Macrophage infiltration plays an important role in obesity-induced insulin resistance. CCAAT enhancer-binding protein-α (C/EBPα) is a transcription factor that is highly expressed in macrophages. To examine the roles of C/EBPα in regulating macrophage functions and energy homeostasis, macrophage-specific C/EBPα knockout (MαKO) mice were created. Chow-fed MαKO mice exhibited higher body fat mass and decreased energy expenditure despite no change in food intake. However, the obese phenotype disappeared after high-fat (HF) diet feeding. Although there was a transient decrease in insulin sensitivity of chow-fed young MαKO mice, systemic insulin sensitivity was protected during HF-feeding due to preserved insulin sensitivity in skeletal muscle. We also found that C/EBPα-deficient macrophages exhibited a blunted response of cytokine-induced expression of M1 and M2 macrophage markers, suggesting that C/EBPα controls both M1 and M2 polarization. Consistent with decreased exercise capacity, mitochondrial respiration rates and signal pathways for fatty acid oxidation were remarkably reduced in the skeletal muscle of chow-fed MαKO mice. Furthermore, expression levels of inflammatory cytokines were reduced in skeletal muscle of HF-fed MαKO mice. Together, these results imply that C/EBPα is required for macrophage activation, which plays an important role in maintaining skeletal muscle energy metabolism.
- CCAAT/enhancer binding protein
- insulin sensitivity
macrophages are resident immune cells found in nearly all tissues (23). Obesity increases macrophage infiltration in white adipose tissue (WAT) and other tissues (14, 17, 22, 29, 30). In addition, macrophages recruited under conditions of obesity appear to be more inflammatory (M1 macrophages), and this leads to reduced insulin sensitivity and glucose metabolism (16, 26, 29). However, under physiological conditions, tissue macrophages are also important in maintaining metabolic homeostasis and tissue structure as they appear to play an anti-inflammatory role (M2 macrophages) (25, 27). For example, macrophage PPARγ deficiency leads to impaired insulin sensitivity in skeletal muscle and liver of chow-fed mice (13). Therefore, macrophages may play distinct roles between lean and obese conditions, either maintaining or impairing metabolism. However, the mechanism underlying the shift of macrophage functions between lean and obese conditions is unknown.
C/EBPα is the prototypical member of the C/EBP transcription factor family, which comprises six members (α-ζ). C/EBP proteins are widely expressed and regulate a variety of cellular and physiological processes, including energy metabolism, immunity, inflammation, hematopoiesis, and adipogenesis (3, 4, 7, 10, 35). C/EBPα is highly expressed in early myeloid progenitors and may regulate monocyte and macrophage development (10–12, 40). However, the role of C/EBPα in macrophage polarization and its effects on metabolism have not been comprehensively examined. Here, using a conditional gene deletion approach, we examined the regulatory effects of C/EBPα on macrophage functions and the role of macrophage C/EBPα in maintaining energy homeostasis and insulin sensitivity in both lean and obese mice.
RESEARCH DESIGN AND METHODS
To delete C/EBPα specifically in macrophages (MαKO), C/EBPα floxed mice were crossed with lysozyme M promoter-directed CRE mice (both in C57BL/6J background and from the Jackson Laboratory, Bar Harbor, ME). Male MαKO and control mice (2–7 mo old) were used for this study and fed either a standard chow diet (Harlan Laboratories, Placentia, CA) or a HF diet (60%, calorie based; Research Diet, New Brunswick, NJ). A HF diet was provided to 3-mo-old mice, and metabolic assays were performed 3 mo later. All mice were maintained under standardized conditions with 12:12-h light/dark cycles, and the procedures conformed to the Association for Assessment and Accreditation of Laboratory Animal Care guidelines with approval of the University of California San Diego Animal Care and Use Committee.
Metabolic data analysis.
Hyperinsulinemic euglycemic clamp studies.
