Galectin-1 accelerates high-fat diet-induced obesity by activation of peroxisome proliferator-activated receptor gamma (PPARγ) in mice

Galectin-1 contains a carbohydrate-recognition domain (CRD) as a member of the lectin family. Here, we investigated whether galectin-1 regulates adipogenesis and lipid accumulation. Galectin-1 mRNA is highly expressed in metabolic tissues such as the muscle and adipose tissues. Higher mRNA expression of galectin-1 was detected in white adipose tissues (WATs) of mice that were fed a high-fat diet (HFD) than in those of mice fed a normal-fat diet (NFD). Protein expression of galectin-1 also increased during adipocyte differentiation. Galectin-1 silencing inhibited the differentiation of 3T3-L1 cells and the expression of lipogenic factors, such as PPARγ, C/EBPα, FABP4, and FASN at both mRNA and protein levels. Lactose, an inhibitor by the binding with CRD of galectin-1 in extracellular matrix, did not affect adipocyte differentiation. Galectin-1 is localized in multiple cellular compartments in 3T3-L1 cells. However, we found that DMI (dexamethasone, methylisobutylxanthine, insulin) treatment increased its nuclear localization. Interestingly, galectin-1 interacted with PPARγ. Galectin-1 overexpression resulted in increased PPARγ expression and transcriptional activity. Furthermore, we prepared galectin-1-knockout (Lgals1−/−) mice and fed a 60% HFD. After 10 weeks, Lgals1−/− mice exhibited lower body weight and gonadal WAT (gWAT) mass than wild-type mice. Fasting glucose level was also lower in Lgals1−/−mice than that in wild-type mice. Moreover, lipogenic genes were significantly downregulated in the gWATs and liver tissues from Lgals1−/− mice. Pro-inflammatory cytokines, such as CCL2, CCL3, TNFα, and F4/80, as well as macrophage markers, were also drastically downregulated in the gWATs and liver tissues of Lgals1−/− mice. In addition, Lgals1−/−mice showed elevated expression of genes involved in thermogenesis in the brown adipose tissue. Collectively, galectin-1 exacerbates obesity of mice fed HFD by increment of PPARγ expression and activation. Our findings suggest that galectin-1 could be a potential therapeutic target for obesity and needed further study for clinical application.


Introduction
Galectins are a family of proteins that bind to β-galactoside. All galectins have a conserved carbohydrate-recognition domain (CRD), consisting of approximately 130 amino acids 1 . Galectins are localized in extra-and intra-cellular compartments, and are involved in cell adhesion, cell cycle progression, apoptosis, inflammation, and cell growth 2 . Most studies on galectins have focused on cancer and inflammatory disorders 3 . Galectins are also involved in metabolic diseases such as obesity and diabetes 4 . For example, the role of galectin-12 4,5 and galectin-3 6 in adipocyte differentiation and obesity has been already investigated.
Galectin-1 has been studied for its involvement in cancer progression and inflammation 7,8 . Galectin-1 overexpression promotes cell transformation by increasing membrane anchorage and oncogenic H-RAS signal transduction 9 . Galectin-1 induces apoptosis in activated CD4 + and CD8 + T cells 10 , and promotes angiogenesis by stimulating vascular endothelial cell proliferation and migration 11 . Interestingly, a previous study was investigated galectin-1 as a novel adipokine during adipocyte differentiation 12 . Galectin-1 is also known as a lipid droplet-associated protein in primary mouse adipocytes 13 . A recent report demonstrated that treating 3T3-L1 cells with thiodigalactoside (TDG), an inhibitor of galectins, retarded adipocyte differentiation. In addition, TDGtreated rats exhibited reduced body weight and white adipose tissue (WAT) mass, suggesting that TDG treatment increases the resistance to high-fat diet (HFD)induced obesity 14 . These findings suggested that galectin-1 might be positively associated with adipocyte differentiation and obesity. However, the underlying molecular mechanisms are not yet fully understood.
In this study, we demonstrated that galectin-1 was highly expressed in the mouse adipose tissue and further increased during adipocyte differentiation and lipid accumulation. Therefore, we explored the mechanism by which galectin-1 regulates adipogenesis and obesity. To this end, preadipocyte 3T3-L1 cells were used, and the metabolic phenotype of galectin-1-deficient (Lgals1 −/− ) mice exposed to an obesogenic HFD was characterized.

