A PPARγ-FGF1 axis is required for adaptive adipose remodelling and metabolic homeostasis

Abstract

Although feast and famine cycles illustrate that remodelling of adipose tissue in response to fluctuations in nutrient availability is essential for maintaining metabolic homeostasis, the underlying mechanisms remain poorly understood. Here we identify fibroblast growth factor 1 (FGF1) as a critical transducer in this process in mice, and link its regulation to the nuclear receptor PPARγ (peroxisome proliferator activated receptor γ), which is the adipocyte master regulator and the target of the thiazolidinedione class of insulin sensitizing drugs. FGF1 is the prototype of the 22-member FGF family of proteins and has been implicated in a range of physiological processes, including development, wound healing and cardiovascular changes. Surprisingly, FGF1 knockout mice display no significant phenotype under standard laboratory conditions. We show that FGF1 is highly induced in adipose tissue in response to a high-fat diet and that mice lacking FGF1 develop an aggressive diabetic phenotype coupled to aberrant adipose expansion when challenged with a high-fat diet. Further analysis of adipose depots in FGF1-deficient mice revealed multiple histopathologies in the vasculature network, an accentuated inflammatory response, aberrant adipocyte size distribution and ectopic expression of pancreatic lipases. On withdrawal of the high-fat diet, this inflamed adipose tissue fails to properly resolve, resulting in extensive fat necrosis. In terms of mechanisms, we show that adipose induction of FGF1 in the fed state is regulated by PPARγ acting through an evolutionarily conserved promoter proximal PPAR response element within the FGF1 gene. The discovery of a phenotype for the FGF1 knockout mouse establishes the PPARγ–FGF1 axis as critical for maintaining metabolic homeostasis and insulin sensitization.

Figure 1. FGF1A is induced in adipose tissue by high-fat diet (HFD)

a, Western blot of FGF1 in gonadal white adipose (gWAT) of chow or HFD-fed wild-type mice (n=3). b, Western blot of FGF1 in whole fat, adipocyte (adip.) and stromal vascular fractions (SVF) of inguinal (ing.) and gonadal (gon.) white adipose of chow-fed wild-type mice (ERK1/2 loading control). c, Diagram depicting three distinct promoters driving the untranslated exons 1A, 1B, and 1D (open bars) of human and mouse FGF1 genes. Alternative splicing of untranslated exons results in identical FGF1 polypeptides. d–f, mRNA tissue distribution in mice for FGF1A (d), FGF1B (e), and FGF1D (f). g, h, mRNA levels of FGF1A (g) and FGF1B (h) in gWAT of chow fed, 2-weeks HFD-fed, or overnight fasted wild-type mice (n = 5). i, Body weight of HFD-fed wild-type and FGF1^−/− mice over 24 weeks. Data expressed as mean ± S.D. * *p<0.01.

Figure 2. Loss of FGF1 results in diet-induced insulin resistance

Metabolic studies on 24 week old male wild-type (open bars) and FGF1^−/− (filled bars) mice fed a HFD for 16 weeks. a, Fasting serum glucose and insulin levels. b, c, Glucose (b) and insulin (c) tolerance tests. d–g, Glucose infusion rate (GIR) (d), glucose disposal rate (GDR) (e), insulin-stimulated GDR (IS-GDR) (f), and percent suppression of hepatic glucose production (HGP) (g) during hyperinsulinemic-euglycemic clamp studies. h, Liver (percent body weight) from chow and HFD-fed mice (n=6–7). i, H&E staining of liver from HFD-fed wild-type (WT) and FGF1^−/− (KO) mice. j, Gonadal white adipose (gWAT) (percent body weight) from chow and HFD-fed mice (n=6–7). k, H&E staining of gWAT from HFD-fed wild-type (WT) and FGF1^−/− (KO) mice. Scale bar = 100 μm. Data expressed as mean ± S.D. *p<0.05, **p<0.01.

Figure 3. Loss of FGF1 results in defects in adipose remodelling during HFD

a, Immunohistochemistry for the macrophage marker, F4/80 in gWAT from HFD-fed wild-type (WT) and FGF1^−/− (KO) mice. b, Trichrome staining for collagen deposition in gWAT from HFD-fed wild-type (WT) and FGF1^−/− (KO) mice. c, Quantitation of adipose cell cross-sectional area from HFD-fed mice, expressed as the ratio of the number of cells from FGF1^−/− (KO) to wild-type (WT) mice defined as small (1000–4000 μm²), medium (4000–10000 μm²) and large (>10000 μm²) cells. d, Fluorescence microscopy of gWAT from HFD-fed wild-type (WT) and FGF1^−/− (KO) mice perfused with microbeads (red) and counterstained with DAPI (blue). e, QPCR verification of induction of genes associated with fat necrosis in HFD-fed wild-type (open bars) and FGF1^−/− (closed bars) mice. f, gWAT depots from HFD to chow converted (HCC) wild-type (WT) and FGF1^−/− (KO) mice. g, H&E staining of gWAT from HCC wild-type (WT) and FGF1^−/− (KO) mice. h, Image and H&E staining of dissociated necrotic WAT taken from peritoneal cavity of HCC FGF1^−/− (KO) mice. Scale bar = 100 μm. Data expressed as mean ± S.D. **p<0.01.

Figure 4. FGF1 is a direct transcriptional target of PPARγ

a, b, Luciferase reporter assays of FGF1A (a) and FGF1B (b) promoters co-transfected with PPARs +/− ligand. c, Sequence alignment of the putative PPRE within the FGF1A promoter from different species. Red indicates nucleotide variations between the PPREs relative to human. d, Species-specific response of the FGF1A promoters to PPARγ +/− ligand using luciferase reporter assays. e, Chromatin immuno-precipitation of PPARγ on the FGF1A promoter in differentiated 3T3-L1 cells (open bars IgG, closed bars PPARγ antibody). f, g, Levels of FGF1A (f) FGF1B (g) mRNA in gonadal white adipose (gWAT) of fed or overnight fasted wild-type mice (n=5) with or w/o TZD (5mg/kg for 3 days i.p). h, Western blot of FGF1, FGF2 in adipocyte (adip.) and stromal vascular (SVF) fractions of gonadal white adipose (gWAT) from chow fed wild-type (WT) and aP2-Cre; PPARγ^fl/fl (adipocyte-specific PPARγ knockout) mice (KO) (ERK1/2 loading control). i, Model depicting the role of the PPARγ-FGF1 axis in adipose remodelling and insulin sensitivity. Data expressed as mean ± S.D. *p<0.05,**p<0.01.