Abstract

Objective

   Brown and white adipose tissue exerts pleiotropic effects on systemic
   energy metabolism in part by releasing endocrine factors. Neuregulin 4
   (Nrg4) was recently identified as a brown fat-enriched secreted factor
   that ameliorates diet-induced metabolic disorders, including insulin
   resistance and hepatic steatosis. However, the physiological mechanisms
   through which Nrg4 regulates energy balance and glucose and lipid
   metabolism remain incompletely understood. The aims of the current
   study were: i) to investigate the regulation of adipose Nrg4 expression
   during obesity and the physiological signals involved, ii) to elucidate
   the mechanisms underlying Nrg4 regulation of energy balance and glucose
   and lipid metabolism, and iii) to explore whether Nrg4 regulates
   adipose tissue secretome gene expression and adipokine secretion.

Methods

   We examined the correlation of adipose Nrg4 expression with obesity in
   a cohort of diet-induced obese mice and investigated the upstream
   signals that regulate Nrg4 expression. We performed metabolic cage and
   hyperinsulinemic-euglycemic clamp studies in Nrg4 transgenic mice to
   dissect the metabolic pathways regulated by Nrg4. We investigated how
   Nrg4 regulates hepatic lipid metabolism in the fasting state and
   explored the effects of Nrg4 on adipose tissue gene expression,
   particularly those encoding secreted factors.

Results

   Adipose Nrg4 expression is inversely correlated with adiposity and
   regulated by pro-inflammatory and anti-inflammatory signaling.
   Transgenic expression of Nrg4 increases energy expenditure and augments
   whole body glucose metabolism. Nrg4 protects mice from diet-induced
   hepatic steatosis in part through activation of hepatic fatty acid
   oxidation and ketogenesis. Finally, Nrg4 promotes a healthy adipokine
   profile during obesity.

Conclusions

   Nrg4 exerts pleiotropic beneficial effects on energy balance and
   glucose and lipid metabolism to ameliorate obesity-associated metabolic
   disorders. Biologic therapeutics based on Nrg4 may improve both type 2
   diabetes and non-alcoholic fatty liver disease (NAFLD) in patients.

   Keywords: Adipose tissue, Brown fat, Nrg4, Adipokine, NAFLD, Diabetes

   Abbreviations: BAT, Brown adipose tissue; WAT, White adipose tissue;
   eWAT, epididymal WAT; Nrg4, Neuregulin 4; HFD, High-fat diet; TAG,
   Triglyceride; WT, Wild type; Tg, Transgenic; KO, Knockout; TNFα, Tumor
   necrosis factor α; NALFD, Non-alcoholic fatty liver disease; VEGFα,
   Vascular endothelial growth factor α; FGF21, Fibroblast growth factor
   21; IL-6, Interleukin-6; BMPs, Bone morphogenetic proteins; UCP-1,
   Uncoupling protein 1; GPR120, G-protein coupled receptor 120; CoA,
   Co-enzyme A

Highlights

     * •
       Nrg4 is a target of pro-inflammatory and anti-inflammatory
       signaling in adipocytes.
     * •
       Transgenic expression of Nrg4 increased energy expenditure and
       glucose metabolism.
     * •
       Nrg4 stimulates hepatic fatty acid oxidation and ketogenesis during
       starvation.
     * •
       Nrg4 promotes a beneficial adipokine profile during obesity.

1. Introduction

   Adipose tissue is central to systemic energy balance and metabolic
   homeostasis. White adipose tissue (WAT) stores lipids and communicates
   with the central nervous system and other peripheral tissues by
   secreting a number of important endocrine hormones [49][1], [50][2],
   [51][3]. WAT also serves as a major site for the integration of immune
   and metabolic signals [52][4], [53][5], [54][6]. The primary role of
   brown adipose tissue (BAT) is fuel oxidation linked to thermogenesis,
   which serves an important function in defense against cold and energy
   excess [55][7], [56][8], [57][9]. Secreted factors are important
   regulators of carbohydrate and lipid metabolism and systemic energy
   homeostasis. Adipose tissue hormones, such as leptin and adiponectin
   [58][1], [59][2], [60][3], gut-derived fibroblast growth factors
   [61][10], [62][11], and myokines [63][12] participate in nutrient
   sensing and coordinate key aspects of nutrient and energy metabolism.
   By integrating mouse tissue and brown adipogenesis transcriptome data,
   a list of brown fat-enriched secreted factors was identified [64][13].
   Among these, Neuregulin 4 (Nrg4) emerged as a BAT-enriched endocrine
   factor that exerts robust effects on hepatic lipid metabolism and
   systemic homeostasis. Brown fat is also known to release factors that
   exhibit a wider expression profile, such as Vascular endothelial growth
   factor α (VEGFα), Fibroblast growth factor 21 (FGF21), Interleukin-6
   (IL-6), and Bone morphogenetic proteins (BMPs) [65][14], [66][15]. As
   such, brown fat is emerging as a source of endocrine factors that
   exerts powerful metabolic effects beyond UCP1-mediated thermogenesis
   [67][14], [68][15].