As previously described (21, 31), two cannulas were inserted into the right jugular vein of each mouse under single-dose anesthesia. The cannulas were tunneled subcutaneously and exteriorized at the back of the neck. After 4–5 days of recovery from surgery, the euglycemic hyperinsulinemic clamp experiments began with a constant infusion [5 μCi/h of d-[3-3H]glucose (DuPont NEN, Boston, MA)]. After 90 min of tracer equilibration, 50% glucose (variable infusion rate; Abbott Laboratories, Abbott Park, IL) and tracer (5 μCi/h) plus insulin (8 mU·kg−1·min−1) were infused into the jugular vein. Small volumes of blood were sampled from the tail vein at 10-min intervals for immediate glucose analysis. Blood samples at t = −10, 0 (basal), 110, and 120 min (end of experiment) were used to determine the specific activity of glucose and the levels of insulin. Blood glucose levels were measured every 10 min, and glucose infusion rate was adjusted accordingly to achieve steady-state glucose levels (120 ± 5 mg/dl) for at least 20 min toward the end of the clamp. All blood samples were centrifuged immediately after collection, and plasma was stored at −80°C for subsequent analyses. Hepatic glucose production (HGP) and glucose disappearance rate (GDR) were calculated in the basal state and during the steady-state phase of the clamp. Tracer-determined rates were quantified using the Steele equation for steady states (36). At steady state, GDR is equal to the rate of endogenous HGPs plus (cold) glucose infusion rate. The insulin-stimulated GDR is equal to the total GDR minus the basal glucose turnover rate.
Biochemical responses to insulin stimulation were evaluated in mice fasted for 6 h prior to anesthesia. After ligation of vessels supplying one side of the leg muscles, one lobe of the liver, and one epididymal fat pad, basal samples of these tissues were harvested. Five minutes after an insulin bolus (1 U/kg injected into the inferior vena cava), the remaining liver, muscle, and fat were harvested and snap-frozen to measure signal transduction markers.
Mitochondrial respiration rate was measured using the Mitocell-MT200 (Strathkelvin Instruments, North Lanarkshire, Scotland) (9). Intact mitochondria (80 μg proteins) isolated from tissues were incubated at 37°C in a chamber equipped with an oxygen electrode. After measurement of basal (state 1) respiration using the endogenous substrates, state 3 respiration and complex 1 activity were measured by adding ADP (100 μM) and the complex 1 substrates glutamate (5 mM) and pyruvate (2.5 mM).
Male mice (12 wk old or 6–7 mo old) were injected intraperitoneally with 3% thioglycollate (41). Peritoneal fluid was collected 3 days later using a 10-ml syringe, and macrophages were cultured in tissue culture plates.
Gene and protein expression analyses.
Total RNA was prepared from epididymal fat, interscapular brown fat (iBAT), liver, and skeletal muscle homogenized in TRIzol (Invitrogen) and converted to cDNA using SuperScript III Reverse Transcriptase and oligo(dT) 12–18 primer (Invitrogen). Real-time PCR was performed using an Mx3000P system (Stratagene) and SYBR Green dye (Molecular Probes, Eugene, OR). For protein and phosphorylation level measurements, homogenized protein samples were separated using NuPAGE gels (Invitrogen) and immunoblotted using the indicated antibodies. Protein was visualized by ECL, and the images were semiquantified using Quantity One software (Bio-Rad).
Data are expressed as means ± SE. The Prism software (GraphPad Software) was used for statistical analysis. Student's t-tests (two-tailed) were used for pairwise comparisons. ANOVA followed by Bonferroni post hoc tests was used for glucose and insulin tolerance tests. Differences were considered significant at P < 0.05.
Generation of macrophage-specific C/EBPα-deficient mice.
