Recent studies suggest that a circulating protein called TRAIL (TNF-related apoptosis inducing ligand) may have an important role in the treatment of type 2 diabetes. It has been shown that TRAIL deficiency worsens diabetes and that TRAIL delivery, when it is given before disease onset, slows down its development. The present study aimed at evaluating whether TRAIL had the potential not only to prevent, but also to treat type 2 diabetes. Thirty male C57BL/6J mice were randomized to a standard or a high-fat diet (HFD). After 4 weeks of HFD, mice were further randomized to receive either placebo or TRAIL, which was delivered weekly for 8 weeks. Body weight, food intake, fasting glucose, and insulin were measured at baseline and every 4 weeks. Tolerance tests were performed before drug randomization and at the end of the study. Tissues were collected for further analyses. Parallel in vitro studies were conducted on HepG2 cells and mouse primary hepatocytes. TRAIL significantly reduced body weight, adipocyte hypertrophy, free fatty acid levels, and inflammation. Moreover, it significantly improved impaired glucose tolerance, and ameliorated non-alcoholic fatty liver disease (NAFLD). TRAIL treatment reduced liver fat content by 47% in vivo as well as by 45% in HepG2 cells and by 39% in primary hepatocytes. This was associated with a significant increase in liver peroxisome proliferator-activated receptor (PPAR) γ (PPARγ) co-activator-1 α (PGC-1α) expression both in vivo and in vitro, pointing to a direct protective effect of TRAIL on the liver. The present study confirms the ability of TRAIL to significantly attenuate diet-induced metabolic abnormalities, and it shows for the first time that TRAIL is effective also when administered after disease onset. In addition, our data shed light on TRAIL therapeutic potential not only against impaired glucose tolerance, but also against NAFLD.

Introduction

TRAIL is an acronym for TNF-related apoptosis inducing ligand, which is a circulating protein belonging to the TNF superfamily. Like other members of this family, such as FasL and TNFα, TRAIL has the ability to induce programmed cell death (apoptosis). However, as compared with them, TRAIL hits preferentially transformed cells, such as cancer cells, while sparing the normal ones [1]. In non-malignant cells, TRAIL actions remain largely unexplored.

Recent experimental evidence suggests that TRAIL has significant metabolic effects [2,3], and might be involved in the regulation of obesity and diabetes mellitus, as well as their complications. The first studies reporting TRAIL beneficial effects on diabetes were carried out in models of type 1 diabetes and showed that TRAIL attenuated disease development and progression [4,5] with partial preservation of islet morphology [6]. Subsequently, we reported that TRAIL delivery significantly reduced the metabolic abnormalities of an experimental model of type 2 diabetes [7]. Consistent with these observations, other groups have shown that genetic lack of TRAIL worsened both forms of diabetes and their associated diseases [8], such as atherosclerosis [8] and non-alcoholic fatty liver disease (NAFLD) [9].

The high-fat diet (HFD) fed mouse is one of the models suitable for the study of type 2 diabetes [10]. After a few weeks, the HFD usually increases body weight and fat mass, leading to peripheral insulin resistance and impaired glucose tolerance, with subsequent hyperglycemia and hyperinsulinemia [10]. Although the HFD-fed mouse does not always develop diabetes, but rather impaired glucose tolerance, it has the advantage of reproducing the human situation more accurately than genetic models of obesity-induced diabetes [11]. Moreover, it allows for the study of common and burdensome diabetic comorbidities, such as obesity, NAFLD, and non-alcoholic steatohepatitis (NASH) [12].

A few years ago, we reported that TRAIL significantly ameliorated the metabolic abnormalities of the HFD-fed mouse. In that paper, TRAIL treatment was started before the development of metabolic abnormalities, following a preventive treatment schedule. In the present study, we aimed at evaluating the therapeutic potential of TRAIL against type 2 diabetes. For this reason, TRAIL treatment was started after the development of metabolic abnormalities, following a therapeutic schedule. Here we report the effects of TRAIL treatment on glucose tolerance and on the tissues regulating it.

Materials and methods

Recombinant human TRAIL

Recombinant histidine-6-tagged human TRAIL (114–281) was produced in transforming bacteria BL21 with a pTrc-His6 TRAIL vector, as described [13] and detailed in the Supplementary materials and methods.

Experimental protocol

As reported in Figure 1A, thirty 8-week-old C57BL/6J male mice (Harlan Laboratories, Udine, Italy) were randomized to receive either a standard diet (SD) (CNT, n=10) or an HFD (n=20) for 12 weeks. After 4 weeks from diet randomization, HFD mice were randomly allocated to receive either saline (HFD, n=10) or recombinant human TRAIL (rhTRAIL) (HFD + TRAIL, n=10) for the remaining 8 weeks. CNT mice received saline (NaCl 0.9%) as well. rhTRAIL (TRAIL) was given at a dose of 10 μg/200 μl per mouse by weekly intraperitoneal (IP) injection. SD provided 22% of calories from protein, 66% of calories from carbohydrate, and 12% of calories from fat, and a digestible energy of 3.0 kcal/g (Teklad Global 16% Protein Rodent Diet, Harlan Laboratories). HFD provided 18.4% of calories from protein, 21.3% of calories from carbohydrate, and 60% of calories from fat, and a digestible energy of 5.1 kcal/g (TD 06.414 adjusted calories diet 60/fat, Harlan Laboratories). Animals were kept (five per cage) in ventilated cabinets (Tecniplast Spa, Buguggiate, Varese, Italy), in specific pathogen-free and temperature-controlled rooms (22°C), with relative humidity of 50–70%, on a 12-h light/12-h dark cycle. They had free access to food and water, and they were fed ad libitum for the length of the study. During the 12-week study period, body weight, food intake, glucose, and insulin were measured at baseline and at 4-week intervals. Tolerance tests were performed before drug randomization and at the end of the study. At the end of the study, animals were anesthetized by an IP injection of tiletamine/zolazepam (80 mg/kg). Blood was collected from the left ventricle, centrifuged, and serum was stored for further analyses (total cholesterol (TC), HDL cholesterol (HDL-C), triglycerides (TGs), free fatty acids, C-reactive protein (CRP)). Pancreas, perigonadal white adipose tissue (pgWAT), liver, and quadriceps were weighed and either snap-frozen or fixed in formalin for further analyses.

Treatment protocol

Figure 1
Treatment protocol

(A) Schematic illustration of the protocol: C57BL/6J mice were fed either a SD (12% fat) or HFD (60% fat) for 12 weeks. Mice on HFD were treated either with saline (NaCl 0.9%) or TRAIL weekly between week 5 and 12. Mice on SD were treated with saline. (*) Tolerance tests before drug randomization; (**) tolerance tests at the end of the study. (B) Glucose and insulin tolerance tests performed before drug randomization. Upper figures show blood glucose and its area under the curve (AUC) during an IP glucose tolerance test (IPGTT), middle figures show serum insulin and its AUC during an IPGTT, and lower figures show blood glucose and its AUC during an IP insulin tolerance test (IPITT). (C) Representative images of Cy5.5-TRAIL clearance from the peritoneal cavity, as assessed by acquisition of fluorescence emission from the peritoneum after an IP injection of 10 μg of Cy5.5-TRAIL. Cy5.5-TRAIL is TRAIL that was labeled with N-hydroxysuccinimide ester of the cyanine 5.5. Fluorescence emission was collected at 700 nm. (D) Circulating human TRAIL after IP injection of 10 μg of TRAIL.

