Fu S, Yang L, Li P, Hofmann O, Dicker L, Hide W, et al. Aberrant lipid metabolism disrupts calcium homeostasis causing liver endoplasmic reticulum stress in obesity. Nature 2011;473:528-531. Available at: www.nature.com (Reprinted with permission.)
The endoplasmic reticulum (ER) is the main site of protein and lipid synthesis, membrane biogenesis, xenobiotic detoxification and cellular calcium storage, and perturbation of ER homeostasis leads to stress and the activation of the unfolded protein response. Chronic activation of ER stress has been shown to have an important role in the development of insulin resistance and diabetes in obesity. However, the mechanisms that lead to chronic ER stress in a metabolic context in general, and in obesity in particular, are not understood. Here we comparatively examined the proteomic and lipidomic landscape of hepatic ER purified from lean and obese mice to explore the mechanisms of chronic ER stress in obesity. We found suppression of protein but stimulation of lipid synthesis in the obese ER without significant alterations in chaperone content. Alterations in ER fatty acid and lipid composition result in the inhibition of sarco/endoplasmic reticulum calcium ATPase (SERCA) activity and ER stress. Correcting the obesity-induced alteration of ER phospholipid composition or hepatic Serca over-expression in vivo both reduced chronic ER stress and improved glucose homeostasis. Hence, we established that abnormal lipid and calcium metabolism are important contributors to hepatic ER stress in obesity. (HEPATOLOGY 2011
The endoplasmic reticulum (ER) is the main site of protein and lipid synthesis, membrane biogenesis, xenobiotic detoxification and cellular calcium storage, and perturbation of ER homeostasis leads to stress and the activation of the unfolded protein response. Chronic activation of ER stress has been shown to have an important role in the development of insulin resistance and diabetes in obesity. However, the mechanisms that lead to chronic ER stress in a metabolic context in general, and in obesity in particular, are not understood. Here we comparatively examined the proteomic and lipidomic landscape of hepatic ER purified from lean and obese mice to explore the mechanisms of chronic ER stress in obesity. We found suppression of protein but stimulation of lipid synthesis in the obese ER without significant alterations in chaperone content. Alterations in ER fatty acid and lipid composition result in the inhibition of sarco/endoplasmic reticulum calcium ATPase (SERCA) activity and ER stress. Correcting the obesity-induced alteration of ER phospholipid composition or hepatic Serca over-expression in vivo both reduced chronic ER stress and improved glucose homeostasis. Hence, we established that abnormal lipid and calcium metabolism are important contributors to hepatic ER stress in obesity.
The endoplasmic reticulum (ER) is the critical organelle for biosynthesis of proteins, lipids (lipogenesis), and biomembranes, and it is central to xenobiotic detoxification and a major locus for calcium storage and intracellular calcium homeostasis. In cells exhibiting high rates of protein synthesis, defective posttranslational modifications can lead to accumulation of unfolded proteins and protein overloading in the ER. This is perceived by the ER and triggers protective pathways collectively termed the unfolded protein response (UPR). The failure of such responses to arrest ongoing protein accumulation activates other molecular switches, in a process known as the ER stress response; these have downstream effects on stress kinase activation, inflammatory pathways, insulin receptor signaling, lipogenesis, and oxidative stress, and can ultimately lead to dismantling of the cell by apoptosis.
