Elevated circulating stearic acid leads to a major lipotoxic effect on mouse pancreatic beta cells in hyperlipidaemia via a miR-34a-5p-mediated PERK/p53-dependent pathway
- 1.7k Downloads
Serum stearic acid (C18:0) is elevated in individuals with hyperlipidaemia and type 2 diabetes. However, the lipotoxicity induced by increased stearic acid in beta cells has not been well described. This study aimed to examine the adverse effects of stearic acid on beta cells and the potential mechanisms through which these are mediated.
Three groups of C57BL/6 mice were fed a normal diet or a high-stearic-acid/high-palmitic-acid diet for 24 weeks, respectively. The microRNA (miR) profiles of islets were determined by microarray screening. Islet injury was detected with co-staining using the TUNEL assay and insulin labelling. A lentiviral vector expressing anti-miRNA-34a-5p oligonucleotide (AMO-34a-5p) was injected into mice via an intraductal pancreatic route.
In both mouse islets and cultured rat insulinoma INS-1 cells, stearic acid exhibited a stronger lipotoxic role than other fatty acids, owing to repression of B cell CLL/lymphoma 2 (BCL-2) and BCL-2-like 2 (BCL-W) by stearic acid stimulation of miR-34a-5p. The stearic-acid-induced lipotoxicity and reduction in insulin secretion were alleviated by AMO-34a-5p. Further investigations in INS-1 cells revealed that p53 was involved in stearic-acid-induced elevation of miR-34a-5p, owing in part to activation of protein kinase-like endoplasmic reticulum kinase (PERK). Conversely, silencing PERK alleviated stearic-acid-induced p53, miR-34a-5p and lipotoxicity.
These findings provide new insight for understanding the molecular mechanisms underlying not only the deleterious impact of stearic-acid-induced lipotoxicity but also apoptosis in beta cells and progression to type 2 diabetes.
KeywordsLipotoxicity MicroRNA Pancreatic beta cells Stearic acid
Activating transcription factor 6
B cell CLL/lymphoma 2
CCAAT/enhancer binding protein homologous protein
Glucose-induced insulin secretion
Lentiviral vector expressing anti-miRNA-34a-5p oligonucleotide
Protein kinase-like endoplasmic reticulum kinase
Saturated fatty acids
Lipotoxicity refers to the cellular dysfunction caused by an elevated level of NEFA. Accumulating evidence indicates that lipotoxicity contributes to beta cell injury and plays a critical role in the development of type 2 diabetes [1, 2, 3]. NEFA is chemically classified into saturated and unsaturated fatty acids. It is widely accepted that lipotoxicity is mostly induced by long-chain saturated fatty acids (SFA), such as palmitic (C16:0) and stearic acid (C18:0), whereas unsaturated fatty acids are, in general, less toxic, and may even be protective [4, 5, 6].
There is growing evidence that chronic consumption of a diet rich in saturated fats increases circulating SFA levels and the risk of type 2 diabetes [7, 8]. It has been observed that NEFA profiles in individuals with obesity or hyperlipidaemia are substantially altered compared with healthy individuals; such changes are regarded as an important risk factor for type 2 diabetes [7, 9]. Palmitic acid is widely employed as a representative fatty acid to investigate the relationship between SFA and type 2 diabetes because it accounts for the highest percentage of dietary NEFA for the human body. However, whether other alterations in NEFA profiles have a detrimental role in beta cell destruction and type 2 diabetes is unknown. Our previous study revealed that the profiles of both fasting and postprandial serum NEFA were significantly altered in patients with hyperlipidaemia . Most strikingly, we observed that only the proportion of stearic acid in the serum increased dramatically in the postprandial state, which implies that patients with hyperlipidaemia have a considerably higher level of serum stearic acid most of the time. In animal models of insulin resistance, the ratio between palmitic and stearic acids was decreased from 3:1 in the normal group to 2:1 in mice fed a high-fat diet . Research in vitro has clearly shown that increasing stearic acid stimulates severe lipotoxicity in hepatocytes and skeletal muscle cells [5, 10, 11]. Added to our previous observation that an increased proportion of stearic acid in the serum is closely related to insulin resistance , we proposed that raising the proportion of stearic acid in the blood makes the greatest contribution to beta cell lipotoxicity relative to other NEFA, and plays a dominant role in type 2 diabetes progression. Moreover, we sought to explore the molecular mechanisms underlying the destructive effect of stearic acid on beta cells.
