Abstract

Cigarette smoking (CS) is known to reduce body weight and this often masks its real effect on insulin action. The present study tested the hypothesis that CS can divert lipid deposition to muscles to offset the supposed benefit of reduced body weight gain on insulin signalling in this major site for glucose tolerance (or insulin action). The study was conducted in mice exposed to chronic CS followed by either a chow (CH) diet or a high-fat (HF) diet. CS increased triglyceride (TG) levels in both plasma and muscle despite a reduced body weight gain and adiposity. CS led to glucose intolerance in CH-fed mice and they retained the glucose intolerance that was induced by the HF diet. In adipose tissue, CS increased macrophage infiltration and the mRNA expression of TNFα but suppressed the protein expression of adipose triglyceride lipase and PPARγ. While CS increased hormone-sensitive lipase and suppressed the mRNA expression of leptin, these effects were blunted in HF-fed mice. These results imply that CS impairs insulin signalling in skeletal muscle via accumulated intramuscular lipids from lipolysis and lipodystrophy of adipose tissues. This may explain why smokers may not benefit from insulin sensitising effects of reduced body weight gain.

Introduction

Today, the number of people that smoke is 1.1 billion worldwide [1]. Cigarette smoking (CS) has widely been recognised as a major cause of a range of non-communicable diseases, such as chronic obstructive pulmonary disease (COPD), cardiovascular disease and multiple types of cancers [2]. However, the effects of CS on lipid metabolism and insulin action appear to be more complicated because it curtails weight gain and obesity [3–5]. It is well known that obesity is associated with dyslipidaemia (particularly hypertriglyceridaemia) and insulin resistance (diminished insulin action on glucose metabolism) which often presents as hyperinsulinemia, increased fasting blood glucose level (or HOME-IR) or glucose intolerance. These metabolic disorders underlie a range of major chronic diseases, including type 2 diabetes (T2D), cardiovascular disease and non-alcoholic fatty liver disease.

Indeed, the effect of obesity on lipid metabolism and insulin action is evidenced by the fact that weight loss generally attenuates dyslipidaemia, insulin resistance and hyperglycaemia [6]. For example, a multicentre clinical trial conducted on 2500 overweight individuals with pre-diabetes showed that moderate weight loss (∼11%) over 8 weeks can significantly reduce insulin resistance and improve the glycaemic and lipidemic profiles [7]. However, despite the reduction in body weight [3–5], smokers are still at risk of developing T2D [8–10]. Experimental study in mice also found no improvement in hyperglycaemia in obese mice that were fed a high-fat (HF) diet, even when CS reduced the gain of body weight [5].

While CS-induced weight loss is largely due to reduced fat mass, muscle is the main site that determines glucose tolerance or insulin resistance at the whole-body level because the organ is responsible for 80% of glucose disposal in response to insulin action [11]. As insulin-stimulated glucose disposal in muscle is mediated by the insulin signalling pathway, insulin resistance in muscle is often caused by disrupted insulin signal transduction. It is widely accepted that the intracellular accumulation of lipids is a key mechanism for muscle insulin resistance because these excess intracellular lipids can disrupt several steps of insulin signal transduction, such as IRS1 and Akt [12,13]. In the state of obesity, muscle insulin resistance often occurs because there is an overflow of lipids from adipose tissue to muscle [12,13].

On the other hand, an inability of adipose tissue to expand for lipid storage (such as lipodystrophy) has been shown to cause severe muscle insulin resistance leading to T2D by redirecting circulating lipids into muscle [14,15]. Interestingly, a chronic infusion of nicotine into rats has been shown to stimulate lipolysis in adipose tissue and possibly divert the released fatty acids towards muscle [16]. Therefore, we hypothesised that CS may divert lipids from adipose tissue to muscle to offset the benefit of a reduction in the gain of body weight to glucose tolerance by impairing insulin signalling in muscle. To test this hypothesis, we examined the effects of CS exposure on body weight gain and insulin action in adipose tissue and skeletal muscle. As obesity and associated insulin resistance are largely due to the overconsumption of energy-rich diet containing large amounts of carbohydrates and/or saturated fat [17], the present study was conducted under the normal condition of chow (CH) diet and during diet-induced obesity in mice.

Materials and methods

Study design

The study was performed in male C57BL/6J mice (9–10 weeks) purchased from the Animal Resources Centre (Perth, Australia). After arriving at RMIT University, the mice were kept at 22 ± 1°C on a 12-h light/dark cycle with free access to water and a standard laboratory CH diet (10% calories from fat). Following 2 weeks of acclimatisation, the mice were exposed to CS as previously described [18] for 14 weeks. The mice were exposed to mainstream smoke generated from four Winfield Red cigarettes per day (two cigarettes/session, two sessions/day for 5 days/week). As the control, the non-smoking mice were exposed to clean filtered air in a separate chamber under a similar procedure. After 2 weeks of CS exposure, both smoking and non-smoking mice were divided into two sub-groups, one group continued with the same CH diet and the other group was placed on a HF diet for 12 weeks. The composition of the HF diet contained 45% calories from animal fat (lard) as previously described [19]. This protocol was designed to represent mild and long-term smokers starting from a relatively early stage of their life where some of them also regularly consume a diet that is rich in fat.

Body weight and food intake were recorded twice a week (Tuesday and Friday). At the end of the experiment, the mice were killed by cervical dislocation. Subcutaneous and testicular fat were collected and weighed using an analytical balance. Muscles (red and white quadriceps) were collected for further assessment. All experiments were approved by the Animal Ethics Committee of the Royal Melbourne Institute of Technology University (Project #1705) in accordance with the guidelines of the National Health and Medical Research Council of Australia.

Determination of plasma triglyceride, free fatty acid, leptin and insulin

For these measurements, 100 μl of blood samples were collected at the end of the animal study from the tail tip using heparin capillary tubes (Hirschmann Laboratory, Germany) after 5–7 h of fasting. Plasma was obtained from the collected blood samples after centrifugation at 15,000 rpm for 1 min and immediately stored in containers at −80°C until use as described previously [20]. The plasma levels of triglyceride (TG) were measured by a colorimetric assay using the Peridochrom triglyceride GPO-PAP reagent according to the manufacturer’s instructions (Roche Diagnostics, Australia). Plasma free fatty acid (FFA) levels were measured using a NEFA-C kit according to the manufacturer’s instructions (Wako Diagnostics, U.S.A.). Plasma insulin levels were determined using a mouse insulin ELISA kit (Thermo Fisher Scientific, U.S.A.) and plasma leptin levels were measured using a commercial leptin ELISA kit (Abcam, Australia).