   Nrg4 expression is more abundant in brown fat than white fat; a similar
   pattern was observed in a recent study on differential gene expression
   in adipose tissue [69][16]. Despite its enrichment in brown fat, Nrg4
   expression is readily detectable in white fat, which likely provides a
   significant source of total Nrg4 in circulation given its relatively
   large mass. An intriguing aspect of Nrg4 regulation is that its
   expression in WAT is markedly downregulated in mouse and human obesity
   [70][13]. These findings raise the possibility that obesity is
   associated with a functional deficit of Nrg4 that exacerbates the
   progression of metabolic disorders. In support of this, mice lacking
   Nrg4 developed more severe insulin resistance and hepatic steatosis
   following high-fat feeding, whereas fat-specific transgenic expression
   of Nrg4 significantly improves metabolic parameters and ameliorates
   diet-induced disruption of homeostasis. At the molecular level, Nrg4
   transduces signals through the ErbB4 and ErbB3 receptor tyrosine
   kinases and elicits an inhibitory effect on hepatic lipogenesis.
   Despite these intriguing findings, the physiological mechanisms through
   which Nrg4 modulates energy balance and glucose and lipid metabolism
   remain incompletely understood. In this study, we performed metabolic
   cage and hyperinsulinemic-euglycemic clamp studies and revealed a
   surprisingly pleiotropic effect of Nrg4 signaling on key aspects of
   systemic energy and glucose metabolism. We identified hepatic fatty
   acid oxidation and ketogenesis as a new metabolic target of endocrine
   signaling by Nrg4. Finally, we found that Nrg4 promotes a beneficial
   adipokine profile during obesity. Together, this work provides
   important insights into the physiological actions of Nrg4 that
   contribute to its beneficial effects on metabolic homeostasis.

2. Materials and methods

2.1. Mouse studies

   All mouse studies were performed according to procedures approved by
   the University Committee on Use and Care of Animals at the University
   of Michigan. The generation of Nrg4 transgenic and knockout mice was
   described previously [71][13]. Mice were maintained under 12/12 h
   light/dark cycles with free access to food and water. Teklad 5001
   laboratory diet was used as a standard chow. For HFD feeding, mice were
   fed a diet containing 60% of calories from fat (D12492, Research
   Diets). Metabolic cage study was performed at the University of
   Michigan Animal Phenotyping Core using CLAMS (Columbus Instruments).
   The measurements were carried out for a total of 72 h. Energy
   expenditure was calculated as the value corresponding to
   3.91VO[2] + 1.1VCO[2]. Body fat and lean mass were measured using an
   NMR analyzer (Minispec LF90II, Bruker Optics). Plasma concentrations of
   TAG, total cholesterol, NEFA, and β-hydroxybutyrate were measured using
   commercial assay kits (Stanbio Laboratory). Plasma insulin and leptin
   concentrations were measured using ELISA (Crystal Chem).

2.2. Hyperinsulinemic-euglycemic clamp

   Mice were fasted overnight (approximately 15 h), and a 2-h
   hyperinsulinemic-euglycemic clamp was conducted in awake mice with a
   primed and continuous infusion of human insulin (150 mU/kg body weight
   priming followed by 2.5 mU/kg/min; Novolin, Novo Nordisk, Plainsboro,
   NJ) [72][17]. To maintain euglycemia, 20% glucose was infused at
   variable rates during clamps. Whole body glucose turnover was assessed
   with a continuous infusion of [3-^3H]glucose (PerkinElmer, Waltham,
   MA), and 2-deoxy-D-[1-^14C]glucose (2-[^14C]DG) was administered as a
   bolus (10 μCi) at 75 min after the start of clamps to measure
   insulin-stimulated glucose uptake in individual organs. At the end of
   the clamps, mice were anesthetized, and tissues were taken for
   biochemical analysis as previously described [73][17].

