Toll-like Receptor-2 Is Essential for the Development of Palmitate-induced Insulin Resistance in Myotubes
-
首席医学网
2008年08月15日 02:01:15 Friday
-
作者：Joseph J. Senn 作者单位：Department of Pediatrics, Charles P. Darby Children‘s Research Institute, Medical University of South Carolina, Charleston, South Carolina 29425
加入收藏夹
【摘要】 Fatty acids can activate proinflammatory pathways leading to the development of insulin resistance, but the mechanism is undiscovered. Toll like receptor 2 (TLR2) recognizes lipids, activates proinflammatory pathways, and is genetically associated with inflammatory diseases. This study aimed to examine the role of TLR2 in palmitate-induced insulin resistance in C2C12 myotubes. Treatment with palmitate rapidly induced the association of myeloid differentiation factor 88 (MyD88) with the TLR2 receptor, activated the stress-linked kinases p38, JNK, and protein kinase C, induced degradation of IB, and increased NF-B DNA binding. The activation of these pathways by palmitate was sensitive and temporally regulated and occurred within the upper physiologic range of saturated fatty acid concentrations in vivo, suggesting a receptor-mediated event and not simple lipotoxicity. When compared with an equimolar concentration of palmitate, fibroblast-stimulating lipopeptide-1, a known TLR2 ligand, was a slightly more potent activator of signal transduction and interleukin (IL)-6 production. Palmitate inhibited insulin signal transduction in C2C12 cells beginning 1-2 h after exposure and reached a maximum at 12-16 h. An antagonist TLR2 antibody, mAb 2.5, led to a 50-60% decrease in palmitate-induced IL-6 production and partially restored insulin signal transduction, whereas an isotype-matched control antibody had no effect. RNA interference-mediated inhibition of TLR2 and MyD88 expression in C2C12 muscle cells resulted in a near complete inhibition of palmitate-induced insulin resistance and IL-6 production. This study provides strong evidence that TLR2 mediates the initial events of fatty acid-induced insulin resistance in muscle.
【关键词】 Tolllike Receptor Essential Development Palmitateinduced Resistance Myotubes
INTRODUCTION
Insulin resistance is recognized as an important component of the metabolic syndrome and type 2 diabetes (1). The prevalence of this condition is increasing rapidly, and it is thought that the spread of the Western diet leading to obesity may be the cause. The mechanism(s) by which obesity couples to insulin resistance has only recently begun to be understood. Free fatty acids and triglycerides are elevated in obese individuals and animals, and they play a critical role in the etiology of the disease (2). Much experimental evidence points to free fatty acids being a primary trigger of insulin resistance in both animal models and humans (3).
Non-esterified fatty acids (NEFAs)2 have been shown experimentally to directly affect insulin sensitivity in adipocytes, muscle, hepatocytes, pancreatic islets, and other non-traditional insulin-responsive cell types (for review, see Ref. 4). NEFA exposure leads to the activation of intracellular signaling via stress-related kinases such as p38, JNK, atypical PKCs, and others. Exposure to elevated NEFAs also caused an elevation in reactive oxygen species, activation of proinflammatory pathways such as the transcription factor NF-B, the production of cytokines, and mitochondrial dysfunction. All of these observations have been correlated with the effects on the insulin receptor signal transduction pathway, leading to decreased activation of signal transduction events (tyrosine phosphorylation of insulin receptor substrate 1 or phosphorylation of AKT) and Glut4 translocation, resulting in inhibited glucose uptake and utilization. However, many questions remained unanswered in this area.
Recent research has focused on the role of the inflammatory signals observed during the development of insulin resistance (5). As briefly mentioned before, NEFAs have been observed to activate inflammatory signal transduction and the production of cytokines in multiple cell types and have also been observed to be elevated before the development of the observed inflammatory signals in animals and humans (5-8). Although several mechanisms have previously been proposed for NEFA-induced inflammatory signals, new research in the field on innate immunity has suggested that saturated fatty acids may serve as a ligand for several members of the Toll-like receptor family (9, 10).
There are at least 11 members of the Toll-like receptor family in humans and mice (11). Toll-like receptors are characterized by an extracellular ligand binding domain consisting of multiple leucine-rich repeats, a single transmembrane domain, and an intracellular domain that contains a highly conserved TollInteracting region (12). Upon ligand binding, the TLR receptor subunits associate, leading to the formation of a complex of Toll-interacting region domain containing adaptor proteins of the MyD88 family. Subsequent downstream signal transduction events lead to the activation of multiple members of the mitogen-activated protein kinase family, in some cell types atypical PKCs and eventually the activation of transcription factors including NF-B, AP-1, and interferon regulatory factor, leading to the transcription of multiple proinflammatory cytokines (13).
The TLRs recognize a diverse repertoire of ligands. TLRs 2 and 4 are hypothesized to recognize components of the bacterial cell wall such as lipopolysaccharide and peptidoglycans and lipopeptides, but also interact with a large number of other lipid containing molecules such as oxidized low density lipopeptide and endogenous proteins like heat shock protein 60 (for review, see Ref. 14). The critical component of lipopolysaccharide involved in stimulating activation of TLRs is the lipid A subunit, which is composed almost entirely of long-chain fatty acids. The specific fatty acid composition (triacyl or diacyl) appears to determine which receptor (TLR-2 or -4) is activated (15). The activation of these receptors on cells of the innate immune system leads to the production of cytokines, chemokines, and the up-regulation of cell surface molecules.
Toll-like receptors are expressed in multiple tissues (16). The predominant site of TLR expression is on cells of the innate immune system, especially macrophage/monocyte cells (for review, see Ref. 17). However, TLRs, particularly TLR2 and -4 are found on a large number of other cells including adipose, muscle, and liver cells where they can activate signal transduction (18-20). Interestingly, recent studies have shown a significantly higher frequency of polymorphisms in the TLR2 gene that seem to correlate to populations at higher risk of insulin resistance and diabetic-related disease (21). The data suggest a novel hypothesis that TLRs, present in insulin-sensitive tissues and macrophage/monocyte cells, are activated by saturated NEFAs and induce proinflammatory pathways leading to NEFA-induced insulin resistance.
This study investigates the ability of palmitate, a saturated fatty acid, to activate the TLR2 receptor pathway in an insulinsensitive model, the C2C12 mouse myotube, and examines the contribution of this receptor pathway to the development of palmitate-induced insulin resistance in these cells. The results suggest that the TLR2 receptor mediates palmitate-induced insulin resistance in this model. These findings provide a rational basis for the development of inflammation and insulin resistance in response to elevations in fatty acids and a novel pharmacologic target to modulate the development of insulin resistance.
