CRISPR/Cas9-based genome-wide screening of Dictyostelium

Abstract

Genome-wide screening is powerful method used to identify genes and pathways associated with a phenotype of interest. The simple eukaryote Dictyostelium discoideum has a unique life cycle and is often used as a crucial research model for a wide range of biological processes and rare metabolites. To address the inadequacies of conventional genetic screening approaches, we developed a highly efficient CRISPR/Cas9-based genome-wide screening system for Dictyostelium. A genome-wide library of 27,405 gRNAs and a kinase library of 4,582 gRNAs were compiled and mutant pools were generated. The resulting mutants were screened for defects in cell growth and more than 10 candidate genes were identified. Six of these were validated and five recreated mutants presented with growth abnormalities. Finally, the genes implicated in developmental defects were screened to identify the unknown genes associated with a phenotype of interest. These findings demonstrate the potential of the CRISPR/Cas9 system as an efficient genome-wide screening method.

Figure 1

Detection of off-target mutations induced by CRISPR/Cas9 stable and transient expression vectors. (a) Percentage of on-target and off-target via three CRISPR vectors. The numbers and percentages of modified and unmodified loci. Outside seed and inside seed represent the presence of mismatches outside and inside the seed region of the target sequence, respectively. (b) Indel frequency for each on-target and off-target sequence. Mismatch sequences within off-targets and PAM sequences are presented in red and green letters, respectively. Seed sequence is underlined in red. Data represent two biological repeats and individual data points are shown. p-values were calculated using a one-way Welch two-sample t-test (*p < 0.05).

Figure 2

Generation of a genome-wide and kinase sgRNA plasmid library. (a) Target gRNA positions in the genome-wide and kinase libraries. Target sequences in genome library were designed within 400 bp of the start codon, while kinase libraries were designed with five gRNAs per 1 kb of each gene. Three representative genes are shown. The kinase domain of each gene is indicated by an orange box. (b) Distribution of designed gRNAs in genes. (c) Deep-sequencing analysis of the gRNAs in plasmid libraries and genome-wide (left) and kinase (right) mutant pools.

Figure 3

Genetic screening using the kinase sgRNA library and genetic validation of candidate genes regulating cell growth. (a) Schematic overview of genetic screening for cell growth with the kinase gRNA library. (b) Candidate genes identified by CRISPR-based screening. Data analysis using MAGeCK revealed under- (left) and over-represented (right) gRNAs and ranked the corresponding genes with RRA scores. Low RRA scores indicate stronger enrichment. (c) Growth curves of recreated CRISPR mutants. Data points represent the mean ± SD of 3–4 biological repeats and individual data points are shown. Samples were compared with Welch t-test; *p < 0.05; **p < 0.01.

Figure 4

Genetic screening using the CRISPR sgRNA libraries and genetic validation of candidate genes regulating development. (a) Schematic overview of genetic screening to identify developmental defective mutants. Mutant pools expressing genome-wide and kinase sgRNA libraries were inoculated on nutrient agar plates with bacteria as a food source and allowed to grow and develop. Isolated plaques exhibiting developmental defects, including aggregation-less and abnormal fruiting bodies, were transferred to 96-well plates. (b) Distribution of target sequences detected in developmental mutants. The numbers indicate the serial numbers of gRNAs for each gene and the bars below the gene indicate the frequency of isolation of mutant strains expressing the gRNAs. (c) Developmental defects in recreated CRISPR mutants. Aggregation-less mutants are indicated using arrowheads. Scale bars, 1 cm.

Figure 5

In-frame indel mutations in tor identified in the screens. (a) Developmental phenotypes of in-frame indel mutations in tor cells. The cells were plated on non-nutrient agar and imaged. Scale bars, 250 µm. (b) Chemoattractant-mediated activation of PKBR1 in 224 N tor mutants. Phosphorylated PkbR1 was analysed by immunoblotting and the band intensity was quantified. Mean values from three independent experiments and individual data points are shown. Error bars represent the SEM.