2.3. Adipocyte studies

   3T3-L1 preadipocytes were cultured in DMEM with 10% bovine growth serum
   until two days post confluence (denoted as day 0). Differentiation was
   induced by adding a cocktail containing 0.5 mM IBMX, 1 μM
   dexamethasone, and 1 μg/mL insulin to DMEM supplemented with 10% FBS.
   Three days after induction, cells were cultured in DMEM containing 10%
   FBS plus 1 μg/mL of insulin for two more days followed by maintenance
   in DMEM supplemented with 10% FBS. Differentiated adipocytes were
   treated with TNFα, VIII compound, Rosiglitazone, GW9508 and DHA for
   17 h in maintenance media before RNA isolation and RT-qPCR analysis.

2.4. qPCR and immunoblotting analyses

   RT-qPCR analysis of gene expression was performed as previously
   described [74][18]. Briefly, total adipocyte and liver RNA was
   extracted using TRIzol method. Total RNA from eWAT and BAT was isolated
   using PureLink RNA isolation kit (ThermoFisher). For RT-qPCR, 2 μg of
   total RNA was reverse-transcribed using MMLV-RT followed by qPCR using
   SYBR Green (Life Technologies). Relative mRNA expression was normalized
   to the levels of ribosomal protein 36B4. Total eWAT lysates were
   prepared by tissue homogenization in a lysis buffer containing 50 mM
   Tris (pH 7.5), 150 mM NaCl, 5 mM NaF, 25 mM β-glycerolphosphate, 1 mM
   sodium orthovanadate, 10% glycerol, 1% tritonX-100, 1 mM
   dithiothreitol, and freshly added protease inhibitors. Immunoblotting
   was performed using specific antibodies against adipsin (Santa Cruz
   Biotech, sc-50419).

2.5. Microarray analysis

   Gene expression profiling was performed using Mouse Gene ST 2.1 array.
   We used a cutoff of normalized array values (log2-transformed
   values > 7.0) for adipose tissue transcripts. Pathway enrichment
   analysis was performed using the Database for Annotation, Visualization
   and Integrated Discovery (DAVID, available at
   [75]http://david.abcc.ncifcrf.gov). Adipose tissue gene expression was
   visualized using TreeView.

2.6. Stimulated Raman scattering (SRS) microscopy

   SRS microscopy was used to evaluate lipid accumulation in hepatocytes
   as previously described [76][19]. Briefly, liver specimens were
   sectioned into a series of 50 μm thickness tissue slices by a
   vibratome. Single slice was sandwiched between two coverslips for SRS
   imaging of lipids with the beating frequency tuned to 2850 cm^−1.

2.7. Hepatic lipid and CoA analyses

   Hepatic lipids were extracted using Bligh-Dyer method in the presence
   of an internal standard (T21:0 TAG, 10 nmol/mg protein) and separated
   on silica gel 60 Å plates that were developed with a nonpolar acidic
   mobile phase (70:30:1, v/v/v, hexane/ethyl ether/acetic acid). Fatty
   acid profiling was performed using procedure described previously
   [77][13]. CoA ester levels were assessed in perchloric acid extracted
   heart tissue using a high-performance liquid chromatography procedure
   described previously [78][20].

2.8. Statistical analysis

   Statistical analysis was performed using GraphPad Prism 7. Statistical
   differences were evaluated using two-tailed unpaired Student's t-test
   or one-way analysis of variance (ANOVA) with appropriate post hoc
   analyses. A p value of less than 0.05 (*p < 0.05, **p < 0.01, and
   ***p < 0.001) was considered statistically significant. Statistical
   methods and corresponding p values for data shown in each panel were
   included in figure legends.

3. Results

3.1. Regulation of adipose Nrg4 expression in obesity

   Nrg4 was identified as a brown fat-enriched endocrine factor that
   improves obesity-associated insulin resistance and hepatic steatosis
   [79][13]. However, the physiological mechanisms underlying the
   metabolic effects of Nrg4 on glucose and lipid metabolism and energy
   balance remain incompletely understood. To address this, we first
   examined adipose tissue expression of Nrg4 in high-fat diet (HFD)
   induced obesity. Wild type C57BL/6J mice gained various amounts of body
   weight following two months of high-fat feeding. Nrg4 mRNA levels in
   epididymal white adipose tissue (eWAT), but not BAT, exhibited a strong
   inverse correlation with body weight and adiposity ([80]Figure 1A, B).
   The profile of Nrg4 expression was remarkably similar to that of
   adiponectin, which has been shown to decrease in eWAT during obesity
   [81][1]. On the contrary, the expression of Ccl2, a chemotactic
   cytokine for macrophages, showed a strong positive correlation with
   body weight. We further examined the association between eWAT Nrg4
   expression and blood glucose and plasma insulin concentrations. As
   shown in [82]Figure 1C, there was a strong inverse correlation of Nrg4
   mRNA levels with blood glucose and plasma insulin concentrations. Mice
   exhibiting low adipose Nrg4 expression tended to have more severe
   HFD-induced hyperglycemia and hyperinsulinemia, suggesting that reduced
   Nrg4 expression may be causally linked to insulin resistance.

