Craniofacial features visualized in frontal/rostral mounts by DAPI staining and confocal microscopy. (A-D) Rostral (A,D), lateral (B) and ventral (C) views of developing zebrafish larvae at 5 days post-fertilization (dpf). (E) Schematic representing the facial anatomical structures visualized in the rostral view, including neuromasts, olfactory placodes, eyes, mouth, frontonasal region (FNP), maxillary region (MX) and mandibular region (MD). Scale bars: 50 µm (A-C), 20 µm (D).

Facial analytics based on a coordinate extrapolation system (zFACE) workflow for morphometric analysis of the developing zebrafish craniofacial region. Schematic representing the steps in zFACE analysis, in which DAPI-stained larvae are mounted, images are acquired, points are placed on the 26 anatomical landmarks, and either feature-based zFACE calculations or overall shape analyses can be performed.

Time series of zebrafish craniofacial development. (A) Zebrafish larvae were analyzed from 2 to 6 dpf. Scanning electron microscope (SEM) images confirmed the soft tissue morphology captured by DAPI staining and confocal microscopy used in zFACE. (B) Changes in zFACE features were followed across developmental time points, and alterations in 28 measurements were determined (representative changes are shown in the graphs). The numbers of individual larvae for each group were as follows: n=30 at 3 dpf, n=31 at 4 dpf, n=28 at 5 dpf and n=28 at 6 dpf. Specifically, 20 zFACE measurements changed between 3 and 4 dpf, 12 changed between 4 and 5 dpf, and 6 nominally changed between 5 and 6 dpf. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001; n.s., not significant (one-way ANOVA). (C) Untransformed zFACE features were subjected to multivariate statistical testing using principal component analysis (PCA). The first two components captured 66% of the variance in the dataset. There were differences in developmental times that were captured across PC1, with no strain-specific differences observed. Scale bars: 50 µm.

zFACE analysis of zebrafish with loss of smarca4a. (A-C) Confocal images of wild-type (WT; A), heterozygous (Het; B) and homozygous mutant (Mut; C) smarca4a larvae. (D) zFACE analysis pinpointed 33 significantly altered facial measurements between the groups, with examples of dimensions that were unaltered, decreased and increased shown. The numbers of larvae analyzed for each group were n=28 for WT, n=34 for Het and n=32 for Mut. ****P<0.0001; n.s., not significant (one-way ANOVA). (E) Untransformed zFACE features were subjected to multivariate statistical testing using PCA. The first two components explained 68% of the variance in the dataset, and the principal component (PC) plot showed clear separation of the homozygous mutant group from the WT and heterozygote groups. (F) Discriminant function analysis after landmark data were Procrustes transformed further supported that smarca4a homozygous mutants had different average facial shapes compared to either WT or heterozygous larvae, whereas WT and heterozygous groups had the same facial shape. Wireframe representations of facial shapes highlight the alterations to landmarks in the upper face, eyes, oral cavity, midface and lower jaw in the homozygous smarca4a mutants [discriminant function analysis (DFA)]. Scale bar: 50 µm.

smarca4a mRNA expression at 5 dpf. (A-D) Fluorescent in situ hybridization showed that smarca4a mRNA is expressed in multiple facial tissues and underlying brain structures in WT and heterozygous zebrafish (A,B), whereas smarca4a^−/− mutants showed dramatically reduced smarca4a mRNA expression (C,D). Scale bars: 50 µm (A,C), 20 µm (B,D).