Assessment of glucose tolerance and insulin sensitivity

A glucose tolerance test (GTT) was performed at week eight of HF feeding after 5–6 h of fasting with an intraperitoneal injection of glucose at 2.5 g/kg average body weight of the CH control (CH-C) group. Blood glucose levels were measured using an Accu-Chek glucometer (Roche Diagnostics, Australia). The homeostatic model assessment of insulin resistance (HOMA-IR) was conducted at week 10 of HF feeding in the same fasting state based on blood glucose and plasma insulin levels. Furthermore, insulin signalling was examined ex vivo in collected fresh muscle and adipose tissue at the end of the experiments. Briefly, testicular fat pads and red quadriceps muscles were excised from the mice after they were culled. Tissues were cut into 2–3 mm-thick slices and placed in a Krebs–Ringer buffer at 37°C gassed with carbogen with an adjusted pH of 7.4. Tissues were then incubated in the buffer with or without 100 nM insulin for 20 min and then immediately frozen and stored at −80°C for the determination of phosphorylation and total of Akt and GSK3β by Western blotting.

Measurement of muscle TG content

At the end of the experiment, the mice were killed by cervical dislocation and the muscle (red and white quadriceps) was freeze-clamped immediately for TG determination. Briefly, muscle samples (40–50 mg) were homogenised in chloroform/methanol (2:1) using a glass pestle tissue grinder (Avantor, Australia). Homogenates were incubated overnight at room temperature on a rotator to solubilise tissue TGs. The next day, 0.6% NaCl was added and samples were then centrifuged at 2000× rpm for 10 min. The lower organic phase containing the solubilised TGs was collected and air-dried at 45°C. The TGs were then resuspended in 100% ethanol and determined by a colorimetric assay using Peridochrom triglyceride GPO-PAP reagent according to the manufacturer’s instructions (Roche Diagnostics, Australia) as described previously [21,22].

Histological evaluation of adipose tissue

Haematoxylin and eosin staining

Subcutaneous fat tissues were fixed in natural formalin for 24 h, and then were paraffin-embedded after dehydration. The tissues were processed into 5 µm thickness sections, warmed in oven at 65°C for 10 min and were directly placed into fresh histolene twice and then in ethanol. The sections were stained with Harris Haematoxylin (Sigma-Aldrich, U.S.A.) for 4 min then washed with water, followed by counterstaining with 1% aqueous eosin (Sigma-Aldrich, U.S.A.) for 1 min. The sections were then coverslipped by DPX mounting medium (Sigma-Aldrich, U.S.A.). Images from each section were captured using Olympus microscopy and Nikon Eclipse TS100 (Olympus, Japan) then examined.

Immunohistochemistry

Mouse subcutaneous fat samples were cut into 5 µm sections and fixed in antigen retrieval buffer (10 mM citric acid, 0.05% Tween 20, pH 6.0) for 20 min at 96°C. Sections were blocked in 5% fetal bovine serum for 1 h at room temperature after washing twice with PBS buffer. The sections were then immunoblotted with primary antibody F4/80 (Monoclonal Antibody, Clone BM8, #14_4801-82, eBioscience) overnight at 4°C and then donkey anti-rat IgG secondary antibody (Cell Signalling, Australia) for 1 h at room temperature after washing twice with PBS buffer. Avidin-biotin complex (ABC)-HRP (Vector Labs) was added and incubated for 30 min. DAB substrate (Sigmafast™ DAB + UREA) (Sigma-Aldrich, U.S.A.) prepared in Milli-Q water was then added to the slides for 1 h at room temperature. Tissue sections were stained using Harris haematoxylin (Sigma-Aldrich, U.S.A.) for 4 min and then coverslipped using DPX mounting medium (Sigma-Aldrich, U.S.A.). Images were captured using the Olympus microscopy and Nikon Eclipse TS100 (Olympus, Japan) then examined.

Western blotting

Freeze-clamped adipose tissues (∼100 mg, subcutaneous and testicular) and muscles (∼40 mg, red and white quadriceps) were homogenised using a RIPA lysis buffer supplemented with phosphatase and protease inhibitor cocktails (Thermo Fisher Scientific, U.S.A.) as described previously [23,24]. Protein concentration was determined by the bicinchoninic acid method (Thermo Fisher Scientific, U.S.A.). Sample proteins (15 mg for adipose tissues and 40 mg for muscle) were resolved by SDS-PAGE and then transferred to PVDF membranes (Bio-Rad Laboratories, U.S.A.). The membranes were blocked in 3% BSA in PBS buffer and incubated with the following primary antibodies (1:1000 dilution, Cell Signalling, Australia): phospho (Ser473) Akt, total Akt pan, phospho (Ser9) GSK3β, total GSK3β, phospho (Ser660) and total hormone-sensitive lipase (HSL), Lipin-1, peroxisome proliferator-activated receptor gamma (PPARγ), adipose triglyceride lipase (ATGL) and GAPDH, overnight at 4°C. Membranes were washed in PBS buffer and then incubated with anti-mouse or anti-rabbit secondary antibodies (1:5000 dilution, Santa Cruz, U.S.A.) that conjugated to horseradish peroxidase. Membranes were developed with an enhanced chemiluminescent reagent (Thermo Fisher Scientific, U.S.A.) and proteins of interest were visualised using a ChemiDoc system. Quantification and analysis of images were performed using Image Lab software (Bio-Rad Laboratories, U.S.A.). The mean value of the normal control group (CH-C) was scaled as 1.0 for quantification of each individual mouse across all groups with the results presented as fold change over the CH-C group.

Quantitative real-time PCR

Total RNA was isolated from ∼20 mg of whole subcutaneous fat tissues according to the manufacturers’ instructions by TRIzol Reagent (Thermo Fisher Scientific, U.S.A.) and the total RNA was extracted using genomic DNA tissue Miniprep System kits (Promega, Australia). Purified RNA was used as a template for cDNA synthesis using a high capacity reaction kit (Applied Biosystems, Foster City, CA) and samples were incubated at 37°C for 60 min. Real-time PCR was performed using pre-designed TaqMan® primer/probe combinations fast advanced master mix (Applied Biosystems, Foster City, CA) according to the manufacturer’s instructions. The gene expression from each sample was analysed in triplicates and normalised against the housekeeper gene 18S. All reactions were performed on a QuantStudio 7 PCR machine (Thermo Fischer Scientific, U.S.A.) and data were quantified relative to normal control group (CH-C, scaled as 1.0) using the 2−ddCT method and presented as fold change over the CH-C group.

Statistical analyses

Data are shown as means ± SE. A two-way ANOVA was used for comparison of the relevant groups. When statistically significant differences were found, a post-hoc Tukey–Kramer multiple comparisons test was applied. Differences at P≤0.05 are considered statistically significant.

