| Obesity is the leading risk factor for development of many life threatening diseases, particularly insulin resistance and type 2 diabetes[181, 182]. The hallmark of obesity is massive expansion of fat mass in order to store excessive amount of energy in the form of triglycerides. Stressed adipocytes secrete multiple chemokines and cytokines, which may contribute to the initial adipose macrophage infiltration and impair adipocyte insulin signaling[30, 46, 113, 183, 184]. At a later stage of obesity development, massive infiltrated and activated macrophages become important source for adipose chemokines and cytokines[111, 112]. It is known that cytokines like TNFαand IL-6 can increase adipocyte lipolysis[185]. Increased circulating FFA levels cause insulin resistance in other organs such as liver and muscle. Approaches aimed to reduce adipose inflammation seem to be effective on improving insulin sensitivity in both animal models and in humans. Decreased macrophage infiltration and reduction of inflammatory gene expression in adipose tissue have also been associated with weight loss in obese subjects[186, 187]. Thiazolidinedione (TZD), a class of insulin sensitizing drug that mainly improves adipose insulin sensitivity of type 2 diabetic patients, also has potent anti-inflammatory effects and suppresses adipose macrophage gene expression in vitro and in vivo[188-194]. The fact that fat mass increases prior to macrophage infiltration indicates that lipid overload in adipocytes is the causal factor for adipose inflammation and insulin resistance[113].Recently, obesity has been shown to cause ER stress in both liver and adipose tissue. Chemical chaperones can effectively alleviate ER stress in liver and adipose tissue as well as reverse systemic insulin resistance and type 2 diabetes in ob/ob mice. Adipose tissue is heterogeneous and contains many cell types. It is unclear whether over-nutrition generates ER stress in adipocytes, a type of cells critical for storage of excessive energy and production of inflammatory cytokines. In order for adipocytes to store excessive amount of lipids, triglycerides derived from diet and de novo synthesis need to be broken down to FFA and glycerol first, then uptaken by adipocytes for reesterification. It is well known that FFA can induce adipocyte inflammation and insulin resistance. However, it remains unclear whether FFA can cause ER stress in adipocytes and whether ER stress plays a role in FFA-induced inflammation and insulin resistance in adipocytes. In the present study, we show that FFA cause ER stress in 3T3-L1 adipocytes. Chemical chaperone can alleviate the adverse effect of FFA on adipocytes. IKKβis an important downstream effecter of FFA-induced ER stress since IKKβoverexpression is sufficient to cause adipocyte inflammation and insulin resistance. Furthermore, activation of IKKβcan also induce ER stress in adipocytes and the expression of endogenous IKKβis elevated in adipocyte of obese mice.1. The effect of FFA in 3T3-L1 adipocytes1.1 FFA activate ER stress pathway and inflammatory pathway, suppress insulin pathway in 3T3-L1 adipocytes.A mixture of FFA was used to treat 3T3-L1 adipocytes and ER stress, inflammation and insulin resistance were examined. FFA treatment caused rapid phosphorylation of both PERK and IRE-1α30 and 60 minutes after treatment (Figure 2.5). Phosphorylation dynamics of both IKKβand JNK is similar to that of PERK and IRE-1α, being strongest at 30 and 60 minutes after FFA treatment and starting to fade away at later time points (Figure 2.6). The expression level of several other ER marker proteins, such as GRP78 and the spliced transcription factor XBP-1 (XBP-1s), peaked at two hours after FFA treatment (P<0.05) (Figure 2.4). In contrast, impairment of insulin signaling, reflected by decreased IR, IRS-1 and Akt phosphorylation, did not reach maximal till three hours post treatment (Figure 2.7). In our previous study, we reported that FFA can induce up-regulation of multiple chemokines in 3T3-L1 adipocytes which peak three hours post treatment under identical conditions (3). These results indicate that FFA can induce ER stress response in 3T3-L1 adipocytes. Activation of ER stress sensor proteins (IRE-1αand PERK) is concurrent with activation of IKKβand JNK but precedes FFA-induced expression of other ER marker proteins (GRP78 and XBP1s), adipocyte chemokine production and maximal impairment of insulin signaling. 