The Cre-loxP approach was employed. C/EBPα allele-floxed mice (C/EBPαf/f) were crossed with lysozyme M promoter-directed CRE transgene mice (LysMCre). To determine the specificity and efficiency of knockout in macrophages, C/EBPα proteins were measured in multiple tissues from C/EBPαf/f-LysMCre and C/EBPα+/+-LysMCre control mice. C/EBPα protein was almost undetectable in peritoneal macrophages from C/EBPαf/f-LysMCre mice. Since C/EBPα and C/EBPβ can bind to the same regulatory region of some target genes (7), we examined C/EBPβ protein levels in macrophages to verify whether C/EBPα deletion induces a compensational upregulation of C/EBPβ. Figure 1 shows that C/EBPβ protein levels in peritoneal macrophages from C/EBPαf/f-LysMCre mice were similar to those of the control mice (Fig. 1). Although this result cannot completely rule out the possibility of functional compensation of C/EBPβ, it indicates that C/EBPα deletion did not increase C/EBPβ protein expression in macrophages. Furthermore, there were no significant changes in C/EBPα protein levels in other tissues such as adipose tissue, skeletal muscle, and liver (data not shown). These results indicate that C/EBPα expression was deficient in macrophages of C/EBPαf/f-LysMCre mice (MαKO).
Macrophage C/EBPα gene knockout increased adiposity and reduced energy expenditure.
To examine the effects of macrophage-specific C/EBPα deficiency on energy homeostasis, we compared body composition between chow-fed MαKO and control mice from weaning to adulthood. Although no differences in body weight and body composition were observed between the two genotypes through early adulthood (from weaning to 3 mo; data not shown), MαKO mice began to demonstrate significantly higher body weight and body fat mass by 6–7 mo of age (Fig. 2, A–D). These results suggest that macrophage C/EBPα gene deficiency alters energy homeostasis and increases adiposity in adulthood.
To study the underlying mechanism of increased adiposity of MαKO mice, we measured food intake and found no difference between MαKO and control mice (Fig. 2E). In addition, the two groups demonstrated similar reductions in food intake and body weight when injected with leptin (data not shown), arguing against a role of central leptin resistance in increasing adiposity in MαKO mice. Using indirect calorimetry, we found a significant reduction in energy expenditure in MαKO mice during both light and dark cycles (Fig. 2F). Together, these data indicate that decreased energy expenditure is a key mechanism that underlies the increased adiposity of MαKO mice.
Chow-fed MαKO mice had transient decrease in insulin sensitivity at early life.
To investigate the effects of macrophage C/EBPα gene knockout on glucose metabolism, we performed glucose and insulin tolerance tests at different ages. We found significantly elevated fasting glucose concentration with impaired glucose and insulin tolerance in 2-mo-old MαKO mice (Fig. 3, A and B). Interestingly, both glucose and insulin tolerance were normalized when these mice were 3 mo or older (Fig. 3, C and D). These results suggest that macrophage C/EBPα deletion induces a transient insulin resistance in the early adulthood of mice.
MαKO mice were protected from diet-induced insulin resistance partly due to preserved insulin sensitivity in skeletal muscle.
Obesity increases macrophage infiltration in peripheral tissues that is closely associated with insulin resistance and defects in energy metabolism. The above studies revealed a transient decrease in insulin sensitivity in chow-fed MαKO mice. We challenged MαKO mice with prolonged HF feeding to further examine the role of macrophage C/EBPα in insulin sensitivity. Similar to most HF feeding studies of C57BL/6 mice, there was significant decrease in insulin sensitivity 3 mo after HF feeding of both genotypes (data not shown). To our surprise, HF-fed MαKO mice are more insulin sensitive than control mice, evidenced by lower fasting insulin levels and lower glucose levels during an insulin challenge test (Fig. 3, E and F), although there was no difference in body weight and body composition between the two genotypes (Fig. 3, G–J).