Figure 1.

   [83]Figure 1
   [84]Open in a new tab

   Regulation of adipose Nrg4 expression. (A) Correlation of eWAT Nrg4,
   Ccl2, and Adipoq mRNA levels with body weight and eWAT mass in mice fed
   HFD for 8 weeks. (B) Correlation of BAT Nrg4 expression with body
   weight. (C) Correlation of eWAT Nrg4 mRNA levels with blood glucose and
   plasma insulin concentrations in HFD-fed mice. (D) qPCR analysis of
   Nrg4 expression in 3T3-L1 adipocytes treated with vehicle (−) or 20 nM
   TNFα without or with 2 μM inhibitor VIII for 17 h. (E) qPCR analysis of
   Nrg4 expression in 3T3-L1 adipocytes treated with vehicle or TNFα
   without or with 10 μM rosiglitazone (Rosi). (F) qPCR analysis of Nrg4
   expression in 3T3-L1 adipocytes treated with vehicle, GW9508, or DHA at
   indicated concentrations (μM) for 17 h. Data represent mean ± sd.
   *p < 0.01, vs. vehicle; one-way ANOVA.

   Obesity is associated with chronic low-grade inflammation in WAT that
   is characterized by increased pro-inflammatory cytokine signaling. We
   next examined how pro-inflammatory and anti-inflammatory stimuli
   modulate Nrg4 expression in cultured adipocytes. Treatment of
   differentiated 3T3-L1 adipocytes with TNFα, a prototypical
   pro-inflammatory cytokine, significantly decreased Nrg4 expression
   through a mechanism that require NF-kB activation ([85]Figure 1D).
   Addition of IKK2 inhibitor VIII, a compound that blocks NF-kB
   activation, abolished the inhibition of Nrg4 expression in response to
   TNFα. The inhibitory effect of TNFα on Nrg4 expression was also
   reversed by rosiglitazone ([86]Figure 1E), which exerts powerful
   anti-inflammatory activities through activation of the nuclear receptor
   PPARγ [87][21]. G-protein coupled receptor 120 (GPR120) was recently
   identified as a receptor for fatty acid ligands, particularly the
   unsaturated fatty acids such as docosahexaenoic acid (DHA) [88][22],
   [89][23]. Interestingly, treatment of 3T3-L1 adipocytes with DHA or the
   synthetic GPR120 agonist GW9508 resulted in robust induction of Nrg4
   expression ([90]Figure 1F). These results indicate that
   pro-inflammatory and anti-inflammatory signals exert opposite effects
   on Nrg4 expression in adipocytes. Further, adipose tissue inflammation
   is likely a driving force in obesity-associated decrease of Nrg4
   expression in adipose tissue.

3.2. Nrg4 protects mice from diet-induced obesity by increasing energy
expenditure

   We previously demonstrated that fat-specific transgenic expression of
   Nrg4 in adipose tissue ameliorates diet-induced obesity and metabolic
   disorders [91][13]. Body composition analysis indicated that, compared
   to wild type (WT) littermate controls, Nrg4 transgenic (Tg) mice had
   significantly less fat mass following seven weeks of high-fat feeding
   ([92]Figure 2A). Whole body lean mass appeared similar between two
   groups. Plasma leptin levels were lower in Tg mice than control,
   suggesting that leptin sensitivity is improved in Tg group
   ([93]Figure 2B). To assess how Nrg4 affects diet-induced weight gain,
   we performed metabolic cage studies in HFD-fed control and transgenic
   mice using Comprehensive Laboratory Animal Monitoring System (CLAMS) to
   assess energy balance. Oxygen consumption rate and energy expenditure
   were significantly elevated in Nrg4 transgenic mice during both dark
   and light phases ([94]Figure 2C, D). While food intake was similar in
   the dark phase, it was moderately higher in Tg mice during the light
   phase ([95]Figure 2E). Transgenic mice also exhibited increased
   locomotor activity during the dark phase ([96]Figure 2F). These results
   strongly suggest that transgenic elevation of Nrg4 levels protects mice
   from diet-induced obesity by stimulating fuel oxidation and increasing
   energy expenditure.

