Di(2-ethylhexyl) phthalate-induced apoptosis in rat INS-1 cells is dependent on activation of endoplasmic reticulum stress and suppression of antioxidant protection

Di(2-ethylhexyl) phthalate (DEHP) is used as plasticizer and is ubiquitously found in the environment. Exposure to DEHP has been linked to an increased incidence of type 2 diabetes. Pancreatic β-cell dysfunction is a hallmark of type 2 diabetes; however, it is unknown whether DEHP exposure contributes to this risk. Here, we aimed to investigate the cytotoxic effects of DEHP on INS-1 cells and to further explore the related underlying mechanisms. INS-1 cells were exposed to 0, 5, 25, 125 or 625 μM DEHP for 24 hrs. Cell viability, glucose-stimulated insulin secretion, reactive oxygen species (ROS) generation, cellular antioxidant response, Ca2+ homoeostasis and the levels of genes and proteins involved in endoplasmic reticulum (ER) stress were measured. The results showed that DEHP decreased insulin secretion and content and induced apoptosis in INS-1 cells in a dose-dependent manner. Furthermore, ROS generation was increased and Nrf2-dependent antioxidant defence protection was dysregulated in INS-1 cells after DEHP exposure. Most importantly, DEHP effectively depleted ER Ca2+ and triggered the ER stress response as demonstrated by the elevated transcription and translation of the ER chaperone GRP78 and GRP94, the increased phosphorylation of protein kinase R-like endoplasmic reticulum kinase (PERK) and its downstream substrate eukaryotic translation initiation factor 2α (eIF2α), as well as the increased levels of activating transcription factor 4 (ATF4) and C/EBP homologous protein (CHOP). Taken together, DEHP exerted toxic effects on INS-1 cells by inducing apoptosis, which is dependent on the activation of the PERK–ATF4–CHOP ER stress signalling pathway and the suppression of Nrf2-dependent antioxidant protection.


Introduction
Di (2-ethyl hexyl) phthalate (DEHP) is one of various ubiquitous environmental endocrine disruptors widely used as plasticizers or solvents in food packaging, medical devices, pharmaceutical formulations, household products and industrial plastic, etc. [1,2]. DEHP is not covalently bound to the plastic matrix; therefore it can easily leach out of products into the environment. Human exposure to DEHP is widespread and frequent. Biomonitoring data continuously collected in the National Health and Nutrition Examination Survey (NHANES) since 1999 showed that a majority of individuals (>6 years old) in the United States had detectable concentrations of DEHP and other metabolites in the urine [3]. In Canada, a national health measure survey (CHMS) from 2007 to 2009 also indicated that DEHP metabolites were detected in >90% of Canadians [4].
DEHP and the occurrence of diabetes in the United State [7]. In animal studies, DEHP was also associated with diabetogenic effects. For example, administration of 10 or 100 mg/kg/day DEHP for 30 days significantly disrupted insulin signal transduction and diminished plasma membrane GLUT4 level in both adipose and skeletal muscle of adult male albino rats, implying that DEHP was able to induce insulin resistance by decreasing glucose oxidation and uptake in peripheral tissue [8,9].
In addition to the increased insulin resistance, progressive pancreatic b-cell dysfunction is a major pathogenic component in type 2 diabetes [10]. b cells express extremely low levels of antioxidant enzymes such as catalase and glutathione peroxidase [11,12], and they are particularly susceptible to reactive oxygen species (ROS)elicited oxidative stress damage. DEHP has been reported to be involved in peroxisome proliferator-activated receptor activation; therefore, it may induce oxidative stress by increasing peroxidase expression and ROS generation [13]. Indeed, NHANES in 1999 reported that higher urinary concentrations of DEHP metabolites were inversely associated with bilirubin [14], a marker of oxidative stress in the prediction of metabolic disease because of its antioxidant activity [15]. Furthermore, animal studies confirmed that DEHP exposure causes oxidative stress in reproductive tissues and cells by disrupting antioxidant defences and increasing ROS [16,17]. Oxidative stress is mostly implicated as a potential mechanism for functional failure of pancreatic b cells in type 2 diabetes. Whether exposure to DEHP impairs b cells through an oxidative stress pathway remains to be elucidated.
Reactive oxygen species is a group of upstream signalling molecules in the endoplasmic reticulum (ER) stress signalling pathway [18,19]. In pancreatic b cells, ER is a crucial site for insulin biosynthesis [20]. ER stress-mediated b-cell dysfunction plays an important role in type 2 diabetes. In a previous study, Wistar rat offspring perinatally exposed to DEHP at 1.25 and 6.25 mg/kg/day showed b-cell defects that was characterized by hypertrophic rough ER, reduced mass and decreased insulin content as well as altered the expression of ER stress genes in islets [21]. ER damage could be another important mechanism underlying DEHP-induced b-cell dysfunction.
In the present study, we treated cells lines derived from a rat insulinoma, INS-1 cells, with DEHP and assessed cellular insulin secretion and content, ROS generation and antioxidant response, as well as activation of the ER stress pathway and apoptosis. This study is the first to identify the potential cytotoxicity of DEHP on b cells and to evaluate the molecular mechanism involved in DEHP-induced b-cell dysfunction and apoptosis.

