Eicosanoids, such as leukotriene B4 (LTB4) and lipoxin A4 (LXA4), may play a key role during obesity. While LTB4 is involved in adipose tissue inflammation and insulin resistance, LXA4 may exert anti-inflammatory effects and alleviate hepatic steatosis. Both lipid mediators derive from the same pathway, in which arachidonate 5-lipoxygenase (ALOX5) and its partner, arachidonate 5-lipoxygenase–activating protein (ALOX5AP), are involved. ALOX5 and ALOX5AP expression is increased in humans and rodents with obesity and insulin resistance. We found that transgenic mice overexpressing ALOX5AP in adipose tissue had higher LXA4 rather than higher LTB4 levels, were leaner, and showed increased energy expenditure, partly due to browning of white adipose tissue (WAT). Upregulation of hepatic LXR and Cyp7a1 led to higher bile acid synthesis, which may have contributed to increased thermogenesis. In addition, transgenic mice were protected against diet-induced obesity, insulin resistance, and inflammation. Finally, treatment of C57BL/6J mice with LXA4, which showed browning of WAT, strongly suggests that LXA4 is responsible for the transgenic mice phenotype. Thus, our data support that LXA4 may hold great potential for the future development of therapeutic strategies for obesity and related diseases.
Obesity has emerged as one of the most important public health concerns of our society, with increasing incidence and morbidity. Obesity is associated with low-grade chronic inflammation of adipose tissue, which has been causally linked to the development of insulin resistance, type 2 diabetes, arthritis, cancer, cardiovascular diseases, asthma, and Alzheimer disease (1,2). In contrast to acute inflammation, in chronic inflammation the initiation phase is not followed by a resolution, and failure of resolution may be responsible for low-grade inflammation in obesity (3–5). During the unresolved inflammation, proinflammatory cytokines released by the adipose tissue act in a paracrine and systemic manner, promoting the development of insulin resistance in metabolic tissues (6,7). However, the mechanisms by which inflammation causes metabolic alterations have not been fully elucidated. The understanding of these processes is essential to the development of future treatments for obesity and type 2 diabetes, and both the prevention of adipose tissue inflammation and the promotion of its resolution may offer new therapeutic opportunities.
Lipid mediators, such as eicosanoids produced from arachidonic acid, play critical roles in inflammation and its resolution (5,6,8). Among them, leukotrienes are well-known proinflammatory molecules (9). These eicosanoids are derived from the enzymatic activity of arachidonate 5-lipoxygenase (ALOX5) and its partner, arachidonate 5-lipoxygenase–activating protein (ALOX5AP or FLAP). The activation of ALOX5/ALOX5AP is a crucial enzymatic step in the transformation of arachidonic acid into the unstable intermediate leukotriene A4 (LTA4), which is subsequently transformed into leukotriene B4 (LTB4) (5S, 12R-dihydroxy-eicosa-6Z, 8E, 10E, 14Z-tetraenoic acid). LTB4, in addition to being a potent chemoattractant for leukocytes, enhances the release of proinflammatory adipokines, such as MCP-1 and interleukin (IL)-6, from obese adipose tissue (10,11). LTB4 levels are elevated in adipose tissue during obesity (11–13), and LTB4 has been described to play a role in the development of insulin resistance and to directly decrease insulin signaling in myocytes and hepatocytes in vitro (5,14,15). In addition, although the activation of the ALOX5/ALOX5AP complex is key to leukotriene synthesis, it can alternatively lead to the biosynthesis of other compounds, such as lipoxins (16). The intermediate LTA4 can be converted to lipoxin A4 (LXA4) (5S, 6R, 15S-trihydroxy-eicosa-7E, 9E, 11Z, 13E-tetraenoic acid) by action of 12-lipoxygenase (ALOX12) in humans (17,18) or in mice by arachidonate 15-lipoxygenase (ALOX15 or ALOX12/15) (19,20).
Lipoxins, which belong to the specialized proresolving mediator (SPM) family, exert anti-inflammatory effects and are involved in the resolution of inflammation (5). Treatment with LXA4 of adipose tissue explants from aging mice—a model of adipose inflammation—leads to a decrease in IL-6 and restoration of GLUT4 and insulin receptor substrate (IRS)1 expression, indicating less inflammation and improved insulin sensitivity (21). In addition, it has recently been reported that treatment of mice with LXA4 protects against high-fat diet (HFD)-induced adipose inflammation and hepatic lipid deposition without affecting glucose tolerance (22).
