a Schematic illustration showing alleles of the sentinel SNP rs72959041 and the tagging SNP rs577721086. b Allelic expression analysis of RSPO3 transcripts was performed by qRT-PCR in abdominal and gluteal AT cDNA from 14 heterozygous carriers at rs72959041 (seven females, seven males). The proportion of total cDNA and gDNA containing the (WHR-increasing) rs577721086-C allele (r² = 0.98 with rs72959041-A) is quantified on the y-axis. Error bars are median values with 95% confidence intervals. Statistical significance was assessed by two-tailed paired Student’s t-tests versus gDNA. cRSPO3 WHRadjBMI GWAS signals co-localise with subcutaneous adipose cis-eQTLs. d, e Allelic expression of RSPO3 transcripts were analysed in abdominal and gluteal isolated adipocyte (ADS) (d) and cultured AP (e) cDNAs from ten heterozygous female carriers at rs72959041 as in b. f Schematic illustration showing alleles of the enhancer SNP rs9491696 and the tagging SNP rs1892172. g–i Allelic expression analysis of RSPO3 in g abdominal and gluteal AT cDNA from 32 individuals (15 females, 17 males) and in h abdominal and gluteal ADS and i cultured AP cDNA from 17 and 19 females, respectively, who are heterozygous at rs9491696. All individuals selected are homozygous for the rs72959041-G (non-risk) allele. The proportion of cDNA and gDNA containing the WHRadjBMI-increasing rs1892172-A allele (r² = 0.84 with rs9491696-G) is quantified on the y-axis. Error bars are median values with 95% confidence intervals. Statistical significance was assessed by two-tailed paired Student’s t-tests versus gDNA. j–m Median adipocyte cell weight calculated from median cell area measured from j abdominal and l gluteal AT histological sections from 14 and 15 pairs, respectively, of age- and BMI-matched females grouped by rs72959041 genotype. Cumulative frequency distribution of the adipocyte cell-surface area are shown for abdominal (k) and gluteal (m) AT. More than 250 cells were measured for each biopsy. Box and whisker plot: centre line, median; box limits, upper and lower quartiles; and whiskers, maximum and minimum values. Statistical significance was assessed by a two-tailed Wilcoxon signed-rank test. Source data are provided as a Source Data file.

Normalised RSPO3 mRNA levels in: a paired SC abdominal (Abdo) versus gluteal (Glut) fat biopsies from healthy women (n = 103) and men (n = 97), b paired visceral versus SC abdominal AT from 11 women and 16 men, and c fractionated AT [cultured APs and isolated mature adipocytes (ADS)] from 59 women. Histogram data are means ± s.e.m. **p < 0.01, ***p < 0.001, statistical significance was assessed by univariate analyses adjusted for age and BMI, and with Bonferroni correction for multiple testing for a, and by two-tailed paired Student’s t-tests for b and c. Source data are provided as a Source Data file.

a–f Effects of RSPO3-KD in immortalised (im) APs. a RSPO3-KD was confirmed by qRT-PCR (Abdo, n = 7 experiments; Glut, n = 9 experiments). b RSPO3-KD results in increased doubling time in imAbdo but not imGlut APs. Doubling time of control (shCON) and RSPO3-KD (shRSPO3) imAbdo (n = 15 experiments) and imGlut APs (n = 14 experiments). c–f RSPO3-KD promotes adipogenesis more robustly in imGlut versus imAbdo cells. Representative micrographs of shCON and shRSPO3 imAbdo and imGlut cells after 14 days of differentiation are shown (c). Scale bar = 500 μm. Adipogenesis was assessed by d AdipoRed staining (four independent experiments, n = 12 replicates each, expressed as relative lipid levels), and e–f qRT-PCR of adipogenic genes (n = 5 experiments). g–j Effects of RSPO3-KD in primary (1°) APs. (g) RSPO3-KD was confirmed by qRT-PCR (n = 6 experiments). RSPO3-KD leads to increased doubling time in 1°Abdo APs (n = 6 experiments) (h) and enhanced adipogenesis in 1° Glut cells (five independent experiments) (i). Representative micrographs of AdipoRed-stained control and shRSPO3 cells at differentiation day 12 are shown (j). Intracellular lipids are stained yellow. Scale bar = 200 μm. k–m Effects of rhRSPO3 treatment on 1° AP (k) proliferation (n = 15 [cells from three female subjects]), and l–m adipogenesis under l serum-free conditions (n = 13 independent experiments [cells from seven female donors]) and m in the presence of serum (FBS) (n = 2 independent experiments, eight replicates each). veh., vehicle; R3 (5), 5 ng ml⁻¹ rhRSPO3; R3 (50), 50 ng ml⁻¹ rhRSPO3. (a, b, d–i, m) Histogram data are means ± s.e.m. k, l Error bars are median values with interquartile ranges. *p < 0.05, **p < 0.01, ***p < 0.001. Statistical significance was assessed by two-tailed Student’s t-test (a, b, e–i, paired; d, m, unpaired, and with Bonferroni correction for m) and a two-tailed Wilcoxon signed-rank test (k, l). Source data are provided as a Source Data file.