Figure 1
Treatment protocol

(A) Schematic illustration of the protocol: C57BL/6J mice were fed either a SD (12% fat) or HFD (60% fat) for 12 weeks. Mice on HFD were treated either with saline (NaCl 0.9%) or TRAIL weekly between week 5 and 12. Mice on SD were treated with saline. (*) Tolerance tests before drug randomization; (**) tolerance tests at the end of the study. (B) Glucose and insulin tolerance tests performed before drug randomization. Upper figures show blood glucose and its area under the curve (AUC) during an IP glucose tolerance test (IPGTT), middle figures show serum insulin and its AUC during an IPGTT, and lower figures show blood glucose and its AUC during an IP insulin tolerance test (IPITT). (C) Representative images of Cy5.5-TRAIL clearance from the peritoneal cavity, as assessed by acquisition of fluorescence emission from the peritoneum after an IP injection of 10 μg of Cy5.5-TRAIL. Cy5.5-TRAIL is TRAIL that was labeled with N-hydroxysuccinimide ester of the cyanine 5.5. Fluorescence emission was collected at 700 nm. (D) Circulating human TRAIL after IP injection of 10 μg of TRAIL.

The Guide for the Care and Use of Laboratory Animals, 8th edition (2011), as well as specific European (86/609/EEC) and Italian (D.L.116/92) laws were followed. In compliance with the principle of reducing as much as possible the number of mice studied, we did not include CNT mice + TRAIL because we have previously observed and reported that repeated injections of TRAIL (providing a cumulative dose of 100 μg/mouse) did not affect glucose, insulin, and/or body weight in normal conditions in vivo [6]. The present study was approved by the Institutional Animal Care and Use Committee of the Cluster in Biomedicine (CBM) and by the Italian Ministry of Health (DM 17/2001 A dd. 02/02/2011). The study period was from June 2015 to May 2017.

Assessment of TRAIL biodistribution

In order to evaluate TRAIL bioavailability when delivered by IP injection, 20-week-old C57BL/6J male mice (n=3) received 10 μg of Cy5.5-TRAIL by IP injection, and fluorescence in the peritoneal cavity was acquired at baseline and after 30 min, 6, and 30 h post injection, as previously reported [14]. The details of TRAIL labeling are in the Supplementary materials and methods. In addition, 20-week-old C57BL/6J male mice (n=8) received 10 μg of TRAIL by IP injection, blood was collected at baseline and after 6, 30, 48, and 72 h, and TRAIL was measured by ELISA (R&D; #DTLR00).

General parameters and biochemistries

Food intake was measured by placing in the cages pellets previously weighed in total. The food that was left over was collected and weighed to find the amount eaten. Energy intake was measured according to the digestible energy provided by both diets. Fasting glucose was measured by glucometer (GlucoMen LX Plus, Menarini). Fasting insulin was measured by ELISA (Millipore; #EZRMI-13K). NEFA concentrations were measured by colorimetric assay (Sigma; #MAK044-1KT). Circulating TC, HDL-C, and TGs were measured with the AU5800 analyzer (Beckman Coulter) by enzymatic colorimetric method, while CRP was measured by immunoturbidimetric method.

Tolerance tests

The IP glucose tolerance test (IPGTT) was performed on day 1 of week 4 and 12 by injecting glucose (2 g/kg) intraperitoneally after an overnight fast. Glucose and insulin were measured at baseline and at 15, 60, and 120 min. The IP insulin tolerance test (IPITT) was performed on day 1 of week 3 and 11 by injecting insulin (1 unit/kg) intraperitoneally after a 6-h fast. Glucose was measured at baseline and at 30, 60, and 120 min.

Tissue stainings

Adipocyte area was measured on pgWAT paraffin sections (4 μm) stained with H&E. pgWAT macrophages were detected by F4/80 immunostaining (1:100 dilution, applied overnight; Abcam #Ab111101) and reported as positive nuclei/100 nuclei. Pancreatic β-cell density and mass were estimated on paraffin sections (4 μm) by insulin immunostaining (1:100 applied overnight; DAKO #A0564), as previously described [15]. Liver steatosis was evaluated on frozen sections (5 μm) stained with Oil Red O, where fat was quantitated as % of positive (red) staining/tissue area. The same method was used for skeletal muscle fat content. For liver fibrosis, paraffin sections (4 μm) were stained with Picrosirius Red, and fibrosis was quantitated as % of positive (red) staining/tissue area [15]. Liver macrophages were detected on frozen sections (5 μm) by CD68 immunostaining (1:50 applied overnight; Serotec #MCA1957S), and they were reported as positive nuclei/frame. Liver peroxisome proliferator-activated receptor (PPAR) γ (PPARγ) co-activator-1 α (PGC-1α) expression was measured on paraffin sections (4 μm) by PGC-1α immunostaining (1:50 applied overnight; Abcam #ab191838) and was reported as the % of positive (brown) staining/tissue area. All the sections were examined by light microscopy (Carl Zeiss – Jenaval) and digitized using a high-resolution camera (Q-Imaging Fast 1394). The % of staining/tissue area was quantitated using ImagePro Plus 6.3 (Media Cybernetics, Bethesda, MD, U.S.A.). Quantitations were performed on 40–100 frames per group.

Liver TG content

For liver TGs, 100 mg of liver were homogenized in 1 ml of 5% NP40. Samples were heated at 95° for 5 min, cooled down at RT for 5 min twice, and then centrifuged to collect the supernatants. TGs were measured with the AU5800 analyzer (Beckman Coulter) by enzymatic colorimetric method (for details, see the Supplementary materials and methods).

Gene expression quantitation by RT-qPCR

Gene expression was determined by real-time quantitative RT-qPCR. In order to isolate mRNA, tissue was homogenized and processed as recently reported [16]. Then, mRNA was treated to eliminate DNA contamination (Ambion DNA-free product #AM-1906), and 3  μg of treated mRNA were subsequently used to synthesize cDNA with Superscript First Strand synthesis system for RT-PCR (Gibco BRL). The gene expression of fatty acid synthase (Fas), sterol regulatory element binding protein (Srebp)1a (Srebp1a), Srebp1c, Acox, carnitine palmitoyl transferase-1a (Cpt1a), Pparα, Irs2, Pepck, Pparγ, hepatocyte nuclear factor 4 (Hnf4), citrate synthase, Pgc-1α, uncoupling protein 2 (Ucp2), sirtuin (Sirt), perilipin (Pln), Gp91phox (is a subunit of NADPH oxidase), interleukin-6 (Il-6), Tnfα, and autophagy-related protein 7 (Atg7) was analyzed by RT-qPCR using the SYBR Green system (Life Technologies). Monocyte chemoattractant protein-1 (Mcp1) was analyzed using the TaqMan system (Life Technologies). Fluorescence for each cycle was quantitatively analyzed by StepOnePlus real-time PCR system (Applied Biosystems). Gene expression was normalized to Rps9 or 18s. Results are reported as ratio compared with the level of expression in untreated controls, which were given an arbitrary value of 1. Primers are reported in Supplementary Table S1.

In vitro studies on HepG2 cells and mouse primary hepatocytes

Human hepatocellular carcinoma cells (HepG2) obtained from ATCC were cultured in high-glucose DMEM, supplemented with 10% (v/v) FBS, l-glutamine (2 mM), penicillin (100 U/ml), and streptomycin (100 μg/ml) in a 5% CO2 atmosphere at 37°C. In order to mimic a HFD milieu, cells were grown in medium supplemented with either BSA-conjugated palmitate (250 mM) or BSA-conjugated oleate (250 mM). Cells were cultured for 24 h with BSA-conjugated oleate with or without TRAIL (1 ng/ml) for cell viability, lipid accumulation, and protein expression studies, while they were cultured for 6 h for gene expression analysis. Cell viability was assessed with the MTT assay (Alfa Aesar; #L11939). Total intracellular lipid content was evaluated by Oil Red O staining. Protein expression of PGC-1α was analyzed by immunofluorescence. Gene expression of PPARγ, PGC-1α, and ATG7 was analyzed by RT-qPCR using the SYBR Green system (Life Technologies). cDNA was synthesized as detailed for our in vivo analyses. Fluorescence for each cycle was quantitatively analyzed by StepOnePlus real-time PCR system (Applied Biosystems). Gene expression was normalized to GADPH or Rpl27, and reported as a ratio compared with the level of expression in untreated controls, which were given an arbitrary value of 1. Primers are reported in Supplementary Table S1.

