Fig. 5

A targeted CRISPR screen in zebrafish identifies genes regulating hypertrophic or hyperplastic adipose morphology. (a) Overview of CRISPR screen structure. In total, 1371 fish were screened (738 Cas9-only controls and 633 mutants) across 25 candidate genes. Two genes (phb and rerea) were lethal before 5 days post fertilisation (dpf). Of the remaining genes, zero altered standard length, one (kazna) significantly increased total adiposity, and three (txnipa, mmp14b and foxp1b) produced significant hypertrophic subcutaneous adipose morphology. (b) Sample sizes per gene. Grey bars indicate Cas9-only controls; blue bars indicate mutants. aspa had the smallest and prrx1b the largest sample size. (c) Left panel: number of experimental replicates per gene. The majority of targets had two replicates: aspa had one replicate, and nav3, prrx1b, tbx15, and txnipa had three. Right panel: replicate consistency scores, representing agreement in effect direction and magnitude across replicates. Genes are ordered by consistency score. (d–f) Forest plots showing effect sizes (% change from Cas9-only controls) for standard length (d, magenta), total adiposity (e, green), and morphology value (f, blue). Small grey dots represent individual mutant fish; medium grey circles represent experiment means; coloured diamonds represent gene-level estimates from linear mixed models (LMMs) with experiment as a random effect. Morphology value represents the deviation in lipid droplet (LD) Feret diameter from that expected given total depot area, with positive values indicating hypertrophy and negative values indicating hyperplasia. Asterisks denote BH-adjusted p<0.05. (g) Scatter plot of depot (subcutaneous adipose tissue [SAT]) lipid area (% change from controls) versus morphology value (LMM estimate, % change from controls) for each gene. Point size reflects sample size. Quadrants indicate the combination of depot size and cell morphology phenotype (e.g. bigger depot with hypertrophic cells). (h) Probability density functions (left) and cumulative distribution functions (right) of morphology values for foxp1b, txnipa, mmp14b, and kazna F0 mutants (blue) compared with matched Cas9-only controls (grey). Dashed lines indicate group means. (i) Representative Nile Red images of foxp1b, txnipa, mmp14b, and kazna F0 mutants showing total adiposity. e, eye. Scale bar: 1 mm. (j) Segmented subcutaneous adipose LDs from representative mutant fish, colour-coded by LD diameter. Scale bar: 200 μm.