a Light microscopy of abdominal and gluteal DFAT stable cell lines after 14 days adipogenic differentiation. Scale bar = 100 μm. b, c Expression of RSPO3 in DFAT stable cell lines at day 15 of adipogenic differentiation following ~ 48-h treatment with 0.05 µg ml⁻¹ doxycycline versus vehicle in hormone-free basal media (n = 5 experiments). d, e Insulin-stimulated glucose uptake in DFAT cells relative to basal glucose uptake following ~ 48-h treatment with 0.05 µg ml⁻¹ doxycycline or vehicle in hormone-free basal media (n = 12, from four independent experiments). f, g Isoproterenol-stimulated glycerol release in DFAT cells relative to basal glycerol release following ~ 48-h treatment with 0.05 μg ml⁻¹ doxycycline or vehicle (n = 12, from four independent experiments). b–g Histogram data are means ± s.e.m. *p < 0.05, **p < 0.01. Statistical significance was assessed by two-tailed paired (b, c) and unpaired (d–g) Student’s t-test comparing doxycycline versus vehicle treatment. h, i Transcriptional profiling of in vitro differentiated abdominal and gluteal DFAT cells following 48-h doxycycline-induced RSPO3-KD. h Number of differentially expressed (DE) genes. FDR p < 0.05. i Gene-set enrichment analysis results of DE genes in gluteal DFAT dox-induced RSPO3-KD cells showing the top 10 GO biological processes. j, k Effect of doxycycline-induced RSPO3-KD in differentiated j abdominal, and k gluteal cells on apoptosis in the presence of indicated concentrations of rhTNFα (ng ml⁻¹). Apoptosis was assayed using Caspase Glo 3/7 reagent. Results are shown as a % of luminescence of non-dox (vehicle)-treated cells in the presence of the same concentration of rhTNFα (n = 7 independent experiments). Statistical significance was assessed by a two-tailed Wilcoxon signed-rank test comparing dox and non-dox treated cells. Solid symbols, tet-shCON cells; open symbols, tet-shRSPO3 cells; circles, no rhTNFα; squares, 10 ng ml⁻¹ rhTNFα; triangles, 100 ng ml⁻¹ rhTNFα. Error bars are median values with interquartile ranges. Source data are provided as a Source Data file.

a Western blots of canonical WNT signalling markers and b qRT-PCR of AXIN2, a WNT/β-catenin target gene, in control (shCON) and RSPO3-KD (shRSPO3) imAbdo (n = 7 experiments) and imGlut APs (n = 9 experiments). c qRT-PCR of AXIN2 in shCON and shRSPO3 1° APs. d RSPO3-KD in imAbdo versus imGlut cells has differential effects on TOPflash promoter activity [W3A (50), 20-h treatment with 50 ng ml^-1 rhWNT3A] (n = 6 replicates/treatment). e Western blots of non-canonical WNT signalling markers in shCON and shRSPO3 imAbdo and imGlut APs. f Protein densitometry for pJNK/tJNK and pCAMK2A/tCAMK2 are shown relative to shCON levels. n = 4 independent experiments. Histogram data are means ± s.e.m. *p < 0.05, **p < 0.01, ***p < 0.001. Statistical significance was assessed by two-tailed paired (b, c, f) and unpaired (d) Student’s t-test. Source data are provided as a Source Data file.

a Schematic illustrating the location of the sa14295 mutation in exon 6 of zebrafish rspo3. b qRT-PCR for rspo3 in abdominal and peripheral AT in wild-type sibling and sa14295 homozygous (rspo3^m/m) adult females (n = 6). c Nile Red staining of adult wild-type and rspo3^m/m females (the blue asterisk indicates abdominal AT, whereas, the orange asterisk indicates peripheral AT that extends ventrally as marked by arrowheads). Scale bars = 1 mm. d Total adiposity, expressed as the % of total adipose-lipid area relative to body area, is significantly increased in rspo3^m/m adult females. e The ratio of abdominal to peripheral SC AT is decreased in rspo3^m/m adult females. f pparγ mRNA is elevated in the abdominal AT of rspo3^m/m adult females. g Adipocyte-localised lipid droplets are significantly larger in rspo3^m/m adult females. h Maximum intensity projection of female abdominal AT stained with LipidTOX (magenta), Hoechst (blue) and EdU (green). Scale bars = 100 μm. i Quantification showing decreased EdU+ nuclei in abdominal AT of rspo3^m/m animals. Two-tailed unpaired Student’s t-test was used for comparisons between genotypes (b, d–g and i). Histogram are means ± s.d. *p < 0.05, **p < 0.01, ***p < 0.001. Source data are provided as a Source Data file.

ZFIN is incorporating published figure images and captions as part of an ongoing project. Figures from some publications have not yet been curated, or are not available for display because of copyright restrictions.