MATERIALS AND METHODS
Reagents-C2C12 myoblasts were obtained from the ATCC (Manassas, VA). All chemicals were obtained from SigmaAldrich. Antibodies to TLR2, monoclonal antibody 2.5, and MyD88 were obtained from Cell Sciences (Canton, MA). Antibodies to Ser(P)-473 AKT, phosphorylated extracellular signalregulated kinase 1/2, phospho PKC, p38, JNK, IB, and E4F2a were purchased from Cell Signaling Technology (Beverly, MA). Palmitate in the free acid form was obtained from Nu-Chek Prep (Elysian, MN). Insulin receptor antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Total PKC antibody was purchased from Upstate Biotech (Lake Placid, NY). Low endotoxin-fatty acid-free bovine serum albumin was purchased from Sigma. ELISA kits for mouse IL-6 and tumor necrosis factor- were purchased from R&D Systems (Minneapolis, MN). NF-B ELISA kit was purchased from Active Motif (San Diego, CA). siRNA to TLR2 and MyD88 and control siRNA to glyseraldehyde-3-phosphate dehydrogenase (Control siRNA) were purchased as predesigned siRNA from Ambion (Austin, TX). FuGENE 6 transfection reagent was purchased from Roche Diagnostics.
Preparation of Palmitate for Use in Experiments-The free fatty acid form of palmitate was dissolved in 100% ethanol to a final stock solution concentration of 75 mM. This solution was then added directly to media in subsequent experiments at the indicated doses by serially diluting this stock in ethanol as a vehicle, and then 10 µl of each solution (to maintain an exact concentration of ethanol) was directly added to cell culture media. Ethanol (vehicle) was added to all control samples. Treatments of cells were consistently made in 1 ml of cell culture media with the addition of 10 µl of palmitate/EtOH (1:100 dilution). Palmitate was used at final concentrations of 0.75 mM to 0.75 nM.
Cell Culture and Differentiation-C2C12 cells were maintained in DMEM (high glucose, glutamine, and pyruvate) with the addition of 10% fetal bovine serum (Invitrogen) for growth at 370 ina5%CO2, humidified atmosphere. C2C12 cells were differentiated into mature myotubes by allowing them to reach confluence and then switching to differentiation media (DMEM, 10% horse serum) for 4 days for use in subsequent experiments.
Induction of Insulin Resistance with Palmitate-Differentiated C2C12 myotubes were placed in DMEM containing 1% fetal bovine serum with the addition of 0.75 µM palmitate for 0-24 h. Cells were serum-starved in DMEM containing 2% NEFA-free BSA plus 0.75 µM NEFA for 3 h before treatment with insulin or control medium. All cells were, including controls, were cultured in the presence of NEFA-free BSA.
Treatment with Monoclonal Antibody TLR 2.5 or Nonspecific IgG-Mature myotubes were treated as described above except that at the time of serum starvation 10 µg/ml concentrations of either nonspecific mouse IgG2a or monoclonal antibody TLR 2.5 were added to the culture media. The cells were then incubated for 1 h at 37°C before exposure to palmitate.
siRNA Knockdown of TLR2 and MyD88-C2C12 myoblasts were transfected with 50 nM siRNA to either TLR2 or MyD88 purchased from Ambion using FuGENE 6 transfection reagent as described by the manufacturer. The treated cells were immediately subjected to differentiation into mature myotubes as described above. Cells were then subjected to the treatments described under "Results" and in the legend to Fig. 5 and analyzed for TLR2 and MyD88 expression by Western blot.
Western Blot and Immunoprecipitation-Mature myotubes were treated with palmitate as described above. At each time point cells were washed 2 times with ice-cold phosphate-buffered saline then lysed in cell lysis buffer (UpstateUpstate Biotechnology, Lake Placid, NY) containing 100 mM phenylmethylsulfonyl fluoride. The lysates was cleared of debris by centrifugation, the protein was normalized using the Bradford assay (22), and normalized samples were utilized for immunoprecipitation or Western blot. Samples were immunoprecipitated with antiTLR2 antibody by using the Pierce seize primary mammalian immunoprecipitation kit. Lysates and immunoprecipitates were placed in Laemmli sample buffer, boiled for 5 min, and separated via SDS-PAGE. The resulting Western blots were probed using the appropriate antibodies at 4 °C overnight then imaged by conjugation with horseradish peroxidase-linked secondary antibody and ECL detection reagent (Amersham Biosciences). All experiments were performed in triplicate with similar results.
IL-6 and Tumor Necrosis Factor- ELISA-Cell culture supernatants from the 0.75 mM treated or 2% NEFA-free BSAtreated control C2C12 differentiated myotubes were harvested at times ranging from 30 min to 24 h. The 0-h treatment time represents media harvested at the conclusion of the experiment (24 h) from BSA-treated cells. Supernatants were stored at -70 and subsequently analyzed by ELISA as per manufacturer's instructions. Three independent experiments were performed in triplicate.
NF-B Activation ELISA-ELISA kits to detect activated NF-B were purchased from Active Motif (San Diego, CA). Cells were treated for the indicated times with 0.75 mM palmitate, nuclear protein extracts were prepared using the technique described by the manufacturer, and the proteins were normalized and subjected to ELISA analysis.
Densitometry and Statistical Analysis-Densitometry was performed using an Epson 1610 Scanner and the associated software. Statistical analysis of the data resulting from ELISA experiments was analyzed using an analysis of variance test on STATAQUEST software package. p values less than 0.05 were considered statistically significant.
RESULTS
Palmitate Treatment Activates TLR2 Signal Transduction in a Time-dependent Manner-A recent report indicated that saturated fatty acids activate TLR2/6 heterodimers when overexpressed in human embryonic kidney 293 cells (9, 10). Several reports also suggested that TLRs, in particular TLR2 and -4, are expressed in muscle cell models (20, 23-27). We chose to utilize the well characterized insulin-sensitive C2C12 mouse myotube model to investigate the effect of palmitate treatment on TLR activation. In the experiment shown in Fig. 1A, we examined the time-dependent association of the adaptor protein MyD88 with the TLR2 receptor after exposure to 0.75 mM palmitate, a concentration of palmitate that has been published previously to induce insulin resistance in this model. As shown in the upper panel of Fig. 1A, MyD88·TLR2 complexes are detected within 15 min of palmitate exposure and persisted for up to 8 h. A slight increase in the absolute protein level of TLR2 is detectable after 15 min of exposure to palmitate, but this increase is does not continue (middle panel) and MyD88 protein (lower panel) levels do not change throughout the duration of this experiment, indicating that the observed association is most probably due to activation of the receptor pathway and not a significant increase in protein expression.