Results
Galectin-1 is upregulated during adipocyte differentiation and predominantly expressed in mouse adipose tissue To determine the in vivo distribution of galectin-1, various mouse organs were examined by RT-PCR analysis (Fig.  1a). Galectin-1 was more highly expressed in metabolic tissues, such as the muscle and adipose tissue, than in other organs. Moreover, the level of galectin-1 mRNA and protein was higher in white adipose tissues from obese mice, which were fed a 60% high-fed diet (HFD) for 12 weeks, compared to those of normal-fat diet (NFD) fed mice (Fig. 1b).
We measured the changes in galectin-1 expression levels during the differentiation of pre-adipocyte 3T3-L1 cells to adipocytes (Fig. 1c). The cells were treated with DMI (dexamethasone, methylisobutylxanthine, insulin) and differentiated for 10 days. Galectin-1 expression, as well as the expression of peroxisome proliferatoractivated receptor gamma (PPARγ) and fatty acid synthase (FASN), two markers of adipocyte differentiation, increased during adipocyte differentiation.
To study the role of galectin-1 in adipocyte differentiation, the corresponding gene was silenced by transfecting 3T3-L1 cells with a galectin-1-specific small interfering RNA (siRNA), followed by cell differentiation for 6 days (Fig. 1d). Galectin-1-depleted 3T3-L1 cells showed significantly delayed adipocyte differentiation and lipid accumulation, compared to control cells (Fig. 1e). Lipid accumulation was also significantly dropped by galecin-1 depletion at day 6 ( Fig. 1f). These results suggested that galectin-1 regulated adipocyte differentiation and lipid accumulation.
Next, the effect of extracellular-localized galectin-1 on adipocyte differentiation was investigated ( Fig. 2d and e). Treatment of 3T3-L1 cells with lactose, an inhibitor of galectin-1 by binding CRD, at concentrations up to 100 mM did not affect cell differentiation. It suggested that galectin-1-mediated regulation of adipocyte differentiation did not occur through an extracellular CRD binding mechanism.

Galectin-1 interacts with PPARγ and increases its expression and transcriptional activation
We examined the subcellular localization of galectin-1 during adipocyte differentiation ( Fig. 3a and b). Total galectin-1 expression was found to increase after DMI treatment. Interestingly, the nuclear localization of galectin-1 significantly increased in a time-dependent manner, as did that of PPARγ (Fig. 3b). Co-localization of galectin-1 and PPARγ was detected in nucleus after DMI treatment (Fig. 3a).
Therefore, we performed immunoprecipitation assays and determined that galectin-1 and PPARγ interacted with each other (Fig. 3c). Overexpression of galectin-1 also increased the protein expression level of endogenous PPARγ (Fig. 3d). Moreover, overexpression of galectin-1 significantly increased the PPARγ transcriptional activation, indicated by the increment of PPAR-response element (PPRE) activation (Fig. 3e). These data suggested that DMI treatment increased galectin-1 expression in nucleus, and that galectin-1 interacted with PPARγ and promoted its expression and transcriptional activation. To confirm the role of galectin-1 in an in vivo mouse model of obesity, male wild-type (Lgals1 +/+ ) and galectin-1 knockout (Lgals1 −/− ) mice (n = 5 per group) were fed a high-fat diet (HFD) containing 60% fat for 10 weeks (Fig. 4a). The body weights were not significantly different between those from Lgals1 −/− mice and wild-type mice fed normal-fat diet (NFD) (Fig. 4b and c). However, Lgals1 −/− mice exhibited smaller body size (Fig. 4b) and lower body weights than those of wild-type mice after exposure to the HFD (Fig. 4c), even though the food intake did not significantly differ between the two groups (Fig. 4d). The weight of mouse tissues involved in fat metabolism, i.e., WAT, brown adipose tissue (BAT), and liver tissues, was also measured (Fig. 4e). Both gWAT and iWAT as per body weight ratio were decreased in HFD-fed Lgals1 −/− mice, compared to those from HFD-fed wild-type mice. Interestingly, the relative BAT weight increased in Lgals1 −/− mice fed a HFD. No differences were observed in liver weight between the groups. Moreover, because obesity is known to increase the risk of type-2 diabetes mellitus (T2D) and hyperlipidemia, the serum levels of glucose, triglycerides, and free fatty acids were determined (Fig. 4f). The fasting glucose level was lower in HFD-fed Lgals1 −/− mice than HFD-fed wild-type mice, although the levels of triglycerides and free fatty acids did not differ.