Although the molecular mechanisms of stearic-acid-induced lipotoxicity are not fully understood, it is evident that enhanced endoplasmic reticulum (ER) stress is a central contributor to lipotoxicity resulting from transcriptional reprogramming , activation of c-Jun N-terminal kinase , CCAAT/enhancer binding protein homologous protein (CHOP) , p53 and mitochondrial apoptosis pathways [15, 16]. Importantly, microRNAs (miRNAs), which are involved in post-transcriptional regulation under ER stress, have emerged as key regulators of hepatocyte and cardiac muscle lipotoxicity [17, 18]. In islet beta cells, the involvement of miRNA in stearic-acid-mediated lipotoxicity is still unclear.
In this study, we hypothesised that increasing stearic acid induces more severe beta cell lipotoxicity than other NEFA and its underlying mechanism is related to miRNA (miR)-34a-5p.
Stock solutions of stearic and palmitic acids (Sigma, St Louis, MO, USA) were prepared as described previously . The oligonucleotides and lentivirus are described in the electronic supplementary material (ESM) Methods.
All experimental procedures followed guidelines for the care and use of animals established by Harbin Medical University and approved by the Animal Experimentation Ethics Committee of Harbin Medical University. Seven-week-old male C57BL/6 mice were purchased from Vital River Laboratories (Beijing, China) and completely randomised to receive a normal diet, a high-stearic-acid diet (HSD) or a high-palmitic-acid diet (HPD) (ESM Table 1). A lentiviral vector expressing anti-miRNA-34a-5p oligonucleotide (AMO-34a-5p) (lenti-AMO-34a-5p) (1 × 108 TU/ml) was slowly injected into mice via the pancreatic duct (100 μl) after feeding for 24 weeks, as described previously . Further measurements were performed at 4 weeks post-lentiviral infection. See ESM Methods. Experimenters were blind to group assignment and outcome assessment.
Serum NEFA profile and glucose and lipid measurement
Blood samples were collected for further detection. See ESM Methods.
miRNA microarray analysis
Oebiotech (Shanghai, China) performed the miRNA microarray work, processing 12 samples from mouse islets (n = 4 per group). See ESM Methods.
Co-staining using TUNEL assay and insulin-labelled islets
Islet injury was detected with the In Situ Cell Death Detection Kit (TUNEL Fluorescence Kit, Roche, Indianapolis, IN, USA). Islets were labelled with anti-insulin antibody (Cell Signaling Technology, Minneapolis, MN, USA), and viewed by confocal microscopy. See ESM Methods.
Pancreatic insulin content
The pancreas was extracted using the acid ethanol method as described previously , and the insulin level was detected.
Intravenous glucose tolerance tests
IVGTTs were performed on overnight-fasted mice. For the measurement of insulin and glucose levels, blood samples were collected at 0, 1, 5, 10, 20, 30 and 60 min after glucose administration (0.75 g/kg) to the tail vein.
Beta cell mass and histochemical analysis
The pancreatic tissues were fixed, embedded in paraffin and stained with anti-insulin and anti-glucagon antibodies. See ESM Methods.
Insulin granules analysis
Electron microscopy was employed to analyse the content of insulin granules in fixed pancreatic tissues as described previously .
Rat insulinoma cell line INS-1 cells were cultured and treated with 600 μmol/l stearic or other fatty acids for 24 h. Mouse pancreatic islets were isolated as described previously . See ESM Methods. Experimenters were blind to group assignment and outcome assessment.