Figure 2.

   [97]Figure 2
   [98]Open in a new tab

   Body composition analysis and metabolic cage study. (A) Body weight,
   fat mass, and lean mass in WT (open, n = 8) and Nrg4 Tg (brown, n = 7)
   mice fed HFD for 7 weeks. (B) Plasma leptin concentrations. (C) Oxygen
   consumption rate. Averaged oxygen consumption rate in dark and light
   phases is indicated on the right. (D) Energy expenditure. (E) Food
   intake. (F) Total activity counts. Data represent mean ± sem.
   *p < 0.05, **p < 0.01, ***p < 0.001, WT vs. Tg; two-tailed unpaired
   Student's t-test.

3.3. Nrg4 improves insulin sensitivity by enhancing peripheral glucose
metabolism

   We next determined how Nrg4 regulates whole body glucose metabolism
   using hyperinsulinemic-euglycemic clamp in awake mice after ten weeks
   of high-fat feeding. Consistent with improved glucose homeostasis, Nrg4
   transgenic mice required a higher glucose infusion rate to maintain
   euglycemia during clamps as compared to WT mice ([99]Figure 3A). Whole
   body glucose turnover and glycolysis were significantly increased in Tg
   mice, whereas clamp hepatic glucose production did not differ between
   Tg and WT mice ([100]Figure 3A). These data indicate that higher
   glucose infusion rate in Tg mice is largely due to increased glucose
   metabolism in peripheral organs, but not in liver. Measurements of
   glucose uptake in individual tissues confirmed that insulin-stimulated
   glucose uptake in skeletal muscle was enhanced in Tg mice
   ([101]Figure 3B). In contrast, glucose uptake in white fat and brown
   fat remained largely unaltered by transgenic expression of Nrg4 in
   adipocytes. Previous studies demonstrated that Nrg4/ErbB4 signaling
   attenuates induction of hepatic lipogenesis in response to HFD feeding
   [102][13]. Consistent with this, measurement of ^3H-glucose
   incorporation in liver triglycerides (TAG) revealed that hepatic
   lipogenic activity was significantly reduced in Nrg4 transgenic mice
   ([103]Figure 3C). These results illustrate that Nrg4 exerts a
   pleiotropic effect on pathways contributing to systemic glucose
   homeostasis. Increased glucose uptake and utilization likely plays a
   dominant role in whole body glucose metabolism despite reduced flux of
   carbohydrates through hepatic lipogenic pathway.

Figure 3.

   [104]Figure 3
   [105]Open in a new tab

   Hyperinsulinemic-euglycemic clamp study. (A) Glucose infusion rate,
   whole body glucose turnover, whole body glycolysis, and hepatic glucose
   production during clamp in WT (open, n = 9) and Tg (brown, n = 8) mice
   fed HFD for ten weeks. (B) Insulin-stimulated glucose uptake in
   skeletal muscle, WAT, and BAT. (C) Incorporation of ^3H-labeled glucose
   into liver lipids. Data represent mean ± sem. *p < 0.05, WT vs. Tg;
   two-tailed unpaired Student's t-test.

3.4. Nrg4 induces a catabolic metabolic state in the liver

   Nrg4 transgenic mice developed less severe hepatic steatosis than
   control following high-fat feeding, whereas mice lacking Nrg4 had more
   pronounced hepatic fat accumulation. This phenotype is at least in part
   due to the inhibitory effect of Nrg4/ErbB4 signaling on hepatic
   lipogenesis. However, whether Nrg4 regulates fatty acid oxidation in
   the liver has not been examined. Interestingly, Nrg4 transgenic mice
   exhibited a significantly higher concentration of β-hydroxybutyrate in
   plasma following overnight fasting ([106]Figure 4A), suggesting that
   Nrg4 may promote hepatic fatty acid oxidation and ketone production
   during starvation. Plasma TAG concentration was significantly lower in
   Tg mice than in control mice, while the concentrations of total
   cholesterol and non-esterified fatty acids (NEFA) were similar between
   two groups. We performed Stimulated Raman Scattering (SRS) microscopy,
   a label-free method to visualize intracellular lipid droplets
   [107][19], to assess fat accumulation in control and Tg mice following
   high-fat feeding. We found that hepatocytes from Tg mouse livers had
   smaller lipid droplets, consistent with ameliorated hepatic steatosis
   ([108]Figure 4B). In contrast, liver glycogen content was low following
   overnight fasting and did not exhibit significant difference between
   two groups (data not shown). Co-enzyme A (CoA) is a coenzyme that plays
   a central role in lipid metabolism. Measurements of CoA levels
   indicated that the concentrations of CoA and acetyl-CoA, but not
   succinyl-CoA, were significantly elevated in Tg liver ([109]Figure 4C),
   likely contributing to increased flux of fatty acid β-oxidation and
   ketogenesis. Surprisingly, malonyl-CoA levels were also elevated in
   Nrg4 Tg livers, likely due to substrate accumulation as a result of
   reduced flux through the de novo lipogenic pathway.