To further examine insulin sensitivity, hyperinsulinemic euglycemic clamps were carried out in HF-fed mice. We found that glucose infusion and GDRs were significantly higher in MαKO mice than in control mice (Fig. 4, A and C). Together, these results indicate that macrophage C/EBPα gene deficiency partially protects mice from HF feeding-induced insulin resistance. The clamp assay also showed that HGP rates were similar between MαKO and control groups in both basal state and after suppression by insulin (Fig. 4, D–F). However, insulin-stimulated GDRs were significantly higher in MαKO mice (Fig. 4B). Since skeletal muscle is responsible for 70–80% of insulin-stimulated glucose disposal (21), these results suggest that insulin sensitivity in skeletal muscle of MαKO mice may be protected in a HF-fed condition. In support of this, the skeletal muscle of MαKO mice exhibited significantly higher phosphorylation levels of Akt and GSK3, two key kinases in insulin signaling (Fig. 4, G and J). However, the phosphorylation levels of Akt were similar in WAT and liver of MαKO and control mice (Fig. 4, H, I, K, and L). These studies indicate that C/EBPα gene deficiency in macrophages protects mice from HF-induced insulin resistance, at least in part, by preserving insulin sensitivity in skeletal muscle.
C/EBPα knockout impaired both M1 and M2 macrophage polarization.
Macrophages can be polarized into two distinct forms depending on stimuli and physiological conditions. M1 macrophages are inflammatory and enriched in HF-induced obese states, whereas M2 macrophages are anti-inflammatory and enriched in lean states (1, 2, 5, 26). Both the number of infiltrated macrophages and their polarizing state are closely associated with systemic energy metabolism and insulin sensitivity (24, 27–29). To investigate whether the metabolic phenotypes of MαKO mice are attributable to macrophage polarization, we examined M1 and M2 markers in macrophages isolated from peritoneal exudates. The expression levels of M1 markers such as TNFα, IL-6, and CD11-c were similar between MαKO and control mice (data not shown). However, LPS-induced M1 polarization as measured by TNFα and IL-6 mRNA levels was significantly attenuated in MαKO macrophages (Fig. 5A), indicating that C/EBPα deficiency impairs M1 macrophage activation. On the other hand, Fig. 5B shows that the mRNA levels of multiple M2 markers were remarkably lower in C/EBPα-deficient macrophages in the basal state. IL-4 has been shown to promote M2 polarization (1). We also confirmed that IL-4 stimulation increased M2 macrophage marker expression in macrophages isolated from wild-type mice (data not shown). After treatment with IL-4, C/EBPα-deficient macrophages showed decreased mRNA levels of M2 markers (such as Arg1, Mgl1, Chi3l3, Mrc1, and IL-10) (Fig. 5C), suggesting that C/EBPα deficiency impairs M2 macrophage activation. Furthermore, adenovirus-mediated C/EBPα overexpression increased the M1 marker TNFα expression and M2 marker Arg1 expression in RAW macrophages cells (Fig. 5D). These results further confirm that C/EBPα is important for both M1 and M2 macrophage activation. However, the gene expression levels of general macrophage markers such as F4/80 and CD68 and inflammatory macrophage marker CD11-c were not changed in WAT, iBAT, and skeletal muscle in MαKO mice (Fig. 5, E–G). In addition, the mRNA expression levels of inflammatory (IL-1β, TNFα, IL-6, and CCL2) and anti-inflammatory (IL-10) markers were not notably changed in WAT (Fig. 5E). Together, our data suggest that C/EBPα deficiency may impair macrophage polarization without significant effects on tissue macrophage infiltration.
Effects of macrophage C/EBPα deficiency on inflammatory cytokines in circulation and in muscle.
We measured serum cytokines to investigate potential effects of macrophage C/EBPα knockout. Interestingly, inflammatory cytokines such as TNFα and IL-1β were elevated in the serum of chow-fed MαKO mice (Fig. 6, A and E), whereas their IL-6 and IL-10 levels were similar to control (Fig. 6, C and G). The increases in TNFα and IL-1β are in line with temporally reduced insulin sensitivity in chow-fed MαKO mice. On the other hand, serum cytokine levels were similar between the two groups after HF diet feeding (Fig. 6, B, D, F, and H). However, the gene expression levels of inflammatory cytokines such as IL-1β and CCL2 were significantly reduced in skeletal muscle (Fig. 6I), but not in WAT (Fig. 6J) and liver (Fig. 6K) of HF-fed MαKO mice, suggesting that macrophage C/EBPα deficiency may ameliorate HF-induced inflammation in skeletal muscle.