Cell viability measurements
Cell viability was determined by assessment of lactate dehydrogenase (LDH) release, propidium iodide (PI) staining, 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay and Hoechst staining. See ESM Methods.
INS-1 cells were transfected with oligonucleotides or recombinant plasmid using Lipofectamine 2000 (Invitrogen), as described previously , and used in detection assays 48 h after transfection.
Luciferase activity assay
The luciferase activity was detected using a luciferase activity assay. See ESM Methods.
Insulin levels in mouse serum and cell culture media were determined by mouse/rat insulin ELISA kit (Linco Research, St Charles, MO, USA). To assess glucose-induced insulin secretion (GSIS), cells were incubated in secretion buffer (see ESM Methods) for an additional 60 min with 2.8 or 20 mmol/l glucose.
Values are expressed as mean ± SD. Multiple groups were analysed with one-way ANOVA followed by a Student–Newman–Keuls test. Two-group-only comparisons were carried out by t test. A two-sided p value < 0.05 was considered statistically significant. Repeated-measures ANOVA was used to analyse the glucose and insulin levels during IVGTT.
Stearic acid induces more severe lipotoxicity than palmitic acid
Stearic acid increases miR-34a-5p levels in both mouse islets and INS-1 cells
Quantitative PCR verified that miR-34a-5p levels increased by 191% in HSD and 97% in HPD islets, respectively, compared with normal islets (Fig. 2b). In INS-1 cells exposed to different ratios of palmitic/stearic acids, miR-34a-5p expression was markedly upregulated compared with controls in cells exposed to a higher proportion of stearic acid, (Fig. 2c).
miR-34a-5p contributes to stearic-acid-induced lipotoxicity in islets and INS-1 cells
miR-34a-5p specifically represses BCL-2 and BCL-W expression
p53 is involved in stearic acid-induced elevation of miR-34a-5p
PERK contributes to stearic-acid-induced upregulation of miR-34a-5p and lipotoxicity
In this study, we provide evidence that, relative to other NEFA, stearic acid is the major contributor to lipotoxicity in beta cells, implying that an increased concentration of serum stearic acid plays a fundamental role in the development of beta cell dysfunction and type 2 diabetes. Our mechanistic studies revealed that miR-34a-5p upregulation modulates stearic-acid-induced lipotoxicity by inhibiting the production of BCL-2 and BCL-W proteins. Additional studies have also demonstrated that the ER-stress-related sensor PERK contributes to stearic-acid-induced upregulation of miR-34a-5p in a p53-dependent signalling pathway.
Although alterations in the NEFA profiles of people with hyperlipidaemia have been widely investigated, the NEFA that is most directly associated with beta cell lipotoxicity remains unclear. Therefore, we first selected NEFA with statistically discrepant levels between healthy and hyperlipidaemic individuals to test their cytotoxicity to pancreatic beta cells. We observed that only stearic and palmitic acids significantly lowered cell viability. Incubating cells in medium containing different ratios of palmitic acid:stearic acid dramatically enhanced cytotoxicity, but insulin secretion was reduced as the proportion of stearic acid was increased. The concentrations of palmitic and stearic acids used in this study were selected to avoid excessive cell death based on previous reports and our previous experiments [7, 26, 27], though this does not happen in human or animal pancreatic beta cells in vivo [26, 27]. The concentrations of palmitic and stearic acids employed in cultured cells were, in fact, lower than in patients with hyperlipidaemia, but induced more severe damage than in the in vivo system, which could be explained by the existence of protective factors (i.e. unsaturated fatty acids [4, 5, 6]). Additionally, we observed that eicosapentaenoic acid (EPA; C20:5), a potential protective factor, was reduced in both HSD and HPD mice, and this might also contribute to high-fat-diet-induced