1.2 Chemical chaperone TUDCA can reduce FFA-induced inflammation and insulin resistance in 3T3-L1 adipocytes.It has been reported that chemical chaperones can ameliorate ER stress and improve insulin sensitivity in obese mice as well as tunicamycin and thapsigargin treated liver cells[158, 164]. We next examined whether TUDCA, a well-established chemical chaperone that has been demonstrated to reduce ER stress, can improve FFA-induced adipocyte inflammation and insulin resistance. TUDCA was used to pre-treat 3T3-L1 adipocytes overnight and added again with FFA mixture next day. Pre-treatment with TUDCA effectively reduced FFA-induced phosphorylation on both IKKβand JNK in 3T3-L1 adipocytes (Figure 2.8), indicating that activation of IKKβand JNK by FFA treatment might be at least partially through turning on ER stress response pathway. TNFαand IL-6 are inflammatory cytokines that have been demonstrated to increase in adipose tissue of obese state and impair adipocyte function. TUDCA pre-treatment completely blocked FFA-induced expression of TNFαand IL-6 in 3T3-L1 adipocytes (P<0.05) (Figure 2.9). Furthermore, TUDCA pre-treatment also improved insulin signaling of 3T3-L1 adipocytes in the presence of FFA as demonstrated by increased tyrosine phosphorylation on IRS-1 and serine/threonine phosphorylation on Akt in comparison to vehicle-pretreatment in the presence of FFA (Figure 2C). These results suggest that ER stress pathway is likely an important mediator of FFA-induced inflammation and insulin resistance in adipocytes.2. The effect of IKKβin adipocytes2.1 The effect of IKKβin 3T3-L1 CAR adipocytes2.1.1 Overexpression of IKKβis sufficient to mimic FFA-induced inflammation and insulin resistance in 3T3-L1 CAR adipocytes. To examine whether IKKβis the critical mediator of FFA-induced inflammation and insulin resistance in 3T3-L1 adipocytes, we next over-expressed both wild type (IKKβWT) and the constitutively active human IKKβ(IKKβSE) by adenovirus- mediated gene transfer in 3T3-L1 CAR adipocytes. Catalytically inactive IKKβ(IKKβK44M) and GFP were also expressed as negative controls. Over- expression of hIKKβSE, not IKKβWT, obviously impaired insulin-stimulated phosphorylation of Akt on both threonine 308 and serine 473 (Figure 3.3). In contrast, IKKβoverexpression did not affect insulin-stimulated tyrosine phosphorylation on either IR or IRS-1 (Supplementary figure 2) despite that IRS-1 has been reported to be a direct target of IKKβfor serine phosphorylation[73]. Overexpression of both IKKβWT and IKKβSE significantly increased expression of inflammatory cytokines TNFαand IL-6 (P<0.05) (Figure 3.6) but dramatically decreased expression of insulin-sensitizing hormones leptin and adiponectin (P<0.05)(Figure 3.7). The constitutively active IKKβhas more robust effect on regulating aforementioned genes compared to the wild type IKKβ. Interestingly, overexpression of both IKKβWT and IKKβSE also significantly increased expression of GRP78 and CHOP(P<0.05)(Figure 3.5), indicating that IKKβis not only a downstream effecter for FFA-induced ER stress but it also can induce ER stress. In contrast, overexpression of JNK1 in 3T3-L1 CAR adipocytes increased IL-6 expression but did not have an effect on TNFαexpression (Supplementary figure 1A). These results are consistent with reported literature that plasma IL-6, not TNFα, was reduced in adipose-specific JNK1 knock out mice fed on high fat diet[168].2.1.2 Overexpression of IKKβhas profound effect on adipocyte lipid metabolism.One potential reason that adipose inflammation develops in response to lipid load might be to limit excessive fat mass expansion. Indeed, TNFαhas been reported to inhibit adipocyte differentiation and induce lipolysis, both can reduce adipocyte lipid content[185, 195]. Since IKKβoverexpression dramatically increased expression of TNF-α, it is very likely that lipid metabolism is also altered. To determine whether IKKβoverexpression has any effect on lipid metabolism, lipolysis and lipid synthesis were evaluated. The expression of hormone sensitive lipase (HSL), adipocyte triglyceride lipase (AGTL), lipoprotein lipase (LPL) and perilipin are all significantly decreased (P<0.05) (Figure 3.11). Glycerol contents in conditioned media collected from hIKKβWT and hIKKβSE over-expressing 3T3-L1 CAR adipocytes are significantly higher than those from adipocytes over-expressing GFP and IKKβKM (P<0.05) (Figure 3.10), suggesting increased lipolysis upon IKKβoverexpression. The decreased expression of lipases might be a feedback mechanism for adipocytes to control dysregulated lipolysis. Expression level of glut4 is also significantly decreased in 3T3-L1 CAR adipocytes over-expressing IKKβcompared to control cells expressing GFP and inactive IKKβ(P<0.05) (Figure 3.4), indicating potential impairment in insulin-stimulated glucose uptake. IKKβoverexpression also significantly reduced expression of many lipogenic genes (P<0.05) (Figure 3.9) such as stearoyl CoA desaturase 1 (SCD1), fatty acid synthase (FAS), acyl CoA carboxylase 1 (ACC1), Acyl CoA carboxylase 2 (ACC2), mitochondrial glycerol-3-phosphate acyltransferase 1 (GPAT1), and diacylglycerol acyltrans- ferase 1 (DGAT1). The reduced triglyceride contents in hIKKβWT and hIKKβSE over-expressing cells are consistent with decreased expression of lipogenic genes. It remains to be studied whether the effect of IKKβon lipid metabolism is directly achieved through modulating expression of relevant genes or indirectly achieved via increase of TNFαand IL-6 expression.2.2 Obesity