The key genes involved in LTB4 and LXA4 formation, Alox5ap and Alox5, are normally expressed in adipocytes and cells from the adipose stromal vascular fraction. Their expression is increased in adipose tissue of obese patients and animals with insulin resistance (11,12,23,24). Nevertheless, mice deficient for ALOX5 present an increase in body weight and body fat content and are more prone to fat accumulation when fed an HFD (13,25,26). In addition, other reports on the effects of the genetic disruption of Alox5 on glucose homeostasis are contradictory (13,25,26). Thus, the role of ALOX5/ALOX5AP in the development of obesity and insulin resistance remains to be clearly established.
To address this issue, we generated transgenic mice overexpressing Alox5ap in adipose tissue. In these mice, ALOX5AP overexpression led to higher LXA4 rather than higher LTB4 levels. Transgenic mice were leaner and presented higher energy expenditure, in part due to browning of white adipose tissue (WAT), and were protected against diet-induced obesity and insulin resistance. Our results suggest that an increase in ALOX5/ALOX5AP activity has beneficial effects on metabolism, through an increase in circulating LXA4, and protects against diet-induced obesity, insulin resistance, hepatic steatosis, and inflammation.
Research Design and Methods
aP2/Alox5ap transgenic mice were generated by the Transgenic Animal Unit of the Center of Animal Biotechnology and Gene Therapy (CBATEG), Universitat Autònoma de Barcelona, by microinjection of a chimeric gene containing the entire coding sequence of the murine Alox5ap gene downstream of the 5.4 kb aP2 promoter and upstream of the SV40 polyadenylation signal into fertilized oocytes from C57BL/6J × SJL mice. Microinjected embryos were then transferred into receptor CD1 females to obtain the aP2/Alox5ap transgenic mice. Mice were kept in a specific pathogen–free facility (Servei d'Estabulari de Ratolins–CBATEG) and maintained under a light-dark cycle of 12 h at 22°C. Mice were fed ad libitum with either a chow diet (Teklad Global 18% Protein Rodent Diet, 2018S; Envigo, Cambridgeshire, U.K.) or an HFD (Teklad Custom Diet, TD88137; Envigo) for 11 weeks. When stated, mice were fasted for 16 h. Experiments were carried out in wild-type and transgenic littermates between 4 and 6 months old. An indirect open-circuit calorimeter (Oxylet; Panlab, Cornellà, Spain) was used to monitor oxygen consumption and carbon dioxide production as previously described (27). Data were taken from the light and dark cycle. For LXA4 stimulation, C57BL/6J male mice were given an intraperitoneal injection of 5 ng/g body wt LXA4 (Merck Millipore, Billerica, MA) or vehicle 24 h and 48 h before sample collection. Animals were anesthetized and killed, and tissues of interest were excised and kept at −80°C or with formalin until analysis. Animal care and experimental procedures were approved by the Ethics Committee in Animal and Human Experimentation of the Universitat Autònoma de Barcelona.
HepG2 cells were purchased from the European Collection of Cell Culture (Salisbury, U.K.) and cultured in minimum essential medium supplemented with 2 mmol/L glutamine, 1% nonessential amino acids, and 10% FCS (Gibco, Thermo Fisher Scientific, Grand Island, NY) at 37°C with 5% CO2. HepG2 cells were later treated with 200 nmol/L LXA4 or vehicle for 4 h.
Pancreata were perfused with 0.1 mg/mL collagenase I/II and thermolysin (Roche, Mannheim, Germany), digested for 19 min at 37°C, and purified from Histopaque 1077 gradient (Sigma-Aldrich, St. Louis, MO). Islets were handpicked under a stereomicroscope (Leica, Wetzlar, Germany) and used for RNA and protein extraction.