lipotoxicity. Unlike the in vitro system, it is unrealistic to establish animal models with only stearic- or palmitic-acid-constructed lipid profiles in blood. However, a well-accepted approach to compare different biological function among the different types of NEFA was achieved by regulating the blood NEFA ratio using various oils as a fat source in the high-fat diet [28, 29]. In this study, models of hyperlipidaemia with high circulating levels of either stearic or palmitic acid were established using lard or palm oil as a fat source, consistent with previous studies [28, 29, 30]. After 24 weeks of feeding, the fasting serum levels of stearic and palmitic acids increased in HSD and HPD mice, respectively. The palmitic acid:stearic acid ratio was 1.7:1 and 6.2:1 in HSD and HPD mice, respectively. Subsequently, the HSD mice showed significant impairment of glucose tolerance and GSIS compared with HPD mice, with an increased number of apoptotic beta cells leading to a decrease in beta cell mass. Our results clearly show that, in vivo, HSD induces more damage in islets than HPD, consistent with the in vitro results. No difference was observed in the total SFA content between the HSD and HPD groups, excluding the possibility that the different extent of lipotoxicity was caused by something other than differences in lipid concentrations. Additionally, these results are also consistent with our previous findings [7, 31], indicating that stearic acid appears to be a more important risk factor for type 2 diabetes than other NEFA. Therefore, great attention should be paid to elevated serum stearic acid, as it appears to have a greater negative impact on beta cell function.
The mechanistic pathways responsible for HSD-induced lipotoxicity in pancreatic beta cells are not fully understood. Recent evidence has provided new clues that miRNAs are involved in the regulation of beta cell failure and type 2 diabetes [32, 33, 34]. In our study, we selected 19 miRNAs that are differentially expressed in islets of HSD mice relative to HPD and normal mice, based on gene-chip microarray analysis. Among these miRNAs, miR-34a-5p displayed the largest fold increase, which was in line with previous observations that a high level of miR-34a-5p was present in the islets of diabetic db/db mice and pancreatic islets exposed to palmitic acid . To demonstrate the potential role of miR-34a-5p in HSD-mediated lipotoxicity in beta cells, endogenous miR-34a-5p was abrogated by AMO-34a-5p in both mice and INS-1 cells. Lenti-AMO-34a-5p injection ameliorated HSD-induced islet apoptosis in vivo. Accordingly, impaired glucose tolerance and GSIS in HSD mice were largely restored. The decrease in beta cell mass seen in HSD mice was also reversed. Insulin secretion and beta cell mass are two important factors for regulating glucose homeostasis through well-balanced beta cell compensation in the early stage of type 2 diabetes, but they gradually worsen [35, 36]. Our results indicate that miR-34a-5p inhibition in HSD islets could probably improve the beta cell dysfunction and further slow the progression of type 2 diabetes. In vitro, AMO-34a-5p relieved the cytotoxic effect of stearic acid and reversed the reduction in insulin secretion. Conversely, overexpression of miR-34a-5p significantly enhanced cytotoxicity and reduced GSIS. These results strongly suggest that miR-34a-5p plays a crucial role in elevating lipotoxicity induced by serum stearic acid. However, the stearic-acid-mediated decrease in GSIS was not completely reversed by miR-34a-5p inhibition, which implies that a non-miR-34a-5p-mediated mechanism operated in GSIS reduction. We also observed that miR-34a-5p levels were higher in HSD islets and stearic-acid-treated INS-1 cells than in HPD islets and the palmitic acid group, respectively, which suggests that the greater adverse effect of stearic over palmitic acid on beta cells, both in vivo and in vitro, is probably attributable to differential regulation of miR-34a-5p.
