Polq-Mediated End Joining Is Essential for Surviving DNA Double-Strand Breaks during Early Zebrafish Development

Abstract

Error-prone repair of DNA double-strand breaks (DSBs) has been postulated to occur through classical non-homologous end joining (NHEJ) in systems ranging from nematode somatic tissues to zebrafish embryos. Contrary to this model, we show that zebrafish embryos mutant for DNA polymerase theta (Polq), a critical component of alternative end joining (alt-EJ), cannot repair DSBs induced by CRISPR/Cas9 or ionizing radiation. In the absence of DSBs, polq mutants are phenotypically normal, but they do not survive mutagenesis and display dramatic differences in the mutation profiles compared with the wild-type. These results show that alt-EJ repair is essential and dominant during the early development of a vertebrate.

Figure 1. Polq is required for survival following Cas9-induced DNA DSBs

All images are of approximately 36 hour post fertilization (hpf) embryos. Additional images are available in Figure S2. Sequences of the three gRNAs used are available in Table S1. a) Homozygous maternal-zygotic Polq (MZpolq) mutant. b) Wild-type (WT) embryo injected with Cas9 protein and gRNA for gria3a gene. c) WT embryo injected with Cas9 protein and gRNA for camk2g gene. d) MZpolq mutant injected with Cas9 protein and gRNA for gria3a gene. e) MZpolq mutant injected with Cas9 protein and gRNA for camk2g gene. f) Survival rates and fraction of deformed embryos observed for MZpolq and wild-type embryos injected with Cas9 protein and gRNAs targeting one of three genes. Embryos were injected at the 1-cell stage and all data was collected at approximately 36 hpf. Depictions of the observed deformities are available in Figure 1 and Figure S2a–c. The experiment was performed two times with n=15–47 for each injection. See also Figure S2 and Table S1.

Figure 2. Polq is required for survival following DSBs induced by ionizing radiation

a) Survival rates and fraction of deformed embryos observed for MZpolq and wild-type (WT) embryos exposed to varying levels of ionizing radiation at three stages of early development (n=14–32 for each condition). This data was collected at approximately 24 hpf. b) Examples of observed deformities in MZpolq embryos compared to wild-type embryos subjected to the same levels of radiation. Additional examples are available in Figure S2d. c) Survival rate observed for MZpolq and wild-type embryos exposed to varying levels of ultraviolet radiation at the shield stage of development. The experiment was performed two-four times per condition with n=14–54 each time. See also Figure S2.

Figure 3. Distribution of mutagenic alleles generated by Cas9-induced DNA repair

The types of alleles generated by Cas9 mutagenesis were determined by deep sequencing and are displayed with a gray arc connecting the bases flanking the deletion. The allele fraction is calculated by dividing the number of reads for the observed allele compared to the total reads for all mutated and wild-type (WT) alleles. The relative arc thickness also corresponds to this number. The PAM sequence for each gRNA is boxed in red. For each experiment 10–20 WT embryos were injected with a gRNA either targeting a) ctgfa, b) camk2g, or c) gria3a. Corresponding plots for single embryos are available in Figure S3a. Knockdown of Lig4 also does not influence the observed alleles (Figure S3b, Figure S3c). The observed mutations are likely guided by microhomologies for camk2g and ctgfa, as well as other sites, shown in Figure S4. Fraction mutagenesis, the number of sequencing reads containing an insertion or deletion induced by Cas9 cleavage and subsequent repair divided by the total number of sequence reads, for individual WT and surviving MZpolq embryos injected with gRNAs targeting d) ctgfa e) camk2g, and f) gria3a. In addition to reduced repair, oligonucleotide-mediated insertions are impaired in MZpolq embryos (Figure S5a, Figure S5b) and homologous recombination rates are not significantly higher (Figure S5c–e). See also Figure S3, S4, and S5.

Figure 4. Types of mutagenic alleles observed in repair of Cas9-induced breaks in MZpolq mutant embryos

Observed distribution of mutagenic alleles for MZpolq embryos injected with the gRNA targeting gria3a (n=6, fraction mutagenesis for each individual embryo is in Figure 3f). The purple arc indicates an insertion between the two ends of the arc (compared to the gray arcs representing deletions, Figure 3) and the inserted sequence is displayed at the top of the arc. The PAM sequence for the gria3a gRNA is boxed in red.