Gene Expression Analysis
Total RNA was extracted from different tissues using isolation reagent (TriPure, Roche, for nonfat tissues and QIAzol, Qiagen, Hilden, Germany, for fat tissues) and an Rneasy Mini kit (Qiagen). For Northern blot analysis, RNA samples were electrophoresed on a 1% agarose gel in MOPS buffer containing 2.2 mol/L formaldehyde. Membranes were hybridized with a 32P-labeled Alox5ap cDNA probe obtained from PCR amplification using specific primers for Alox5ap cDNA (FW-5′AGCGTGGTCCAGAATGCG3′; RV-5′GATCCGCTTG CCGAAGATGTA3′) (GE Healthcare, Buckinghamshire, U.K.). For quantitative RT-PCR analysis, total RNA (1 μg) was retrotranscribed using the Transcriptor First Strand cDNA Synthesis kit (Roche). Quantitative PCR was performed in a LightCycler (Roche) using the LightCycler 480 SyBr Green I Master Mix (Roche) and specific primers as described in Supplementary Table 2. Results were analyzed using the mathematical model of Pfaffl (28). All samples were processed in triplicate, and a mean Ct was calculated. The Ct for each transcript of interest was normalized by the Ct obtained for the reference gene RplpO.
Western Blot Analysis
Tissues were homogenized in protein lysis buffer. Western blot analysis was performed by standard procedures from total cellular homogenates of WAT, brown adipose tissue (BAT), liver, and islets as previously described (27). Briefly, tissues were homogenized in protein lysis buffer containing protease inhibitors. Protein expression was analyzed by electrophoresis of 15–30 μg different protein extracts in 10–12% SDS-PAGE gels transferred to polyvinylidene fluoride membranes and then incubated with Ponceau staining. Proteins were detected using 1:1,000 dilution of primary antibodies: rabbit polyclonal anti–uncoupling protein (UCP)1 (ab10983; Abcam, Cambridge, U.K.), rabbit polyclonal anti–5-lipoxygenase activating protein (FLAP) (bs-7556R; Bioss, Woburn, MA), rabbit polyclonal anti–liver X receptor (LXR)β (PA1-333; Thermo Fisher Scientific), and rabbit polyclonal anti–α-tubulin (ab4074; Abcam). Detection was performed using 1:10,000 dilution of horseradish peroxidase–labeled swine anti-rabbit immunoglobulins (P0217; Dako, Glostrup, Denmark) and Western blotting detection reagent (ECL Plus; Amersham, Freiburg, Germany).
Tissues were fixed for 24 h in formalin, embedded in paraffin, and sectioned. Sections were deparaffinized and stained with hematoxylin-eosin.
Glucose and Insulin Tolerance Tests
For the glucose tolerance test, awake mice, fasted overnight (16 h) with free access to water, were given an injection of 1 g glucose/kg body wt i.p., and glucose concentration was determined in blood samples at indicated time points using a Glucometer Elite analyzer (Bayer, Leverkusen, Germany). For the insulin tolerance test, 0.75 units of insulin/kg body wt i.p. (Humulin regular; Eli Lilly, Indianapolis, IN) was injected into awake fed mice, and glucose concentration was determined at indicated time points.
In Vivo Glucose-Stimulated Insulin Release
For insulin release determination, fasted mice were given an injection of 3 g glucose/kg body wt i.p. and venous blood was collected at indicated time points in chilled tubes and immediately centrifuged, and the plasma was stored at −20°C.
Ex Vivo LTB4 and LXA4 Release
Epididymal WAT (eWAT) and BAT explants from wild-type and transgenic mice were washed and incubated in 1 mL Krebs-Ringer bicarbonate HEPES buffer, pH 7.4, with either PD146176 (P4620; Sigma) or vehicle for 30 min, at 37°C at 300 rpm. LTB4 and LXA4 secretion in the media was then analyzed by ELISA.
