Fig. S3

Related to Figure 4: Pmel is a substrate for Bace2 but is not responsible for the wanderlust phenotype.

(A) zebrafish Bace2 cleaves Pmela. We constructed a fusion protein of zebrafish Pmela and a C-terminal Flag tag to detect cleavage by immunoblotting using Flag antibody. Expression of the fusion protein in HEK 293T cells reveals an immature core-glycosylated P1 form of Pmela (~150 kDa), the Mβ (~47 kDa), and the CTF (~18 kDa) when treated with γ secretase inhibitor DAPT (1μM), in a similar manner to human PMEL protein. Inhibition of human BACE2 using BACE inhibitor IV (1μM) increases the relative abundance of Mβ and decreases the abundance of the CTF, suggesting Bace2 cleaves Mβ and gives rise to CTF. Co-expression of the fusion protein and zebrafish Bace2 (zBACE2) leads to accumulation of cleavage product CTF, suggesting zebrafish Bace2 is sufficient to cleave zebrafish Pmela. (B) Electron microscopy of 72hpf embryos tail melanophores showed a difference in melanin distribution as well as melanosome (M) shape. Melanosomes in WT are round with a smooth outline and homogeneously filled with melanin (yellow arrow). bace2-/- melanophores have a darker hue in the cytosol (red arrow) and irregular melanosome membrane (yellow arrow), suggesting a leakage of melanin out of melanosomes into the cytosol. (C) Knockout of pmela using CRISPR-Cas9 in WT zebrafish leads to pale but not dendritic melanophores, similar to previously published fading vision mutant. This suggests that loss of pmela does not result in hyperdendritic melanophores. Melanophore membrane structure is visualized using Tg(tyrp1b: membrane- mCherry) strain at 3dpF. (D-E) Knockdown of pmelb alone (D) or co-knockdown of pmela and pmelb (E) using morpholinos in WT embryos does not lead to the hyperdendritic melanophore phenotype as seen in the bace2-/- mutant. Scale bars, 100 nM (B bottom panel), 1μM (B top panel), 100μM (C-E).