Figure 4.

   [110]Figure 4
   [111]Open in a new tab

   Effects of Nrg4 transgenic expression on lipid metabolism. (A) Plasma
   lipid concentrations in HFD-fed WT (open, n = 10) and Tg (brown, n = 9)
   mice following overnight starvation. (B) SRS imaging of liver sections.
   Scale bar = 50 μm. (C) Concentration of CoA metabolites in the liver.
   Data represent mean ± sem. *p < 0.05, WT vs. Tg; two-tailed unpaired
   Student's t-test.

   We next determined whether Nrg4 deficiency impairs hepatic fatty acid
   oxidation in the fasting state. Consistent with transgenic studies,
   plasma concentration of β-hydroxybutyrate was significantly lower in
   Nrg4 knockout (KO) mice following overnight starvation
   ([112]Figure 5A). While plasma TAG and NEFA levels remained similar,
   total cholesterol level was moderately elevated in KO group. The
   differential effects of Nrg4 transgenic overexpression and deficiency
   on plasma TAG and cholesterol levels suggest that distinct compensatory
   mechanisms may be activated in gain- and loss-of-function mouse models.
   SRS microscopy revealed that Nrg4 KO mice developed more severe hepatic
   steatosis following high-fat feeding, as shown by the presence of large
   lipid droplets in hepatocytes ([113]Figure 5B). Accordingly, fatty acid
   profiling of liver lipids indicated that the levels of major fatty acid
   species, including palmitic acid (C16:0), linoleic acid (C18:2n-6), and
   oleic acid (C18:1n-9), were significantly elevated in Nrg4 KO mouse
   liver ([114]Figure 5C). These results suggest that Nrg4 may exert its
   effects on hepatic fat content by reducing lipogenesis and augmenting
   fatty acid oxidation.

Figure 5.

   [115]Figure 5
   [116]Open in a new tab

   Effects of Nrg4 deficiency on lipid metabolism. (A) Plasma lipid
   concentrations in HFD-fed WT (open, n = 9) and Nrg4 KO (blue, n = 8)
   mice following overnight starvation. (B) SRS imaging of liver sections.
   Scale bar = 50 μm. (C) Fatty acid profiling analysis of total liver
   lipids. Data represent mean ± sem. *p < 0.05, ***p < 0.001, WT vs. KO;
   two-tailed unpaired Student's t-test.

3.5. Regulation of adipose tissue gene expression by Nrg4

   White fat plays a central role in the regulation of fuel storage and
   also serves as an important source of endocrine hormones. In addition,
   adipose tissue inflammation has been causally linked to
   obesity-associated metabolic disorders. Previous Nrg4 binding studies
   were inconclusive for adipose tissue due to poor preservation of
   histology on frozen fat sections [117][13]. Despite abundant Nrg4
   expression in adipose tissue, it remains unknown whether Nrg4 may
   directly or indirectly regulate adipose tissue function. To test this,
   we performed gene profiling in eWAT from HFD-fed WT and Tg mice using
   microarray. We identified two clusters of genes that exhibit
   significant downregulation (81 genes) or upregulation (73 genes) in
   response to transgenic expression of Nrg4 ([118]Figure 6A and
   [119]Supplementary Table S1). Pathway analysis of these two clusters
   indicated that genes downregulated by Nrg4 were enriched for those
   involved in immune and inflammatory response. Genes encoding proteins
   predicted to contain immunoglobulin-like domain and secreted
   glycoproteins were also enriched in this cluster. The upregulated
   cluster was enriched for genes involved in mitochondrial function and
   energy metabolism. QPCR analysis indicated that the expression of
   several genes involved in inflammatory signaling, including Ccl4, Tnfα,
   Cd68, Il10, and Emr1, was significantly lower in transgenic eWAT
   ([120]Figure 6B). On the contrary, mRNA expression for
   inflammation-associated genes was elevated in eWAT from HFD-fed Nrg4 KO
   mice ([121]Figure 6C). The expression of adipogenic genes (Pparγ,
   Fabp4) remained similar. These results are consistent with recent
   studies that demonstrate anti-inflammatory activities of Nrg4/ErbB4
   signaling in adipose tissue and colon [122][24], [123][25].

