Fig 2: GPR120 activation promotes beige adipocyte differentiation and increases FGF21 expression and release.For a–d, iWAT precursors from mice were differentiated in the presence of the differentiation media (DM, white bars), supplemented with rosiglitazone to drive beige differentiation (DM+rosiglitazone, brown bars; n=4) or treated with GW9508 instead of rosiglitazone (DM+GW9508, blue bars; n=5; see the Methods section). (a) Representative optical microscopy images at the end of differentiation (day 7; scale bar, 200 μm). (b) Relative mRNA expression levels of browning-related and general adipogenic genes. (c) FGF21 protein levels in the cell culture medium (4 day accumulation). (d) Cell culture temperature. For e–h, iWAT precursors were differentiated and treated during 24 h with GW9508 (100 μM, blue bars; n=5) or nor treated (n=3). (e) mRNA expression levels of UCP1, PGC-1α, Glut1, COXIV, FGF21, Sirt3 and FABP4. (f) FGF21 protein levels in culture media (24 h accumulation). (g) Glucose oxidation rate. (h) Cell culture temperature. Bars are means+s.e.m. (*P<0.05, **P<0.01 and ***P<0.001 versus DM (b–d) or versus controls (e–h); for b–d, analysis of variance with Tukey's post hoc test was performed; for e–h, two-tailed unpaired Student's t-test).

Fig 3: FGF21 gene invalidation reduces the effects of GW9508 treatment in mice.Wild-type and FGF21-null mice were fed a control diet (white and grey bars, respectively) or supplemented with GW9508 (blue and black bars, respectively) for 7 days (n=5). (a) Relative mRNA levels of thermogenesis-related and adipogenic genes in iBAT (c) and iWAT. Representative optical microscopy images of H&E staining (scale bar, 125 μm), the relative lipid content and UCP1 protein levels in iBAT (b) and iWAT (d), and the percentage of multilocular adipocytes in iWAT. Bars are means+s.e.m. (*P<0.05, **P<0.01 and ***P<0.001 relative to untreated control mice of each genotype; +P<0.05, ++P<0.01 and +++P<0.001 relative to same treatment of the wild-type group; analysis of variance with Tukey's post hoc test).

Fig 4: GPR120 gene invalidation compromises thermoregulation and iWAT browning in association with a reduction in FGF21 levels.Wild-type (Wt, white bars) mice and GPR120−/− mice (grey bars) mice were exposed to cold (4 °C) for 7 days (n=5). (a) Body temperature on days 1 and 7 of cold exposure. (b) Representative images of H&E-stained iWAT (scale bar, 200 μm), the relative lipid droplet content, and the percentage of multilocular adipocytes. (c) Relative expression levels of thermogenic and adipogenic genes in iWAT. (d) Circulating levels of FGF21. Bars are means+s.e.m. (*P<0.05, **P<0.01 and ***P<0.001 relative to wild-type animals exposed to cold; two-tailed unpaired Student's t-test).

Fig 5: Protein restriction protects against age-related functional decline via FGF21.a Latency to fall from an accelerating rod (Rotarod) at 21 months of age in WT and Fgf21 KO mice on control or LP diets (n = 8–12 mice/group; geno*diet p = 0.008). b Grip Strength at 21 months of age (n = 8–12 mice/group; geno*diet p = 0.04). c Grip Strength normalized to body mass (n = 8–12 mice/group; geno*diet p = 0.01). d Ambulatory activity as measured in the metabolic chambers at 20 months of age (n = 8–12 mice/group geno*diet p = 0.39). e Locomotor activity as measured in the metabolic chambers at 20 months of age (n = 8–12 mice/group geno*diet p = 0.19). Statistical analyses were conducted using two-way ANOVA. All values are mean ± SEM, with significant posthoc comparison within the diet*genotype interaction noted as *p < 0.05 compared with respective control. Source data are provided as a Source Data file.