Results

Effects on whole-body metabolism related to metabolic syndrome

The mice that were assigned to four different groups showed similar body weight at the start of the study (∼23 g, P>0.5, Table 1). The HF diet fed control mice (HF-C) for 12 weeks gained about 15 g of weight (65% above the baseline). This weight gain was much greater (P<0.01) compared with the non-smoking CH-fed mice (CH-C) which gained 5 g (20% above the baseline). CS reduced the gain of body weight by about 4 g in both the CH-fed (CH-S) and the HF-fed (HF-S) mice. A reduction of 14% in CH-S (from 28.7 to 24.6 g) and 11% in the HF-S (from 38.6 to 34.4 g) groups compared with the corresponding non-smoking groups (both P<0.01).

Table 1
Effects on whole-body metabolic parameters
VariableCH-CCH-SHF-CHF-S
Starting body weight (g) 23.7 ± 0.3 23.5 ± 0.5 23.4 ± 0.1 23.9 ± 0.2 
Final body weight (g) 28.7 ± 0.5 24.6 ± 0.7** 38.6 ± 0.5** 34.4 ± 0.4**## 
Food intake per group (g) 192 152 171 155 
Plasma TG (μM) 0.65 ± 0.01 0.89 ± 0.10** 1.10 ± 0.04** 1.35 ± 0.04**# 
Plasma FFA (μM) 0.47 ± 0.0 0.51 ± 0.0 0.61 ± 0.0** 0.59 ± 0.02** 
Plasma leptin (ng/ml) 6.2 ± 0.7 16.7 ± 3.6* 18.8± 2.4** 21.4 ± 6.1**## 
Blood glucose (mM) 8.2 ± 0.3 7.0 ± 0.2* 12.8 ± 0.6** 11.0 ± 0.5*# 
Plasma insulin (μIU/ml) 200 ± 8 236 ± 19 314 ± 29** 288 ± 30* 
HOMA-IR 71.6 ± 5.0 72.6 ± 4.1 175 ± 9** 136 ± 14*# 
GTT iAUC (mM × 90 min) 601 ± 41 808 ± 32** 942 ± 88** 897 ± 93** 
VariableCH-CCH-SHF-CHF-S
Starting body weight (g) 23.7 ± 0.3 23.5 ± 0.5 23.4 ± 0.1 23.9 ± 0.2 
Final body weight (g) 28.7 ± 0.5 24.6 ± 0.7** 38.6 ± 0.5** 34.4 ± 0.4**## 
Food intake per group (g) 192 152 171 155 
Plasma TG (μM) 0.65 ± 0.01 0.89 ± 0.10** 1.10 ± 0.04** 1.35 ± 0.04**# 
Plasma FFA (μM) 0.47 ± 0.0 0.51 ± 0.0 0.61 ± 0.0** 0.59 ± 0.02** 
Plasma leptin (ng/ml) 6.2 ± 0.7 16.7 ± 3.6* 18.8± 2.4** 21.4 ± 6.1**## 
Blood glucose (mM) 8.2 ± 0.3 7.0 ± 0.2* 12.8 ± 0.6** 11.0 ± 0.5*# 
Plasma insulin (μIU/ml) 200 ± 8 236 ± 19 314 ± 29** 288 ± 30* 
HOMA-IR 71.6 ± 5.0 72.6 ± 4.1 175 ± 9** 136 ± 14*# 
GTT iAUC (mM × 90 min) 601 ± 41 808 ± 32** 942 ± 88** 897 ± 93** 

Mice (male C57BL/6J, at 9–10 weeks) were exposed to cigarette smoke (CS, twice daily) for 14 weeks. After 2 weeks of CS, they were fed either a chow (CH) or a high fat (HF) diet for 12 weeks. Plasma triglycerides (TG), free fatty acid (FFA), leptin and insulin blood glucose were measured toward the end of the study. The homeostatic model assessment of insulin resistance (HOMA-IR) was calculated based on blood glucose (mM) x plasma insulin (μIU/ml) /22.5. Intraperitoneal glucose tolerance test (ipGTT) was performed at week 8 and expressed as the incremental area under the curve (iAUC) of blood glucose. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05, ##P<0.01 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

Table 1 shows the effects of CS on metabolic features. As expected, HF-C mice showed significant increases in plasma TG (by 70%), FFA (by 30%), leptin (3-fold), insulin (by 1.6-fold) and blood glucose (by 56%) (all P<0.05, compared with CH-C). At the same time, CS elevated plasma levels of TG (by 37% in CH-S and by 22% in HF-S) and leptin (2.7 fold in CH-S and by 14% in HF-S) compared with the corresponding non-smoking groups (all P<0.05). Interestingly, CS moderately decreased blood glucose in both CH-fed and HF-fed mice (by 13–15%, P<0.05) and retained HF-induced hyperinsulinemia. The calculated HOMA-IR was twice as high in both the HF-C and HF-S groups compared with the CH-C group (P<0.05). However, there was no significant difference in HOMA-IR between the CH-C and CH-S groups or between the HF-C and HF-S groups. Consistent with this, HF diet and CS induced glucose intolerance, respectively, as indicated by the incremental area under the curve (iAUC) in GTT (57% and 34%, respectively, P<0.01 vs CH-C). There was no improvement in glucose intolerance in HF-S mice with the iAUC of GTT, remaining 50% higher than the CH-C group (P<0.01) and similar to that of HF-C group (P>0.05).

Effects on adiposity and markers in adipose tissue

A HF diet alone increased the weight of subcutaneous and testicular fat depots by ∼3-fold (both P<0.01, compared with CH-C) (Figure 1A,B). This was reflected by the enlarged sizes of adipocytes (Figure 1C). In contract, CS reduced the adiposity in both the CH-S and HF-S mice by ∼20% (both P<0.05, Figure 1A,B) and the size of adipocytes (Figure 1C). To assess whether there was a dysregulation in adipocytes, we measured the mRNAs of leptin and apelin, the adipokines expressed exclusively in adipocytes. As shown in Figure 1D,E, the HF-C group showed approximately a 7-fold increase in the mRNA expression levels of leptin and apelin (all P<0.01 vs CH-C). In CH-fed mice, CS inhibited the mRNA expression of leptin and apelin by 50% (all P<0.05) where a similar degree of decrease in apelin expression was found in HF-fed mice exposed to CS (P<0.05, vs HF-C).