Materials and methods
Cell culture and treatment INS-1 cells (between passages 38 and 67), were purchased from Biohermes (Shanghai, China) and were cultured in RPMI-1640 medium supplemented with 10% foetal bovine serum (FBS), 50 lM b-mercapto-ethanol, 25 mM HEPES, 2 mM L-glutamine, 5.6 mM glucose, 100 U/ml penicillin and 100 mg/ml streptomycin at 37°C under 5% CO 2 . Culture medium, FBS and supplements were purchased from Invitrogen (Carlsbad, CA, USA). The cells were seeded into 96-, 24-or 6-well plates depending on the experiment. After a 24-hr pre-culture, the cells were incubated with serial concentrations of DEHP (0, 5, 25, 125 or 625 lM) for 24 hrs. DEHP was dissolved in dimethyl sulfoxide (DMSO); the final concentration of DMSO in medium was 0.1% (v/v) in all experiments.

Glucose-stimulated insulin secretion (GSIS)
INS-1 cells were cultured in 24-well plates and exposed to DEHP for 24 hrs. After exposure, the medium was removed and the cells were incubated with Krebs-Ringer bicarbonate HEPES buffer (KRBH) supplemented with 0.5% BSA and 3.0 mM glucose for 1 hr. Afterwards, the cells were challenged to secrete insulin in KRBH buffer containing 3.0, 5.6 or 16.7 mM glucose for an additional 1 hr. Supernatants were collected and measured with a Rat Insulin ELISA kit (Millipore, Billerica, MA, USA).

Cell viability assay
INS-1 cells were seeded into 96-well plates and stimulated with DEHP for 24 hrs. Then, the medium was removed and the cells were incubated with 0.5 mg/ml MTT (Sigma-Aldrich, St. Louis, MO, USA) for 4 hrs at 37°C. The resulting formazan crystals were solubilized by DMSO and the absorbance was read at 490 nm on a SpectraMAX M5 microplate reader (Molecular Devices, Sunnyvale, CA, USA). The OD values of controls were taken set at 100% and the values for DEHP-treated cells were expressed as % of controls.

Determination of ROS production
Intracellular ROS generation was measured using a peroxide-sensitive fluorescent probe: 2,7-dichlorofluorescein diacetate (DCFH-DA; Sigma-Aldrich). In brief, cells were seeded in 6-well plates and exposed to various concentrations of DEHP for 24 hrs. Then, the cells were washed with PBS and co-incubated with 10 lM DCFH-DA for 30 min. at 37°C. After incubation, the cells were washed and re-suspended with ice-cold PBS. Intracellular ROS was imaged by the LSM 710 Laser Scanning Confocal Microscope (Zeiss, Jena, Germany) using a condition of 488 nm for excitation and 525 nm for emission. The image density was quantified with an Image Pro Plus version 6.0 software (Media Cybernetics, Bethesda, MD, USA). Five images per treatment were captured: one image in each of the four quadrants and one in the centre of the well [22]. Equivalent adjustments for brightness and contrast were applied to each image.

Immunofluorescence analysis of Nrf2 localization
Cells were seeded in sterile glass slides and exposed to various concentrations of DEHP for 24 hrs. After exposure, the cells were fixed with 4% paraformaldehyde solution for 15 min., and were then treated with permeation solution (0.2% Triton X-100 and 0.2% BSA in PBS) for 10 min. on ice. The cells were blocked with 0.02% Triton X-100 and 5% BSA in PBS for 2 hrs and incubated overnight with nuclear factor erythroid 2-related factor 2 primary antibodies (Nrf2, Q16236; 1:250; Epitomics, CA, USA) at 4°C. In the next day, the cells were incubated with Alexa Fluor â 555 Donkey Anti-Rabbit IgG (A31572; 1:500; Invitrogen) for 1.5 hrs. Finally, the slides were mounted with DAPI (1 lg/ml; Roche) for 10 min. and the subcellular localization of Nrf2 was imaged using an LSM 710 Laser Scanning Confocal Microscope.