Lgals1 −/− mice exhibit decreased adiposity and altered expression of genes involved in lipid metabolism and thermogenesis The gWATs of HFD-fed Lgals1 −/− mice exhibited smaller size and weight than those of wild-type mice (Fig. 5a). We observed that adipocytes in gWAT were smaller in Lgals1 −/− mice than those in wild-type mice (Fig. 5b). HFD-fed Lgals1 −/− mice also showed lower iWAT weight than that of their wild-type counterparts (Fig. 5e). The expression levels of genes, which were involved in fatty acid uptake, and lipogenesis were found to be significantly decreased in both gWATs and iWATs from HFD-fed Lgals1 −/− mice ( Fig. 5c and f). Obesity is known to cause macrophage infiltration and elevated levels of pro-inflammatory cytokines in WATs 15 . Therefore, the expression of F4/80, a macrophage marker, and pro-inflammatory cytokines, such as TNFα, CCL2 and CCL3 was measured ( Fig. 5d and g). It revealed that they were significantly downregulated in both gWATs and iWATs from HFD-fed Lgals1 −/− mice. The level of galectin-1 mRNA and protein was analyzed by quantitative RT-PCR and western blotting in gonadal white adipose tissues (gWATs) and inguinal (iWATs) from mice fed a normal-fat diet (NFD) or a high-fat diet (HFD) for 12 weeks. Data are presented as the mean ± standard error of the mean (SEM). *p < 0.05 for the NFD group (n = 5) vs. the HFD group (n = 6). c The protein level of galectin-1 and lipogenic proteins, such as PPARγ and FASN was analyzed by western blotting during DMI (dexamethasone, methylisobutylxanthine, insulin) induced differentiation of preadipocyte 3T3-L1 cells. β-actin expression was used as a loading control. d The expression level of galectin-1 mRNA and protein was analyzed by RT-PCR (two upper panels) and western blotting (two lower panels) after galectin-1 knockdown by a specific siRNA in 3T3-L1 cells. The mRNA and protein expression levels were normalized to those of β-actin. e Detection of adipocyte differentiation of 3T3-L1 cells following galectin-1 silencing. The cells were treated with galectin-1 siRNA and differentiation was induced by DMI treatment for 6 days. Oil red o (ORO) staining was performed to detect lipid accumulation (upper figures) and the staining level was taken by picture. Cell morphology was detected (lower figures) by microscopy. Bars indicated 50 μm. f Measurement of lipid accumulation. The ORO dye (used for staining) was eluted with 100% isopropanol and the optical density at 500 nm (OD 500 ) was detected. ***p < 0.001 for the NC group vs. the galectin-1 siRNA group.
Interestingly, no difference was observed in BAT weight between Lgals1 −/− mice and wild-type mice when both the groups were fed NFD and HFD (Fig. 5h). However, the BAT weight ratio per body weight was higher in HFD-fed Lgals1 −/− mice than that of HFD-fed wild-type mice (Fig. 4e). Thermogenesis, fatty acid oxidation, and lipid accumulation and synthesis were previously reported to regulate the fat mass 16 . Therefore, the expression levels of genes involved in thermogenesis and fatty acid oxidation were measured in gWAT, iWAT and BAT. Not only in BAT of HFD-fed Lgals1 −/− mice, these genes were significantly upregulated in both gWAT and iWAT (Fig. 5c, f, and i). It suggested that depletion of galectin-1 increased thermogenic gene expression to lead resistance to obesity.
The expression of genes promoting hepatic gluconeogenesis and lipogenesis is decreased in Lgals1 −/− mice Because fatty liver diseases are often exhibited in obesity, we conducted histological analysis of the liver. No phenotypic differences were found in livers from both NFD-fed Lgals1 −/− mice and wild-type mice ( Fig. 6a and  b). However, while a remarkable fat accumulation was observed in the liver of HFD-fed wild-type mice, this phenomenon was substantially attenuated in HFD-fed Lgals1 −/− mice. However, differences of liver weights were detected slightly in the between Lgals1 − / − mice and wildtype mice, after HFD exposure (Fig. 6a). The expression of genes involved in gluconeogenesis, such as glucose-6phosphatase (G6PC) and phosphoenolpyruvate carboxykinase (PCK), were significantly reduced in the livers of HFD-fed Lgals1 −/− mice (Fig. 6c). The expression of FASN, ACC1, and SCD1 (genes involved in lipogenesis) were also downregulated in HFD-fed Lgals1 −/− mice (Fig. 6c). Moreover, macrophage markers (F4/80) and proinflammatory cytokines (TGFβ, TNFα, CCL2 and CCL3) were found to be significantly downregulated in the livers of HFD-fed Lgals1 −/− mice compared to their Lgals1 +/+ counterparts (Fig. 6d). It suggested that depletion of galectin-1 significantly reduced fat accumulation and inflammatory cytokine secretion in liver. Data are presented as the mean ± SEM (n = 3 for each lane). ***p < 0.001 for the NC group vs. the Lgals1 siRNA treated group. c The protein expression level of adipogenic and lipogenic factors was analyzed by western blotting in 3T3-L1 cells, with or without galectin-1 silencing. β-actin expression was used as a loading control. d Detection of adipocyte differentiation of 3T3-L1 cells in the absence or presence of lactose (25, 50, or 100 mM). Adipocyte differentiation was induced by treating the cells with DMI and lactose for 6 days. Cell morphology was detected by microscopy. Bars indicated 50 μm. e Measurement of lipid accumulation. The ORO dye (used for staining) was eluted with 100% isopropanol and the OD 500 was measured.