In parallel with HepG2 in vitro studies, mouse primary hepatocytes were isolated from C57BL/6J male mice following a protocol adapted from Severgnini et al. [17]. After digestion with collagenase IV, liver suspension was passed through a cell strainer. Primary hepatocytes were washed four times and seeded on precoated plates with a plating medium (Williams E medium supplemented with CM3000, Thermo Fisher Scientific) for 5 h in a 5% CO2 atmosphere at 37°C. Then, this medium was changed to a maintenance solution (CM4000 Thermo Fisher Scientific) for 24 h before starting the treatment. As before, mouse primary hepatocytes were cultured for 24 h with BSA-conjugated oleate (250 mM) or BSA alone, with or without TRAIL (1 ng/ml). Total intracellular lipid content was evaluated by Oil Red O staining.

Statistics

Results are expressed as means ± S.E.M. Differences in the mean amongst groups were analyzed using one-way or two-way ANOVA. Pairwise multiple comparisons were made using Bonferroni post-hoc analysis. A threshold of P<0.05 was considered statistically significant.

Results

Animal model and in vivo bioavailability of TRAIL

Mice were put on a high-fat dietary regime to induce obesity and associated metabolic comorbidities (Figure 1A). After 4 weeks, before drug randomization, HFD mice had increased in weight (6.3 ± 0.6 g) as compared with CNT mice (3.7 ± 0.6 g; P<0.05 compared with HFD). HFD mice weighed 30.4 ± 0.5 g and CNT mice weighed 28.9 ± 0.7 g. There were no differences between the groups in terms of fasting glucose and insulin. Nevertheless, the glucose and insulin tolerance tests showed that HFD mice had developed impaired glucose tolerance as compared with CNT mice (Figure 1B). At this stage, HFD mice were randomized to receive either TRAIL or saline by weekly IP injection for 8 weeks. This schedule was chosen based on our previous study [7]. Mice tolerated this treatment well, and they did not show any sign of distress during the study or gross abnormalities at necroscopic examination as compared with those treated with saline, in line with previous reports [18]. When we then looked at TRAIL bioavailability, we found that TRAIL was detectable after 6 h from the IP injection and then disappeared 24 h later. These data were in line with our imaging experiment showing that the highest fluorescence signal of Cy5.5-TRAIL was detected 6 h after the IP injection, and then tended to decrease over time (Figure 1C,D).

TRAIL treatment significantly reduced obesity, as well as adipose and systemic inflammation

At the end of the study, HFD mice became obese as compared with CNT mice. TRAIL treatment significantly reduced body weight gain in HFD mice (Figure 2A), without affecting their food and energy intake. In particular, HFD mice ate 2.7 ± 0.2 g/day with an energy intake of 13.7 ± 1.2 kcal/day, as compared with HFD + TRAIL mice, who ate 2.6 ± 0.3 g/day with an energy intake of 13.4 ± 1.3 kcal/day. Second, HFD mice exhibited WAT hypertrophy, as assessed by pgWAT weight (Figure 2B), which is considered one of the largest visceral WAT depots [19]. This was associated with an increase in the adipocyte area (Figure 2C–E) and the number of infiltrating macrophages (Figure 2E,F). TRAIL treatment significantly reduced WAT hypertrophy, adipocyte area, and macrophage accumulation in the WAT (Figure 2B–F). Third, HFD mice exhibited higher levels of NEFA and CRP (Figure 2G–I), in line with the concept that excess fat is associated with high levels of NEFA [20] and a low-grade systemic inflammatory state [21]. TRAIL treatment significantly reduced both NEFA and CRP (Figure 2G–I). Interestingly, TRAIL treatment was associated with a significant increase in TC, driven by an increase in HDL-C.

Body weight, adipose tissue, and inflammation

Figure 2
Body weight, adipose tissue, and inflammation

(A) Body weight throughout the study. The gray area corresponds to the treatment period. (B) pgWAT weight at the end of the study. (C,D) Frequency distribution and adipocyte area. (E) Representative images of pgWAT H&E staining (upper panel; 12.5×, scale bar 50 μm) and F4/80+ macrophages (lower panel; 25×, scale bar 50 μm).(F) Quantitation of F4/80+ cells (brown nuclei)/100 cells. (G) NEFA levels at the end of the study. (H) TC, HDL-C, and TGs at the end of the study. (I) CRP levels at the end of the study. Results are presented as mean ± S.E.M.; *P<0.05 compared with CNT; #P<0.05 compared with HFD. (F) Quantitation of F4/80+ cells (brown nuclei)/100 cells.

Figure 2
Body weight, adipose tissue, and inflammation

(A) Body weight throughout the study. The gray area corresponds to the treatment period. (B) pgWAT weight at the end of the study. (C,D) Frequency distribution and adipocyte area. (E) Representative images of pgWAT H&E staining (upper panel; 12.5×, scale bar 50 μm) and F4/80+ macrophages (lower panel; 25×, scale bar 50 μm).(F) Quantitation of F4/80+ cells (brown nuclei)/100 cells. (G) NEFA levels at the end of the study. (H) TC, HDL-C, and TGs at the end of the study. (I) CRP levels at the end of the study. Results are presented as mean ± S.E.M.; *P<0.05 compared with CNT; #P<0.05 compared with HFD. (F) Quantitation of F4/80+ cells (brown nuclei)/100 cells.

TRAIL treatment significantly reduced impaired glucose tolerance

The tolerance tests (Figure 3A–F) performed at the end of the study showed that the HFD resulted in impairment of glucose clearance, leading to significant hyperglycemia and hyperinsulinemia 2 h after a glucose load (Figure 3D,E). This was associated with the inability of insulin to lower glucose levels to the levels of CNT mice (Figure 3F). At fasting, glucose did not differ between the groups, and HFD mice displayed only a significant hyperinsulinemia as compared with CNT mice (Figure 3G), indicating the presence of insulin resistance with impaired glucose tolerance. These data were consistent with the lack of significant changes in β-cell mass and density (Figure 3H; Supplementary Figure S1). TRAIL treatment significantly reduced both impaired glucose tolerance and insulin resistance (Figure 3A–G).

Glucose metabolism

Figure 3
Glucose metabolism

(AF) Glucose and insulin tolerance tests performed at the end of the study. (A) Blood glucose during an IPGTT. (B) Serum insulin during an IPGTT. (C) Blood glucose during an IPITT. (D) Area under the curve (AUC) of the glucose levels during the IPGTT. (E) AUC of the insulin levles during the IPGTT. (F) AUC of the glucose levels during the IPITT. (G) Fasting glucose and insulin at the end of the study. (H) Representative images of pancreatic sections immunostained for insulin (25×, scale bar 50 μm). Results are presented as mean ± S.E.M.; *P<0.05 compared with CNT; #P<0.05 compared with HFD.

Figure 3
Glucose metabolism

(AF) Glucose and insulin tolerance tests performed at the end of the study. (A) Blood glucose during an IPGTT. (B) Serum insulin during an IPGTT. (C) Blood glucose during an IPITT. (D) Area under the curve (AUC) of the glucose levels during the IPGTT. (E) AUC of the insulin levles during the IPGTT. (F) AUC of the glucose levels during the IPITT. (G) Fasting glucose and insulin at the end of the study. (H) Representative images of pancreatic sections immunostained for insulin (25×, scale bar 50 μm). Results are presented as mean ± S.E.M.; *P<0.05 compared with CNT; #P<0.05 compared with HFD.

TRAIL treatment significantly reduced NAFLD

To understand why TRAIL reduced impaired glucose tolerance, we looked at the tissues regulating glucose metabolism, focussing on the liver and skeletal muscle [22]. At the end of the study, the HFD regime did not increase the deposition of fat (lipid droplets) in the skeletal muscle, as assessed by Oil Red O staining (results not shown). By contrast, when we looked at the liver, HFD mice developed significant steatosis (lipid accumulation >10% of tissue area), which was assessed by Oil Red O staining and by TG content quantitation (Figure 4A–C). The accumulation of more than 5–10% of fat in the liver, without any primary cause such as viral hepatitis, alcoholic disease, or drug-induced liver injury is what defines histologically NAFLD [23]. In our study, HFD mice also exhibited a significant increase in liver macrophage infiltration (Figure 4C,D), but no fibrosis, which should appear after approximately 50 weeks of HFD [12]. TRAIL treatment significantly reduced liver steatosis as well as inflammation, therefore ameliorating HFD-induced NAFLD.