Intracellular Ca 2+ concentration ([Ca2+] i ) measurement
Intracellular Ca 2+ concentrations [Ca 2+ ] i were measured using Ca 2+ -sensitive fluorescent Fluo-3/acetoxymethyl (AM; AAT Bioquest, Sunnyville, CA, USA). After 24 hrs of DEHP treatments, cells were incubated with Fluo-3/AM in Ca 2+ -and Mg 2+ -free HBSS (Invitrogen) at 37°C for 30 min. Then, the cells were washed with Ca 2+ -and Mg 2+ -free HBSS to remove the unloaded probes and incubated with the same saline solution supplemented with 1.0% (v/v) FBS at 37°C for 40 min. to completely de-esterify the dye. Finally, the cells were washed and re-suspended in Ca 2+ -and Mg 2+ -free HBSS until the fluorescence was measured. For imaging, INS-1 cells were monitored by the Laser Scanning Confocal Microscope (Zeiss), using a 506 nm laser for excitation and a 526 filter for emission. Free [Ca 2+ ] i were determined using a microplate reader (Molecular Devices) and the following equation: where F is the fluorescence intensity of the indicator at experimental calcium levels, F min is the fluorescence intensity in the absence of calcium and F max is the fluorescence intensity of the calcium-saturated probe. The dissociation constant (K d ) for Fluo-3 is assumed to be 325 nM. To estimate the ER Ca 2+ concentration ([Ca 2+ ] ER ), the treated cells were loaded with Fluo-3/AM as described above and stimulated with 1 lM thapsigargin (Tg; Sigma-Aldrich Total RNA preparation and real-time PCR Total RNA was extracted from INS-1 cells using a High Pure RNA kit (Roche) and was reverse-transcribed with the PrimeScript TM RT-PCR Kit (Takara, Dalian, China). Real-time PCR was performed with a SYBR â Premix Ex Taq TM II kit (Takara) according to the manufacturer's instructions using Light Cycler 240 Instrument (Roche Applied Science). Relative mRNA concentrations were calculated using the 2 ÀDDCt method, where Ct was the mean threshold cycle value and 36B4 was used to normalize. The primers sequences are listed in Table 1.

Statistical analyses
The data are presented as means AE SEM. Statistical analyses were performed with SPSS 17.0 software (SPSS Inc., Chicago, IL, USA) using one-way ANOVA followed by Bonferroni's post hoc test. Data were considered significant when P < 0.05 (*), <0.01 (**) or <0.001 (***) in the comparison test.  Figure 1A showed that insulin release at basal glucose concentration (3.0 mM) was not affected by DEHP in INS-1 cells. In addition, 5 and 25 lM DEHP did not significantly alter the insulin secretion response evoked by 5.6 mM glucose. However, insulin secretion in response to 5.6 mM glucose of the cells incubated with 125 or 625 lM DEHP for 24 hrs was reduced to 54.8% and 46.2% of untreated control cells respectively. The inhibitory effect of DEHP on insulin secretion was higher in response to 16.7 mM glucose than in response to 5.6 mM glucose. Insulin secretion in the presence of 16.7 mM glucose was decreased to 81.1, 52.8, 7.4 and 3.2% of that detected in untreated control cells after 24 hrs of incubation with 5, 25, 125 or 625 lM DEHP respectively (Fig. 1A). At the molecular level, the mRNA amount of insulin was not affected by 5 lM DEHP in INS-1 cells, but was significantly decreased after exposure to 25, 125 or 625 lM DEHP (Fig. 1B). Compared with the untreated control cells, insulin protein levels were found to be markedly decreased in the cells exposed to 125 or 625 lM DEHP (Fig. 1C). No difference was detected in the level of insulin protein between 5 or 25 lM DEHPexposed and the control cells (Fig. 1C).