Enzyme, Metabolite, and Hormone Assays
To determine pancreatic insulin content, whole pancreata were removed from the mice, weighted, and homogenized in 20 v/v of cold acidic ethanol (75% ethanol and 1.5% concentrated HCl) followed by 48 h of agitation at 4°C. Afterward, insulin was quantified in the supernatants of the samples diluted in phosphate buffer by radioimmunoassay (CIS Biointernational). Tissue triglyceride content was determined by extracting total lipids from liver samples with chloroform-methanol (2:1 v/v) as previously described (29). Triglycerides, total cholesterol, HDL cholesterol, and LDL cholesterol were quantified spectrophotometrically using an enzymatic assay kit (Horiba ABX, Montpellier, France). For determination of bile acid (BA) excretion rate, feces from individually housed mice were collected during a 24-h period and dried for 1 h at 70°C. The extraction was performed in 2 mL chloroform:methanol (2:1) following a method previously described (30). BAs from feces and from plasma were measured enzymatically (Randox Laboratory, Crumlin, U.K.). Serum nonesterified fatty acids (FFAs) were measured by the acyl-CoA synthase and acyl-CoA oxidase methods (Wako Chemicals GmbH, Neuss, Germany). All biochemical parameters were determined using a Pentra 400 Analyzer (Horiba-ABX). Glucose was determined using a Glucometer Elite analyzer (Bayer), and insulin levels were measured using the Rat Insulin ELISA kit (Crystal Chem, Chicago, IL). Leptin concentration was determined using the Mouse leptin ELISA kit (Crystal Chem). Serum MCP-1 levels were determined using the Mouse MCP-1 ELISA kit (eBioscience, San Diego, CA). Serum LTB4 levels were measured with the LTB4 ELISA kit (Enzo Life Sciences, Farmingdale, NY), and LXA4 levels were measured in serum samples and in WAT extracts using the Mouse LXA4 ELISA kit (CUSABIO BIOTECH Co. Ltd., Wuhan, China).
All values are expressed as the mean ± SEM. Differences between groups were compared by Student t test. Statistical significance was considered if P < 0.05.
Adipose Tissue ALOX5AP Overexpression Leads to Increased Production of LXA4 Rather Than LTB4
For evaluation of the role of the Alox5ap gene in adipose tissue, transgenic mice overexpressing Alox5ap under the control of the aP2 promoter were generated. The aP2 promoter has mostly been used to drive transgene expression in adipose tissue, although it has also been reported to allow expression in other tissues to a lesser extent (31–33). Two transgenic lines (Tg2 and Tg4) expressing high levels of Alox5ap mRNA in abdominal eWAT, measured either by Northern blot or by quantitative PCR (Fig. 1A and B), were analyzed. Since similar results have been obtained in both lines for a number of analyses, and to avoid an unnecessary increase in the number of mice studied, Tg2 was randomly selected for a complete phenotyping. In Tg2 mice, expression was 200-fold increased in BAT and 50-fold higher in subcutaneous inguinal WAT (scWAT) (Fig. 1C and D). In nonadipose tissues such as liver or pancreatic islets, a slight increase in Alox5ap expression has also been observed, but to a lesser extent than in adipose tissue, while expression remained unchanged in heart (Fig. 1D). ALOX5AP/FLAP protein levels were also higher in eWAT and BAT from transgenic mice (Fig. 1E). Measurements of LXA4 and LTB4 production by adipose tissue explants showed that BAT from Tg2 released higher levels of LXA4 but not LTB4 compared with wild-type mice (Fig. 1F and G). A similar tendency was also observed in scWAT (Fig. 1F and G). Accordingly, circulating levels of LXA4 were increased approximately twofold in Tg2 mice (Fig. 1H), while circulating LTB4 levels were similar in both groups (Fig. 1I). In addition, specific inhibition of ALOX15 by PD146176 led to a reduction in LXA4 production by BAT from Tg2 without affecting LTB4 release (Supplementary Fig. 1A). This suggests that LXA4 production in transgenic mice results from the sequential action of ALOX5 and ALOX15. However, expression of Alox5 and Alox15 remained significantly unchanged in scWAT and eWAT from Tg2 (Supplementary Fig. 1B). Moreover, adipose mRNA levels of Irs1 and Slc2a4 (GLUT4 gene), two genes whose expression increases after LXA4 treatment (21), were higher in transgenic mice (Fig. 1J). Altogether, these results indicate that ALOX5AP overexpression leads to increased production of LXA4 rather than LTB4 in adipose tissue.