Liver changes

Figure 4
Liver changes

(A) Liver TG content. (B) Percentage of hepatic area positive for fat (stained red) with Oil Red O (ORO). (C) Representative images of liver ORO staining (25×, scale bar 50 μm), Picrosirius Red (Sirius Red) staining (12.5×, scale bar 100 μm), CD68+ macrophages (12.5×, scale bar 100 μm), and PGC-1α staining (12.5×, scale bar 100 μm). (D) Quantitation of CD68+ cells (brown nuclei)/frame liver tissue. (E) Liver mRNA expression of Il-6. mRNA expression is reported as fold induction standardized to the mRNA expression in CNT mice. (F) Liver mRNA expression of Pgc-1α. mRNA expression is reported as fold induction standardized to the mRNA expression in CNT mice. (G) Quantitation of PGC-1α + staining (brown)/area of liver tissue. Results are presented as mean ± S.E.M.; *P<0.05 compared with CNT; #P<0.05 compared with HFD.

Figure 4
Liver changes

(A) Liver TG content. (B) Percentage of hepatic area positive for fat (stained red) with Oil Red O (ORO). (C) Representative images of liver ORO staining (25×, scale bar 50 μm), Picrosirius Red (Sirius Red) staining (12.5×, scale bar 100 μm), CD68+ macrophages (12.5×, scale bar 100 μm), and PGC-1α staining (12.5×, scale bar 100 μm). (D) Quantitation of CD68+ cells (brown nuclei)/frame liver tissue. (E) Liver mRNA expression of Il-6. mRNA expression is reported as fold induction standardized to the mRNA expression in CNT mice. (F) Liver mRNA expression of Pgc-1α. mRNA expression is reported as fold induction standardized to the mRNA expression in CNT mice. (G) Quantitation of PGC-1α + staining (brown)/area of liver tissue. Results are presented as mean ± S.E.M.; *P<0.05 compared with CNT; #P<0.05 compared with HFD.

TRAIL significantly increased PGC-1α expression in the liver

In parallel studies, we quantitated liver mRNA encoding transcription factors and metabolic enzymes involved in de novo lipogenesis, fatty acid oxidation, glucose metabolism, and mitochondrial function, as well as pro-oxidative and proinflammatory molecules. In the liver, HFD significantly increased the gene expression of Fas, Pparγ, and Il-6, while it decreased that of Cpt1α, Pepck, and Pgc-1α, as compared with CNT mice (Table 1). In HFD mice, TRAIL significantly decreased the gene expression of Fas and Il-6, while it increased that of Pparγ and Pgc-1α (Table 1 and Figure 4E,F). The increase in Pparγ and Pgc-1α gene expression that followed TRAIL treatment was observed only in the liver (Supplementary Figure S2A). Given that PGC-1α regulates energy metabolism and could explain part of our findings, we further evaluated it by immunostaining. In HFD mice, PGC-1α protein expression increased and changed pattern of distribution as compared with CNT mice, where it was located in the nuclei. TRAIL further increased PGC-1α protein expression in the liver (Figure 4C,G). In addition, we quantitated liver mRNA encoding for Atg7, which is a protein essential for autophagy, a process that has been previously implicated in lipid metabolism regulation [24]. Interestingly, an increase in Atg7 was observed in the liver of HFD + TRAIL mice (Table 1 and Supplementary Figure S2A).

Table 1
Liver gene expression analysis
 CNT HFD HFD + TRAIL HFD compared with CNT HFD + TRAIL compared with CNT HFD compared with HFD + TRAIL 
De novo lipogenesis 
Fas 1.00 ± 0.10 1.60 ± 0.11 1.20 ± 0.07 <0.01 NS <0.05 
Srebp1a 1.00 ± 0.04 1.17 ± 0.10 1.16 ± 0.11 NS NS NS 
Srebp1c 1.00 ± 0.20 0.70 ± 0.06 1.17 ± 0.21 NS NS NS 
Fatty acid oxidation 
Aox 1.00 ± 0.08 0.77 ± 0.06 0.83 ± 0.07 NS NS NS 
Cpt1a 1.00 ± 0.06 0.51 ± 0.07 0.55 ± 0.07 <0.0001 <0.0001 NS 
Pparα 1.00 ± 0.11 0.75 ± 0.14 0.85 ± 0.15 NS NS NS 
Gluconeogenesis, insulin signaling, insulin sensitivity 
Irs2 1.00 ± 0.09 0.79 ± 0.08 0.74 ± 0.12 NS NS NS 
Pepck 1.00 ± 0.10 0.30 ± 0.09 0.32 ± 0.06 <0.0001 <0.0001 NS 
Pparγ 1.00 ± 0.10 1.82 ± 0.21 2.57 ± 0.20 <0.01 <0.0001 <0.05 
Hnf4 1.00 ± 0.11 0.99 ± 0.11 1.07 ± 0.09 NS NS NS 
Mitochondrial function 
Cit synt 1.00 ± 0.09 1.00 ± 0.11 1.12 ± 0.08 NS NS NS 
Pgc-1α 1.00 ± 0.05 0.41 ± 0.05 0.65 ± 0.03 <0.0001 <0.0001 <0.01 
Pln5 1.00 ± 0.21 0.72 ± 0.07 0.81 ± 0.08 NS NS NS 
Ucp2 1.00 ± 0.11 0.69 ± 0.07 0.79 ± 0.11 NS NS NS 
Sirt-1 1.00 ± 0.07 1.06 ± 0.10 1.03 ± 0.06 NS NS NS 
Oxidative stress and inflammation 
Gp91phox 1.00 ± 0.13 0.83 ± 0.15 0.86 ± 0.16 NS NS NS 
Il-6 1.00 ± 0.18 1.74 ± 0.14 1.16 ± 0.10 <0.01 NS <0.05 
Mcp1 1.00 ± 0.20 1.95 ± 0.25 1.99 ± 0.38 NS NS NS 
Tnfα 1.00 ± 0.14 0.82 ± 0.12 1.14 ± 0.23 NS NS NS 
Autophagy 
Atg7 1.00 ± 0.06 1.33 ± 0.12 1.53 ± 0.11 <0.05 <0.01 NS 
 CNT HFD HFD + TRAIL HFD compared with CNT HFD + TRAIL compared with CNT HFD compared with HFD + TRAIL 
De novo lipogenesis 
Fas 1.00 ± 0.10 1.60 ± 0.11 1.20 ± 0.07 <0.01 NS <0.05 
Srebp1a 1.00 ± 0.04 1.17 ± 0.10 1.16 ± 0.11 NS NS NS 
Srebp1c 1.00 ± 0.20 0.70 ± 0.06 1.17 ± 0.21 NS NS NS 
Fatty acid oxidation 
Aox 1.00 ± 0.08 0.77 ± 0.06 0.83 ± 0.07 NS NS NS 
Cpt1a 1.00 ± 0.06 0.51 ± 0.07 0.55 ± 0.07 <0.0001 <0.0001 NS 
Pparα 1.00 ± 0.11 0.75 ± 0.14 0.85 ± 0.15 NS NS NS 
Gluconeogenesis, insulin signaling, insulin sensitivity 
Irs2 1.00 ± 0.09 0.79 ± 0.08 0.74 ± 0.12 NS NS NS 
Pepck 1.00 ± 0.10 0.30 ± 0.09 0.32 ± 0.06 <0.0001 <0.0001 NS 
Pparγ 1.00 ± 0.10 1.82 ± 0.21 2.57 ± 0.20 <0.01 <0.0001 <0.05 
Hnf4 1.00 ± 0.11 0.99 ± 0.11 1.07 ± 0.09 NS NS NS 
Mitochondrial function 
Cit synt 1.00 ± 0.09 1.00 ± 0.11 1.12 ± 0.08 NS NS NS 
Pgc-1α 1.00 ± 0.05 0.41 ± 0.05 0.65 ± 0.03 <0.0001 <0.0001 <0.01 
Pln5 1.00 ± 0.21 0.72 ± 0.07 0.81 ± 0.08 NS NS NS 
Ucp2 1.00 ± 0.11 0.69 ± 0.07 0.79 ± 0.11 NS NS NS 
Sirt-1 1.00 ± 0.07 1.06 ± 0.10 1.03 ± 0.06 NS NS NS 
Oxidative stress and inflammation 
Gp91phox 1.00 ± 0.13 0.83 ± 0.15 0.86 ± 0.16 NS NS NS 
Il-6 1.00 ± 0.18 1.74 ± 0.14 1.16 ± 0.10 <0.01 NS <0.05 
Mcp1 1.00 ± 0.20 1.95 ± 0.25 1.99 ± 0.38 NS NS NS 
Tnfα 1.00 ± 0.14 0.82 ± 0.12 1.14 ± 0.23 NS NS NS 
Autophagy 
Atg7 1.00 ± 0.06 1.33 ± 0.12 1.53 ± 0.11 <0.05 <0.01 NS 