activates IKKβin primary adipocytes.Our IKKβoverexpression study in L1-CAR adipocytes demonstrated that IKKβis an important player in adipocyte inflammation and insulin resistance. To assess whether expression and phosphorylation status of endogenous IKKβcan be increased in adipocytes of obese mice, primary adipocytes were isolated from 9-week old ob/ob mice and littermate lean controls. Expression of phospho-IKKβwas significantly increased by 147% in adipocytes isolated from ob/ob mice compared to the level in lean mice (Figure 3.12, 3.13). The increase of phospho-IKKβis mainly due to elevated expression of IKKβprotein, which was increased by 151% in adipocytes from ob/ob mice (Figure 3.12, 3.13). In contrast, expression of phospho-IKKαdecreased by 53% in adipocytes from ob/ob mice despite increased expression of IKKαprotein by 17.7% (Figure 3.12, 3.14). These results indicate that IKKβmay play an important role in adipocyte inflammation and insulin resistance in obese state.Understanding the link between lipid overload and development of adipocyte inflammation and insulin resistance will likely provide useful information for potential new therapeutic directions to treat metabolic syndrome. It is well known that FFA can induce inflammation and insulin resistance in adipocytes but the molecular mechanism is not very clear. ER stress has recently been linked to obesity-related insulin resistance. Despite the solid evidence that ER stress exists in obese adipose tissue, it remained to be studied whether ER stress occurs in adipocyte in response to increased lipid accumulation. In the present study, we have demonstrated that FFA can activate two major ER stress sensing pathways, PERK and IRE-1α. Attenuation of FFA-induced ER stress by chemical chaperone effectively reduced FFA-induced expression of inflammatory cytokines and improved insulin signaling, which was accompanied by reduced IKKβand JNK phosphorylation. These results indicate that ER stress pathway is involved in FFA-induced production of inflammatory factors and impairment of insulin signaling.Both JNK and IKKβhave been demonstrated to be critical mediators of obesity-related inflammation and insulin resistance. JNK1 deficient mice are protected from the development of obesity-related systemic insulin resistance [77]. Absence of JNK1 in macrophages can also modestly change inflammatory profile in adipose tissue[196]. Heterozygous IKKβknock out mice are protected from developing insulin resistance induced by high fat diet and leptin deficiency[65]. Overexpression of the constitutively active IKKβin the liver causes hepatic and systemic insulin resistance[72].Constitutive activation of IKKβin the hypothalamus impairs central insulin and leptin signaling and suppression of IKKβin the hypothalamus protects from development of obesity and glucose intolerance[197]. Deletion of IKKβin hepatocytes improved high fat diet-induced liver insulin resistance while deficiency of IKKβin myeloid cells renders global insulin sensitivity upon high fat diet[71]. Gain-of-funtion approach was used next to explore whether overexpression of IKKβand JNK can mimic the effect of FFA on causing adipocyte inflammation and insulin resistance. Due to the fact that the constitutively active form of JNK is not available, only wild type JNK1 was over-expressed in 3T3-L1 CAR adipocytes, which increased expression of IL-6 but not TNFα. Overexpression of JNK1WT did not impair insulin signaling (supplementary figure 1B, 1C). In contrast, over- expression of IKKβWT significantly increased expression of both TNFαand IL-6 but was still not sufficient to impair insulin signaling. Overexpression of the constitutively active form of IKKβnot only further increased expression of inflammatory cytokines but also reduced insulin-stimulated Akt phosphorylation. Interestingly, IKKβactivation in adipocytes also induced ER stress as demonstrated by increased expression of GRP78 and CHOP. These data indicate that a certain threshold level of proinflammatory cytokines might be required to damage insulin signaling in adipocytes and constitutive activation of IKKβis sufficient to mimic the effect of FFA on inducing adipocyte inflammation and insulin resistance.In addition to increasing expression of proinflammatory cytokines, activation of IKKβdrastically repressed expression of leptin and adiponectin, both are important hormones that can modulate systemic insulin sensitivity and energy homeostasis. Overexpression of IKKβalso had profound effect on adipocyte lipid metabolism, such as increasing lipolysis and decreasing triglyceride synthesis. Furthermore, activation of endogenous IKKβhas been observed in primary adipocytes isolated from obese mice. Activation of IKKβin adipocytes during the development of obesity not only likely to influence adipose tissue inflammation profile and lipid storage but also may affect whole body energy metabolism by modulating expression of adipocyte-derived hormones.
|
PDF Full Text Request