aP2/Alox5ap Transgenic Mice Are Leaner and Present Increased Energy Expenditure and Thermogenesis
Both lines of aP2/Alox5ap transgenic mice were smaller than wild-type mice, with a shorter naso-anal length and lower body weight (Fig. 2A and B and Supplementary Fig. 2A and B). The weight of several main organs was also reduced (Supplementary Table 1). However, when corrected by total body weight, only the weight of BAT and WAT adipose depots was decreased in transgenic mice (Fig. 2C and Supplementary Fig. 2C). A smaller adipocyte size was also observed in the eWAT depot (Fig. 2D). This decrease in adiposity was not due to changes in food intake (Fig. 2E and Supplementary Fig. 2D) but, rather, to higher energy expenditure, as measured by indirect calorimetry (Fig. 2F and Supplementary Fig. 2E). In addition, the respiratory quotient (RQ) was reduced in transgenic mice, suggesting an increase in fat substrate oxidation (Fig. 2G and Supplementary Fig. 2F). Higher energy expenditure may reflect changes in thermogenesis. The mitochondrial UCP1 in BAT is essential to this function. BAT from transgenic mice expressed higher levels of Ucp1 mRNA compared with wild-type mice and showed a decrease in the expression of Twist1, a negative regulator of Ucp1 gene transcription (34) (Fig. 2H). Under certain conditions, adipocytes within scWAT have the capacity to express UCP1 and may contribute to thermogenesis (35,36). These adipocytes share characteristics with brown adipocytes, in particular, their multilocular lipid droplet morphology, their high mitochondrial content, and the expression of a set of markers such as Ucp1, Cidea, or Ppargc1a (36). In aP2/Alox5ap transgenic mice, enhanced expression of Ucp1, Cidea, and Ppargc1a mRNA was observed in scWAT (Fig. 2I). In addition, expression of two negative regulators of Ucp1, namely, Twist1 and Nr1h3, encoding for LXRα, was decreased (Fig. 2I). Higher UCP1 levels were also detected by Western blot in scWAT of transgenic mice (Fig. 2J). In addition, histological analysis of transgenic scWAT depots showed numerous clusters of multilocular adipocytes interspersed within the classical unilocular white adipocytes, indicating browning of scWAT (Fig. 2K). To determine whether LXA4 is responsible for these effects, C57BL/6J mice were treated with LXA4 for 48 h. Treated mice presented increased levels of Ucp1 and browning markers in scWAT, such as Cidea, Ppargc1a, or Dio2 (Fig. 2L). Thus, ALOX5AP overexpression, and the resulting increase in LXA4, led to decreased adiposity, increased energy expenditure, and induction of browning of WAT.
aP2/Alox5ap Transgenic Mice Present Higher Levels of BAs in Circulation
Serum parameters reflecting lipid metabolism were examined. Circulating levels of triglycerides and FFAs remained unchanged in transgenic mice, whereas they presented higher levels of total cholesterol, reflecting an increase in HDL cholesterol without changes in LDL cholesterol (Fig. 3A–E).
In addition, circulating levels of BAs were higher in transgenic mice, as was the fecal BA excretion rate (Fig. 3F and G). The genes encoding for two enzymes involved in BA synthesis, namely, cholesterol 7α-hydroxylase (CYP7A1) and oxysterol 7α-hydroxylase (CYP7B1) (37–39), were upregulated in the liver of transgenic mice, and this may have led to higher plasma BA levels (Fig. 3H). The expression of LXR, a transcription factor key to the regulation of Cyp7a1 expression (40), was examined next. Hepatic expression of both isoforms of LXR, LXRα (Nr1h3 gene) and LXRβ (Nr1h2 gene), was increased in transgenic animals (Fig. 3I), as were protein levels of LXRβ (Fig. 3J and K). A direct effect of LXA4 on these genes has been further demonstrated in vitro. HepG2 cells treated with LXA4 presented increased CYP7A1 expression and a tendency to increase NR1H2 expression compared with cells treated with vehicle (Fig. 3L). BAs have been shown to increase brown fat activity and energy expenditure through the G-protein–coupled BA receptor 1 (GPBAR1) (also called TGR5) and the activation of type 2 iodothyronine deionidase (DIO2) (41,42). In BAT from aP2/Alox5ap mice, Gpbar1 expression levels were higher and Dio2 expression showed a tendency to increase (Fig. 3M). These results suggest that overexpression of ALOX5AP in adipose tissue leads to upregulation of hepatic LXR, which in turn enhances Cyp7a1 expression, thus increasing BA synthesis, contributing to higher thermogenesis and energy expenditure.