We next examined the activation of several downstream components of the TLR2 signal transduction pathway, the stress-linked kinases p38, JNK, and PKC and the NF-B pathway. As shown in Fig. 1B, p38 and JNK (top and bottom blots, respectively) are rapidly phosphorylated (3 min) in response to palmitate treatment in C2C12 cells with no change in total protein mass. PKC has been suggested to be activated by palmitate in several models, and a pan phosphor-PKC antibody indicated that PKC was weakly phosphorylated after 5 min of palmitate exposure (middle panel, Fig. 1B). It is possible that the specific PKC isoform previously suggested to be activated is not recognized by the pan-PKC antibody or that it is activated after longer term exposure to palmitate. Additionally, the upper panel of Fig. 1C illustrates that palmitate induced the degradation of IB, an NF-B inhibitor, within 30 min, and it decreased to near undetectable levels by 12-24 h. These changes were not associated with changes in protein levels (Fig. 1C, lower panel).
Finally, we analyzed NF-B DNA binding activity and observed a significant increase in NF-B DNA binding 1 h after palmitate treatment (Fig. 2D). This activity reached a maximum at 2 h but remained detectable until 8 h, similar to previously published reports (28). These data suggest that the TLR2 signal transduction pathway is rapidly activated in response to palmitate treatment in C2C12 myotubes, possibly leading to the production of inflammatory cytokines such as IL-6 (13).
Previous reports have indicated that treatment of human myotubes or the L6 muscle cell line with palmitate resulted in the production of IL-6 (28, 29). We, therefore, investigated whether palmitate treatment induced a similar increase in IL-6 production from C2C12 mouse myotubes. Treatment of C2C12 myotubes for 2 h resulted in an increase in IL-6 secretion (Fig. 1B) from near undetectable levels to 20 pg/ml, correlating with the onset of significant NF-B activation. IL-6 production continued to increase in the presence of palmitate and reached a maximum of 200-300 pg/ml after 24 h. We also investigated the production of tumor necrosis factor- but found no significant tumor necrosis factor- production induced by palmitate over the time course observed (data not shown). After exposure to 0.75 mM palmitate for 24 h no significant toxicity was observed.
Palmitate Dose-dependently Activates TLR2 Signal Transduction in C2C12 Myotubes-To investigate whether the activation of TLR2-related signaling events in response to palmitate was due to exposure to toxic or "pharmacologic" doses of palmitate, we treated C2C12 myotubes with multiple doses of palmitate ranging from 0.75 nM to 0.75 mM. As shown in Fig. 2A, treatment with doses of palmitate as low as 0.75 nM were capable of inducing a weak but detectable association of TLR2 and its adaptor protein, which increased in abundance with increasing palmitate concentration, reaching a maximum at 0.75 µM as detected by immunoprecipitation and Western blot. This occurred with no change in the absolute level of MyD88 protein or TLR2 receptor mass (lower panel and data not shown). A phosphorylation of JNK (Fig. 2B, upper panel) and a degradation of the IB protein (Fig. 2B, middle panel) were also observed after treatment with 75 nM to 75 µM palmitate with no change in total protein levels (Fig. 2B, lower panel). The activation of these specific pathways occurs relatively dose-responsively, reaching a maximum level of activation and stabilizing. This experiment suggested that palmitate was capable of activating the TLR2 pathway at concentrations which could be found in vivo.
FIGURE 1.
Palmitate treatment temporally activates TLR-related signal transduction pathways in C2C12 myotubes. C2C12 myoblasts were grown to confluence in DMEM (high glucose) with 10% fetal bovine serum, differentiated in DMEM (high glucose) containing 10% horse serum, then cultured in NEFA-free media (DMEM, high glucose) with the addition of 2% NEFA-free BSA before exposure to 750 µM palmitate for the times indicated. A, at each time point indicated cellular protein was harvested, and protein was normalized and subjected to immunoprecipitation (IP) with anti-TLR 2.5 monoclonal antibody. The subsequent precipitates were separated using SDS-PAGE and Western-blotted (IB) to determine whether there was association of the MyD88 adaptor protein with the receptor as indicated. B, C2C12 myotubes were exposed to 750 µM (0.75 mM) palmitate as indicated, and cellular protein was analyzed by Western for the activation of p38 (upper panel), PKC (middle panel), and JNK (lower panel) kinases using phospho-specific antibodies. C, lysates were analyzed by Western for the presence of IB to analyze the activation of the NF-B pathway (upper) or for protein loading using E4F2a (lower panel). D, nuclear protein was isolated from C2C12 myotubes treated for the indicated times with palmitate and analyzed using a commercially available ELISA kit for the DNA binding/activation state of NF-B. E, cell culture supernatant was harvested from C2C12 myotubes treated with palmitate for the indicated times and analyzed for IL-6 production. All blots represent three independent experiments and for IL-6 production (n = 9, control versus treated (*, p p-JNK, phosphorylated JNK.
FIGURE 2.
Palmitate dose-dependently actives signal transduction in C2C12 myotubes. Differentiated C2C12 myotubes were exposed to palmitate at concentrations ranging from 0 to 75 µM for 10 min. At this time cellular protein was harvested, and protein was normalized for use in immunoprecipitation (IP; A) of TLR2 receptor analysis of intracellular signal transduction (JNK phosphorylation (upper panel) or degradation of IB (lower panel) by Western blot (B). All data represent at least three independent experiments. p-JNK, phosphorylated JNK.
FSL-1, a Known TLR2:6 Ligand, Activates a Similar TLR2related Pathway in C2C12 Myotubes at Concentrations Comparable with Palmitate-FSL-1 is a synthetic ligand similar to the known TLR2 lipoprotein ligand macrophage-activating lipopeptide-2. We examined if FSL-1 was capable of activating TLR2 signal transduction in C2C12 myotubes and compared this activation with that observed for palmitate. FSL-1 rapidly (10 min) and dose-dependently induced association of TLR2 with its adaptor protein MyD88 at concentrations of 1 nM and increased steadily to a maximum at 10 µM FSL-1 (Fig. 3A, upper panel). This association occurred with no detectable changes in TLR2 mass (Fig. 3A, lower panel).