Effects on adiposity and adipocyte markers

Figure 1
Effects on adiposity and adipocyte markers

(A) Weight of subcutaneous fat (in the trunk) as a percentage (%) of body weight (BW). (B) Weight of testicular fat as a percentage (%) of BW. (C) Representative images of H&E staining in subcutaneous fat at 10× (upper panel) and 20× (lower panel) magnifications, respectively. (D) Leptin mRNA expression in subcutaneous fat. (E) Apelin mRNA expression in subcutaneous fat. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05 vs HF-C. Data are shown as mean ± SE of 6–7 mice per group.

Figure 1
Effects on adiposity and adipocyte markers

(A) Weight of subcutaneous fat (in the trunk) as a percentage (%) of body weight (BW). (B) Weight of testicular fat as a percentage (%) of BW. (C) Representative images of H&E staining in subcutaneous fat at 10× (upper panel) and 20× (lower panel) magnifications, respectively. (D) Leptin mRNA expression in subcutaneous fat. (E) Apelin mRNA expression in subcutaneous fat. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05 vs HF-C. Data are shown as mean ± SE of 6–7 mice per group.

Effects on macrophages infiltration and inflammation in adipose tissue

CS has been suggested to impair the lung function by inducing inflammation as indicated by macrophage infiltration and increased expression of inflammatory cytokines in the lung [25]. Compared with the CH-C mice, the HF-C mice appeared to display an increase in F4/80 cells in adipose tissue (Figure 2A) associated with a 4-fold increase (P<0.01 vs CH-C) in the mRNA of the C-C motif chemokine ligand 2 (CCL2, also known as MCP1), which is mainly expressed in macrophages (Figure 2B). Consistent with this, the mRNA of TNFα increased by 3-fold (both P<0.01 vs CH-C) (Figure 2C). In CH-fed mice, CS also showed similar effects inducing F4/80 cells (Figure 2A) and the mRNA expression of CCL2 and TNFα (by 2- to 2.8-fold, both P<0.05, Figure 2B,C). Interestingly, the effects of CS on these inflammatory markers were additive to the effects of the HF diet (Figure 2B,C). Compared with the CH-C mice, the mRNA levels of CCL2 and TNFα in the adipose tissue of HF-S mice increased by approximately 12- and 7-fold, respectively (all P<0.01). The mRNA expression level of TNFα was positively correlated with the mRNA expression of CCL2 (marker of macrophage infiltration) (Figure 2D).

Effects on macrophage infiltration and inflammation in adipose tissue

Figure 2
Effects on macrophage infiltration and inflammation in adipose tissue

(A) Representative images of immunohistochemistry staining with the antibody (F4/80) specific for macrophages (40× magnification). (B) C-C motif chemokine ligand 2 (CCL2, a macrophage marker) mRNA expression. (C) TNFα (an inflammation marker) mRNA expression. (D) Relation between mRNA expression levels of TNFα and CCL2 determined by a Spearman’s correlation. Results were obtained from subcutaneous fat. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group, #P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

Figure 2
Effects on macrophage infiltration and inflammation in adipose tissue

(A) Representative images of immunohistochemistry staining with the antibody (F4/80) specific for macrophages (40× magnification). (B) C-C motif chemokine ligand 2 (CCL2, a macrophage marker) mRNA expression. (C) TNFα (an inflammation marker) mRNA expression. (D) Relation between mRNA expression levels of TNFα and CCL2 determined by a Spearman’s correlation. Results were obtained from subcutaneous fat. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group, #P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

Effects on adipose lipases and adipogenesis

Figure 3 shows the changes in HSL and ATGL, two key enzymes essential in TG metabolism. A HF diet alone did not show significant changes in the phospho-HSL protein (Figure 3A,C) or ATGL (Figure 3B,C), in subcutaneous and testicular adipose tissues. In CH-fed mice, CS increased the p-HSL protein in both subcutaneous and testicular adipose tissues (by approximately 1-fold, both P<0.01 vs CH-C, Figure 3A). On the contrary, ATGL protein levels decreased by ∼50% in both subcutaneous and testicular adipose tissues when exposed to CS (P<0.05 vs CH-C, Figure 3B). A similar effect of CS on ATGL was also observed in subcutaneous adipose tissue of HF-fed mice (∼50% reduction vs HF-C, P<0.05, Figure 3B). In testicular adipose tissue, HF feeding reduced ATGL by ∼50% compared with CH-fed mice (P<0.05 vs CH-C) but CS did not show a significant effect on ATGL in HF-fed mice (Figure 3B).

Effects on adipose lipases

Figure 3
Effects on adipose lipases

Subcutaneous fat (left) and testicular fat (right) from mice were immunoblotted for (A) phosphorylated and total hormone-sensitive lipase (p-HSL and t-HSL), and (B) adipose triglyceride lipase (ATGL). Data were normalized by the corresponding t-HSL and GAPDH, respectively. (C) Representative images of proteins. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

Figure 3
Effects on adipose lipases

Subcutaneous fat (left) and testicular fat (right) from mice were immunoblotted for (A) phosphorylated and total hormone-sensitive lipase (p-HSL and t-HSL), and (B) adipose triglyceride lipase (ATGL). Data were normalized by the corresponding t-HSL and GAPDH, respectively. (C) Representative images of proteins. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

To examine the effects on adipogenesis, we measured the protein levels of lipin-1 and PPARγ (Figure 4). In CH-fed mice, CS decreased the lipin-1 protein in both subcutaneous and testicular adipose tissues by 50–60% (all P<0.05) (Figure 4A,C). A similar inhibitory effect of CS on PPARγ was observed in testicular adipose tissue (P<0.01, Figure 4B,C). Compared with CH-fed mice, HF diet did not affect lipin-1 levels (Figure 4A,B) but only decreased PPARγ (by ∼60%) in testicular adipose tissue (P<0.01, Figure 4B). In HF-fed mice, CS showed no significant effect on lipin-1 nor on PPARγ.

Effects on adipogenesis

Figure 4
Effects on adipogenesis

Subcutaneous (left) and testicular (right) adipose tissues from mice were immunoblotted for (A) total lipin-1 and (B) peroxisome proliferation-activating protein γ (PPARγ). Data were normalized by GAPDH. (C) Representative images of proteins. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group. Data are shown as mean ± SE of 6–7 mice per group.

Figure 4
Effects on adipogenesis

Subcutaneous (left) and testicular (right) adipose tissues from mice were immunoblotted for (A) total lipin-1 and (B) peroxisome proliferation-activating protein γ (PPARγ). Data were normalized by GAPDH. (C) Representative images of proteins. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group. Data are shown as mean ± SE of 6–7 mice per group.