Adipose Overexpression of ALOX5AP Leads to Impaired Insulin Secretion and Glucose Homeostasis
An intraperitoneal glucose tolerance test showed that transgenic mice displayed impaired glucose disposal after a glucose load (Fig. 4A and Supplementary Fig. 3A). Glucose intolerance may result from defects in insulin sensitivity and/or in inappropriate insulin secretion. When an insulin tolerance test was performed, transgenic mice showed an insulin response similar to that in wild-type mice, indicating preserved insulin sensitivity and suggesting a defect in insulin secretion (Fig. 4B and Supplementary Fig. 3B). Accordingly, circulating insulin levels showed a tendency to be lower (Fig. 4C) and the in vivo glucose-stimulated insulin release was altered in transgenic mice (Fig. 4D and Supplementary Fig. 3C). In particular, the first peak of insulin secretion was blunted (Fig. 4D and Supplementary Fig. 3C). In addition, pancreatic insulin content was increased, supporting the hypothesis that ALOX5AP overexpression led to a defect in insulin secretion without altering insulin synthesis (Fig. 4E). Since LXRβ may play a role in insulin secretion (43), its protein levels were examined in islets isolated from transgenic and wild-type mice. In islets from transgenic mice, LXRβ expression was 50% decreased, and this reduction may have contributed to the impairment of insulin secretion (Fig. 4F and G). Furthermore, in islets isolated from LXA4-treated C57BL/6J mice, a tendency toward a decrease in LXRβ gene (Nr1h2) expression was also observed (Fig. 4H), suggesting a direct involvement of LXA4.
Adipose Overexpression of ALOX5AP Prevents Diet-Induced Obesity and Insulin Resistance
The effects of ALOX5AP overexpression in adipose tissue were then examined under obesogenic conditions. In transgenic mice fed an HFD for 11 weeks, Alox5ap expression was increased in eWAT compared with wild-type mice, and LXA4 circulating levels were also higher (Fig. 5A and B). In addition, whereas wild-type mice gained ∼35% of their initial body weight, transgenic mice only gained 20% (Fig. 5C). Accordingly, the absolute and relative weights of epididymal white fat pad were lower in transgenic mice and adipocyte size was reduced (Fig. 5D–F). However, food intake was similar in both genotypes and the calculated food conversion efficiency was 50% lower in transgenic mice (Fig. 5G and H). This reduction may be partially explained by a tendency toward higher energy expenditure in transgenic mice (Fig. 5I), although no differences were observed in RQ between transgenic and wild-type mice (Fig. 5J). In addition, transgenic mice on an HFD showed high levels of UCP1 mRNA and protein in scWAT, indicating browning of WAT (Fig. 5K and L). In contrast, in BAT, Ucp1 mRNA was not increased in transgenic mice despite a decrease in the expression of Twist1 and Nr1h2 (Fig. 5M), both negative regulators of Ucp1 expression. This may have been due to the observed decrease in Ppargc1a expression (Fig. 5M), which is critical for Ucp1 transcription (34,44). Furthermore, the decrease in adipocyte size may also be caused by changes in WAT gene expression. In HFD-fed transgenic mice, the decrease in expression of lipoprotein lipase (Lpl), and the increase in Cpt1a expression (Fig. 5N), may have contributed to the smaller size of fat cells. In addition, enhanced expression of Nr1h2 and -3, encoding, respectively, for LXR-β and -α, and previously described as possible regulators of lipolysis (45), may also have led to a decrease in WAT lipid deposition (Fig. 5N). In HFD-fed transgenic mice, Slc2a4 expression was also upregulated, both in scWAT and eWAT (Fig. 5K and N). This may have resulted from the increased levels of LXRα and LXA4 (21,46,47).
In addition, when challenged with a glucose load, transgenic mice fed an HFD presented an improvement in glucose tolerance with respect to wild-type mice, although they were more glucose intolerant than mice fed a standard diet (Fig. 6A). On an HFD, transgenic mice also showed better whole-body insulin sensitivity than wild-type mice (Fig. 6B). In agreement with the higher insulin sensitivity observed in transgenic mice fed an HFD, insulin levels were lower in these mice than in wild-type mice (Fig. 6C). In addition, the histological analysis of liver sections and the measurement of triglyceride content showed fewer fat droplets in the liver of transgenic mice and decreased hepatic triglyceride content (Fig. 7A and B). In addition, transgenic mice showed hepatic downregulation of fatty acid synthase (Fasn) and Lipin1 (Fig. 7C). Moreover, expression of sterol regulatory element–binding protein 1c (Srebf1) tended to decrease in transgenic mice, similar to Nr1h3, while Nr1h2 expression was higher (Fig. 7C). Thus, adipose ALOX5AP overexpression prevents not only diet-induced weight gain but also hepatic steatosis and insulin resistance. In addition, transgenic mice on an HFD showed a tendency toward increased levels of BAs and upregulation of hepatic Cyp7a1 and Cyp7b1 (Fig. 7D and E). Gpbar1 and Dio2 expression levels were also increased in BAT from transgenic mice compared with wild-type mice (Fig. 7F). Since HFD-induced obesity is associated with adipose tissue inflammation, the eWAT mRNA levels of the macrophage marker F4/80 and the cytokine IL-6, both markers of inflammation, were examined next. Expression of both markers was reduced in transgenic mice, suggesting protection against diet-induced inflammation (Fig. 7G). In addition, circulating MCP-1 and leptin levels were lower in transgenic mice fed an HFD than in wild-type mice (Fig. 7H and I). Altogether, these results suggest that overexpression of ALOX5AP prevents HFD-induced obesity, through browning of WAT, and protects against insulin resistance and inflammation.