Treatment with FSL-1 for 10 min at the indicated concentrations resulted in a significant phosphorylation of both JNK and p38 kinases (Fig. 3B, upper panel and middle panel, respectively) that occurred at concentrations of 10 nM and increased to a maximum activation at 1 µM then subsequently showed a slight decrease at the highest concentration, possibly suggesting toxicity. FSL-1 also increased the degradation of IB over a similar dose range (Fig. 3B, lower panel). We next compared the ability of equimolar concentrations of FSL-1 and palmitate to induce TLR2-dependent signal transduction in the C2C12 myotube. Fig. 3C shows that treatment of C2C12 cells with 0.75 µM FSL-1 or palmitate for 10 min induced the association of MyD88 with TLR2 as determined by immunoprecipitation (upper panel). However, the association induced by FSL-1 qualitatively appeared greater than that induced by palmitate at identical molar concentrations. Similarly, 0.75 µM FSL-1 activated JNK and p38 phosphorylation and the degradation of IB to a greater extent than palmitate at equimolar concentrations (Fig. 3D). This experiment suggests that although both FSL-1 and palmitate are capable of activating similar pathways, the C2C12 myotube is less sensitive to activation with palmitate.
Treatment of C2C12 Cells with a TLR2 Antagonistic Antibody Blocks Palmitate-induced Signal Transduction and IL-6 Production-A recently characterized monoclonal antibody to the extracellular region of TLR2, monoclonal antibody 2.5, has been shown to antagonize the production of cytokines, such as IL-6 and the activation of TLR2 signal transduction by its known ligands (30). We utilized this antibody to investigate the contribution of TLR2 to the production of IL-6and TLR2related signal transduction events in C2C12 cells. Cells were pretreated with TLR2 antibody or control IgG for 1 h before exposure to a more physiologic concentration of palmitate, 0.75 µM palmitate, as described previously and then treated with 0.75 µM palmitate for 10 min. Western blots for phosphorylated JNK indicated that palmitate alone or in the presence of control IgG activated JNK (Fig. 4A, upper panel). However, cells preincubated with TLR2 2.5 antibody showed a significant decrease in palmitate-induced JNK phosphorylation (Fig. 4A), indicating that the antagonistic antibody blocked the effect of palmitate on JNK phosphorylation. After 24 h in the presence of palmitate, control IgG and no antibody groups showed a significant increase in IL-6 production from roughly 40 pg/ml in BSAtreated control cells to 300 pg/ml by those treated with 0.75 µM palmitate (Fig. 4B). However, cells that were incubated in the presence of TLR2 antagonist antibody throughout the duration of the experiment displayed a 60% decrease in palmitate-induced IL-6 production. These data suggest a critical role for TLR2 in palmitate-induced IL-6 secretion in C2C12 cells.
FSL-1 and Palmitate-induced Signal Transduction and IL-6 Production Is Blocked by siRNA to TLR2 and/or MyD88-Although the previous experiments indicated TLR2 was at least partially responsible for the response of C2C12 myotubes to palmitate, it did not eliminate the possibility that alternate TLRs expressed in C2C12 cells may be involved. To determine the specific contribution of TLRs to palmitate-induced responses in C2C12 myotubes, we utilized siRNA to MyD88, a common TLR adaptor, to broadly inhibit TLR signal transduction in these cells and in a separate group of cells specifically blocked TLR2 through the use of TLR2 siRNA.
To determine the effect of this decrease in TLR components on palmitate and FLS-1-induced responses we exposed these cells to 0.75 µM palmitate or 0.75 µM FSL-1 for 5 min and examined the phosphorylation of JNK. As indicated in Fig. 5A, cells transfected with control, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) siRNA, exhibited a detectable increase in JNK phosphorylation in response to treatment with both FSL-1 and palmitate. Expression of TLR2 and MyD88 siRNA significantly decreased the detectable JNK phosphorylation, indicating that the loss of either TLR2 or its adaptor MyD88 affects the ability of both ligands to activate signal transduction events.
Cell lysates were probed by Western blot for the presence of MyD88 protein (Fig. 5B, upper panel) or TLR2 (Fig. 5B, lower panel) protein. This experiment indicated that transfection with either siRNA was capable of decreasing the levels of both proteins, whereas control siRNA did not effect the expression of either protein. Finally, C2C12 myotubes transfected with the indicated siRNAs were exposed to palmitate (0.75 µM) or FSL-1 (0.75 µM) overnight as described previously. As indicated in Fig. 5, C-D, control siRNA had no significant effect on palmitate or FSL-induced IL-6 production in C2C12 cells. However, TLR2 siRNA decreased palmitate induced IL-6 production by 50% (untreated, 6.64-fold ± 1.39 increase versus TLR2 siRNAtreated, 1.84-fold ± 2.6; p = 0.019252). The response to FSL was also blunted by 30% in these cells (FSL-1, 39.14 ± 4.22 versus TLR2 siRNA, 29.8 ± 2.64; p = 0.0354). These results correlate well with previously published data suggesting that palmitate and FSL can interact with the TLR2/6 heterodimer, but FSL is putatively capable of activating inflammatory signals in the absence of TLR2 (Fig. 5, C-D) (31). Importantly, with the inhibition of MyD88, we observed a near complete loss of palmitate-induced IL-6 production (Control, 6.64-fold ± 1.39 versus MyD88 siRNA, 1.14-fold ± 0.33; p = 0.001) and 70-80% decrease in FSL-stimulated IL-6 production (Control, 39.143-fold ± 4.22 versus MyD88 siRNA, 11.81-fold ± 1.65; p = 0.0001), indicating that a TLR pathway is critical to the proinflammatory response to both ligands in this cell type (Fig. 5, C-D).
FIGURE 3.
Comparison of the potency of FSL-1 (known TLR2 ligand) versus palmitate to activate TLR2 signal transduction. C2C12 myotubes were treated with FSL-1, a potent TLR2 agonist, at concentrations of 0-7.5 µM for 10 min. At this point cellular protein was harvested and normalized. A, association of MyD88 with the TLR2 receptor after treatment with FLS-1 by immunoprecipitation (IP) as described in Fig. 1. B, phosphorylated JNK (p-JNK; upper panel), p38 (middle panel), and the degradation of IB (lower panel) were measured by Western blot (WB) from the cellular lysates described. C2C12 myotubes were treated with 0.75 µM FSL-1 or palmitate to analyze the relative potency of these ligands to activate TLR2 signal transduction. C, comparison of FSL-1 (0.75 µM) and palmitate (0.75 µM) induced association of TLR2 with MyD88. D, comparison of FSL and palmitate induced JNK (upper) and p38 (middle) phosphorylation or IB degradation (lower). Total protein levels were also measured but not shown. All data represent three independent experiments.
