Lipid storage by adipose tissue macrophages regulates systemic glucose tolerance

Abstract

Proinflammatory pathways in adipose tissue macrophages (ATMs) can impair glucose tolerance in obesity, but ATMs may also be beneficial as repositories for excess lipid that adipocytes are unable to store. To test this hypothesis, we selectively targeted visceral ATMs in obese mice with siRNA against lipoprotein lipase (LPL), leaving macrophages within other organs unaffected. Selective silencing of ATM LPL decreased foam cell formation in visceral adipose tissue of obese mice, consistent with a reduced supply of fatty acids from VLDL hydrolysis. Unexpectedly, silencing LPL also decreased the expression of genes involved in fatty acid uptake (CD36) and esterification in ATMs. This deficit in fatty acid uptake capacity was associated with increased circulating serum free fatty acids. Importantly, ATM LPL silencing also caused a marked increase in circulating fatty acid-binding protein-4, an adipocyte-derived lipid chaperone previously reported to induce liver insulin resistance and glucose intolerance. Consistent with this concept, obese mice with LPL-depleted ATMs exhibited higher hepatic glucose production from pyruvate and glucose intolerance. Silencing CD36 in ATMs also promoted glucose intolerance. Taken together, the data indicate that LPL secreted by ATMs enhances their ability to sequester excess lipid in obese mice, promoting systemic glucose tolerance.

Keywords: adipose tissue macrophages; foam cells; insulin resistance; obesity; siRNA.

Fig. 1.

Formation of lipid-laden macrophages in epididymal adipose tissue (AT) of obese mice. Stromal vascular fraction (SVF) from epididymal AT (visceral AT; VAT) of WT and ob/ob mice was isolated and analyzed by flow cytometry. A: representative flow cytometry dot plots and mean fluorescence intensity (MFI) of AT macrophages (ATMs) stained with bodipy. B: percentage of macrophages stained with bodipy; n = 10. C: microscopy of SVF isolated from ob/ob VAT stained with antibodies against F4/80 (red) and bodipy (green); ×20 (scale bar, 50 μm). D: TEM of whole VAT of ob/ob mice; A, adipocyte; M, macrophage; arrows, lipid droplet; scale bar, 2 μm. E: gene expression measured by RT-PCR in SVF isolated from VAT of WT and ob/ob mice; n = 5. Results are mean fold change (FC) ± SE. F: gene expression measured by RT-PCR in macrophages isolated using CD11b antibody bound to magnetic beads; n = 5. All graphs are expressed as means ± SE. Statistical significance was determined by Student's t-test. ***P < 0.001, **P < 0.01, *P < 0.05.

Fig. 2.

Glucan-encapsulated siRNA particle (GeRP)-mediated ATM LPL silencing decreases formation of lipid-laden macrophages in VAT of obese mice. Peritoneal macrophages (1 × 10⁶) were treated with particles made with a mixture of 160 pmol siRNA and 3 nmol EP. Forty-eight hours after treatment, mRNA levels were measured by RT-PCR (A) and protein by Western blot (B). C: outline of GeRP treatment given to mice. Briefly, mice were injected for 5 consecutive days with 5.6 mg GeRPs/kg body wt loaded with 262 μg siRNA/kg body wt and 2.1 mg EP/kg body wt. On day 6, tissues were isolated. D: mRNA expression of LPL in VAT and liver from mice treated with SCR- or LPL-GeRPs; n = 14–15. E: representative LPL Western blot and densitometry using epididymal SVF lysates from mice treated with SCR- or LPL -GeRPs. Actin was used as a loading control. Statistical significance was determined by Student's t-test. **P < 0.01, *P < 0.05. F: representative flow cytometry dot plots and MFI of ATMs stained with bodipy. G: Percentage of macrophages expressing bodipy; n = 10. All graphs are expressed as means ± SE. Statistical significance was determined by Student's t-test. ***P < 0.001, **P < 0.01, *P < 0.05. H: TEM of whole VAT from ob/ob mice treated with SCR- or LPL-GeRPs; A, adipocyte; arrows, GeRPs. Scale bar, 5 μm.

Fig. 3.

LPL silencing in ATMs increases plasma FFA. Serum triglyceride (TG; A) and LDL/VLDL and HDL cholesterol (B) in mice treated with scrambled (SCR)- or LPL-GeRPs; n = 10. C: serum FFA levels; n = 14–15. D: adipose triglyceride lipase (ATGL) and HSL expression in adipocytes isolated from ob/ob mice treated with SCR- or LPL-GeRPs; n = 14–15. E: glycerol levels in media of isolated adipocytes in untreated (UNT) or treated with 10 μM isoproterenol (ISO) for 2 h; n = 5. Gene expression measured by RT-PCR in SVF (n = 14–15; F) and ATMs containing GeRPs sorted by FACS (n = 5; G). H: representative dot-plot and percentage of ATMs expressing bodipy in fed and fasted states in VAT; n = 10. All graphs are expressed as means ± SE. Statistical significance was determined by Student's t-test. ***P < 0.001, *P < 0.05; n/s, not significant.

Fig. 4.

LPL silencing in ATMs exacerbates glucose intolerance in ob/ob mice. A: H&E liver sections of mice treated with SCR- or LPL GeRPs; ×20 magnification images; scale bar, 5 μm. B: liver TG content; n = 10. C: PEPCK and G6Pase expression in liver of ob/ob mice treated with SCR- or LPL-GeRPs; n = 10. D: pyruvate tolerance test and area under the curve (AUC); n = 10. E: glucose tolerance tests (GTT) and AUC, n = 14–15, performed on mice fasted for 18 h. F: GTT performed on ob/ob mice treated with SCR- or CD36-GeRPs; n = 5. G: GTT performed on 8-wk-old C57BL6 lean mice treated with SCR- or LPL-GeRPs; n = 5. H: serum FFA levels in lean mice treated with SCR- or LPL-GeRPs; n = 5. Results are means ± SE. Statistical significance was determined by t-test or ANOVA and Tukey's posttest. ***P < 0.001, *P < 0.05.

Fig. 5.

LPL silencing in ATMs increases aP2 production by adipocytes. A: gene expression measured by RT-PCR in VAT (A) and adipocytes (B) isolated from of ob/ob mice treated with SCR- or LPL-GeRPs; n = 14–15. C: serum aP2 levels; n = 14–15. D: aP2 protein levels in media of intact adipocytes measured by ELISA. Adipocytes were treated with media containing serum or conditioned media from SVF; n = 4. Results are means ± SE. Statistical significance was determined by t-test or ANOVA and Tukey's posttest. *P < 0.05.

Fig. 6.

Hypothetical model for how LPL produced by ATMs can regulate whole body glucose metabolism. LPL silencing in ATMs secondarily decreases the expression of CD36 and DGAT2 by mechanisms not understood. This results in reduced lipid uptake (via CD36) and esterification (via DGAT2) by ATMs and a concomitant increase in circulating FFA. In parallel, silencing LPL in ATMs is accompanied by an increase in aP2 expression and secretion by adipocytes. The increases in circulating aP2 and FFA may enhance hepatic gluconeogenesis, thus contributing to the impairment of glucose tolerance observed in ob/ob mice following LPL silencing in ATMs. Short vertical arrows indicate effects (arrow up, increase; arrow down, decrease) on specific components or pathways in response to silencing LPL in ATMs.