Discussion
The prevalence of obesity is rapidly growing worldwide. Obesity is known to increase the rate of mortality through an elevated risk of developing several chronic conditions, such as T2D, cardiovascular disease, osteoarthritis, dementia, and cancer, as well as to exacerbate the severity of various acute diseases 17 . Here, we addressed the role of galectin-1 in obesity. Galectins are involved in various aspects of biological homeostasis and their role in cancer and inflammatory diseases has been extensively investigated [18][19][20][21] . Galectins also represent a potential therapeutic target for obesity 17 . The role of galectin-1 in obesity has also been also explored 14,22 . However, the exact function of galectin-1 in the context of obesity is still unknown.
In this study, we observed that higher expression of galectin-1 was detected in adipose tissues of HFD-fed mice than in those of NFD-fed mice. Galectin-1 also increased during adipocyte differentiation and lipid accumulation 12 . It was reported that treatment with thiogalactoside (TDG) 14 or lactulose 22 reduced adipocyte differentiation and obesity in murine fed a HFD. TDG and lactulose is a non-metabolized disaccharide that can bind to galectins, and inhibit the function of intracellular and extracellular galectins 23 . In this study, the inhibition of extracellular galectins by lactose did not affect adipocyte differentiation. It was reported that extracellular galectin-1 stimulates angiogenesis in vitro and in vivo, and these effects were inhibited by lactose treatment 11 . Our finding suggested that the regulation of adipocyte differentiation by galectin-1 was based on an intracellular mechanism. Next, we determined that the expression of galectin-1 protein increased in both the nucleus and cytoplasm after induction of adipocyte differentiation. The increased nuclear localization of galectin-1 during adipocyte differentiation further supports this conclusion. Moreover, we provided the first evidence of an interaction between galectin-1 and PPARγ. Galectin-1 overexpression elevated PPARγ expression and PPRE activity. Based on these results, we suggested that nuclear galectin-1 may promote adipocyte differentiation by interacting with PPARγ, which is a major transcription factor modulating several biological processes that are perturbed in obesity, including inflammation, lipid and glucose metabolism, and overall energy homeostasis 24 . The high relevance of these processes explains the strong interest in PPAR agonists and antagonists in medical research and drug discovery 25 . Thus, our data identify galectin-1 as a potential regulator of PPARγ. It is also suggested that targeting galectin-1 may represent a new way to treat a wide spectrum of metabolic diseases by modulating PPARγ activity.
We further determined the metabolic phenotype in galectin-1-deficient (Lgals1 −/− ) mice that were fed a 60% high-fat diet (HFD). Lgals1 −/− mice showed decreased body weights and adipose tissues mass compared to those of wild-type (Lgals1 +/+ ) mice, following HFD administration. Downregulation of genes involved in adipogenesis, fatty acid uptake, and lipogenesis were detected in both gWATs and iWATs of Lgals1 −/− mice. Interestingly, the expression of genes involved in thermogenesis and fatty acid oxidation were increased in WATs and BATs from Lgals1 −/− mice. Obesity causes decreased thermogenesis and fatty acid oxidation 26,27 . Therefore, these two processes are potential therapeutic targets for obesity [28][29][30] . In Lgals1 −/− mice, decreased lipogenesis, as well as increased thermogenesis and fatty acid oxidation, resulted in a synergistic inhibition of body weight gain. Further studies should be conducted to establish whether the increase in thermogenesis induced by galectin-1 knockdown involves the β-adrenergic signaling, which typically regulates the expression of thermogenic genes 28,29 .
