Coordinated regulation of nutrient and inflammatory responses by STAMP2 is essential for metabolic homeostasis

Abstract

Metabolic and inflammatory pathways crosstalk at many levels, and, while required for homeostasis, interaction between these pathways can also lead to metabolic dysregulation under conditions of chronic stress. Thus, we hypothesized that mechanisms might exist to prevent overt inflammatory responses during physiological fluctuations in nutrients or under nutrient-rich conditions, and we identified the six-transmembrane protein STAMP2 as a critical modulator of this integrated response system of inflammation and metabolism in adipocytes. Lack of STAMP2 in adipocytes results in aberrant inflammatory responses to both nutrients and acute inflammatory stimuli. Similarly, in whole animals, visceral adipose tissue of STAMP2(-/-) mice exhibits overt inflammation, and these mice develop spontaneous metabolic disease on a regular diet, manifesting insulin resistance, glucose intolerance, mild hyperglycemia, dyslipidemia, and fatty liver disease. We conclude that STAMP2 participates in integrating inflammatory and metabolic responses and thus plays a key role in systemic metabolic homeostasis.

Figure 1. Regulation of STAMP2 expression

(A) Regulation of adipose tissue STAMP2 mRNA expression in ob/ob and ob/ob-TNF α-/- treated with vehicle or thiazolidinediones, assessed by Northern blot. (B) Northern blot showing tissue distribution of STAMP2 mRNA in wt mice. (C) STAMP2 expression by Northern blot during differentiation of 3T3-L1 preadipocytes into adipocytes. (D) STAMP2 expression in adipocyte and stromal-vascular fractions of adipose tissue. (E) Activation of the STAMP2 promoter. Responsiveness of the −2kb STAMP2 promoter to transcription factors CEBPα, LXRα, and PPARγ was analyzed by luciferase reporter assays in HeLa cells. T0901317 is a synthetic LXR agonist and pioglitazone (Pio) is an agonist of PPARγ. (F) Regulation of STAMP2 expression in response to 4mM or 25 mM glucose or 6-hour treatments of 100 nM insulin, 300 μM oleic acid, 20% serum, or 10 ng/ml TNFα, assessed by quantitative RT-PCR (qPCR). Data indicates mean ± s.e.m. ** p<0.005. (G) Regulation of STAMP2 expression by in WAT in fed and fasted conditions (Northern blot). (F=Fed; S=Starved). (H) Immunostaining for adipose-tissue STAMP2 protein in fed and fasted mice in samples from viscerol depot.

Figure 2. STAMP2 deficiency results in elevated inflammatory gene expression and reduced insulin action in cultured adipocytes

(A) Efficient knockdown of STAMP2 mRNA confirmed by qPCR following electroporation of control and STAMP2-specific siRNAs into 3T3-L1 adipocytes under low and high glucose (Glc). (B) Knockdown of STAMP2 protein confirmed by immunofluorescence. (C) Knockdown of Flag-STAMP2 expressed in 3T3-F442A adipocytes confirmed by Western blot analysis. (D) IL-6 expression measured by qPCR after overnight incubation in high glucose serum-free medium, and stimulation with TNFα (10 ng/ml, 6h). (E) IL-6 expression in adipocytes in high and low glucose conditions (F) IL-6 secretion measured by ELISA in cell supernatants collected under the same conditions as panel E. (G) Insulin-stimulated ³H-2-deoxyglucose uptake in cultured adipocytes upon suppression of STAMP2 (H) Insulin stimulated myc-Glut4-GFP translocation to the plasma membrane assessed by immunocytochemistry and confocal microscopy (I) Quantitation of experiments shown in panel H in 3 independent experiments. Data represents mean ± s.e.m. Open bars-control siRNA; closed bars-STAMP2 siRNA. * indicates p<0.05; ** indicates p<0.005.

Figure 3. Adipose tissue of STAMP2-/- mice exhibits elevated expression of inflammatory genes, accumulation of mononuclear cells and oxidative stress

Tissues were harvested from 5- to 6-month-old mice on standard diet. (A) STAMP2 mRNA expression in the adipose tissue of STAMP2-/- (KO) and wild type (WT) mice examined by Northern blot analysis (B) Expression of STAMP family members in VWAT examined by qPCR. (C) Inflammatory gene expression in VWAT and SWAT, examined by qPCR. (Open bars- WT; closed bars-STAMP2-/-). (D) Tissue sections from WT and STAMP2-/- mice were stained with hemotoxylin and eosin (E) F4/80 antigen positivity in WT and STAMP2-/- VWAT, detected by immunohistochemistry (F) Gene expression related to oxidative stress was evaluated in VWAT (open bars- WT; closed bars- STAMP2-/-) (G) Levels of TBARS in VWAT (n=5 animals/ genotype). All data presented as mean ± s.e.m. For all qPCR experiments, 5-6 animals in each genotype were examined, and mRNA expression of each gene was normalized to 18S rRNA levels. * indicates p< 0.05; ** p<0.005; *** p<0.0005.