Palmitate Rapidly Inhibits Insulin-induced Signal Transduction in C2C12 Myotubes, and This Inhibition Is Significantly Reversed by Blockade of TLR2 Signal Transduction-The data described above showed that palmitate rapidly activates signal transduction in the C2C12 myotube in a dose-dependent manner. We aimed to determine whether or not palmitate could also rapidly block insulin signal transduction in this model. In Fig. 6A, we examined that activation of insulin signal transduction in the presence of 0.75 µM palmitate for 0-24 h. As expected, 10 nM insulin treatment for 5 min increased the tyrosine phosphorylation of the insulin receptor (Fig. 6A, upper panel), but the increase in tyrosine phosphorylation was greatly diminished after 2 h of palmitate exposure and reached a maximum inhibition 12-24 h similar to previously published results. Similarly, the insulininduced phosphorylation of AKT on Ser-473 was also inhibited by exposure to palmitate. This inhibition was detectable as early as 1-2 h post-exposure to palmitate (Fig. 6A, middle panel). The decrease in phosphorylated insulin receptor and AKT was not due to loss of total protein levels as the mass of both the insulin receptor and AKT remained constant throughout the duration of this experiment (Fig. 6A, lower panels). Insulin-dependent IR autophosphorylation and phosphorylation of AKT are inhibited by the presence of palmitate with similar kinetics, but IR autophosphorylation appears to rebound slightly at 4 h of exposure to palmitate, whereas AKT phosphorylation remains inhibited.
We next investigated if the inhibition of TLR2 through the use of the TLR2 2.5 antagonist antibody characterized earlier could reverse palmitate-induced inhibition of insulin-induced AKT phosphorylation. As shown in Fig. 6B, cells treated with insulin alone or insulin and antibody exhibited a significant increase in phosphorylation of Ser-473 on AKT. This increase was significantly inhibited by nearly 80% in the presence of 0.75 µM palmitate. The presence of nonspecific IgG control antibody had little effect on this inhibition (Fig. 6B). However, palmitate-induced inhibition of AKT phosphorylation recovered in C2C12 cells preincubated with TLR2 2.5 antagonist antibody by 50% (Fig. 6B), indicating that the inhibition of TLR2 activation by palmitate played a significant role in palmitate-induced inhibition of insulin signal transduction.
FIGURE 4.
Antagonist TLR2 antibody TLR2.5 inhibits palmitate-induced events in C2C12 myotubes. C2C12 myotubes were cultured as described previously with the addition of no antibody, control isotype-specific IgG2a, or TLR 2.5 antagonistic antibody (mAb) for 1 h before exposure to palmitate (0.75 µM) for 10 min (A), and cellular protein was isolated and analyzed by Western for the presence of phosphorylated JNK (p-JNK). B, cells were incubated in the presence or absence of the indicated antibodies for 1 h and treated with palmitate and antibody for an additional 24 h, and cell culture media was isolated and analyzed for IL-6 production by ELISA. All data represent three (A) or nine (B) independent experiments. *, p versus treated); #, p versus TLR2 antibody-treated cells).
The final set of experiments aimed to investigate if a decrease in the expression of either TLR2 or MyD88 could alter the ability of palmitate to induce insulin resistance in this model. As expected from previous experiments, palmitate and a TLR2/6 ligand, FSL, were capable of inhibiting insulin-induced AKT phosphorylation in untransfected cells and those transfected with control siRNA (Fig. 6C). Expression of either control or TLR2 siRNA did slightly, but significantly inhibit insulin-induced AKT phosphorylation, but this decrease did not significantly affect the absolute decrease caused by palmitate or FSL-1 on insulin signal transduction. Strikingly, TLR2-specific siRNA almost completely reversed palmitate-induced inhibition of insulin signal transduction and returned insulin-induced AKT phosphorylation to near 80% of normal, whereas FSL was still capable of inhibiting insulin-induced AKT phosphorylation (Fig. 6, C and D). These data correlate well with the observed effect of this treatment on C2C12 IL-6 production and signal transduction. However, siRNA to TLR2 substantially decreased FSL-1-induced JNK activation. It is possible that TLR2 plays a specific role in this aspect of FSL-1 signal transduction, whereas other mechanisms may be involved in IL-6 production. Importantly, MyD88 siRNA completely restored insulin-induced AKT phosphorylation to normal levels in palmitate and FSLtreated cells to near control cell levels (Fig. 6, C and D). These data provide conclusive evidence that the TLR pathway and, more specifically, TLR2 may be critical to the development of insulin resistance and palmitate-induced inflammatory responses in this muscle cell model.
DISCUSSION
Multiple studies have correlated elevations in circulating fatty acids, experimentally induced or associated with high fat diet and obesity, with the development of insulin resistance (2). Several hypotheses on the role of fatty acids in the development of insulin resistance suggest the primary effect of fatty acids is to alter cellular metabolism or function leading to secondary events that effect insulin signal transduction. No mechanism has been conclusively proven to underlie the effect of elevated fatty acids on insulin responses in vivo or in isolated cell models. Although it is quite possible that several mechanisms play roles in this process, recent research has strongly implied that an inflammatory process may be the initiating event in the development of fatty acid induced-insulin resistance (8, 32).