ER stress is an important disease mechanism in type 2 diabetes (T2D) where it may contribute critically to loss of pancreatic beta cell mass,1 in ulcerative colitis where it affects intestinal epithelium,2 and in some liver diseases where it affects hepatocytes.3 The latter are exemplified by alpha-1-antitrypsin deficiency, in which ER stress appears to be a pivotal part of the pathogenesis of cirrhosis and hepatocellular carcinoma.4 As recently reviewed in HEPATOLOGY,3 ER stress may also contribute to disease mechanisms in alcoholic liver disease,5 drug-induced liver injury,6 and obesity-related nonalcoholic fatty liver disease (NAFLD). With regard to the latter, it has been suggested that ER stress could be a pathway to cell death and inflammation in some, but not all, experimental forms of nonalcoholic steatohepatitis (NASH).7-9 There is evidence that some tissues, and particularly fat,10 are under ER stress in obese humans. Yet, to date, the evidence for operation of hepatic ER stress in human NAFLD/NASH or in animal models showing both the metabolic determinants of NASH and steatohepatitis pathology is equivocal.11
As depicted in Fig. 1, ER stress results from imbalance between the load of unfolded proteins that enter the ER and the capacity of that organelle to respond by lowering protein synthesis, activating chaperones for unfolded proteins (glucose-regulated protein 78 [Grp78]), selenoprotein S1 [SEPS1]) or the ER protein-folding machinery (eukaryotic translation initiation factor 2α [eIF2α]). Together, these comprise the UPR. The core of this response is activation of a triad of stress-sensing proteins: inositol-requiring protein-1 (IRE1), activating transcription factor-6 (ATF6), and protein kinase RNA (PKR)-like ER kinase (PERK). The resultant cellular changes include: activation of the c-Jun N-terminal kinase (JNK), a likely mediator of the serine/threonine phosphorylation of insulin receptor substrate 1 (IRS1) that impairs insulin receptor signaling; oxidative stress and activation of nuclear factor-kappaB (NF-κB), which are proinflammatory pathways; and induction of cell death. Apoptosis particularly involves C/EBP-homologous protein (CHOP), which transcriptionally suppresses antiapoptotic Bcl-2 and induces proapoptotic Bim.12
Although the broad framework of molecular events in ER stress is partially mapped (Fig. 1), what has been less clear is whether increased de novo protein synthesis (so-called “client loading”) into the ER or relative impairment of ER protein folding and degradation is what leads to the ER stress in tissues of obese mice and humans. The article by Fu et al.13 from the laboratory of Dr. Hotamisligil, a pioneer in this area, provides new insights into this question. Using as their model obese mice with a null mutation in the appetite-regulating peptide, leptin (ob/ob mice), they showed by polysome profiling that ER-associated protein synthesis was down-regulated in obese liver compared with lean (leptin replete, wild-type [WT]) counterparts. They then used a proteomic approach, confirmed by immunoquantitation of individual proteins when possible, to characterize the profile of ER proteins in livers from ob/ob versus WT mice. Among 120 differentially up-regulated proteins, those involved with metabolic regulation were overrepresented. The latter particularly included genes involved with lipogenesis (fatty acid synthase [FAS], stearoyl coenzymeA desaturase 1 [SCD-1], diacylglycerol acyltransferase 2 [DGAT2]) and phospholipid synthesis (choline-phosphate cytidyltransferase 1a [PCYT1a], phosphatidylethanolamine N-methyltransferase [PEMT]). Conversely, among numerous proteins down-regulated in the obese liver “ER proteome,” those mediating protein synthesis and transport functions were overrepresented. Interestingly, there was not a broad change in the quantity of ER chaperones, a finding that seems counterintuitive to the present focus of liver ER stress in obesity.
Given the overall reduction in ER-associated protein synthesis, the investigators postulated that hepatic ER stress in obesity may not be driven by protein overloading; rather, the changes in lipid metabolism may mediate a compromise in protein-folding capacity. Others had earlier shown that both palmitic acid and cholesterol,14, 15 which are prime contenders among potential lipotoxic molecules in NASH,16, 17 can induce ER stress in some cell types, and this occurs after incorporation of these lipids into the membrane of the ER.14, 15 This consideration prompted a detailed lipid analysis of ER from livers of obese versus lean (WT) mice, which produced several sets of interesting findings.