Obesity has become a major public health problem in recent decades. Thus, finding new therapeutic approaches based on a better understanding of the pathophysiology of the disease is a key issue for our society. In this work, we demonstrate that overexpression of ALOX5AP leads to increased production of LXA4 rather than LTB4, decreased adiposity, and protection against HFD-induced obesity and insulin resistance.
The role of ALOX5/ALOX5AP in eicosanoid production has been recognized mainly for LTB4 synthesis, although there is accumulating evidence that LXA4, an SPM, is also a key product of this pathway (9,48,49). The mechanisms controlling the balance between LTB4 and LXA4 production are, however, as yet unclear. Surprisingly, in aP2/Alox5ap transgenic mice, ALOX5AP overexpression led to higher levels of LXA4, whereas other reports have associated the increase in ALOX5/ALOX5AP in adipose tissue from rodent models of obesity with higher levels of LTB4 (11,12,14,24). It has been reported that resolvin D1, another SPM, may increase LXA4 and decrease LTB4 production in macrophages (50). In addition, LXA4 itself is able to decrease LTB4 synthesis (50). Interestingly, obese adipose tissue displays lower resolvin D1 levels than lean tissue (51), and this may explain the increase in adipose tissue LTB4 content observed in obesity.
The lean phenotype of ALOX5AP-overexpressing mice may at least partially be the result of an increase in energy expenditure as a consequence of the activation of thermogenesis in both BAT and scWAT. In agreement with this, ALOX5AP-overexpressing mice showed appearance of brown-like, multilocular adipocytes in scWAT, accompanied by upregulation of UCP1 expression and increased expression in markers of brown cells, indicating an enhancement of WAT browning. Furthermore, the resulting activation of thermogenesis may have also contributed to the improvement of insulin resistance observed in mice fed an HFD. In line with our results, the promotion of BAT activity or the browning of WAT has been reported to result in increased energy expenditure and protection against obesity and type 2 diabetes (35,36,52,53). These preclinical observations together with the recent discovery that browning of scWAT occurs in adult humans make this cell type an attractive therapeutic target for the treatment of obesity and type 2 diabetes (35,53,54). Our results suggest that molecules arising from the ALOX5/ALOX5AP pathway may have potential therapeutic effects. Here we demonstrate that one of these factors, LXA4, is able to promote browning of scWAT in vivo. Moreover, other molecules, such as eicosapentanoic acid, a precursor of various factors belonging to the same family as LXA4 (SPM family), are also able to induce the thermogenic capacity of subcutaneous adipocytes in vitro (55).