Effects on insulin signalling in adipose tissue

As lipodystrophy can promote insulin resistance [26], we assessed the insulin signalling in adipose tissue by examining the phosphorylation of Akt and the substrate GSK3β (Figure 5A-C). First, this was examined in testicular adipose tissue collected directly from the mice at the plasma level of insulin (shown in Table 1). Under resting conditions, CS alone was not sufficient to suppress the ratio of phospho-Akt to total-Akt, but it significantly decreased the ratio of phospho-GSK3β to total-GSK3β (by ∼70% P<0.01) (Figure 5A, upper panel). As the levels of insulin were different when adipose tissue was collected, we further evaluated the ratio of phospho-Akt or phospho-GSK3β against plasma insulin levels (Figure 5A, lower panel). A HF diet alone significantly reduced both phospho-Akt/insulin and phospho-GSK3β/insulin (by ∼40–50%, P<0.05 vs CH-C). While CS alone did not significantly inhibit phospho-Akt/insulin, it reduced phospho-GSK3β/insulin by ∼50% (P<0.01 vs CH-C) and this inhibition was additive to the effect of a HF diet (P<0.01 vs HF-C). Similar patterns of changes in phospho-Akt/insulin and phospho-GSK3β/insulin were observed in subcutaneous adipose tissue (Supplementary Figure S1). To further examine insulin signalling in response to maximal insulin stimulation, we isolated the fresh testicular adipose tissue and stimulated it with insulin ex vivo. The results showed (Figure 5B) that CS alone markedly inhibited insulin-stimulated phosphorylation of Akt (by ∼60%) and GSK3β (by ∼70% vs CH-C). While a HF diet alone did not significantly alter the phosphorylation of Akt, the insulin-stimulated phosphorylation of GSK-3β in the adipose tissue was suppressed by ∼40% (p<0.05 vs CH-C). The inhibitory effects of a HF diet and CS on insulin-stimulated phosphorylation of GSK3β were additive (P<0.05 vs HF-C and CH-C).

Effects on insulin signalling in adipose tissue

Figure 5
Effects on insulin signalling in adipose tissue

Testicular adipose tissue from mice were immunoblotted for (A) phosphorylated and total levels of Akt and GSK3β under basal conditions, Data were normalised by corresponding total level (upper panel) and basal plasma insulin level as shown in Table 1 (lower panel), respectively. (B) phosphorylated and total levels of Akt and GSK3β stimulated with insulin ex vivo. Data were normalised by corresponding total level. (C) Representative images of proteins. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05, ##P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

Figure 5
Effects on insulin signalling in adipose tissue

Testicular adipose tissue from mice were immunoblotted for (A) phosphorylated and total levels of Akt and GSK3β under basal conditions, Data were normalised by corresponding total level (upper panel) and basal plasma insulin level as shown in Table 1 (lower panel), respectively. (B) phosphorylated and total levels of Akt and GSK3β stimulated with insulin ex vivo. Data were normalised by corresponding total level. (C) Representative images of proteins. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05, ##P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

Effects on insulin signalling and triglyceride content in muscle

Skeletal muscle accounts for 80% of insulin-stimulated glucose disposal [11] under regulation of normal insulin signalling including the phosphorylation of the key downstream proteins Akt and GSK3β [12,13]. Therefore, we first examined the phosphorylation of Akt in red quadriceps muscles under the basal conditions and in response to insulin stimulation (Figure 6A-C). As shown in Figure 6A, under the basal conditions, CS alone inhibited the phosphorylation of Akt and the ratio of phospho-Akt/insulin by 50–60% (both P<0.05 vs CH-C). In comparison, the inhibitory effect of a HF diet alone was only reflected in the ratio of phospho-Akt/insulin (∼50% reduction, P<0.01 vs CH-C) which was further supressed by CS (P<0.05 between HF-C and HF-S). Further examination also showed that CS alone was substantially suppressed (by ∼60%, P<0.05 vs CH-C) in the phosphorylation of Akt stimulated by maximal insulin stimulation in isolated red quadriceps muscles (Figure 6B). Whereas a HF diet alone did not significantly inhibit insulin-stimulated phosphorylation of Akt, CS still retained its suppression of insulin-stimulated phosphorylation of Akt in HF-fed mice (∼50%, P<0.05, Figure 6B).

Effects on insulin signalling and triglyceride content in muscle

Figure 6
Effects on insulin signalling and triglyceride content in muscle

Red quadriceps muscles from mice were immunoblotted for (A) phosphorylated and total Akt under basal conditions. Data were normalised by total Akt (left panel) and basal plasma insulin level as shown in Table 1 (right panel), respectively. (B) Phosphorylated and total levels of Akt stimulated with insulin ex vivo. Data were normalised by total Akt. (C) Representative images of immunoblotting. (D) Triglyceride (TG) content in red quadriceps muscles at the end of the study. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

Figure 6
Effects on insulin signalling and triglyceride content in muscle

Red quadriceps muscles from mice were immunoblotted for (A) phosphorylated and total Akt under basal conditions. Data were normalised by total Akt (left panel) and basal plasma insulin level as shown in Table 1 (right panel), respectively. (B) Phosphorylated and total levels of Akt stimulated with insulin ex vivo. Data were normalised by total Akt. (C) Representative images of immunoblotting. (D) Triglyceride (TG) content in red quadriceps muscles at the end of the study. CH-C, CH control (without CS); HF-C, HF control (without CS); CH-S, CH with CS; HF-S, HF with CS. *P<0.05, **P<0.01 vs CH-C group and #P<0.05 vs HF-C group. Data are shown as mean ± SE of 6–7 mice per group.

We next examined the TG level in red quadriceps muscles because excess intramuscular lipids may impair insulin signalling leading to muscle insulin resistance [12,13]. As expected, there was a ∼3-fold increase in TG in HF-fed mice (P<0.01 vs CH-C) whereas CS alone also moderately increased muscle TG levels (by ∼25% vs CH-C, P<0.05). Interestingly, the combination of CS and HF feeding additively increased TG levels further (∼5-fold, P<0.01vs CH-C, P<0.05 vs HF alone) (Figure 6D). Similar patterns of changes in phospho-Akt and muscle TG content were also observed in white quadriceps muscles (Supplementary Figure S2).

Discussion

The present study investigated the metabolic effects of CS under normal conditions and during dietary fat-induced obesity, with a major focus on the hypothesis of lipid shifting from adipose tissue to muscle. The study was designed to mimic humans who start CS at a young age and with subsequent concurrent overconsumption of diets containing low and high amounts of fat. The present study confirms previous findings of the effect of CS on weight gain in both CH- and HF-fed mice without improving glucose metabolism [5]. Importantly, our study revealed that CS leads to adipose dysfunction, hypertriglyceridemia, lipid accumulation and impairment of insulin signalling in muscle. Our findings suggest that CS may re-direct lipids from adipose tissue to muscle to impair muscle insulin sensitivity which averts the otherwise beneficial effect of lesser weight gain on glucose tolerance.

