Long-term exposure to a high-fat diet results in the development of glucose intolerance and insulin resistance in interleukin-1 receptor I-deficient mice

Abstract

Emerging evidence has demonstrated that saturated fatty acids prime pro-IL-1β production and inflammasome-mediated IL-1β activation is critical in obesity-associated insulin resistance (IR). Nonetheless, IL-1 receptor I-deficient (IL-1RI(-/-)) mice develop mature-onset obesity despite consuming a low-fat diet (LFD). With this apparent contradiction, the present study evaluated whether IL-1RI(-/-) mice were protected against long-term (6 mo) high-fat diet (HFD)-induced IR. Male wild-type and IL-1RI(-/-) mice were fed LFD or HFD for 3 or 6 mo, and glucose and insulin tolerance tests were performed. Adipose insulin sensitivity, cytokine profiles, and adipocyte morphology were assessed. The adipogenic potential of stromal vascular fraction was determined. Hepatic lipid accumulation and insulin sensitivity were characterized. IL-1RI(-/-) mice developed glucose intolerance and IR after 6 mo HFD compared with 3 mo HFD, coincident with enhanced weight gain, hyperinsulinemia, and hyperleptinemia. The aggravated IR phenotype was associated with loss of adipose functionality, switch from adipocyte hyperplasia to hypertrophy and hepatosteatosis. Induction of adipogenic genes was reduced in IL-1RI(-/-) preadipocytes after 6 mo HFD compared with 3 mo HFD. Obese LFD-IL-1RI(-/-) mice exhibited preserved metabolic health. IL-1RI(-/-) mice develop glucose intolerance and IR after 6 mo HFD intervention. While mature-onset obesity is evident in LFD-IL-1RI(-/-) mice, the additional metabolic insult of HFD was required to drive adipose inflammation and systemic IR. These findings indicate an important interaction between dietary fat and IL-1, relevant to optimal metabolic health.

Keywords: adipose inflammation; hepatic steatosis.

Fig. 1.

Lack of IL-1 receptor I (IL-1RI) results in development of glucose intolerance and insulin resistance after 6 mo high-fat diet (HFD). Male wild-type (WT) and IL-1 receptor I-deficient (IL-1RI^−/−) mice (8–10 wk old) were placed on HFD for 3 or 6 mo before evaluation of glucose homeostasis after a 6-h fast by glucose tolerance test (GTT, 1.5 g/kg glucose ip, A) (n = 16–18 mice) and insulin tolerance test (ITT, 0.75 U/kg insulin ip, B) (n = 16–21). KO, knockout. C: weight of mice over time on HFD (n = 9–34). D and E: plasma levels of insulin were monitored after overnight fast (n = 8–20, D), and insulin secretion (E) in response to a glucose load (1.5 g/kg glucose ip) was evaluated (n = 6–10). F: plasma leptin was measured after a 6-h fast (n = 8–20). G: mice were injected ip ± insulin (1.5 U/kg) and were killed after 15 min. Insulin-induced phosphorylation of AKT in liver protein samples was analyzed by immunoblotting, and densitometry was calculated using β-actin loading control (n = 7). Representative blot presented. H: liver triacylglyceride (TAG) levels were quantified enzymatically (n = 8). Statistics: *P < 0.05, **P < 0.01, and ***P < 0.001, WT vs. knockout (KO); #P < 0.05, ##P < 0.01, and ###P < 0.001, 3 vs. 6 mo HFD; +P < 0.05 and +++P < 0.001, baseline vs. HFD.

Fig. 2.

Adipose morphology of IL-1RI^−/− mice switched from hyperplasia after 3 mo HFD to hypertrophy after 6 mo HFD. A: adipocyte morphology was monitored in paraffin-embedded adipose tissue samples by hematoxylin and eosin (H&E) staining. B: adipose explants were cultured for 24 h in serum-containing media (50 mg/ml), and levels of IL-6 secreted in media were measured by ELISA (n = 8). C: percentage of M1 macrophages (F480⁺/CD11B⁺/CD11C⁺) in the stromal vascular fraction (SVF) of adipose tissue was monitored by flow cytometry and correlated to the weight of individual mice (P < 0.001, WT r² = 0.49; IL-1RI^−/− r² = 0.46). D: insulin (100 nM)-stimulated [³H]glucose uptake in adipose explants was monitored immediately after death. Fold increase in [³H]glucose transport in response to insulin over basal (noninsulin stimulated) for each individual mouse was calculated and is presented (n = 8). E: adipose tissue RNA was harvested, and mRNA levels of GLUT4 were measured by real-time PCR analysis (n = 4–6). Statistics: **P < 0.01 and ***P < 0.001, WT vs. KO; ###P < 0.001, 3 vs. 6 mo HFD; ++P < 0.01 and +++P < 0.001, baseline vs. HFD. F–I: adipose-derived SVF cells were isolated from mice after 3 or 6 mo HFD and were cultured for 7 days to enrich for preadipocytes. Cells were then incubated ± differentiation media for 24 h, and mRNA expression of adipose differentiation-related protein (Adfp, F), forkhead box protein O1 (FOXO1, G), adiponectin (H), and peroxisome proliferator-activated receptor (PPAR)-γ (I) was assessed by real-time PCR. Expression was normalized for GAPDH, and fold change in expression relative to individual nondifferentiated control is presented [*P < 0.05 and **P < 0.01, control (Ctr) vs. differentiated (diff 24 h), n = 4].