Lgals1 −/− mice exhibited lower fasting glucose levels than those in the wild-type counterparts, supposed to be resulted in an improvement of diabetes. The expression of macrophage markers and pro-inflammatory cytokines were also decreased in the adipose tissues of Lgals1 −/− mice. Insulin resistance in obesity is known to be associated with an increased inflammation in adipose tissue 31,32 . The reduced inflammation observed in the adipose tissues of Lgals1 −/− mice suggested that galectin-1 targeting improved insulin resistance. A previous study conducted in non-obese diabetic mice showed that treatment with soluble galectin-1 promotes the apoptosis of pathological Th1 cells, causing pancreatic β-cell destruction 33 . Galectin-1 deficiency results in pro-diabetic effects, suppressing the function of pancreatic B-cells and increasing inflammation of the adipose tissues. In this study, we focused on lipogenesis and lipid accumulation in the adipose tissue, but did not explore the Fig. 4 Lgals1 −/− mice were lean and showed reduced WAT masses. Both Lgals1 +/+ and Lgals1 −/− mice were fed a normal-fat diet (NFD) or a high-fat diet (HFD) (60% fat) for 10 weeks (n = 5 per group). a Expression of galectin-1 mRNA was analyzed by RT-PCR in gonadal white adipose tissue (gWAT), inguinal white adipose tissue (iWAT), brown adipose tissue (BAT), and liver samples from a NFD fed Lgals1 +/+ and Lgals1 −/− mice. b Body sizes and weights were measured and presented as a picture and a body weight graph. c The increment of body weight for 10 weeks was presented as a graph. d Daily food intake of Lgals1 +/+ and Lgals1 −/− mice fed a NFD or a HFD. e Ratio of gWAT, iWAT, and BAT weights to body weights of Lgals1 +/+ and Lgals1 −/− mice fed a NFD or a HFD. f Levels of serum glucose, triglycerides, and free fatty acids in Lgals1 +/+ and Lgals1 −/− mice fed a NFD or a HFD. Data are presented as the mean ± SEM *p < 0.05, **p < 0.01, and ***p < 0.001 for the Lgals1 +/+ mice vs. the Lgals1 −/− mice.
function of pancreatic β-cells or insulin resistance. Further studies focusing on the role of galectin-1 in glucose homeostasis will be necessary to clarify the impact of galectin-1 on diabetes.
We also found that Lgals1 −/− mice liver have lower expression of genes involved in gluconeogenesis, lipogenesis, and inflammation. Furthermore, we found that hepatic triglyceride accumulation was also reduced in Lgals1 −/− mouse livers. It is unclear whether this result reflects direct regulation by galectin-1 or an additional effect of obesity resistance. A recent report showed that the interaction of galectin-1 with neuropilin-1 promotes liver fibrosis by activating the hepatic stellate cells 34 . In addition, galectin-1 silencing was shown to inhibit the activation and proliferation of mouse hepatic stellate cells in a mouse model of liver fibrosis 35 . These results suggest that galectin-1 may have a direct impact on both obesity and liver disease.
Taken together, our findings suggest that galectin-1 expression increases during adipogenesis and accelerates adipocyte differentiation and lipid accumulation. We demonstrated that galectin-1 interacts with PPARγ and promoted its expression and transcriptional activity. Deletion of galectin-1 attenuated the effects of HFD on blood glucose levels and inflammation and reduced the body mass. In light of our results, the development of Both Lgals1 +/+ and Lgals1 −/− mice were fed a NFD or a HFD for 10 weeks (n = 5 per group) and their gWATs (a-d), iWATs (e-g) and BATs (h and i) were prepared for experiments. a Size and weight of gWATs in Lgals1 +/+ and Lgals1 −/− mice fed a NFD or a HFD were measured and presented as a picture and a graph. b Slide sections of gWAT from Lgals1 +/+ and Lgals1 −/− mice fed a NFD or a HFD were prepared and stained with H&E. Morphology of adipocytes was presented as a picture. Bars indicated 100 μm. Size of adipocytes was measured using Image J software and presented as a graph. c The expression of genes involved in fatty acid uptake, such as CD36 and FABP4, in lipogenesis, such as PPARγ, FASN, SREBF1, in thermogenesis, such