First, obese ER was enriched with monounsaturated fatty acids (MUFAs). The authors attribute this to hepatic lipogenesis, because SCD-1, the rate-limiting enzyme catalyzing the first desaturation of saturated fatty acids (SFA) required for their esterification into triglycerides (TG) is coregulated, with FAS, by sterol regulatory element binding protein 1c (SREBP1c). In turn, SREBP1c is activated by insulin, and such activation is a feature of hepatic lipid metabolism in insulin resistance.18 However, it seems equally possible that that up-regulation of SCD-1 is a specific feature of the leptin-deficient ob/ob model rather than a hallmark of obesity, because leptin levels increase in most forms of human and murine obesity, and leptin suppresses SCD-1 independently of insulin or SREBP1c.19 The authors also speculate that desaturation of SFAs to MUFAs contributes to the decrease in polyunsaturated fatty acid (PUFA) content in the ER, which “may limit its reducing capacity and contribute to ER stress.” This proposal is not supported by any evidence presented here. Actually, within the ER, PUFAs are substrates for cytochrome P450 reductase, a process which can form lipoperoxides that would potentially accentuate oxidative stress. If this alternative suggestion is correct, reduced content of PUFAs could arguably be a protective mechanism against ER stress. Similarly, because SFAs are potent inducers of ER stress,15 it can be reasoned that a relative decrease in SFAs with concomitant increased content of MUFAs may be more protective than deleterious in the ER biomembrane. Indeed, in the methionine- and choline-deficient model of NASH (where leptin levels are vanishingly low),20 and in Scd-1−/− mice, decreased MUFA levels leading to decreased TG synthesis and very low density lipoprotein excretion associate with ER stress, as well as energy uncoupling. Similarly, loss of SCD-1 worsens diabetes in leptin-deficient ob/ob mice,21 which would not be expected if SCD-1 induction, resulting in increase of MUFAs, contributes to ER stress.
Notwithstanding the debatable implications of up-regulated SCD-1 expression and changes in SFA, MUFA, and PUFA content for causation of ER stress, a fascinating discovery from the present work is that there are quantitative alterations in ER phospholipid content in ob/ob mouse liver ER. Specifically, there are higher levels of phosphatidylcholine (PC) and lower levels of phosphatidylethanolamine (PE), substantially altering the PC:PE ratio. This relative increase in PC is likely related to up-regulation of two key genes, PCYT1a, which is involved with PC synthesis, and PEMT, which catalyzes conversion of PE to PC. The authors then showed that the change in PC:PE ratio found in the obese liver ER is associated with reduced enzyme activity of the ER Ca2+ translocase, sarco/endoplasmic reticulum calcium adenosine triphosphatase (SERCA). This association of altered phospholipid content and SERCA activity was shown to be causal by an elegant in vitro experiment in which PC was added to WT hepatic microsomes so as to increase ER PC content and hence the PC:PE ratio, a change that simulated that of ob/ob mice. This phospholipid compositional change reproduced the identical impairment of SERCA activity.
Reduced SERCA activity could be seminal to the ER stress response (Fig. 1). The resultant depletion of the ER calcium store would impair protein folding enzymes that are known to be Ca2+-dependent, thereby producing or exacerbating the UPR response and ER stress; in fact, this is the pathway used to provoke ER stress experimentally with agents such as cyclopiazonic acid and thapsigargin, which are known inhibitors of SERCA. In the study by Fu et al., a second approach to altering the PC:PE ratio was also employed, overexpressing PEMT1 in WT liver, and this produced identical results.13 Furthermore, the calcium transport activity of microsomes prepared from livers of obese mice appeared to be lower than those from lean animals. To test the contention that the abnormal PC:PE ratio in ob/ob livers could be responsible for ER stress, knockdown of PEMT expression was then performed. Use of a short hairpin RNA construct achieved a 50%-70% decrease in PEMT expression, and this was sufficient to decrease PC content of ER by 7%, with a corresponding increase in PE. This effectively restored the PC:PE ratio to that of lean mice, with only minor changes in ER fatty acid composition. Indicators of ER stress were suppressed by this restoration of “lean” phospholipid composition. It also appeared to decrease steatosis and hepatic TG levels in the leptin-deficient ob/ob model, albeit any changes in a different, dietary (high-fat diet [HFD]) model appear less impressive from the data presented (supporting figure 5 in Fu et al.13). It was associated with improvement in glucose disposal and lowering of serum insulin levels. Consistent with the PEMT knockdown experiment, SERCA overexpression in ob/ob liver restored SERCA activity, corrected ER stress marker expression and lowered blood glucose levels.
