Systematic genome editing of the genes on zebrafish Chromosome 1 by CRISPR/Cas9

Abstract

Genome editing by the well-established CRISPR/Cas9 technology has greatly facilitated our understanding of many biological processes. However, a complete whole-genome knockout for any species or model organism has rarely been achieved. Here, we performed a systematic knockout of all the genes (1333) on Chromosome 1 in zebrafish, successfully mutated 1029 genes, and generated 1039 germline-transmissible alleles corresponding to 636 genes. Meanwhile, by high-throughput bioinformatics analysis, we found that sequence features play pivotal roles in effective gRNA targeting at specific genes of interest, while the success rate of gene targeting positively correlates with GC content of the target sites. Moreover, we found that nearly one-fourth of all mutants are related to human diseases, and several representative CRISPR/Cas9-generated mutants are described here. Furthermore, we tried to identify the underlying mechanisms leading to distinct phenotypes between genetic mutants and antisense morpholino-mediated knockdown embryos. Altogether, this work has generated the first chromosome-wide collection of zebrafish genetic mutants by the CRISPR/Cas9 technology, which will serve as a valuable resource for the community, and our bioinformatics analysis also provides some useful guidance to design gene-specific gRNAs for successful gene editing.

Figure 1.

GC-content distribution in the 12-nt seed sequence of all the tested target sites (including 1086 positive sites and 1191 negative sites). (A) GC percentage of positive and negative target sites at the seed region, respectively. (B) GC percentage distribution of the seed region in all the tested target sites. GC count: number of G or C nucleotides in the 12-nucleotide (nt) seed sequence of the target sites. (C) Correlation of the positive rate of target sites with different GC percentages of the seed region. GC count: number of G or C nucleotides in the 12-nt seed sequence of the target sites.

Figure 2.

Distribution of nucleotide motifs in the 12-nt seed region of all the tested target sites. (A) The statistical data showing the distribution of each single nucleotide in the seed region of the target sites. (B) The statistical data showing the distribution of 2-nt motifs in the seed region of the target sites. (C) The statistical data showing the distribution of 3-nt motifs in the seed region of the target sites.

Figure 3.

Characterization of plrg1 mutant generated by the CRISPR/Cas9 system. (A) Whole-mount in situ hybridization (WISH) showing the expression of plrg1 at different developmental time points from the one-cell stage to 36 h postfertilization (hpf). (B) The comparison of genomic DNA sequences between wild type (WT) and plrg1 mutants with 10-base pair deletion. (C) The plrg1 mutants and morphants showed severe developmental defects, with black head and small body compared to the siblings and control embryos, respectively, at 24 hpf. (D) WISH showing the expression of gsc and ntl at the 50% epiboly stage, sox17 at 90% epiboly stage, and ntl at bud stage in the offspring of plrg1 heterozygous parents. The right panels show the magnified images, and the black arrowheads indicate corresponding expression of gsc at the dorsal margin, ntl at anterior axial hypoblast, forerunner cell group, and margin, and sox17 at endoderm and forerunner cells. (E) WISH showing the expression of lmo2, gata1, and scl at lateral plate mesoderm in the siblings and plrg1 mutants. (F) Overexpression of the plrg1 full-length mRNA can rescue plrg1 mutants until 4 dpf. The body defects of mutants were rescued efficiently by mRNA overexpression, but there is still a black head at 48 hpf (arrowhead) in plrg1 mutants. (G) TUNEL assay displays that there are obvious apoptotic signals at 48 hpf in plrg1 mutants injected with plrg1 mRNA. (H) Injection of tp53 morpholino can rescue the developmentally defective phenotype of plrg1 mutants efficiently at 24 hpf. There are three subtypes of defective embryos, and we describe the siblings as normal, plrg1 mutants as severe (S), and partial rescued mutants as mild (T1) and mild (T2). (I) The quantification of plrg1 mutant embryos in different treatment groups shown in H.

Figure 4.

Phenotypic comparison between klf3 mutants and morphants. (A) WISH showing the expression of hbbe2 (also known as βe2-globin) in klf3 WT, morphants, and mutants. The black arrowheads indicate hbbe2 expression in erythroid lineages. (B) Quantitative real-time PCR showing the expression of Klf members klf1, klf3, klf4b, klf6a, and klf8 in klf3 mutants.