The experiments detailed above indicate that palmitate rapidly and dose-dependently activated signal transduction in the C2C12 myotube cell model. Palmitate at high concentrations used in previous publications rapidly induced the association of MyD88 with the TLR2 receptor (6). Additionally, palmitate activated several stress-related kinases that have been previously been suggested to be involved in NEFA-induced insulin resistance. Contrary to previously published studies involving long term exposure to NEFAs, the rapid activation of these pathways suggests that these events may not be due to altered cellular metabolism, development of reactive oxygen species, or alterations in membrane fluidity but instead indicate the possibility of a receptor-mediated event. This suggestion is further exemplified by the sensitive and temporally regulated activation of these signaling components. It has been reported that circulating NEFA concentrations in human populations can reach roughly 500-1000 µM in the plasma and, with palmitate the most prevalent of saturated NEFAs present, can reach 100-200 µM (33). Data from this study indicated that concentrations as low as 10 nM were capable of eliciting the association of TLR2 and MyD88 as well as the activation of intracellular signaling pathways in isolated cell cultures. However, we did not observe a significant inhibition of insulin signal transduction at these concentrations (data not shown). In fact, inhibition of insulin signal transduction was observed at concentrations of 75 nM and above with the maximal inhibition observed at 750 µM. The concentrations, which activate signal transduction, are slightly below what has been observed in circulating plasma (400-600 µM, total FFA and 100-200 µM, palmitate) but are well within the range for the concentrations observed in muscle tissue, where muscle palmitoyl-CoA has been observed in the range of 1nM in type 2 diabetics (33-35). These concentrations appear to fit well within range of the concentrations observed in vivo, suggesting that these observations are relevant to the study of high fat dietand obesity-related insulin resistance and inflammation (33, 35).
FIGURE 5.
siRNA to TLR2 and MyD88 blocks partially inhibits palmitate and FSL-mediated responses in C2C12 myotubes. C2C12 myoblasts were transfected with siRNAs indicated prior to differentiation. The differentiated myotubes were then treated with either 0.75 µM FSL-1 or palmitate for 5 min (B) or for 24 h (C). A, cellular protein was analyzed by Western blot (IB) for the presence of TLR2 and MyD88 following exposure to siRNA for 72 h. GAPDH, glyceraldehyde 3-phosphate dehydrogenase B, C2C12 cells transfected with the siRNA indicated was treated with FSL-1 or palmitate for 5 min and protein harvested and analyzed by Western for phosphorylated JNK. C2C12 cells transfected with the indicated siRNAs were treated with either palmitate (C) or FSL-1 (D) for 24 h and culture medium analyzed for IL-6 protein by ELISA and represented as the average fold increase over base line. All data represent four independent experiments. *, p versus FSL-1 or palmitate-treated control); #, p versus siRNAtreated cells).
We have also shown that a known TLR2 ligand, FSL-1, is capable of dose-dependently activating similar pathways to those activated by palmitate. Although palmitate was less potent at activating intracellular signal transduction and IL-6 production in the C2C12 myotubes, it was capable of activating these responses at similar molar concentrations, indicating that, indeed, palmitate may be an endogenous ligand for the TLR2 receptor pathway. Similarly, FSL-1 was also capable of inhibiting insulin-induced signal transduction in this model. These data indicate that TLR activation by multiple factors may play important roles in the development of insulin resistance in conditions such as sepsis, cancer, viral infection, and perhaps obesity and diabetes.
Data presented above show that palmitate treatment of differentiated C2C12 myotubes results in a time-dependent inhibition of insulin-activated signal transduction, specifically tyrosine phosphorylation of the insulin receptor and the phosphorylation of AKT. Additionally, palmitate induced the production of significant amounts of IL-6. This evidence is in agreement with several recently published studies that indicated IL-6 was produced in isolated human myotubes and in the L6 rat muscle cell model (28). IL-6 has been shown to directly affect insulin sensitivity in other insulin-sensitive tissues such as liver and adipocytes and, as an important modulator of insulin signal transduction, provides an interesting area of exploration in the development of insulin resistance in vivo (36-38). Although, IL-6 effects on muscle cell insulin sensitivity are controversial, it has been shown to directly impact insulin signal transduction in muscle in vivo (37). This does not preclude the fact that other untested cytokines/factors may play important roles in regulating insulin sensitivity in these cells after TLR2 activation. IL-6 has not conclusively been proven to directly affect C2C12 myotubes and, therefore, may simply serve as an indication of the production of multiple NF-B-responsive genes.
FIGURE 6.
Palmitate rapidly inhibits insulin signal transduction in a TLR-dependent manner. C2C12 myotubes were treated with 0.75 µM palmitate for 0-24 h and then briefly (10 min) stimulated with 10 nM insulin to analyze insulin signal transduction. A, palmitate treatment rapidly inhibits insulin-induced IR autophosphorylation and phosphorylation (p) of AKT as analyzed by Western blot (IB). IP, immunoprecipitation. pY, phosphorylated tyrosine residue. B, C2C12 myotubes were pretreated with TLR 2.5 antibody (Ab), IgG control, or no antibody for 1 h and then exposed to palmitate overnight. Insulin-induced AKT phosphorylation on Ser-473 was analyzed by Western blot and quantitated using densitometry. The graph below represents the percent of maximum (control cell insulin stimulated) AKT phosphorylation observed ± S.D. C, C2C12 cells transfected with the indicated siRNAs were subsequently exposed to palmitate overnight as described previously. After a 10-min exposure to insulin (Ins), protein was harvested, and AKT Ser-473 phosphorylation was analyzed by Western blot. D, the data represent an analysis of the densitometry data from the Western blots performed in C and graphed as the percent maximum AKT phosphorylation of the control cells. (n = 4). *, p versus TLR2 antibody-treated cells). #, p versus TLR2/MyD88-treated cells).
The rapid activation of TLR2 signal transduction led us to hypothesize that this receptor pathway may play an important role in initiating the response of C2C12 myotubes to palmitate. Treatment of cells with a monoclonal antagonist antibody TLR 2.5 blocked the subsequent induction of IL-6 production and restored insulin-induced AKT phosphorylation. We also utilized siRNA technology to block the expression of both TLR2 and MyD88 in this model. The decrease in TLR2 expression was accompanied by a near complete loss of palmitate-induced IL-6 production and a restoration of normal insulin-induced AKT phosphorylation in the presence of palmitate. The inhibition of TLR2 also significantly blocked the ability of FSL-1 to induce the expression of IL-6, although this did not significantly increase insulin sensitivity. The data presented indicate a slight discrepancy in the sensitivity of FSL-1 signaling to TLR2 blockade. The antagonistic antibody is fairly potent in blocking both FSL-1 effects on insulin signal transduction, whereas siRNA to TLR2 is not. The TLR2 antagonist antibody most likely inhibits the activation of TLR2 by FSL-1 more completely through a competition for binding to the receptor, whereas siRNA reduction of TLR2 simply lowers the absolute level of the TLR2 receptor but does not affect the individual receptor activity. This conclusion is congruent with the preceding data that siRNA to TLR2 minimally inhibited FSL-1-induced IL-6 production in the C2C12 model. These data are also in line with previous data regarding the signal transduction events activated by FSL-1 which have suggested that this ligand can signal directly through TLR6 or possibly other receptors through its peptide portion (31, 39). However, the inhibition of MyD88, a general adaptor for all TLR signal transduction, blocked IL-6 production in response to both palmitate and FSL-1 and completely restored insulin sensitivity in this model, indicating that although TLR2 plays a significant role in the development of palmitate-induced insulin resistance, alternate TLRs such as TLR4 may be involved.