The observed effects of ALOX5AP overexpression on energy homeostasis may also rely on BA. Indeed, the increased levels of BAs in Alox5ap transgenic mice may have led to activation of BAT and higher energy expenditure. In this regard, it has been reported that BAs exert pleiotropic effects on metabolism, including activation of BAT leading to increased energy expenditure, which has prevented obesity and insulin resistance during HFD feeding (41). In humans, it has recently been reported that BAs activate BAT and increase energy expenditure (42). These effects of BA are mediated by the GPBAR1 (TGR5) receptor and dependent on the induction of DIO2 (42), expression of which was increased in BAT from aP2/Alox5ap transgenic mice. The observed increase in BA levels was probably due to hepatic Cyp7a1 upregulation. Consistent with our findings, Cyp7a1-overexpressing transgenic mice presented higher BA synthesis; were protected against HFD-induced obesity, insulin resistance, and fatty liver; and presented higher levels of energy expenditure, associated with an increase in Ucp1, Dio2, and Gpbar1 expression in BAT (56). In aP2/Alox5ap transgenic mice, upregulation of Cyp7a1 was most likely due to an increase in hepatic LXR expression, a main regulator of Cyp7a1 gene expression (40). Our results suggest that LXR mediates the effects of LXA4 and plays a central role in the phenotype of aP2/Alox5ap mice. In particular, we show that LXA4 induces Nr1h2 expression in hepatoma cells and decreases it in islets. In agreement with this, it has been recently reported that LXA4 controls LXRα expression and cholesterol metabolism in macrophages (57). Since LXR has raised much interest for the treatment of atherosclerosis, diabetes, inflammation, Alzheimer disease, and cancer (58,59), a better understanding of its regulation by LXA4 would be particularly relevant.
Despite the beneficial effects of ALOX5AP overexpression on energy metabolism, transgenic mice also presented glucose intolerance when fed a standard diet, probably due to a deficiency in insulin secretion. The observed decrease in LXRβ expression in islets may have contributed to this defect, as reported in mice deficient for LXRβ (43). LXRβ knockout mice were glucose intolerant due to impaired glucose-induced insulin secretion but displayed normal insulin sensitivity (43). Similarly, ALOX5AP-overexpressing mice presented alterations in glucose homeostasis but were not resistant to insulin and presented improved insulin sensitivity compared with wild-type mice when fed an HFD. This improvement may be, at least partially, due to the leaner phenotype of these mice, associated with decreased lipid accumulation in nonadipose tissues and decreased inflammation in WAT. The increase in Cyp7b1 hepatic expression observed in transgenic mice may also have contributed to their protection against diet-induced glucose intolerance and hepatic steatosis, as suggested by the beneficial effects of hepatic Cyp7b1 overexpression on these parameters previously reported (60).
In summary, our results clearly demonstrate that an increase in ALOX5AP expression may have beneficial effects on energy metabolism during obesity, increasing LXA4 rather than LTB4 levels and thus protecting against diet-induced obesity, through browning of WAT, and preventing insulin resistance, hepatic steatosis, and inflammation (Fig. 8). In agreement with our results, blocking LTA4 hydrolase, the enzyme responsible for converting LTA4 to LTB4, leads to anti-inflammatory effects mediated by an increase in LXA4 production (61,62). Our data also support the idea that LXA4 deficiency plays a role in the development of obesity and its complications. In line with this, LXA4 levels have been reported to be lower in adipose tissue of obese mice and to decrease with age (22,49). In addition, in humans, decreased LXA4 levels accompanied by augmented levels of LTB4 have been reported to be a hallmark of peri-wound fat from patients with a BMI >25 kg/m2 (63).
Altogether, these results underscore the importance of acquiring a better understanding of the pathways involved in lipoxins and leukotrienes synthesis and their regulation, as they would be crucial for the development of new treatments for obesity and its comorbidities.
Acknowledgments. The authors thank Drs. Malcolm Watford (Rutgers University) and Virginia Haurigot (Universitat Autònoma de Barcelona) for helpful discussion; Mireia Zaguirre, Marta Moya, Jennifer Barrero, and Lidia Hernandez (Universitat Autònoma de Barcelona) for technical assistance; and Dr. Anna Pujol and Anna Arbós (Universitat Autònoma de Barcelona) for transgenic mice generation.
Funding. This work was supported by a grant from Ministerio de Economía y Competitividad of Spain, Instituto de Salud Carlos III (PI07/0313) to S.F. and from Ministerio de Economía y Competitividad, Plan Nacional I+D+I (SAF2014-54866-R), and Generalitat de Catalunya (Agència de Gestió d’Ajuts Universitaris i de Recerca, 2014-SGR1669, and Institució Catalana de Recerca i Estudis Avançats Academia) to F.B.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. I.E. and S.F. designed the study, conducted the experiments, and wrote the manuscript. T.F., L.V., A.C., and M.G. conducted experiments and supported the work with key suggestions. S.M., M.M., J.A., C.R., and J.R. conducted experiments. F.B. designed the study, participated in discussion and editing, and secured funding. S.F. directed the project and secured funding. S.F. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.