Fas, Srebp1a, and Srebp1c are transcription factors and enzymes involved in lipogenesis. Acyl-CoA oxidase (Aox), Cpt1a, and Pparα are transcription factors and enzymes involved in fatty acid oxidation. Irs2 regulates insulin signaling; Pepck regulates gluconeogenesis; Pparγ sensitizes to insulin. Hnf4 is a transcriptional regulator of gluconeogenic genes. Citrate synthase (Cit synt), which is the pace-making enzyme of Krebs cycle, and Ucp2 are localized in mitochondria, while Pgc-1α, perilipin 5 (Pln5), and sirtuin-1 (Sirt-1) regulate mitochondrial functions. Gp91phox is a subunit of NADPH oxidase, which is involved in oxidative stress, Il-6, Mcp1, and Tnfα are proinflammatory mediators. Atg7 is a molecule involved in autophagy.

NS: not significant

TRAIL significantly reduced lipid droplet accumulation in hepatocytes cultured in an HFD milieu

To determine whether TRAIL had direct effects on hepatocytes, we used an in vitro model of NAFLD [9], and cultured HepG2 cells with either palmitate or oleate. It must be noted that although HepG2 cells are derived from hepatocellular carcinoma and they express TRAIL receptors (Supplementary Figure S2B), they are normally resistant to TRAIL-induced apoptosis [25]. MTT assay showed that palmitate significantly impaired HepG2 cell viability, while oleic acid had no effect on it (Figure 5A). In HepG2 cells cultured with oleic acid, TRAIL treatment promoted cell viability (Figure 5A), which is consistent with earlier observations that TRAIL can also activate survival pathways [26]. Most importantly, in these cells, TRAIL treatment reduced lipid droplet accumulation by 45% (Figure 5B,C). This effect was confirmed in primary hepatocytes, where TRAIL treatment reduced lipid droplet accumulation by 39% (Figure 5B,C). In addition, in HepG2 cells cultured with oleic acid, TRAIL up-regulated PPARγ, PGC-1α, and ATG7 gene expression (Figure 6A–C), as well as PGC-1α protein expression (Figure 6D), while it had no effect on specific proinflammatory markers (Supplementary Figure S2C).

Hepatic lipid droplet accumulation in vitro

Figure 5
Hepatic lipid droplet accumulation in vitro

(A) HepG2 cell viability after exposure to palmitic acid or oleic acid in presence or absence of TRAIL. (B) HepG2 and mPH percentage of cell surface positive for lipid droplets (dark red staining) with Oil Red O (ORO). Cells were incubated with oleic acid (250 μM) in presence or absence of TRAIL (1 ng/ml). HepG2 is for HepG2 cells and mPH is for mouse primary hepatocytes. (C) Representative images of HepG2 cells (upper panel) and mouse primary hepatocytes (lower panel) stained with ORO (20×, scale bar 100 μm). Results are presented as mean ± S.E.M.; *P<0.05 compared with control; #P<0.05 compared with oleic acid without TRAIL.

Figure 5
Hepatic lipid droplet accumulation in vitro

(A) HepG2 cell viability after exposure to palmitic acid or oleic acid in presence or absence of TRAIL. (B) HepG2 and mPH percentage of cell surface positive for lipid droplets (dark red staining) with Oil Red O (ORO). Cells were incubated with oleic acid (250 μM) in presence or absence of TRAIL (1 ng/ml). HepG2 is for HepG2 cells and mPH is for mouse primary hepatocytes. (C) Representative images of HepG2 cells (upper panel) and mouse primary hepatocytes (lower panel) stained with ORO (20×, scale bar 100 μm). Results are presented as mean ± S.E.M.; *P<0.05 compared with control; #P<0.05 compared with oleic acid without TRAIL.

In vitro studies on HepG2 cells

Figure 6
In vitro studies on HepG2 cells

(A) HepG2 cell mRNA expression of PPARγ; (B) PGC-1α; (C) ATG7. (D) Representative images of PGC-1α immunostaining on HepG2 cells. Results are presented as mean ± S.E.M.; *P<0.05 compared with control without TRAIL.

Figure 6
In vitro studies on HepG2 cells

(A) HepG2 cell mRNA expression of PPARγ; (B) PGC-1α; (C) ATG7. (D) Representative images of PGC-1α immunostaining on HepG2 cells. Results are presented as mean ± S.E.M.; *P<0.05 compared with control without TRAIL.

Discussion

The first new finding of the present study is that TRAIL significantly ameliorated diet-induced metabolic abnormalities even when it was administered after the development of impaired glucose tolerance. At the end of the study, HFD mice became obese and insulin resistant, displaying impaired glucose tolerance as well as an increase in NEFA and CRP levels. By comparison, the group of HFD + TRAIL mice showed a significant reduction in body weight gain, NEFA and CRP levels, as well as a significant amelioration of insulin resistance and impaired glucose tolerance.

Looking at the adipose tissue, TRAIL reduced body weight gain, which was associated with a decrease in fat weight, adipocyte size, and pgWAT macrophage infiltration. These findings are consistent with the earlier observations that TRAIL had significant effects on body weight. As a matter of fact, genetic TRAIL deficiency was found to be associated with increased body weight [8], and we reported that TRAIL treatment significantly reduced the adiposity of HFD-fed mice as assessed by EchoMRI [7]. More recently, it has been shown that TRAIL had the ability to inhibit adipogenic differentiation through caspase activation [27]. When interpreting our results in light of this finding, it is also by reducing fat mass that TRAIL could have ameliorated insulin resistance, impaired glucose tolerance, as well as circulating NEFA and CRP in HFD mice. Excess fat mass has been associated not only with insulin resistance [20], but also with high levels of NEFA [20] and a low-grade systemic inflammatory state [21]. Moreover, experimental studies have shown that insulin sensitivity and systemic inflammation improve following adipose tissue removal [28].

In the present study, HFD mice developed also hepatic steatosis and inflammation, corresponding to human NAFLD. At present, NAFLD is found to be a frequent comorbid factor in the setting of type 2 diabetes [29,30]. It is estimated that approximately 70% of obese patients with diabetes have NAFLD and as many as 30–40% have NASH, which is characterized by hepatic steatosis with inflammation and/or necrosis [30]. Both NAFLD and NASH are conditions leading to hepatic cirrhosis, end-stage liver disease, and hepatocellular carcinoma [29]. Given that NAFLD is reaching epidemic proportions in diabetic patients [29,30], it is predicted that cirrhosis related to NASH will surpass HCV-related cirrhosis as the most common indication for liver transplantation in the United States [31].