Figure 6.

   [124]Figure 6
   [125]Open in a new tab

   Regulation of eWAT gene expression by Nrg4. (A) Two clusters of genes
   downregulated (I) and upregulated (II) by over 1.6-fold in eWAT from
   Nrg4 Tg mice (top). Enrichment scores of biological processes for the
   genes in two clusters (bottom). (B) qPCR analysis of eWAT gene
   expression in HFD-fed WT (open, n = 10) and Nrg4 transgenic (brown, n =
   9) mice. (C) qPCR analysis of eWAT gene expression in HFD-fed WT (open,
   n = 9) and Nrg4 KO (blue, n = 8) mice. Data represent mean ± sem.
   *p < 0.05, vs. WT; two-tailed unpaired Student's t-test.

3.6. Nrg4 promotes a healthy profile of adipokine secretion

   Adipose tissue releases diverse secreted factors that act locally or
   function as endocrine regulators. To determine whether Nrg4 plays a
   role in modulating the adipocyte secretome, we examined microarray
   expression levels of the genes that are included in the mouse secretome
   list and predicted to encode secreted proteins. We identified a list of
   genes encoding secreted proteins that exhibited transcriptional
   regulation in Nrg4 transgenic and knockout mouse eWAT in the opposite
   manner ([126]Figure 7A). Heat map representation revealed that
   transgenic expression of Nrg4 increased mRNA expression of several
   adipokines that have been demonstrated to exert beneficial metabolic
   effects, including Adipsin (Complement factor d, Cfd), Adiponectin, and
   Vascular endothelial growth factor a (Vegfa). Adipsin has been shown to
   improve β cell function in diabetes and its expression is downregulated
   in obesity [127][26], [128][27]. While adiponectin exerts its
   beneficial effects on diverse tissues including skeletal muscle and the
   liver [129][28], [130][29], VEGFα appears to improve metabolic
   physiology by acting locally to enhance adipose tissue vascularization
   and function [131][30], [132][31]. The expression of genes involved in
   the classical complement activation pathway, including C1qa, C1qb, and
   C1qc, were downregulated in eWAT from Nrg4 transgenic mice. Remarkably,
   Nrg4 deficiency resulted in elevated expression for these factors. C1qa
   expression has been shown to be elevated in visceral adipose tissue
   from obese patients and its deficiency protects mice from HFD-induced
   insulin resistance [133][32], [134][33]. We confirmed the expression of
   adipsin using qPCR ([135]Figure 7B). Further, while plasma adipsin
   levels were similar between chow-fed WT and Tg mice, transgenic
   expression of Nrg4 greatly increased adipsin levels in circulation
   following high-fat feeding ([136]Figure 7C). Similarly, no difference
   was observed between WT and KO mice under chow-fed condition. Nrg4
   deficiency resulted in lower adipsin levels in diet-induced obese mice,
   but not chow-fed lean mice ([137]Figure 7D). These results strongly
   suggest that Nrg4 promotes a healthy adipokine profile that likely
   contributes to the amelioration of obesity-associated metabolic
   disorders.

Figure 7.

   [138]Figure 7
   [139]Open in a new tab

   Regulation of eWAT secretome by Nrg4. (A) Reciprocal regulation of
   secreted factor gene expression in eWAT from Nrg4 transgenic and
   knockout mice. Shown are genes that exhibit an increase or decrease in
   expression by over 1.4-fold based on microarray expression values. (B)
   qPCR analyses of adipsin gene expression in eWAT from a cohort of
   HFD-fed WT (open, n = 10) and Nrg4 Tg (brown, n = 9) mice and a cohort
   of WT (open, n = 9) and Nrg4 KO (blue, n = 8) mice. Data represent
   mean ± sem. *p < 0.05, vs. WT; two-tailed unpaired Student's t-test.
   (C) Immunoblotting of plasma samples from chow or HFD-fed WT and Tg
   mice. (D) Immunoblotting of plasma samples from chow or HFD-fed WT and
   Nrg4 KO mice.