It is interesting to note that the critical portion of many ligands for these TLRs contain a similar fatty acid component. Lipopolysaccharide, for example, contains a core lipid A molecule that is composed of a small carbohydrate backbone bonded to 2-3 long chain fatty acids (40). In addition, TLR2-specific ligands macrophage-activating lipopeptide-2 and FSL-1 are composed entirely of a short peptide sequence linked to two palmitate or stearic acid molecules (41, 42). In the absence of this lipid component these molecules do not act as potent ligands. This suggests that an important component of TLR ligand recognition involves the recognition of a fatty acid side chain.
Cell culture models have indicated that oxidized low density lipopeptide molecules can induce the activation of TLR4 receptors in monocytes and endothelial cells (43). However, studies in clinical populations have revealed no genetic link between TLR4 polymorphisms and the development of insulin resistance or type 2 diabetes (44). Interestingly, a recent study has indicated that polymorphisms in the TLR2 receptor gene are linked to populations at greater risk of the development of type 2 diabetes (45, 46). Additionally, studies have indicated a pronounced correlation between polymorphisms in the NOD (nuclear oligomerization domain) gene family hypothesized to regulate TLR signal transduction and inflammatory diseases such as Crohn disease and rheumatoid arthritis (47, 48). These data strongly suggest that polymorphisms affecting the regulation of either of these gene families may be important to the development of insulin resistance.
The link between TLRs, NEFAs, and insulin resistance is strengthened by recent advances in diabetes research suggesting that the underlying cause of insulin resistance may be the development of a chronic inflammatory state. The cause of this inflammatory state is unknown, but it is well know that NEFA elevations even in healthy subjects can transiently induce insulin resistance and that NEFAs exposure in cells and animals leads to the activation of inflammatory-linked pathways such as NF-B and PKCs leading to the production of inflammatory factors that may in turn inhibit insulin signal transduction (49, 50). These data taken together suggest that elevations in NEFAs may activate an inflammatory pathway, and the TLRs provide an excellent candidate for such a pathway.
Finally, it has been observed that monocytes may play a critical role in the inflammatory processes associated with the development of insulin resistance (51-53). These cells are known to contain multiple TLRs, and therefore, a significant component of NEFA-induced inflammation may in fact result from the activation of monocyte TLRs either resident to insulin-sensitive tissues or circulating, and these activated monocytes may in turn produce inflammatory factors that modulate insulin sensitivity in tissues such as the muscle, liver, and fat.
The preceding study indicates a novel pathway in the development of palmitate-induced insulin resistance, TLR2/MyD88, and also provides for the first time detailed evidence that TLR2 pathways are present and functional in a muscle cell model. These data are of critical importance to both the study of the development of insulin resistance and possibly to the study of atherosclerosis and other diseases associated with elevated lipids and inflammation. TLRs provide an excellent therapeutic target for the treatment of obesity and high fat diet-induced insulin resistance, and further investigations will elucidate the role of these receptors in vivo.
ACKNOWLEDGMENTS
I gratefully acknowledge Drs. Richard W. Furlanetto, James A. Cook, and Maria Buse for assistance in the preparation of this manuscript and critical discussion of the data.
【参考文献】
Bray, G. A. (2004) J. Clin. Endocrinol. Metab. 89, 2583-2589
Boden, G. (2003) Exp. Clin. Endocrinol. Diabetes 111, 121-124
Liu, R. H., Mizuta, M., Kurose, T., and Matsukura, S. (2002) Int. J. Obes. Relat. Metab. Disord. 26, 318-326
Evans, J. L., Goldfine, I. D., Maddux, B. A., and Grodsky, G. M. (2002) Endocr. Rev. 23, 599-622
Perseghin, G., Petersen, K., and Shulman, G. I. (2003) Int. J. Obes. Relat. Metab. Disord. 27, Suppl. 3, 6-11
Ajuwon, K. M., and Spurlock, M. E. (2005) J. Nutr. 135, 1841-1846
Ghanim, H., Aljada, A., Hofmeyer, D., Syed, T., Mohanty, P., and Dandona, P. (2004) Circulation 110, 1564-1571
Jove, M., Planavila, A., Laguna, J. C., and Vazquez-Carrera, M. (2005) Endocrinology 146, 3087-3095
Lee, J. Y., Plakidas, A., Lee, W. H., Heikkinen, A., Chanmugam, P., Bray, G., and Hwang, D. H. (2003) J. Lipid Res. 44, 479-486
Lee, J. Y., Zhao, L., Youn, H. S., Weatherill, A. R., Tapping, R., Feng, L., Lee, W. H., Fitzgerald, K. A., and Hwang, D. H. (2004) J. Biol. Chem. 279, 16971-16979
Takeda, K., and Akira, S. (2001) Genes Cells 6, 733-742
Akira, S., and Sato, S. (2003) Scand. J. Infect. Dis. 35, 555-562
Fan, H., and Cook, J. A. (2004) J. Endotoxin Res. 10, 71-84