Therefore, the second important finding of the present study is that TRAIL treatment markedly reduced NAFLD in the HFD-fed mouse, where it decreased liver fat content by 47%. Given that NAFLD generally promotes metabolic abnormalities [29,30], it is also by reducing liver fat content that TRAIL might have ameliorated insulin resistance and subclinical inflammation in HFD-fed mice. To date, only a few studies have addressed the relationship between TRAIL and NAFLD [9,32]. By comparison, this is the first study describing the direct effect of TRAIL in an experimental model of NAFLD. Nevertheless, our data are consistent with the finding that TRAIL deficiency worsens NAFLD [9].

There seem to be several mechanisms underlying TRAIL actions on the liver. First, the reduction in body weight gain, fat mass, and circulating NEFA, which were induced by TRAIL, could have ameliorated NAFLD in the HFD + TRAIL group. It has been argued that excess storage of hepatic TGs comes mostly from an excess of circulating NEFA [33], which is usually associated with visceral obesity [34] or adipose tissue insulin resistance/inflammation [34,35]. Second, the significant reduction in insulin, which was observed in the HFD + TRAIL group, could have also contributed to NAFLD amelioration. Hyperinsulinemia usually contributes to liver steatosis. Insulin promotes the synthesis and inhibits the degradation of lipids [36]. It stimulates key lipogenic genes in the liver, such as FAS, while reducing CPT1, which is the transporter of NEFA into mitochondria, thereby reducing fatty acid oxidation [36]. FAS inhibitors have proven useful to reduce liver TG content [37]. In the present study, HFD mice displayed an up-regulation of Fas and a down-regulation of Cpt1a in the liver. TRAIL significantly reduced HFD-induced Fas up-regulation, consistent with the reduction in insulin and liver steatosis. Third, the present study clearly shows that TRAIL also has direct actions on hepatocytes, where it significantly decreased fat content.

In the present study, TRAIL significantly reduced lipid droplet accumulation in both HepG2 cells and primary hepatocytes cultured with oleate. Our results are consistent with the observation that TRAIL treatment reduces palmitate-induced lipid uptake by 30% in hepatocytes [9]. Interestingly, in the present study, TRAIL promoted hepatocyte cell viability and did not have anti-inflammatory effects in vitro, suggesting that the in vivo reduction in liver inflammation might be secondary to the amelioration of steatosis. In addition, TRAIL showed a direct stimulatory effect on liver Pparγ and PGC-1α expression in vivo, which was confirmed in vitro, where TRAIL significantly increased PPARγ and PGC-1α expression in HepG2 cells. PPARγ is a nuclear receptor expressed in adipose tissue, muscle, and liver. Several studies have shown that PPARγ agonists significantly reduce hepatic TG content and NAFLD in patients with diabetes [38,39]. There is a functional interaction between PPARγ and PGC-1α, which might explain the parallel increase in their gene expression induced by TRAIL [40,41]. PGC-1α, which is a transcriptional co-activator of nuclear receptors, is currently considered as a key component of regulatory networks that control cellular actions to adapt to higher cellular demands, such as mitochondrial function, gluconeogenesis and glucose transport, glycogenolysis, and fatty acid oxidation [42]. Interestingly, PGC-1α polymorphisms have been associated with obesity and increased risk of diabetes [43,44]. Moreover, both PGC-1α [45] and PGC-1α-responsive genes are co-ordinately down-regulated in human diabetes [46]. When looking at the liver, it has been shown that PGC-1α-deficient mice [47] and liver-specific PGC-1α heterozygous mice [48] develop hepatic steatosis. Moreover, PGC-1α seems to have a suppressive effect on liver inflammation [49]. Therefore, our results suggest that TRAIL effects on the liver and glucose metabolism might involve PPARγ and/or PGC-1α actions.

In addition, based on our in vitro studies, where TRAIL increased ATG7, a possible unifying hypothesis explaining TRAIL effects on liver steatosis and PGC-1α is that they might be due to TRAIL-induced liver autophagy [50]. Autophagy is a mechanism by which TG content, lipid droplet number and size is regulated. In this process, lipids are sequestered in autophagosomes where they are degraded [24]. This should stimulate mitochondrial β-oxidation [24]. Moreover, autophagy seems to regulate the flux of cholesterol out of the cell to APOA-I [51], which is the major component of HDL and could explain the increase in HDL-C that we found in HFD + TRAIL mice. Further studies are needed to evaluate the mechanisms underlying TRAIL effect on liver steatosis and PGC-1α, as well as to test in detail the hypothesis that they might include TRAIL actions on liver autophagy.

In conclusion, the present study shows that TRAIL was effective in reducing HFD-induced metabolic abnormalities even when it was administered after the development of impaired glucose tolerance. Second, the present study shows that TRAIL markedly improved NAFLD in HFD-fed mice, and that this was associated with a significant increase in PGC-1α. Overall, our results shed light on the therapeutic potential of TRAIL against diabetes, NAFLD, and NASH.

Clinical perspectives

  • Experimental evidence suggests that a circulating protein called TRAIL protects against type 2 diabetes. The present study was designed to evaluate whether TRAIL had the potential not only to prevent, but also to treat diet-induced type 2 diabetes.

  • Our in vivo results show that TRAIL had the ability to attenuate the metabolic abnormalities induced by a HFD, also when TRAIL was given after disease onset (after 4 weeks of HFD). In particular, TRAIL treatment significantly reduced body weight, impaired glucose tolerance, and liver steatosis, all of which are frequently associated with type 2 diabetes.

  • Our in vitro results show that TRAIL had direct effects on hepatocytes, where it reduced lipid droplet accumulation. These data shed light on TRAIL therapeutic potential against impaired glucose tolerance and NAFLD.

We thank Dr Barbara Dapas (Department of Life Sciences, Università degli Studi di Trieste, Italy) for her technical support in in vitro experiments, as well as the Foundation ‘Dott. Carlo Fornasini’ (Poggio Renatico, Ferrara, Italy).

Competing interests

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

Funding

This work was supported by the Italian Ministry of Health [grant numbers GR-2013-02357830 (to S.B.), GR-2010-2310832 (to V.T.)]; and the Università degli Studi di Trieste [grant number FRA-2012 (to B.F.)].

Author contribution

S.Bernardi contributed to the conception and design, acquisition, analysis, and interpretation of data, and drafting of the article. B.T. contributed to the acquisition, analysis, and interpretation of data and drafting of the article. V.T., F.B., S.Biffi, and A.L. contributed to the acquisition, analysis, and interpretation of data. G.Z. and P.S. contributed to the conception, interpretation of data, and article revision for important intellectual content. B.F. contributed to the conception and design, and article revision for important intellectual content. All the authors have read and approved the submission of this manuscript.