Preventing obesity or body weight gain is usually beneficial for alleviating the metabolic defects in T2D, including dyslipidaemia, insulin resistance and hyperglycaemia [27]. However, smokers are still at a risk to develop T2D [8–10] even though they often have less body weight gain [5]. Interestingly, CS can co-exist with the consumption of a diet rich in fat, another contributing factor to obesity, dyslipidaemia, insulin resistance and hyperglycaemia [17]. The present study shows that CS twice per day over 14 weeks reduces the gain of body weight (by 11–14%) and adiposity (by ∼20%) in both CH- and HF-fed mice, which are comparable to the levels reported in humans [5]. It has been reported in humans that a 6% weight loss (∼10% fat loss) in the insulin resistant offspring has been shown to reduce fasting plasma insulin by 40% and insulin resistance by 30% by improving insulin sensitivity in muscles (∼40%) due to reduced intramuscular lipids (∼30%) [6]. However, in the present study CS caused dyslipidaemia, hyperinsulinemia, HOMA-IR index and glucose intolerance despite lesser weight gain in the CH-fed mice. Furthermore, our results also demonstrated significant exacerbation of these metabolic phenotypes by CS in HF-fed mice against the benefits of reduced gain of body weight, a similar phenomenon reported in humans [28].

Glucose homeostasis is controlled by its disposal into muscle (as the major component) and hepatic glucose production (to a lesser degree). Studies using a hyperinsulinaemic–euglycaemic clamp in combination with a glucose tracer have identified muscle (not the liver) as the site responsible for improved glucose homeostasis resultant from weight reduction [6]. The present study showed that CS supressed the insulin signalling in muscle (as indicated by the reduced phosphorylation of Akt and GSK3β) and the effect of CS on the phosphorylation of GSK3β was more severe in HF-fed mice. Interestingly, these effects followed the patterns of changes in hyperlipidaemia and accumulation of TG in muscle (red quadriceps). These results suggest that CS-induced exacerbated glucose intolerance is probably due to insulin resistance in muscles as a result of an accumulation of intramuscular lipids. This interpretation is consistent with the mechanism of lipid-induced insulin resistance via the inhibition of the insulin signal transduction (particularly at the sites of IRS1 and Akt) [12,13] and the reported strong relationship between CS and muscle insulin resistance in humans [28].

The levels of lipids in muscles are mainly determined by the balance of the influx of fatty acids and their utilisation. The finding of CS-induced hyperlipidaemia indicates that increased influxes of lipids is likely to be responsible for the lipid accumulation in muscles resulting from CS. The lipid levels in muscle is known to be heavily influenced by capacity and functions of adipose tissue for lipid storage. In the present study, CS reduced adiposity (fat mass) was reflected by the shrinking of adipocytes, indicating a diminished capacity of adipocytes for lipid storage. Thus, CS-induced hyperlipidaemia and impaired insulin signalling in muscle share similarities with lipodystrophic conditions where muscle insulin resistance is caused by redistributed lipids from adipose tissue by its diminished storage and/or increased release [26,29]. To investigate the mechanism responsible for CS-induced lipodystrophic phenotypes, we examined the effects of CS in the lipolytic pathway by its effects on the lipolytic enzymes HSL and ATGL. Our results showed that CS increased levels of HSL in both subcutaneous and testicular adipose tissues. HSL is a lipolytic enzyme predominately located in the endothelium and it plays a key role in facilitating FFA entry into various tissues [30]. Further, activated HSL has been observed in adipose tissue during obesity, insulin resistance and dyslipidaemia [14]. These data collectively suggest an increased lipolytic pathway which may result in a lipodystrophic phenotype following CS exposure. Our results further showed that CS reduced ATGL, which catalyses TG hydrolysis for processing within adipose tissue [31]. Indeed, it has been shown that while stimulating adipose tissue lipolysis nicotine actually inhibits the pro-lipogenic lipase (such as lipoprotein lipase) in adipose tissue [16]. Consistent with this earlier report, our findings suggest that CS is likely to inhibit the ability of adipose tissue to process lipids for storage.

The role of the adipogenic pathway in lipodystrophy syndrome is indicated by the expression levels of genes including leptin and apelin in adipose tissue as biomarkers. In the present study, CS decreased mRNA expression of leptin and apelin in CH-fed mice albeit a reduced apelin mRNA expression was only evident in HF-fed mice. We next measured the protein level of lipin-1 because its deficiency is an important cause of lipodystrophy [14,32]. Indeed, there is evidence showing CS suppresses lipin-1 levels in adipose tissue in CH-S mice and PPARγ (a transcription factor controlling adipocyte differentiation [33]) in HF-S mice. Our studies further showed that CS suppressed insulin-stimulated Akt phosphorylation in adipose tissue. As insulin promotes both adipogenesis and lipogenesis, impaired insulin signalling in adipose tissue may underlie, at least in part, the suppressed adipogenic pathway by CS. This interpretation is also consistent with the fact that lipodystrophy is often associated with impaired insulin signalling in adipose tissue [26,29].

To further investigate the mechanism responsible for the deleterious effects of CS, we examined adipose tissue inflammation which has been reported to be upregulated under insulin resistant state [34] and CS has been shown to induce inflammation in tissues such as lung [25]. Indeed, CS increased the mRNA expression of the inflammatory cytokine TNFα in adipose tissue and this effect was additive to the effect of a HF diet. As TNFα is known to inhibit the insulin signalling pathway [35], it is possible that the impaired insulin signalling in adipose tissue was due to inflammation of this tissue by CS. Notably, TNFα has also been reported to induce lipolysis via suppression of ATGL [35] and inhibit adipogenesis via reduction in PPARγ [36,37]. Taken together, our findings suggest that TNFα may be a key mediator for the effects of CS in adipose tissue.

As adipocytes per se have limited capacity of producing inflammatory cytokines, adipose inflammation has been suggested to be related to macrophage infiltration. Our further analyses showed that mice who had been exposed to CS had up-regulated mRNA expression of the macrophage marker CCL2 in subcutaneous adipose tissue compared with the CH mice. In fact, CS-induced increases in the mRNA expression of TNFα was closely correlated to CCL2. Activation of CCL2 mRNA expression has been reported as a main factor to the development of obesity-induced inflammation, lipolysis and macrophage infiltration in humans [38–40]. We speculate that CS-induced increase in TNFα in adipose tissue in the present study is likely to come from the infiltrated macrophages.