Fig. 3.

IL-1RI^−/− mice develop mature-onset obesity on a low-fat diet (LFD) but have preserved glucose homeostasis and negligible adipose tissue inflammation. A: weight of male WT and IL-1RI^−/− mice (8–10 wk old) after 6 mo on HFD or LFD (n = 9–34). Glucose homeostasis was monitored by GTT (n = 13–31, B) and ITT (n = 9–23, C). D: insulin secretion in response to a glucose load (1.5 g/kg glucose ip) was evaluated after 6 mo HFD or 6 mo LFD (n = 6–10). E: adipocyte morphology was monitored in paraffin-embedded adipose tissue samples by H&E staining. F: adipose explants were stimulated ± insulin (100 nM) for 60 min ex vivo, and levels of phosphorylated AKT and β-actin were measured by immunoblot analysis (n = 3). G: insulin (100 nM)-stimulated [³H]glucose uptake in adipose explants was monitored immediately after death. Fold increase in [³H]glucose transport in response to insulin over basal (noninsulin stimulated) for each individual mouse was calculated and is presented (n = 6–11). H: IL-6 secretion levels from adipose explants was measured after 24 h incubation in complete media (n = 8). Statistics: #P < 0.05, ##P < 0.01, and ###P < 0.001, HFD vs. LFD groups; *P < 0.5, **P < 0.01, and ***P < 0.001, WT vs. KO.

Fig. 4.

Adipose gene and protein expression analysis after 6 mo LFD and HFD in WT and IL-1RI^−/− mice. A–C: adipose tissue RNA was harvested, and mRNA levels of Glut4 (A), insulin receptor substrate (IRS)-1 (B), CD36 (C), fatty acid-binding protein (FABP)-4, IL-6, diglyceride acyltransferase (DGAT), adiponectin (adipoQ), lipoprotein lipase (LPL), hormone sensitive lipase (HSL), and patatin-like phospholipase domain containing 1 (PNPLA1) were measured by real-time PCR (***P < 0.001, WT vs. KO; ##P < 0.01 and ###P < 0.001, LFD vs. HFD, n = 4–8). D: protein levels of CD36 and β-actin loading control were measured in adipose by immunoblot analysis; representative blot and densitometry analysis are shown (#P < 0.05, LFD vs. HFD, n = 3). E: average weekly caloric intake per mouse on LFD and HFD (**P < 0.01, WT vs. KO; ###P < 0.001, LFD vs. HFD). Intake measured weekly and averaged over 6 wk (HFD) or 25 wk (LFD), n = 3 cages/group.

Fig. 5.

Characterization of hepatic lipotoxicity and insulin sensitivity after 6 mo intervention with either LFD or HFD. A: liver morphology of WT and IL-1RI^−/− mice was monitored after 6 mo LFD and HFD by H&E staining. B: liver triacylglyceride (TAG) levels were quantified enzymatically (n = 8). C: plasma alanine aminotransferase (ALT) levels were monitored by ELISA (n = 5–14). D: mice were injected ip ± insulin (1.5 U/kg) and were killed after 15 min. Insulin-induced phosphorylation of AKT in liver protein samples was analyzed by immunoblotting, and densitometry was calculated relative to β-actin loading control (n = 10–13). Representative blot presented. E: liver mRNA was isolated and reverse transcribed before real-time PCR analysis of lipogenic genes (n = 12–14). 18S was used as an internal control, and genes are expressed relative to 6 mo WT LFD group. F: citrate synthase activity was monitored in WT and IL-1RI^−/− liver lysates after LFD or HFD and compared with baseline levels as a biomarker of mitochondrial function (n = 6). Statistics: #P < 0.05, ##P < 0.01, and ###P < 0.001, LFD vs. HFD; *P < 0.05, **P < 0.01, and ***P < 0.001, WT vs. KO.