Abbreviations

     
  • Acox

    Acyl-CoA oxidase

  •  
  • Atg7

    autophagy-related protein 7

  •  
  • Cpt1a

    carnitine palmitoyl transferase-1a

  •  
  • CRP

    C-reactive protein

  •  
  • CNT

    control

  •  
  • FAS

    fatty acid synthase

  •  
  • HFD

    high-fat diet

  •  
  • Il-6

    interleukin-6

  •  
  • IP

    intraperitoneal

  •  
  • Irs2

    insulin receptor substrate 2

  •  
  • NAFLD

    non-alcoholic fatty liver disease

  •  
  • NEFA

    non esterified fatty acids

  •  
  • NASH

    non-alcoholic steatohepatitis

  •  
  • PGC-1α

    peroxisome proliferator-activated receptor-γ co-activator-1 α

  •  
  • pgWAT

    perigonadal white adipose tissue

  •  
  • PPAR

    peroxisome proliferator-activated receptor

  •  
  • PPARγ

    peroxisome proliferator-activated receptor-γ

  •  
  • Pepck

    phosphoenolpyruvate carboxykinase

  •  
  • rhTRAIL

    recombinant human TRAIL

  •  
  • SD

    standard diet

  •  
  • Srebp

    sterol regulatory element binding protein

  •  
  • TC

    total cholesterol

  •  
  • TG

    triglyceride

  •  
  • TRAIL

    TNF-related apoptosis inducing ligand

  •  
  • TNF

    tumor necrosis factor

References

References
1
Ashkenazi
A.
and
Herbst
R.S.
(
2008
)
To kill a tumor cell: the potential of proapoptotic receptor agonists
.
J. Clin. Invest.
118
,
1979
1990
2
Harith
H.H.
,
Morris
M.J.
and
Kavurma
M.M.
(
2013
)
On the TRAIL of obesity and diabetes
.
Trends Endocrinol. Metab.
24
,
578
587
3
Bossi
F.
,
Bernardi
S.
,
Zauli
G.
,
Secchiero
P.
and
Fabris
B.
(
2015
)
TRAIL modulates the immune system and protects against the development of diabetes
.
J. Immunol. Res.
2015
,
680749
4
Mi
Q.S.
,
Ly
D.
,
Lamhamedi-Cherradi
S.E.
,
Salojin
K.V.
,
Zhou
L.
,
Grattan
M.
et al. 
(
2003
)
Blockade of tumor necrosis factor-related apoptosis-inducing ligand exacerbates type 1 diabetes in NOD mice
.
Diabetes
52
,
1967
1975
5
Lamhamedi-Cherradi
S.E.
,
Zheng
S.
,
Tisch
R.M.
and
Chen
Y.H.
(
2003
)
Critical roles of tumor necrosis factor-related apoptosis-inducing ligand in type 1 diabetes
.
Diabetes
52
,
2274
2278
6
Zauli
G.
,
Toffoli
B.
,
di Iasio
M.G.
,
Celeghini
C.
,
Fabris
B.
and
Secchiero
P.
(
2010
)
Treatment with recombinant tumor necrosis factor-related apoptosis-inducing ligand alleviates the severity of streptozotocin-induced diabetes
.
Diabetes
59
,
1261
1265
7
Bernardi
S.
,
Zauli
G.
,
Tikellis
C.
,
Candido
R.
,
Fabris
B.
,
Secchiero
P.
et al. 
(
2012
)
TNF-related apoptosis-inducing ligand significantly attenuates metabolic abnormalities in high-fat-fed mice reducing adiposity and systemic inflammation
.
Clin. Sci. (Lond.)
123
,
547
555
8
Di Bartolo
B.A.
,
Chan
J.
,
Bennett
M.R.
,
Cartland
S.
,
Bao
S.
,
Tuch
B.E.
et al. 
(
2011
)
TNF-related apoptosis-inducing ligand (TRAIL) protects against diabetes and atherosclerosis in Apoe (-)/(-) mice
.
Diabetologia
54
,
3157
3167
9
Cartland
S.P.
,
Harith
H.H.
,
Genner
S.W.
,
Dang
L.
,
Cogger
V.C.
,
Vellozzi
M.
et al. 
(
2017
)
Non-alcoholic fatty liver disease, vascular inflammation and insulin resistance are exacerbated by TRAIL deletion in mice
.
Sci. Rep.
7
,
1898
10
Winzell
M.S.
and
Ahren
B.
(
2004
)
The high-fat diet-fed mouse: a model for studying mechanisms and treatment of impaired glucose tolerance and type 2 diabetes
.
Diabetes
53
,
S215
S219
11
King
A.J.
(
2012
)
The use of animal models in diabetes research
.
Br. J. Pharmacol.
166
,
877
894
12
Ito
M.
,
Suzuki
J.
,
Tsujioka
S.
,
Sasaki
M.
,
Gomori
A.
,
Shirakura
T.
et al. 
(
2007
)
Longitudinal analysis of murine steatohepatitis model induced by chronic exposure to high-fat diet
.
Hepatol. Res.
37
,
50
57
13
MacFarlane
M.
,
Ahmad
M.
,
Srinivasula
S.M.
,
Fernandes-Alnemri
T.
,
Cohen
G.M.
and
Alnemri
E.S.
(
1997
)
Identification and molecular cloning of two novel receptors for the cytotoxic ligand TRAIL
.
J. Biol. Chem.
272
,
25417
25420
14
Tisato
V.
,
Garrovo
C.
,
Biffi
S.
,
Petrera
F.
,
Voltan
R.
,
Casciano
F.
et al. 
(
2014
)
Intranasal administration of recombinant TRAIL down-regulates CXCL-1/KC in an ovalbumin-induced airway inflammation murine model
.
PLoS ONE
9
,
e115387
15
Toffoli
B.
,
Bernardi
S.
,
Candido
R.
,
Sabato
N.
,
Carretta
R.
,
Corallini
F.
et al. 
(
2011
)
Osteoprotegerin induces morphological and functional alterations in mouse pancreatic islets
.
Mol. Cell. Endocrinol.
331
,
136
142
16
Bernardi
S.
,
Tikellis
C.
,
Candido
R.
,
Tsorotes
D.
,
Pickering
R.J.
,
Bossi
F.
et al. 
(
2015
)
ACE2 deficiency shifts energy metabolism towards glucose utilization
.
Metabolism
64
,
406
415
17
Severgnini
M.
,
Sherman
J.
,
Sehgal
A.
,
Jayaprakash
N.K.
,
Aubin
J.
,
Wang
G.
et al. 
(
2012
)
A rapid two-step method for isolation of functional primary mouse hepatocytes: cell characterization and asialoglycoprotein receptor based assay development
.
Cytotechnology
64
,
187
195
18
Secchiero
P.
,
Candido
R.
,
Corallini
F.
,
Zacchigna
S.
,
Toffoli
B.
,
Rimondi
E.
et al. 
(
2006
)
Systemic tumor necrosis factor-related apoptosis-inducing ligand delivery shows antiatherosclerotic activity in apolipoprotein E-null diabetic mice
.
Circulation
114
,
1522
1530
19
Sanchez-Gurmaches
J.
,
Hung
C.M.
and
Guertin
D.A.
(
2016
)
Emerging complexities in adipocyte origins and identity
.
Trends Cell Biol.
26
,
313
326
20
Heymsfield
S.B.
and
Wadden
T.A.
(
2017
)
Mechanisms, pathophysiology, and management of obesity
.
N. Engl. J. Med.
376
,
254
266
21
Visser
M.
,
Bouter
L.M.
,
McQuillan
G.M.
,
Wener
M.H.
and
Harris
T.B.
(
1999
)
Elevated C-reactive protein levels in overweight and obese adults
.
JAMA
282
,
2131
2135
22
Defronzo
R.A.
(
2009
)
Banting Lecture. From the triumvirate to the ominous octet: a new paradigm for the treatment of type 2 diabetes mellitus
.
Diabetes
58
,
773
795
23
Haga
Y.
,
Kanda
T.
,
Sasaki
R.
,
Nakamura
M.
,
Nakamoto
S.
and
Yokosuka
O.
(
2015
)
Nonalcoholic fatty liver disease and hepatic cirrhosis: comparison with viral hepatitis-associated steatosis
.
World J. Gastroenterol.
21
,
12989
