Schematic of zebrafish Adgrg6 alleles and the LOPAC screening pipeline. (A) Schematic of the Adgrg6 protein and the following domains: Signal peptide (SP), complement C1r/C1s, Uegf, BMP1 (CUB), pentraxin (PTX), sperm protein, enterokinase, and agrin (SEA), hormone receptor (HormR), GPCR autoproteolysis-inducing (GAIN), Stachel sequence (S), and 7-pass transmembrane (7TM) (for details, see references in.8 The GPCR autoproteolytic site (GPS) is shown (black arrow). The position of the amino acid change in the tb233c allele is shown,10 together with other alleles3, 10, 16 discussed in the text. Hypomorphic missense alleles and the amino acid changes are shown above in black; truncating nonsense mutations are shown in red. (B) Schematic of the screening pipeline. The LOPAC library was screened on homozygous adgrg6tb233c−/− hypomorphic mutant zebrafish embryos. Initially, all compounds were tested using the primary vcanb screen (blue) followed by two retests. Confirmed hits and some additional selected compounds (see text) were then tested in an optimized counter screen for mbp expression (orange). Images show the expression of vcanb in the developing inner ear (lateral views with anterior to the left and dorsal to the top, taken from Diamantopoulou et al.23) and mbp expression in the posterior lateral line ganglia (arrowheads) and nerves (dorsal views of the embryo; anterior to the left). Scores shown below the images were assigned as described previously.23 Scale bars: vcanb images, 50 μm; mbp images, 100 μm.

A primary drug screen of the LOPAC1280 library reveals 48 hit compounds that consistently down-regulate otic vcanb expression in adgrg6tb233c−/− embryos. (A) Overview of vcanb staining scores in the primary screening assay. Green, hits (score <6); grey, non-hits (score 6–9); red, toxic/corrosive compounds. (B) Total vcanb staining scores (summed from 3 embryos) from all compounds in the primary screen, including 64 initial hits (green). Library compounds are ordered along the x axis based on structural similarity. Red data points, toxic/corrosive compounds; grey, no effect on vcanb expression (non-hits); green, reduced vcanb expression (initial hits); blue, additional compounds selected for retest based on structural similarity with hits. (C) Comparison of lipophilicity of initial hit (green) and non-hit (grey) primary screen compounds. ****p < 0.0001, Welch's two-tailed t test. (D) Retesting 82 compounds revealed 48 (green and blue) that consistently down-regulated vcanb expression after two retests.

A counter screen reveals 17 hit compounds that can also rescue mbp expression in adgrg6tb233c−/− mutant embryos. (A) Overview of LOPAC library screen results. Hit compounds were grouped according to the strength of vcanb

Hit compounds that mediate rescue of both vcanb and mbp expression in mutant embryos, and are possible modulators of the Adgrg6 pathway. Lateral images of the inner ear at 90 hpf stained for vcanb and dorsal images of embryos at 78 hpf stained for mbp. Anterior to the left in all images. (A) Images of control group embryos, including wild types incubated in E3 and adgrg6tb233c−/− mutants incubated in DMSO or IBMX (positive control). IBMX restores wild-type levels of vcanb (inner ear) and mbp (PLLg) expression in adgrg6 mutant embryos. The dotted rectangle marks the region of interest (ROI) enclosing the left PLLg, enlarged in the inserts in the top right of each mbp panel. (B–D) Images of adgrg6tb233c−/− mutants treated with dihydropyridine hit compounds from the LOPAC screen (B), previously known hit compounds from the Spectrum and Tocris screen also identified in the LOPAC screen (C) or novel hit compounds from the LOPAC screen (D). Hits mediate partial or full restoration of vcanb expression in the inner ear, and mbp expression around the PLLg. Note the over-expression of mbp in the PNS of forskolin-treated embryos (C, lower panels). Compounds were tested at 25 and 15 μM in the vcanb and mbp assays, respectively. The IBMX control was tested at 100 and 50 μM in the vcanb and mbp assays, respectively. Arrow, mbp expression restored around the PLLg; asterisk, expression restored along the supraorbital lateral line nerve; c, expression restored along nerves innervating the inner ear cristae.

Performance of selected hits in dose–response assays on adgrg6tb233c−/− mutants. Ivermectin (A) and ebastine (B) were tested in a 1.5-fold dilution series on groups of nine adgrg6tb233c−/− hypomorphic mutant embryos. Charts show the number of mutant embryos (columns 2–11) that scored 0, 1, 2, or 3 in the vcanb and mbp assays, in comparison with untreated wild-type embryos (column 1). The blue shades represent the in situ hybridization staining score for each gene. For vcanb, a pale shade indicates full rescue (reduction) of otic expression to wild-type levels (vcanb score 0; see scale), whereas the darkest shade indicates no rescue (vcanb score 3). For mbp, the darkest shade indicates full rescue (restoration) of expression around the PLLg (mbp score 3), and pale shades no rescue (mbp scores 1 or 0). Toxic effects are highlighted in red (pale, toxic (T); dark, corrosive (C; no embryos in well)). (A) Example images show lateral view of an ear from a mutant embryo treated with 22.5-μM ivermectin in the vcanb assay (A I′, score 0), and dorsal view of a mutant embryo treated with 10-μM ivermectin in the mbp assay (A II′, score 2). (B) Example images show lateral view of an ear from a mutant embryo treated with 10-μM ebastine in the vcanb assay (B I′, score 1) and dorsal image of an embryo treated with 6.67-μM ebastine in the mbp assay (B II′, score 2). Arrows mark restoration of mbp expression around the PLLg.