Decreased mitochondria respiration and oxidative gene expression in skeletal muscle of MαKO mice.
Current studies show that macrophage C/EBPα knockout reduces energy expenditure in adult mice. To further study the underlying mechanisms for the decreased energy expenditure, we compared the expression of key metabolic genes in selected tissues between MαKO and control mice.
Due to the high energy dissipation rate, BAT may play an important role in maintaining energy homeostasis. Our study found that the iBAT from MαKO mice showed no significant changes in tissue mass, histology (data not shown), or mRNA levels of key transcription factors and genes related to fatty acid (FA) oxidation (Fig. 7A). Protein levels of UCP1, phospho-PGC-1α, and cytochrome c in iBAT were comparable between iBAT of MαKO and control mice (Fig. 7B). These results suggest that the reduction of energy expenditure in MαKO mice is most likely not due to changes in iBAT functions.
The liver and WAT are also important in energy metabolism. Using the same approach, we compared MαKO and control mice for the expression of genes closely relate to FA oxidation (including PPARα, CPT1a, and CPT1b) and lipogenesis (including SREBP1c, SCD1, PPARγ, FAS, and ACC1) in liver and epididymal fat. No significant differences were detected (data not shown).
Skeletal muscle accounts for ∼40% of the total body mass and 20–30% of overall oxygen consumption at rest (38, 42). In addition, UCP3-mediated thermogenesis in skeletal muscle has been shown to contribute to energy expenditure (6, 20, 38). These data indicate a great potential of skeletal muscle in maintaining energy homeostasis. By comparing mRNA levels, we found several remarkable alternations in gene expression profile in skeletal muscle of MαKO mice (Fig. 8A). The expression levels of genes related to FA oxidation, such as PPARα, PPARδ, and CPT1α, were significantly decreased in MαKO skeletal muscle (Fig. 8A). In addition, the expression of genes related to energy consumption, such as PGC-1β and UCP3, was also decreased in MαKO skeletal muscle (Fig. 8A). Furthermore, the expression levels of markers of slow oxidative type I fiber (MHC1) and fast oxidative type II fiber (MHC2a) were significantly reduced (Fig. 8A) without changes in markers of fast glycolytic fiber (MHC2b and MHC2x) (Fig. 8A). At the protein level, phosphorylation levels of AMPK and ACC, key factors for FA oxidation in skeletal muscle (15), were decreased (Fig. 8B). Collectively, these data suggest that FA oxidation in the skeletal muscle was impaired in MαKO mice.
Muscle oxidative capacity is essential for exercise capacity (19, 39). To further examine whether the defect in FA oxidation in MαKO muscle affects muscle function, we used a treadmill to assess exercise capacity. We found that the treadmill running time was significantly shorter in MαKO mice than in the control group (Fig. 8C), indicating that macrophage-specific C/EBPα knockout decreases muscle exercise capacity.
Mitochondria play predominant roles in energy metabolism by producing ATP using various fuel sources and by dissipating chemical energy through thermogenesis (34). To examine mechanisms for reduced FA oxidation capacity, we isolated mitochondria from skeletal muscle and measured their respiration rates. We found that MαKO mice exhibited significant decreases in mitochondria respiration rate both in the basal levels (state 1 respiration) and after adding the complex I substrates glutamate and pyruvate (Fig. 9, A and B), indicative of mitochondrial dysfunction. Given the importance of skeletal muscle in energy expenditure (6, 20, 38), we postulated that the reduction in FA oxidation in MαKO mice might be due to decreased mitochondrial function in skeletal muscle.