Zuany-Amorim, C., Hastewell, J., and Walker, C. (2002) Nat. Rev. Drug Discov. 1, 797-807
Muller, S. D., Muller, M. R., Huber, M., Esche Uv, U., Kirschning, C. J., Wagner, H., Bessler, W. G., and Mittenbuhler, K. (2004) Int. Immunopharmacol. 4, 1287-1300
Muzio, M., Polentarutti, N., Bosisio, D., Manoj Kumar, P. P., and Mantovani, A. (2000) Biochem. Soc. Trans. 28, 563-566
Muzio, M., and Mantovani, A. (2000) Microbes Infect. 2, 251-255
Netea, M. G., van der Graaf, C., Van der Meer, J. W., and Kullberg, B. J. (2004) J. Leukocyte Biol. 75, 749-755
Lin, Y., Lee, H., Berg, A. H., Lisanti, M. P., Shapiro, L., and Scherer, P. E. (2000) J. Biol. Chem. 275, 24255-24263
Lang, C. H., Silvis, C., Deshpande, N., Nystrom, G., and Frost, R. A. (2003) Shock 19, 538-546
Yim, J. J., Ding, L., Schaffer, A. A., Park, G. Y., Shim, Y. S., and Holland, S. M. (2004) FEMS Immunol. Med. Microbiol. 40, 163-169
Bradford, M. M. (1976) Anal. Biochem. 72, 248-254
Frost, R. A., Nystrom, G., Burrows, P. V., and Lang, C. H. (2005) Alcohol Clin. Exp. Res. 29, 1247-1256
Nishimura, M., and Naito, S. (2005) Biol. Pharm. Bull. 28, 886-892
Yang, X., Coriolan, D., Murthy, V., Schultz, K., Golenbock, D. T., and Beasley, D. (2005) Am. J. Physiol. Heart Circ. Physiol. 289, 1069-1076
Jimenez, R., Belcher, E., Sriskandan, S., Lucas, R., McMaster, S., Vojnovic, I., Warner, T. D., and Mitchell, J. A. (2005) Proc. Natl. Acad. Sci. U. S. A. 102, 4637-4642
Frost, R. A., Nystrom, G. J., and Lang, C. H. (2004) Am. J. Physiol. Cell Physiol. 287, 1605-1615
Sinha, S., Perdomo, G., Brown, N. F., and O'Doherty, R. M. (2004) J. Biol. Chem. 279, 41294-41301
Weigert, C., Brodbeck, K., Staiger, H., Kausch, C., Machicao, F., Haring, H. U., and Schleicher, E. D. (2004) J. Biol. Chem. 279, 23942-23952
Wang, J. E., Warris, A., Ellingsen, E. A., Jorgensen, P. F., Flo, T. H., Espevik, T., Solberg, R., Verweij, P. E., and Aasen, A. O. (2001) Infect. Immun. 69, 2402-2406
Okusawa, T., Fujita, M., Nakamura, J., Into, T., Yasuda, M., Yoshimura, A., Hara, Y., Hasebe, A., Golenbock, D. T., Morita, M., Kuroki, Y., Ogawa, T., and Shibata, K. (2004) Infect. Immun. 72, 1657-1665
Chinetti, G., Fruchart, J. C., and Staels, B. (2000) Inflamm. Res. 49, 497-505
Fraser, D. A., Thoen, J., Rustan, A. C., Forre, O., and Kjeldsen-Kragh, J. (1999) Rheumatology (Oxford) 38, 948-952
Bajaj, M., Suraamornkul, S., Romanelli, A., Cline, G. W., Mandarino, L. J., Shulman, G. I., and DeFronzo, R. A. (2005) Diabetes 54, 3148-3153
Belfort, R., Mandarino, L., Kashyap, S., Wirfel, K., Pratipanawatr, T., Berria, R., Defronzo, R. A., and Cusi, K. (2005) Diabetes 54, 1640-1648
Tomas, E., Kelly, M., Xiang, X., Tsao, T. S., Keller, C., Keller, P., Luo, Z., Lodish, H., Saha, A. K., Unger, R., and Ruderman, N. B. (2004) Proc. Nutr. Soc. 63, 381-385
Kim, H. J., Higashimori, T., Park, S. Y., Choi, H., Dong, J., Kim, Y. J., Noh, H. L., Cho, Y. R., Cline, G., Kim, Y. B., and Kim, J. K. (2004) Diabetes 53, 1060-1067
Senn, J. J., Klover, P. J., Nowak, I. A., and Mooney, R. A. (2002) Diabetes 51, 3391-3399
Nakao, Y., Funami, K., Kikkawa, S., Taniguchi, M., Nishiguchi, M., Fukumori, Y., Seya, T., and Matsumoto, M. (2005) J. Immunol. 174, 1566-1573
Rietschel, E. T., Brade, L., Brandenburg, K., Flad, H. D., de Jong-Leuveninck, J., Kawahara, K., Lindner, B., Loppnow, H., Luderitz, T., and Schade, U. (1987) Rev. Infect. Dis. 9, Suppl. 5, 527-536
Fujita, M., Into, T., Yasuda, M., Okusawa, T., Hamahira, S., Kuroki, Y., Eto, A., Nisizawa, T., Morita, M., and Shibata, K. (2003) J. Immunol. 171, 3675-3683
Muhlradt, P. F., Kiess, M., Meyer, H., Sussmuth, R., and Jung, G. (1997) J. Exp. Med. 185, 1951-1958
Miller, Y. I., Chang, M. K., Binder, C. J., Shaw, P. X., and Witztum, J. L. (2003) Curr. Opin. Lipidol. 14, 437-445
Illig, T., Bongardt, F., Schopfer, A., Holle, R., Muller, S., Rathmann, W., Koenig, W., Meisinger, C., Wichmann, H. E., and Kolb, H. (2003) Diabetes 52, 2861-2864
Schroder, N. W., and Schumann, R. R. (2005) Lancet Infect. Dis. 5, 156-164
Park, Y., Park, S., Yoo, E., Kim, D., and Shin, H. (2004) Ann. N. Y. Acad. Sci. 1037, 170-174
Abreu, M. T., Fukata, M., and Arditi, M. (2005) J. Immunol. 174, 4453-4460
Rook, G. A., Martinelli, R., and Brunet, L. R. (2003) Curr. Opin. Allergy Clin. Immunol. 3, 337-342
Roden, M., Price, T. B., Perseghin, G., Petersen, K. F., Rothman, D. L., Cline, G. W., and Shulman, G. I. (1996) J. Clin. Investig. 97, 2859-2865
Itani, S. I., Ruderman, N. B., Schmieder, F., and Boden, G. (2002) Diabetes 51, 2005-2011
Sewter, C. P., Digby, J. E., Blows, F., Prins, J., and O'Rahilly, S. (1999) J. Endocrinol. 163, 33-38
Yamakawa, T., Tanaka, S., Yamakawa, Y., Kiuchi, Y., Isoda, F., Kawamoto, S., Okuda, K., and Sekihara, H. (1995) Clin. Immunol. Immunopathol. 75, 51-56
Arkan, M. C., Hevener, A. L., Greten, F. R., Maeda, S., Li, Z. W., Long, J. M., Wynshaw-Boris, A., Poli, G., Olefsky, J., and Karin, M. (2005) Nat. Med. 11, 191-198
订阅登记：
请您在下面输入常用的Email地址、职业以便我们定期通过邮箱发送给您最新的相关医学信息，感谢您浏览首席医学网！