miRNAs in the rice circRNA mutants as compared to the WT background. (b) Quantification for the numbers of differentially expressed protein‐encoding genes in the rice circRNA mutants as compared to the WT background. (c) Pathway enrichment analyses for the differentially expressed genes in four rice circRNA mutants.

Figure 6

Os06cic02797 negatively regulates OsMIR408 expression. (a) Phenotypes of 7‐day‐old WT plants and os06circ02797∆1 mutants. (b). Quantification of chlorophyll A/B content in the WT plants and os06circ02797∆1 mutants. Error bars represent standard deviations (n = 3). (c) Quantification for the height of seedlings in the 7‐day‐old WT plants and os06circ02797∆1 mutants. Error bars represent standard deviations (n = 30). (d) Quantification for the length of 1^st leaf sheath in the 7‐day‐old WT plants and os06circ02797∆1 mutants. Error bars represent standard deviations (n = 30). (e) Relative expression of miR408‐5p and 9 potential targeting genes in the WT and os06circ02797∆1 mutant backgrounds. Error bars represent standard deviations (n = 3). (f) Bioinformatic analysis of seven OsMIR408‐binding sites in the Os0circ02797 circRNA using a web tool (http://www.rna‐society.org/raid/; The parameters are set as follows: Number of (sub)optimal interactions: 10, Suboptimal interaction overlap: can overlap in query, and Others: default). (g) A model depicting the Os06circ02797 circRNA functions as a sponge for OsMIR408 to negatively regulate its function in rice. Statistical significance is indicated by asterisks (* indicative of a P‐value < 0.05 by Student’s t‐test; ** indicative of a P‐value < 0.01 by Student’s t‐test).