Fig. S2

Efficiency and specificity of sox4 MOs and sox4 CRISPR phenotype. (A-D′) EF1α-EGFP plasmids containing the sox4aor sox4b MO binding site were co-injected with either control MO or sox4 MO. At 24 hpf, control MO injected embryos showed ubiquitous GFP expression (A-B′), whereas sox4 MO injected embryos showed no GFP expression (C-D′). (E) Quantification of the percentage of injected embryos with GFP expression at 24 hpf. In control MO injected embryos, 91.94±1.93% had GFP expression (n=107/116). In sox4 MO injected embryos, none of them showed any GFP expression (n=0/227). (F) Sox4a and sox4b are both required for choroid fissure closure. Two different doses of sox4a and sox4b MO were injected into zebrafish embryos either separately or combined. A synergistic effect on the incidence of coloboma was observed when sox4a and sox4bMO were co-injected at both doses (Fisher′s exact test, P<0.0001). (G-J) A second set of non-overlapping sox4a/b MOs also resulted in coloboma at similar proportions to sox4 MO1 (40.85±8.39%, n=183/448). Bracket and arrows in (H) and (I) indicate the open choroid fissure and extrusion of the retina at the coloboma region. (K-L) Representative images of eyes from uninjected control (UIC) and tyrosinase (tyr)sgRNA/Cas9 injected embryos at 48 hpf. The tyr sgRNA/Cas9 injected embryos displayed a reduction in retinal pigmentation but did not have coloboma. (M-N) Representative images of eyes from sox4a and sox4bsgRNA/Cas9 injected embryos at 48 hpf that displayed coloboma (asterisk). (O) Quantification of the percentage of embryos with coloboma at 48 hpf. (P-Q) HRMA analysis detected the presence of mutant alleles in individual sox4a/b sgRNA/Cas9 injected embryos (arrows), which were confirmed by sequencing (not shown).