In summary, the present study demonstrates that CS can lead to moderate hyperlipidaemia and glucose intolerance irrespective of the types of diets and the reduced body weight gain. As hypothesised, this is due, at least in part, to increased adipose lipolysis and suppressed adipogenesis which shifts lipids to muscle to cause muscle insulin resistance. Adipose inflammation from macrophage infiltration may play an important role in CS-induced adipose dysfunction (Figure 7). Overall, our findings provide a mechanistic explanation on why the metabolic risk in smokers remains high despite weight loss or reduced body weight gain.

Proposed mechanism underlying the lack of improvement in glucose tolerance from reduced body weight by cigarette smoking

Figure 7
Proposed mechanism underlying the lack of improvement in glucose tolerance from reduced body weight by cigarette smoking

Cigarette smoking causes inflammation, suppresses adipogenesis and stimulates lipolysis in adipose tissue. The liberated fatty acids from lipolysis shifts lipids from adipose tissue to muscle, leading to or exacerbating lipid-induced insulin resistance in muscle. As a result, glucose intolerance persists despite the reduction in body weight gain.

Figure 7
Proposed mechanism underlying the lack of improvement in glucose tolerance from reduced body weight by cigarette smoking

Cigarette smoking causes inflammation, suppresses adipogenesis and stimulates lipolysis in adipose tissue. The liberated fatty acids from lipolysis shifts lipids from adipose tissue to muscle, leading to or exacerbating lipid-induced insulin resistance in muscle. As a result, glucose intolerance persists despite the reduction in body weight gain.

Clinical perspectives

  • Background as to why the study was undertaken

    Cigarette smoking is a risk factor of Type 2 diabetes but its detrimental effect on insulin action is often masked by associated body weight loss.

  • A brief summary of the results

    Cigarette smoking suppresses insulin signalling in muscle to block the benefit of reduced body weight gain to glucose tolerance due to a redistribution of lipids to muscle as a result of dysfunctional adipose tissue.

  • The potential significance of the results to human health and disease

    Currently, more than 1 billion people smoke worldwide. The present study explains why they are still at a risk of developing Type 2 diabetes despite reduced body weight gain from cigarette smoking.

Competing Interests

The authors declare that there are no competing interests associated with the manuscript.

Funding

This work was supported in part by funding from Diabetes Australia Research [Program Grant GG12314] and National Health and Medical Research Council of Australia [Project Grant APP1138915]. Anwar was supported by Saudi Arabia Cultural Mission.

Author Contribution

J.M.Y. and R.V. conceived the study. J.M.Y., S.F. and A.K. designed experiments. A.K. and S.F. performed most of the experiments. S.C., A.M. and C.R. contributed to the experiments and tissue assays. A.K. and X.Z. analysed data and drafted the manuscript. J.M.Y. and R.V. reviewed the data and revised the manuscript.

Acknowledgements

The authors would like to thank Mr Hueijiunn Seow for his technical support.