Macrophages are versatile immune cells that exhibit distinct responses to different local and systemic stimuli. The increase in proinflammatory macrophage infiltration in WAT and other tissues has been suggested as a key mechanism underlying obesity-induced insulin resistance (18, 24, 29, 30). C/EBPα is highly expressed in macrophages (12). To understand the role of C/EBPα in macrophage activation and its effects on systemic energy homeostasis, we studied a metabolic phenotype of macrophage-specific C/EBPα knockout mice. Our results indicate that C/EBPα may regulate M1 macrophage activation as well as M2 macrophage polarization. When the mice were fed a chow diet, C/EBPα deficiency in macrophages increased adipose tissue mass through reducing energy expenditure, which may be associated with impaired oxidation capacity in skeletal muscle. However, macrophage-specific C/EBPα deletion appears to protect mice from HF-induced insulin resistance at least partially by reducing inflammatory gene expression in skeletal muscle. Therefore, our study demonstrates that C/EBPα plays a unique role in regulating macrophage functions. It also suggests that macrophages may play an active role in regulating energy metabolism.
Macrophage polarization is regulated at the transcriptional level in that PPARγ and PPARδ stimulate the polarization toward the M2 anti-inflammatory phenotype and improve insulin resistance through the effects on peripheral tissues such as liver and skeletal muscle (27, 28). Our study shows that C/EBPα deficiency in macrophages downregulates both inflammatory and anti-inflammtory macrophage markers (Fig. 5), indicating that C/EBPα regulates both M1 and M2 activation through a mechanism(s) that remains to be investigated.
In contrast to the effect on insulin signaling, the regulatory effects of macrophages on energy homeostasis have not been fully investigated. Macrophage-specific PPARγ deficiency predisposes the mice to HF diet-induced obesity (27). PPARδ deficiency impairs M2 activation of Kupffer cells, which is related to increased fat mass in HF-fed mice (28). Our study showed that macrophage C/EBPα deficiency increased body fat in a chow-fed condition (Fig. 2, A–C), revealing an important role of macrophages in energy homeostasis under physiological conditions.
Positive energy balance is the hallmark of fat accumulation. Consistent with increased body fat, MαKO mice exhibited significantly reduced energy expenditure with no changes in central leptin sensitivity or food intake. These results led us to propose that decreased energy expenditure induces the obese phenotype of MαKO mice. By measuring the expression levels of UCP1 and other key metabolic genes in iBAT (Fig. 7), our study excludes BAT in causing obesity in MαKO mice. Due to its bulk and capacity to consume energy, skeletal muscle is another key tissue that controls energy expenditure (38, 42). Our study found that skeletal muscle of MαKO mice had significantly reduced expression of key regulators of FA oxidation (Fig. 8, A and B). Furthermore, exercise capacity was also decreased in MαKO mice (Fig. 8C). Therefore, we postulate that decreased oxidative metabolism in skeletal muscle may contribute to the reduced energy expenditure in MαKO mice. Further studies will be necessary to examine whether reduced exercise capacity is caused from fiber type changes in skeletal muscle of MαKO mice.
Both obesity and obesity-induced insulin resistance are associated with mitochondria dysfunction (32, 33, 37). Obesity increases macrophage infiltration in various tissues (29). However, the effects of macrophages on the mitochondrial function of surrounding cells are still unclear. Our study found that macrophage-specific C/EBPα deficiency impaired mitochondrial respiration in skeletal muscle (Fig. 9), suggesting a regulatory effect of macrophages on energy metabolism of surrounding cells. In support of this conclusion, M2 macrophages have been shown to regulate FA metabolism in peripheral tissues, including skeletal muscle and liver (27, 28). Our results suggest that the mechanism by which C/EBPα-null macrophages regulate mitochondrial functions in skeletal muscle involves reduced expression of transcription factors PPARα, PPARδ, and PGC-1β (Fig. 8A) and increased inflammatory cytokines in circulation (Fig. 6, A and E).
Another intriguing observation of this study is that macrophage C/EBPα deletion impairs metabolism predominantly in skeletal muscle. We found that macrophage C/EBPα deficiency reduced the expression of genes related to FA oxidation and energy expenditure in skeletal muscle (Fig. 8), but not in iBAT and WAT (Fig. 7 and data not shown).