Previous studies have indicated that miR-34a-5p is closely associated with apoptosis through the BCL-2 family in cancer cells, tubular cells and cardiomyocytes [37, 38, 39]. The BCL-2 family includes both anti-apoptosis members (BCL-2, BCL-2-like 1 [BCL-XL] and BCL-W) and pro-apoptosis members (BCL-2-associated X protein [BAX], BCL-2-antagonist/killer [BAK] and BCL-2-associated agonist of cell death [BAD]) [40, 41]. In our study, we confirmed that BCL-2 and BCL-W are direct targets of miR-34a-5p using luciferase activity assays. However, how stearic acid regulates miR-34a-5p levels is still unclear. Accumulating evidence supports p53 as a classic transcription factor regulating miR-34a-5p. To further investigate the involvement of p53 in miR-34a-5p upregulation after long-term incubation of INS-1 cells with stearic acid, we silenced p53 and then investigated whether miR-34a-5p levels were reversed, with an associated reduction in lipotoxicity, which would suggest that the mechanism underlying HSD-induced stimulation of miR-34a-5p uses a p53-dependent pathway.
Enhanced ER stress, associated with beta cell lipotoxicity, transduces apoptotic signals in the early stage of type 2 diabetes . Three major signalling pathways separately initiated by ER transmembrane sensor proteins, including IRE1, PERK and ATF6, are involved in ER stress . We subsequently knocked down Ire1, Perk and Atf6 with siRNAs in stearic-acid-treated INS-1 cells, and found that inhibition of Ire1 and Atf6 failed to inhibit stearic-acid-induced stimulation of miR-34a-5p. However, transfection of siRNA for Perk attenuated lipotoxicity and miR-34a-5p levels on stearic acid treatment. These findings demonstrate that miR-34a-5p might be promoted by PERK in conditions of ER stress. We also discovered that PERK exhibited an inhibitory effect on p53 protein expression, suggesting that a PERK–p53–miR-34a-5p axis is mechanistically involved in stearic-acid-induced lipotoxicity in beta cells.
Additionally, the inflammatory response was the common reaction during prolonged exposure to NEFA, which in turn results in the aggravation of lipotoxicity. Chronic exposure to inflammatory conditions leads to beta cell dysfunction and apoptosis [32, 44, 45]. A recent investigation has revealed that the level of miR-34a increased concomitantly in cytokine-treated MIN6 cells and human pancreatic islets, and contributed to beta cell dysfunction . Therefore, we proposed that the HSD-induced increase in miR-34a-5p and further pancreatic injury probably represented an important inflammatory response pathway. This needs to be clarified in future studies.
In conclusion, we provide evidence that elevated levels of circulating stearic acid induce severe lipotoxicity in pancreatic beta cells. It appears that upregulation of miR-34a-5p and the associated inhibition of BCL-2/BCL-W play a critical role in mediating stearic acid cytotoxicity. Our results also suggest a regulatory role for PERK in miR-34a-5p expression and that p53 is a mediator of PERK-induced miR-34a-5p upregulation. Collectively, these findings allow us to propose a novel signalling pathway linking stearic acid to lipotoxicity in beta cells: stearic acid–PERK–p53-miR-34a-5p–BCL-2/BCL-W. These findings provide a novel insight into stearic-acid-induced lipotoxicity and progression of type 2 diabetes, whereby lowering miR-34a-5p might be an effective strategy for improving lipotoxicity in beta cells.
We thank B. X. Wu (Department of Nutrition and Food Hygiene, Public Health College, Harbin Medical University) for helping with serum glucose and lipid measurement and Q. N. Qu and Y. Y. Liu (Department of Nutrition and Food Hygiene, Public Health College, Harbin Medical University) for help with animal feeding.
This work was supported by the State Key Program of National Natural Science of China (grant number 81130049), and the National Natural Science Foundation of China (grant numbers 81202187, 81202188 and 81202191).
Duality of interest
The authors declare that there is no duality of interest associated with this manuscript.
CS, LN and SLi conceived the experiments, analysed and interpreted the data and reviewed and edited the manuscript. HL and SLi were responsible for designing and performing the study, acquisition and analysis of data and writing the draft of the manuscript. HL, LH, SLin, LL, YC, HC, TZ, XC and LN performed the study, collected data and interpreted and drafted the manuscript. All authors approved the final version of the manuscript. CS is the guarantor of this work.