Figure 4. Impaired insulin action in the visceral WAT of STAMP2-/- mice

(A) PBS or intralipid and glucose (LG) were injected intraperitoneally into wt and STAMP2-/- mice followed by measurement of IL-6 and SOCS-3 mRNA levels in VWAT by qPCR. Data pooled from 2 independent experiments, shown as mean ± s.e.m. (B) Glucose transport in primary adipocytes. (C) Insulin stimulated Akt phosphorylation in VWAT and SWAT. Representative animals are shown in the Western blot, with each lane representing one animal. Phosphorylation of Akt normalized to total Akt protein was quantified and both absolute (see supplementary figure 3) and fold insulin-stimulated Akt phosphorylation graphed, with data pooled from 2 independent experiments, represented as mean ± s.e.m. Expression of (D) metabolic and (E) macrophage-specific genes in VWAT was determined by qPCR. For each gene 6-9 animals/ genotype were examined, shown as mean ± s.e.m. For B-E, open bars- WT, closed bars- STAMP2-/-. * indicates p<0.05; ** p< 0.005; *** p< 0.0005.

Figure 5. Development of metabolic disease in STAMP2-/- mice

(A) Wild type (n=11) and STAMP2-/- (n=9) mice were weighed weekly until 20 weeks of age. (B) Percent body fat was measured in a separate group of mice aged 2 and 5 months using DEXA analysis. (C) Every 4 weeks, blood was collected after a 6 hour fast. At 4, 12, and 20 weeks of age, serum insulin measured by ELISA. (D) Blood glucose was measured every 4 weeks after a 6 hour fast. (E) Serum triglycerides and cholesterol were measured and (F) distribution of lipoprotein particles profiled at 12 weeks of age. (G) Glucose and (H) insulin tolerance tests were performed on mice aged 12 and 17 weeks respectively. All data is presented as mean ± s.e.m. In A, D, G & H- squares- wt, triangles- STAMP2-/-; in B, C, & E, open bars- wt, closed bars- STAMP2-/-. * p<0.05; ** p<0.005; *** p<0.0005.

Figure 6. Liver insulin action and hyperinsulinemic-euglycemic clamp

(A) Insulin action in the liver following insulin injection into the portal vein of anesthetized mice. Each lane represents one animal. (B) Hyperinsulinemic-euglycemic clamp experiments. Rate of glucose infusion was higher in wild type than in STAMP2-/- mice over the 2 h course of the clamp. (C) Average GIR and Rd were calculated. (D) Hepatic glucose production at basal and clamp conditions, as well as (E) percent suppression of hepatic glucose production by hyperinsulinemic clamp were calculated. (F) Muscle and (G) adipose tissue glucose uptake during the clamp were assessed by measuring uptake of a ¹⁴C-2-deoxyglucose tracer. * p<0.05; ** p<0.005. Data graphed as mean ± s.e.m. For parts C-G, open bars- wt; closed bars- STAMP2-/-. For part B, open circles- wt; closed squares- STAMP2-/-.

Figure 7. Liver lipid accumulation in STAMP2-/- mice, and absence of STAMP2 exacerbates metabolic phenotype of ob/ob mice

(A) Liver sections were stained with hemotoxylin and eosin. (B) Quantitation of liver triglyceride, with ob/ob liver used as a control. (C) FAS and SCD-1 expression determined by qPCR in liver at 6 months of age. (D) Generation of ob/ob (n=4) and STAMP2-/-ob/ob (n=5) mice. In 12-week-old mice, (D) body weight was similar between genotypes, though (E) body fat was significantly higher in mice lacking STAMP2. (F) Blood glucose after a 6 hour fast was significantly higher in the STAMP2-/-ob/ob mice. (G) H & E staining of liver sections showed higher lipid accumulation with larger lipid droplets in the mice lacking STAMP2. All bar graph data represents mean ± s.e.m.