12995
24
Liu
K.
and
Czaja
M.J.
(
2013
)
Regulation of lipid stores and metabolism by lipophagy
.
Cell Death Differ.
20
,
3
11
25
Sun
J.
,
Yu
X.
,
Wang
C.
,
Yu
C.
,
Li
Z.
,
Nie
W.
et al. 
(
2017
)
RIP-1/c-FLIPL induce hepatic cancer cell apoptosis through regulating tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)
.
Med. Sci. Monit.
23
,
1190
1199
26
Toffoli
B.
,
Bernardi
S.
,
Candido
R.
,
Zacchigna
S.
,
Fabris
B.
and
Secchiero
P.
(
2012
)
TRAIL shows potential cardioprotective activity
.
Invest. New Drugs
30
,
1257
1260
27
Zoller
V.
,
Funcke
J.B.
,
Keuper
M.
,
Abd El Hay
M.
,
Debatin
K.M.
,
Wabitsch
M.
et al. 
(
2016
)
TRAIL (TNF-related apoptosis-inducing ligand) inhibits human adipocyte differentiation via caspase-mediated downregulation of adipogenic transcription factors
.
Cell Death Dis.
7
,
e2412
28
Pitombo
C.
,
Araujo
E.P.
,
De Souza
C.T.
,
Pareja
J.C.
,
Geloneze
B.
and
Velloso
L.A.
(
2006
)
Amelioration of diet-induced diabetes mellitus by removal of visceral fat
.
J. Endocrinol.
191
,
699
706
29
Tilg
H.
,
Moschen
A.R.
and
Roden
M.
(
2017
)
NAFLD and diabetes mellitus
.
Nat. Rev. Gastroenterol. Hepatol.
14
,
32
42
30
Cusi
K.
(
2016
)
Treatment of patients with type 2 diabetes and non-alcoholic fatty liver disease: current approaches and future directions
.
Diabetologia
59
,
1112
1120
31
Ge
P.S.
and
Runyon
B.A.
(
2016
)
Treatment of patients with cirrhosis
.
N. Engl. J. Med.
375
,
767
777
32
Idrissova
L.
,
Malhi
H.
,
Werneburg
N.W.
,
LeBrasseur
N.K.
,
Bronk
S.F.
,
Fingas
C.
et al. 
(
2015
)
TRAIL receptor deletion in mice suppresses the inflammation of nutrient excess
.
J. Hepatol.
62
,
1156
1163
33
Donnelly
K.L.
,
Smith
C.I.
,
Schwarzenberg
S.J.
,
Jessurun
J.
,
Boldt
M.D.
and
Parks
E.J.
(
2005
)
Sources of fatty acids stored in liver and secreted via lipoproteins in patients with nonalcoholic fatty liver disease
.
J. Clin. Invest.
115
,
1343
1351
34
Tchkonia
T.
,
Thomou
T.
,
Zhu
Y.
,
Karagiannides
I.
,
Pothoulakis
C.
,
Jensen
M.D.
et al. 
(
2013
)
Mechanisms and metabolic implications of regional differences among fat depots
.
Cell Metab.
17
,
644
656
35
Perry
R.J.
,
Camporez
J.P.
,
Kursawe
R.
,
Titchenell
P.M.
,
Zhang
D.
,
Perry
C.J.
et al. 
(
2015
)
Hepatic acetyl CoA links adipose tissue inflammation to hepatic insulin resistance and type 2 diabetes
.
Cell
160
,
745
758
36
Weickert
M.O.
and
Pfeiffer
A.F.
(
2006
)
Signalling mechanisms linking hepatic glucose and lipid metabolism
.
Diabetologia
49
,
1732
1741
37
Wu
M.
,
Singh
S.B.
,
Wang
J.
,
Chung
C.C.
,
Salituro
G.
,
Karanam
B.V.
et al. 
(
2011
)
Antidiabetic and antisteatotic effects of the selective fatty acid synthase (FAS) inhibitor platensimycin in mouse models of diabetes
.
Proc. Natl. Acad. Sci. U.S.A.
108
,
5378
5383
38
Mayerson
A.B.
,
Hundal
R.S.
,
Dufour
S.
,
Lebon
V.
,
Befroy
D.
,
Cline
G.W.
et al. 
(
2002
)
The effects of rosiglitazone on insulin sensitivity, lipolysis, and hepatic and skeletal muscle triglyceride content in patients with type 2 diabetes
.
Diabetes
51
,
797
802
39
Cusi
K.
,
Orsak
B.
,
Bril
F.
,
Lomonaco
R.
,
Hecht
J.
,
Ortiz-Lopez
C.
et al. 
(
2016
)
Long-term pioglitazone treatment for patients with nonalcoholic steatohepatitis and prediabetes or type 2 diabetes mellitus: a randomized trial
.
Ann. Intern. Med.
165
,
305
315
40
Fujisawa
K.
,
Nishikawa
T.
,
Kukidome
D.
,
Imoto
K.
,
Yamashiro
T.
,
Motoshima
H.
et al. 
(
2009
)
TZDs reduce mitochondrial ROS production and enhance mitochondrial biogenesis
.
Biochem. Biophys. Res. Commun.
379
,
43
48
41
Puigserver
P.
,
Wu
Z.
,
Park
C.W.
,
Graves
R.
,
Wright
M.
and
Spiegelman
B.M.
(
1998
)
A cold-inducible coactivator of nuclear receptors linked to adaptive thermogenesis
.
Cell
92
,
829
839
42
Wu
H.
,
Deng
X.
,
Shi
Y.
,
Su
Y.
,
Wei
J.
and
Duan
H.
(
2016
)
PGC-1alpha, glucose metabolism and type 2 diabetes mellitus
.
J. Endocrinol.
229
,
R99
R115
43
Ek
J.
,
Andersen
G.
,
Urhammer
S.A.
,
Gaede
P.H.
,
Drivsholm
T.
,
Borch-Johnsen
K.
et al. 
(
2001
)
Mutation analysis of peroxisome proliferator-activated receptor-gamma coactivator-1 (PGC-1) and relationships of identified amino acid polymorphisms to type II diabetes mellitus
.
Diabetologia
44
,
2220
2226
44
Hara
K.
,
Tobe
K.
,
Okada
T.
,
Kadowaki
H.
,
Akanuma
Y.
,
Ito
C.
et al. 
(
2002
)
A genetic variation in the PGC-1 gene could confer insulin resistance and susceptibility to type II diabetes
.
Diabetologia
45
,
740
743
45
Patti
M.E.
,
Butte
A.J.
,
Crunkhorn
S.
,
Cusi
K.
,
Berria
R.
,
Kashyap
S.
et al. 
(
2003
)
Coordinated reduction of genes of oxidative metabolism in humans with insulin resistance and diabetes: potential role of PGC1 and NRF1
.
Proc. Natl. Acad. Sci. U.S.A.
100
,
8466
8471
46
Mootha
V.K.
,
Lindgren
C.M.
,
Eriksson
K.F.
,
Subramanian
A.
,
Sihag
S.
,
Lehar
J.
et al. 
(
2003
)
PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes
.
Nat. Genet.
34
,
267
273
47
Leone
T.C.
,
Lehman
J.J.
,
Finck
B.N.
,
Schaeffer
P.J.
,
Wende
A.R.
,
Boudina
S.
et al. 
(
2005
)
PGC-1alpha deficiency causes multi-system energy metabolic derangements: muscle dysfunction, abnormal weight control and hepatic steatosis
.
PLoS Biol.
3
,
e101
48
Estall
J.L.
,
Kahn
M.
,
Cooper
M.P.
,
Fisher
F.M.
,
Wu
M.K.
,
Laznik
D.
et al. 
(
2009
)
Sensitivity of lipid metabolism and insulin signaling to genetic alterations in hepatic peroxisome proliferator-activated receptor-gamma coactivator-1alpha expression
.
Diabetes
58
,
1499
1508
49
Barroso
W.A.
,
Victorino
V.J.
,
Jeremias
I.C.
,
Petroni
R.C.
,
Ariga
S.K.K.
,
Salles
T.A.
et al. 
(
2017
)
High-fat diet inhibits PGC-1alpha suppressive effect on NFkappaB signaling in hepatocytes
.
Eur. J. Nutr.
,
50
Nikoletopoulou
V.
,
Markaki
M.
,
Palikaras
K.
and
Tavernarakis
N.
(
2013
)
Crosstalk between apoptosis, necrosis and autophagy
.
Biochim. Biophys. Acta
1833
,
3448
3459
51
Ouimet
M.
,
Franklin
V.
,
Mak
E.
,
Liao
X.
,
Tabas
I.
and
Marcel
Y.L.
(
2011
)
Autophagy regulates cholesterol efflux from macrophage foam cells via lysosomal acid lipase
.
Cell Metab.
13
,
655
667