Little is known about the role of macrophages in muscle energy metabolism. Although more studies are necessary for better understanding of macrophage-to-muscle cross talk, we speculate that skeletal muscle-specific effects of macrophages on FA and energy metabolism may be achieved by total number of macrophages infiltrated or distinct activation state of macrophages or macrophage-secretory factors in skeletal muscle. However, considering similar macrophage marker expression in WAT, BAT, and skeletal muscle between MαKO and control mice (Fig. 5, E–G), it is less likely that the differences in total number of macrophages infiltrated make the specific effects on skeletal muscle. It may be possible that macrophages in different tissues may have their own unique characteristics to regulate tissue metabolism such as Kupffer cells and microglia. Therefore, some functions of macrophages in skeletal muscle may be dependent on C/EBPα.
It is well known that macrophage-mediated inflammation impairs insulin signaling in the obese state (26, 29) and that this effect depends on the polarization state of the infiltrated macrophages (25, 26). Our study showed that macrophage C/EBPα deficiency reduces insulin sensitivity in young mice while also protecting mice from HF-induced insulin resistance. Although further mechanistic studies are required, our data suggest that the impaired macrophage polarization in MαKO mice may contribute to these different phenotypes of insulin sensitivity between lean and obese conditions. In the lean condition, macrophages are predominantly in the anti-inflammatory M2 state, which promotes insulin sensitivity (26–28). Our in vitro study revealed that C/EBPα is required for IL-4 to induce M2 macrophage polarization (Fig. 5C). Therefore, C/EBPα deficiency may impair anti-inflammatory M2 macrophage activation in the lean condition, which could lead to insulin resistance in chow-fed young MαKO mice. In contrast, the obese state is enriched with inflammatory M1 macrophages in tissues (25, 26). The impaired response to LPS-induced M1 marker expression in C/EBPα deficient macrophages (Fig. 5A) prompted us to speculate on impaired HF feeding-induced M1 macrophage activation in skeletal muscle. In support of this, the expression levels of inflammatory genes such as IL-1β and CCL2 were significantly decreased in skeletal muscle of MαKO mice (Fig. 6I). The reduction in M1 macrophage activation and expression of inflammatory cytokines may protect MαKO mice from HF-induced insulin resistance. Similarly, a recent study reported that a distinct population of inflammatory macrophages infiltrates into skeletal muscle in HF-induced obese mice, which is linked to systemic insulin resistance (8). On the other hand, skeletal muscle from CCL2 KO mice on HF diet did not gain inflammatory macrophages and maintained insulin sensitivity (8). Therefore, inhibiting inflammatory macrophage activation in skeletal muscle may be beneficial for improving systemic insulin sensitivity in obese conditions.
All our results taken together, our study demonstrates that C/EBPα plays an important role in macrophage activation. Our study also indicates that macrophage activation has a significant impact on energy metabolism of surrounding cells in addition to insulin signaling. Further studies are required to investigate the underlying mechanisms of macrophage-controlled energy metabolism and its relationship with obesity.
This work is supported by National Institutes of Health Grants DK-080418 (J. Shao), HD-069634 (J. Shao), DK-095132 (J. Shao), R01 DK-075916 (G.-S. Feng), and US Dept. of Veterans Affair Grant I01-BX000702 (N.-W. Chi).
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: B.L., L.Q., J.M.O., and J. Shao conception and design of research; B.L., L.Q., M.L., H.s.Y., and W.W.C. performed experiments; B.L., N.-W.C., and J. Shao analyzed data; B.L., L.Q., R.M., J. Schaack, G.-S.F., N.-W.C., J.M.O., and J. Shao interpreted results of experiments; B.L. prepared figures; B.L. and J. Shao drafted manuscript; B.L. and J. Shao edited and revised manuscript; J. Shao approved final version of manuscript.
We thank Dr. Simon Schenk (UC San Diego) for helpful suggestions and technical support of the endurance test. Dr. Jianhua Shao is the guarantor of this work.