Together, the results presented by Fu et al.13 support the concept that under conditions of increased de novo lipogenesis (particularly with leptin deficiency), MUFA and PC content of the ER membrane increases. The phospholipid compositional changes alter SERCA function, most likely by steric changes in enzyme activity. An alternative proposal published recently is that ob/ob mouse liver has decreased expression of SERCA, molecular restoration of which corrects ER stress and improves insulin sensitivity.22 Whichever of these two suggestions is correct, they both point to a reduction in SERCA activity with resultant depletion ER calcium stores as pivotal to the activation of all three branches of the ER stress response (Fig. 1). Fu et al. also propose that correction of such ER stress improves insulin sensitivity. Others have shown that ER stress contributes to stimulation of lipogenesis (via insulin-activated SREBP1),23 thereby linking a vicious cycle of worsening metabolic state and more extensive steatosis; their concept is portrayed in Fig. 1. In our minds, a missing link is whether decreased steatosis is the reflection of decreased de novo lipogenesis in the liver or results instead from other modifications of bodily lipid metabolism. In obese patients with NAFLD,24 most fatty acids accumulating in the liver are derived from the periphery, not from de novo lipogenesis. It has also been shown repeatedly by the Hotamisligil group25 and other workers23 that inhibition of hepatic ER stress also improves very low density lipoprotein secretion and peripheral insulin resistance, thereby reducing delivery of fatty acids from peripheral tissues (adipose, muscle) to the liver.
We think it important to consider whether the present results, as well as much-detailed earlier study of the potential importance of ER stress to the fatty liver disease in T2D and obesity, could be unduly influenced by the model usually chosen to study these phenomena. As reviewed elsewhere,26, 27 leptin-deficient ob/ob mice show the liver pathology of simple steatosis, not NASH. Fu et al. did reproduce some results similar to those in ob/ob mice upon knockdown of PMET in WT mice fed an HFD, but the extent to which HFD altered hepatic TG levels, lipogenesis or PC:PE ratio or SERCA activity was not shown for those experiments.13 This is a critical consideration, because several nutritional models of obesity exhibit less pronounced de novo lipogenesis than observed in ob/ob livers, and expression of ER stress and NASH pathology varies with these models.9 For example, as opposed to ob/ob mice with steatosis (but not NASH) that show ER stress, HFD-fed C57 mice (with steatosis or NASH) have variable ER stress.28 In addition, although there is direct evidence for ER stress in the liver and adipose tissue of obese humans,10 published data for individual UPR markers and ER stress indicators in those with NAFLD or NASH do not provide a coherent picture of its unambiguous operation in causing either steatosis or steatohepatitis.11 Finally, the true relevance of ER stress as a disease mechanism comes from in vivo studies that used chemical chaperones to block its operation.29 One of these chaperones is tauroursodeoxycholic acid. It may therefore be salient to note that this is the agent that we are most confident has no therapeutic efficacy in patients with NASH.30, 31
To conclude, relationships between hepatic ER stress, lipogenesis, insulin resistance, and steatosis in obesity/metabolic syndrome have been the subject of intense scrutiny for the last decade. Work such as that by Fu et al. forms a benchmark for the molecular detail that can be gleaned from contemporary reductionist science and molecular manipulations. It shows how ER Ca2+ homeostasis is central to the operation of cellular protective responses in the stressed liver. However, much more work needs to be done in models that are more relevant to obese humans with metabolic syndrome, as well as careful studies in affected humans, before we can conclude that ER stress has particular relevance to the pathogenesis of NASH itself.