as CIDEA and PGC1α, and in fatty acid oxidation, such as ACOX1, CPT1 and PPARα in gWATs from Lgals1 +/+ and Lgals1 −/− mice fed a HFD was analyzed by quantitative RT-PCR. d The expression of genes, included F4/80, a macrophage marker and pro-inflammatory cytokines, such as TNFα, CCL2 and CCL3 in gWATs from Lgals1 +/+ and Lgals1 −/− mice fed a HFD was analyzed by quantitative RT-PCR. Data are presented as the mean ± SEM ***p < 0.001 for the Lgals1 +/+ mice vs. the Lgals1 −/− mice. e Weight of iWATs in Lgals1 +/+ and Lgals1 −/− mice fed a NFD or a HFD was measured and presented as a graph. f The expression of genes involved in adipogenesis, such as PPARγ, CEBPα and CEBPβ, in fatty acid uptake, such as CD36 and FABP4, in lipogenesis, such as SREBP, ACC1, ACLY and SCD1, in thermogenesis, such as UCP1, PGC1α, PRDM16, CIDEA, ADIPOQ, Elovl3 and COX2, and in fatty acid oxidation, such as PPARα and Mcad in iWATs from Lgals1 +/+ and Lgals1 −/− mice fed a HFD was analyzed by quantitative RT-PCR. g The expression of genes, included F4/80, TNFα, CCL2 and CCL3 in iWATs from Lgals1 +/+ and Lgals1 −/− mice fed a HFD was analyzed by quantitative RT-PCR. Data are presented as the mean ± SEM *p < 0.05 and **p < 0.01 for the Lgals1 +/+ mice vs. the Lgals1 −/− mice. h Weight of BATs in Lgals1 +/+ and Lgals1 −/− mice fed a NFD or a HFD were measured and presented as a graph. i The expression of genes involved in thermogenesis, such as UCP1, PGC1α, PRDM16, CIDEA, Elovl3 and COX2, and in fatty acid oxidation, such as PPARα and Mcad in iBATs from Lgals1 +/+ and Lgals1 −/− mice fed a HFD was analyzed by quantitative RT-PCR. Data are presented as the mean ± SEM *p < 0.05 and **p < 0.01 for the Lgals1 +/+ mice vs. the Lgals1 −/− mice.
galectin-1 inhibitors should be considered as a therapeutic strategy for obesity and other metabolic diseases. Further studies are needed on this subject for their development.
with protein A/G agarose beads (Santa Cruz Biotechnology), and precleared by incubation at 4°C for 30 min in a rotor. After centrifugation (4°C for 25 min at 13 200 rpm), anti-Flag beads (Sigma-Aldrich) were added to the supernatants, followed by an overnight incubation at 4°C in a rotor. The immunoprecipitates were washed twice in immunoprecipitation buffer, mixed with 2× SDS sample buffer, and boiled at 100°C for 5 min. After centrifugation (4°C for 2 min at 13 200 rpm), the supernatants were analyzed by western blotting.

Mouse studies
The animal studies were approved by the Yonsei University Health System Institutional Animal Care and Use Committee (Permission number for animal experiments: 2015-0104). Galectin-1 knockout (Lgals1 −/− C57BL/6 mice were purchased from the Knockout Mouse Project Repository (Oakland, CA, USA). Wild-type (Lgals1 +/+ ) and galectin-1 knockout (Lgals1 −/− ) mice were bred as heterozygotes in house and maintained on a C57BL/6 background. Heterozygous mice were bred with each other to generate Lgals1 +/+ and Lgals1 −/− mice. Sevenweek-old male Lgals1 +/+ and Lgals1 −/− mice (5 per cage) were fed a NFD or a HFD (Research Diets, Inc., New Brunswick, NJ, USA) for 10 weeks. Littermate-matched mice were randomly divided into 4 groups. All mice were provided free access to food and water and kept on a 12 h light/12 h dark cycle. Body weight and food intake were measured once a week. After 10 weeks, mice were euthanized with CO 2 gas. Mouse tissues and serum were stored at -80°C before analysis. Portions of liver and WAT were fixed in 10% formalin for histological analysis.

Immunohistochemical analysis of WATs and liver tissues
Mouse WATs and livers were fixed in 10% formalin and embedded in paraffin. Paraffin-embedded WAT or liver sections were subjected to hematoxylin and eosin (H&E) staining. H&E-stained sections were analyzed using Image J software 41 (NIH, Bethesda, Maryland, USA).

Statistical analysis
The unpaired (two-sample) t-test was used to determine the p-values. P-values < 0.05 were considered to reflect statistically significant differences. Statistical analysis was performed using Prism 5 (GraphPad software, La Jolla, CA, USA).