Abbreviations

     
  • ABC

    avidin-biotin complex

  •  
  • ATGL

    adipose triglyceride lipase

  •  
  • CCL2

    C-C motif chemokine ligand 2

  •  
  • CH

    chow

  •  
  • CH-C

    chow control

  •  
  • CH-S

    chow with cigarette smoking

  •  
  • COPD

    chronic obstructive pulmonary disease

  •  
  • CS

    cigarette smoking

  •  
  • FFA

    free fatty acid

  •  
  • GTT

    glucose tolerance test

  •  
  • HF

    high fat

  •  
  • HF-C

    high fat control

  •  
  • HF-S

    high fat with cigarette smoking

  •  
  • HOMA-IR

    homeostatic model assessment of insulin resistance

  •  
  • HSL

    hormone-sensitive lipase

  •  
  • iAUC

    incremental area under the curve

  •  
  • PPARγ

    peroxisome proliferator-activated receptor gamma

  •  
  • T2D

    type 2 diabetes

  •  
  • TG

    triglyceride

References

References
1.
Bilano
V.
et al.
(
2015
)
Global trends and projections for tobacco use, 1990-2025: an analysis of smoking indicators from the WHO Comprehensive Information Systems for Tobacco Control
.
Lancet
385
,
966
976
[PubMed]
2.
WHO
. (
2017
)
WHO Report on the Global Tobacco Epidemic 2017
,
World Health Organization
3.
Freathy
R.M.
et al.
(
2011
)
Genetic variation at CHRNA5-CHRNA3-CHRNB4 interacts with smoking status to influence body mass index
.
Int. J. Epidemiol.
40
,
1617
1628
[PubMed]
4.
Morris
R.W.
et al.
(
2015
)
Heavier smoking may lead to a relative increase in waist circumference: evidence for a causal relationship from a Mendelian randomisation meta-analysis. The CARTA consortium
.
BMJ Open
5
,
e008808
5.
Chen
H.
et al.
(
2007
)
Detrimental metabolic effects of combining long-term cigarette smoke exposure and high-fat diet in mice
.
Am. J. Physiol. Endocrinol. Metab.
293
,
E1564
E1571
[PubMed]
6.
Petersen
K.F.
et al.
(
2012
)
Reversal of muscle insulin resistance by weight reduction in young, lean, insulin-resistant offspring of parents with type 2 diabetes
.
Proc. Natl. Acad. Sci. U.S.A.
109
,
8236
8240
[PubMed]
7.
Christensen
P.
et al.
(
2018
)
Men and women respond differently to rapid weight loss: Metabolic outcomes of a multi-centre intervention study after a low-energy diet in 2500 overweight, individuals with pre-diabetes (PREVIEW)
.
Diabetes Obes. Metab.
20
,
2840
2851
[PubMed]
8.
Xie
X.T.
et al.
(
2009
)
Impact of cigarette smoking in type 2 diabetes development
.
Acta Pharmacol. Sin.
30
,
784
787
[PubMed]
9.
Willi
C.
et al.
(
2007
)
Active smoking and the risk of type 2 diabetes: a systematic review and meta-analysis
.
JAMA
298
,
2654
2664
[PubMed]
10.
Chang
S.A.
(
2012
)
Smoking and type 2 diabetes mellitus
.
Diabetes Metab. J.
36
,
399
403
[PubMed]
11.
DeFronzo
R.A.
and
Tripathy
D.
(
2009
)
Skeletal muscle insulin resistance is the primary defect in type 2 diabetes
.
Diabetes Care
32
,
S157
S163
[PubMed]
12.
Eckardt
K.
,
Taube
A.
and
Eckel
J.
(
2011
)
Obesity-associated insulin resistance in skeletal muscle: role of lipid accumulation and physical inactivity
.
Rev. Endocrine Metabolic Disorders
12
,
163
172
[PubMed]
13.
Petersen
M.C.
and
Shulman
G.I.
(
2018
)
Mechanisms of Insulin Action and Insulin Resistance
.
Physiol. Rev.
98
,
2133
2223
[PubMed]
14.
Garg
A.
(
2011
)
Lipodystrophies: genetic and acquired body fat disorders
.
J. Clin. Endocrinol. Metab.
96
,
3313
3325
15.
Mann
J.P.
and
Savage
D.B.
(
2019
)
What lipodystrophies teach us about the metabolic syndrome
.
J. Clin. Invest.
130
,
4009
4021
16.
Sztalryd
C.
et al.
(
1996
)
Alterations of lipolysis and lipoprotein lipase in chronically nicotine-treated rats
.
Am. J. Physiol.
270
,
E215
E223
[PubMed]
17.
Cummins
T.D.
et al.
(
2014
)
Metabolic remodeling of white adipose tissue in obesity
.
Am. J. Physiol.-Endocrinol. Metab.
307
,
E262
E277
[PubMed]
18.
Vlahos
R.
et al.
(
2006
)
Differential protease, innate immunity, and NF-κB induction profiles during lung inflammation induced by subchronic cigarette smoke exposure in mice
.
Am. J. Physiol.-Lung Cell. Mol. Physiol.
290
,
L931
L945
[PubMed]
19.
Ye
J.-M.
et al.
(
2003
)
PPARα/γ ragaglitazar eliminates fatty liver and enhances insulin action in fat-fed rats in the absence of hepatomegaly
.
Am. J. Physiol.-Endocrinol. Metab.
284
,
E531
E540
[PubMed]
20.
Turner
N.
et al.
(
2009
)
Enhancement of muscle mitochondrial oxidative capacity and alterations in insulin action are lipid species dependent: potent tissue-specific effects of medium-chain fatty acids
.
Diabetes
58
,
2547
2554
[PubMed]
21.
Ren
L.-P.
et al.
(
2012
)
Differing endoplasmic reticulum stress response to excess lipogenesis versus lipid oversupply in relation to hepatic steatosis and insulin resistance
.
PLoS ONE
7
,
e30816
[PubMed]
22.
Li
S.
et al.
(
2016
)
Dietary cholesterol induces hepatic inflammation and blunts mitochondrial function in the liver of high-fat-fed mice
.
J. Nutr. Biochem.
27
,
96
103
[PubMed]
23.
Chan
S.M.
et al.
(
2013
)
Activation of PPARα ameliorates hepatic insulin resistance and steatosis in high fructose–fed mice despite increased endoplasmic reticulum stress
.
Diabetes
62
,
2095
2105
[PubMed]
24.
Cleasby
M.E.
et al.
(
2004
)
Metformin prevents the development of acute lipid-induced insulin resistance in the rat through altered hepatic signaling mechanisms
.
Diabetes
53
,
3258
3266
[PubMed]
25.
Bozinovski
S.
et al.
(
2015
)
Innate cellular sources of interleukin-17A regulate macrophage accumulation in cigarette-smoke-induced lung inflammation in mice
.
Clin. Sci.
129
,
785
796
[PubMed]
26.
Fiorenza
C.G.
,
Chou
S.H.
and
Mantzoros
C.S.
(
2011
)
Lipodystrophy: pathophysiology and advances in treatment
.
Nat. Rev. Endocrinol.
7
,
137
[PubMed]
27.
Eckel
R.H.
et al.
(
2010
)
The metabolic syndrome
.
Lancet
375
,
181
183
[PubMed]
28.
Bergman
B.C.
et al.
(
2012
)
Novel and reversible mechanisms of smoking-induced insulin resistance in humans
.
Diabetes
61
,
3156
3166
[PubMed]
29.
Villena
J.A.
et al.
(
2008
)
Resistance to high-fat diet–induced obesity but exacerbated insulin resistance in mice overexpressing Preadipocyte Factor-1 (Pref-1): a new model of partial lipodystrophy
.
Diabetes
57
,
3258
3266
[PubMed]
30.
Young
S.G.
and
Zechner
R.
(
2013
)
Biochemistry and pathophysiology of intravascular and intracellular lipolysis
.
Genes Dev.
27
,
459
484
[PubMed]
31.
Morak
M.
et al.
(
2012
)
Adipose triglyceride lipase (ATGL) and hormone-sensitive lipase (HSL) deficiencies affect expression of lipolytic activities in mouse adipose tissues
.
Mol. Cell. Proteomics
11
,
1777
1789
[PubMed]
32.
Pelosi
M.
et al.
(
2017
)
Normal human adipose tissue functions and differentiation in patients with biallelic LPIN1 inactivating mutations
.
J. Lipid Res.
58
,
2348
2364
[PubMed]
33.
Tontonoz
P.
and
Spiegelman
B.M.
(
2008
)
Fat and beyond: the diverse biology of PPARγ
.
Annu. Rev. Biochem.
77
,
289
312
[PubMed]
34.
Ahlin
S.
et al.
(
2013
)
Macrophage gene expression in adipose tissue is associated with insulin sensitivity and serum lipid levels independent of obesity
.
Obesity
21
,
E571
E576
[PubMed]
35.
Lee
M.J.
and
Fried
S.K.
(
2012
)
Glucocorticoids antagonize tumor necrosis factor-alpha-stimulated lipolysis and resistance to the antilipolytic effect of insulin in human adipocytes
.
Am. J. Physiol. Endocrinol. Metab.
303
,
E1126
E1133
[PubMed]
36.
Valyasevi
R.W.
et al.
(
2001
)
Effect of tumor necrosis factor-alpha, interferon-gamma, and transforming growth factor-beta on adipogenesis and expression of thyrotropin receptor in human orbital preadipocyte fibroblasts
.
J. Clin. Endocrinol. Metab.
86
,
903
908
[PubMed]
37.
Xu
H.
and
Hotamisligil
G.S.
(
2001
)
Signaling pathways utilized by tumor necrosis factor receptor 1 in adipocytes to suppress differentiation
.
FEBS Lett.
506
,
97
102
[PubMed]
38.
Arner
E.
et al.
(
2012
)
Adipose tissue microRNAs as regulators of CCL2 production in human obesity
.
Diabetes
61
,
1986
1993
[PubMed]
39.
Michaud
A.
et al.
(
2013
)
Markers of macrophage infiltration and measures of lipolysis in human abdominal adipose tissues
.
Obesity
21
,
2342
2349
[PubMed]
40.
Ahmad
R.
et al.
(
2018
)
The synergy between palmitate and TNF-α for CCL2 production is dependent on the TRIF/IRF3 pathway: implications for metabolic inflammation
.
J. Immunol.
200
,
3599
3611
[PubMed]

Supplementary data