4. Discussion

   Adipose tissue releases a plethora of secreted factors that act locally
   or on distal tissues to influence metabolic physiology. Nrg4 was
   recently identified as a brown fat-enriched endocrine factor that
   ameliorates HFD-induced insulin resistance and hepatic steatosis
   [140][13]. Despite these intriguing findings, the physiological
   mechanisms through which Nrg4 regulates systemic energy balance and
   glucose and lipid metabolism have not been elucidated. In this study,
   we provide the novel demonstration that Nrg4 reduces HFD-induced weight
   gain in mice by enhancing basal metabolic rate and energy expenditure.
   Transgenic activation of Nrg4 signaling resulted in increased whole
   body glucose turnover and glycolysis that was in part due to augmented
   skeletal muscle glucose utilization. We also identified hepatic fatty
   acid β-oxidation and ketogenesis as a new downstream target of Ngr4 in
   hepatocytes, likely contributing to the protective effects of Nrg4 on
   hepatic steatosis. Finally, Nrg4 promotes a beneficial adipokine
   profile in white fat that was associated with reduced expression of
   genes involved in inflammatory response. This study illustrates the
   surprisingly pleiotropic effects of Nrg4 on energy and nutrient
   metabolism ([141]Figure 8).

Figure 8.

   [142]Figure 8
   [143]Open in a new tab

   A model depicting the pleiotropic action of Nrg4 and its role in
   metabolic homeostasis. Nrg4 expression is repressed by pro-inflammatory
   signaling in white adipose tissue. Nrg4 preserves insulin sensitivity
   and ameliorates NAFLD by attenuating hepatic lipogenesis, augmenting
   fuel oxidation and energy expenditure, and maintaining a healthy
   adipokine profile.

   Neuregulin proteins belong to a small family of extracellular ligands
   that includes Nrg1, Nrg2, Nrg3, and Nrg4 [144][34], [145][35],
   [146][36]. Among these, Nrg1, 2, and 3 are highly expressed in the
   central nervous system. Nrg4 is the only member that exhibits high
   levels of expression in brown and white fat. Remarkably, Nrg4 mRNA
   expression in white adipose tissue inversely correlates with adiposity
   in both rodents and humans [147][13]. These observations strongly
   suggest that defective Nrg4 secretion may lead to a hormonal
   insufficiency during obesity that exacerbates the progression of
   metabolic disorders. In support of this, plasma Nrg4 levels were
   inversely associated with the risk for metabolic syndrome in obese
   Chinese adults [148][37]. While adipocyte Nrg4 expression is suppressed
   by pro-inflammatory cytokines, anti-inflammatory agents such as
   rosiglitazone and agonists for GPR120 exert an opposite effect. These
   results illustrate that the balance between pro-inflammatory and
   anti-inflammatory signals is likely a key driver of adipose tissue Nrg4
   expression.

   Because Nrg4 binding appears to be restricted to hepatocytes, the
   pleiotropic effects of Nrg4 on glucose and lipid metabolism came
   somewhat as a surprise. It is likely that these metabolic effects are
   due to a combination of direct Nrg4 action and secondary consequences
   of Nrg4 manipulations. For example, adipose tissue expresses very low
   levels of ErbB3 and ErbB4, two receptors that mediate Nrg4 signaling.
   It is possible that the shift in adipokine profile in response to
   transgenic expression and deficiency of Nrg4 occurs as a result of Nrg4
   signaling in other tissues. The molecular nature of such a regulatory
   circuitry remains unknown.

   Previous studies demonstrated that ErbB receptors are expressed in
   skeletal myocytes and that a fragment corresponding to the EGF-like
   domain of Nrg1 was capable of stimulating glucose uptake in cultured
   C2C12 myotubes [149][38], [150][39], [151][40]. Nrg1 treatment improved
   whole body glucose metabolism in db/db mice and diabetic rats
   [152][41], [153][42]. Whether Nrg4 may act directly on skeletal muscle
   to promote glucose uptake and utilization remains an important question
   to be addressed in future studies. Beyond peripheral tissues, ErbB4 is
   highly expressed in the central nervous system, including the
   hypothalamus [154][43], [155][44], [156][45]. Importantly, ErbB4
   polymorphism was recently shown to strongly associate with body mass
   index in a genome-wide association study [157][46]. Because Nrg4
   transgenic mice exhibited reduced fat mass and elevated energy
   expenditure, our findings raise the possibility that Nrg4/ErbB4
   signaling may be a genetic factor for human obesity and its associated
   metabolic disorders. Given the pleiotropic metabolic benefits elicited
   by Nrg4, biologic therapeutics targeting this pathway may provide an
   effective treatment that simultaneously targets type 2 diabetes and
   NAFLD in patients.

Acknowledgments