DEHP decreases viability and induces apoptosis in INS-1 cells
To test the cytotoxic effect of DEHP, we initially examined the cell viability by MTT assay. Figure 2A showed that cell viability was slightly higher in the 5 lM DEHP-exposed cells than that in the control, but this effect did not reach statistical significance. Conversely, DEHP treatment significantly decreased cell viability to 64.2 (25 lM), 43.3 (125 lM) or 35.1% (625 lM) of the control respectively ( Fig. 2A). The adverse effects of DEHP on cell viability may be related to the imbalance of cell proliferation and apoptosis. Therefore, the cell-proliferative effect was then determined by the assay using BrdU. Figure 2B showed that exposure to DEHP led to changes in BrdU incorporation in a non-monotonic dose-response manner. Treatment with 25, 125 and 625 lM DEHP significantly reduced BrdU incorporation, compared with the control cells, whereas 5 lM DEHP moderately increased it (Fig. 2B). To evaluate b-cell apoptosis, cells were stained with Annexin V-PI, and the results demonstrated that treatment of INS-1 cells with 25, 125 or 625 lM DEHP led to a marked and dose-dependent increase in the amount of early apoptotic cells compared with the controls cells (Fig. 2C). Moreover, 125 and 625 lM DEHP also increased the amount of advanced apoptotic cells (Fig. 2C). There was no difference in the extent of early apoptosis and apoptosis induction by 5 lM DEHP compared with the control (Fig. 2C). Consistently, full-length caspase-3 was decreased and cleaved caspase-3 (19 kD) was increased in 25, 125 or 625 lM DEHP-exposed cells compared with controls (Fig. 2D), indicating that DEHP caused caspase-3 activation and induced INS-1 cells apoptosis at higher concentrations.

DEHP stimulates ROS production and alters Nrf2dependent antioxidant response in INS-1 cells
Intracellular ROS levels were measured using a DCFH-DA probe and the results showed that DCFH-DA fluorescence was increased significantly in the cells exposed to 25, 125 or 625 lM DEHP compared with controls (Fig. 3A). As excess ROS generation may be related to an inadequate of antioxidant defence response, we next measured the nuclear accumulation of Nrf2 as well as the transcription of Nrf2 and its target genes. Figure 3B-D showed that higher concentrations of DEHP significantly inhibited the Nrf2-dependent antioxidant response. Both the protein levels of Nrf2 in the nucleus and the cytosol and the mRNA amounts of Nrf2 and its downstream antioxidant enzyme genes, Nqo1, Gclm, Hmox1 and Gpx4, were significantly decreased in the cells exposed to 25, 125 or 625 lM DEHP when compared with the control. In addition, mRNA of the Cat and Gclc were also decreased after exposure to 625 lM DEHP (Fig. 3D). In contrast, 5 lM DEHP activated the Nrf2-mediated adaptive response in INS-1 cells. Figure 3B showed that the nuclear Nrf2 was increased but cytosolic Nrf2 was decreased in the 5 lM DEHP-exposed cells compared with the control. Similar results were observed in immunofluorescence analysis of Nrf2 localization, showing that 5 lM DEHP treatment slightly increased perinuclear localization and nuclear translocation of Nrf2 (Fig. 3C). Apart from nuclear translocation, 5 lM DEHP also induced transcriptional up-regulation of Nrf2 and many Nrf2-target genes such as Nqo1, Gclc, Gclm, Cat, Hmox1 and Gpx4 in INS-1 cells (Fig. 3D).

DEHP induces the ER stress response in INS-1 cells
Endoplasmic reticulum stress pathway signalling and the levels of downstream target genes were measured to test the hypothesis that early apoptosis occurred in INS-1 cells treated with DEHP as a result of ER stress response. Figure 4 showed that both the mRNA and protein levels of an ER chaperones, GRP78 were dose-dependently upregulated in cells exposed to 25, 125 or 625 lM DEHP. Similarly, the level of GRP94 protein was significantly increased in 25 or 625 lM DEHP-exposed cells and its mRNA levels were increased by exposure  Figure 4A showed that the amounts of phosphorylation of PERK and its substrate eIF2a were significantly increased after exposure to 25, 125 or 625 lM DEHP. Phosphorylated PERK and eIF2a were also slightly elevated in cells exposed to 5 lM DEHP; however, the effects did not reach statistical significance (Fig. 4A). Levels of ATF4 mRNA and protein, known to be induced by eIF2a   phosphorylation, were significantly elevated in the cells after exposure to 25, 125 or 625 lM DEHP (Fig. 4). Apart from eliciting the ER stress signalling pathways, 125 or 625 lM DEHP dramatically increased transcription and translation of a pro-apoptotic factors, CHOP (Fig. 4). mRNA levels of several target genes of CHOP that encode pro-apoptotic functions, such as Gadd34 and Trib3, were also up-regulated in the cells exposed to 125 or 625 lM DEHP compared with the control (Fig. 4B). The amount of CHOP-dependent Ero1a, which can generate oxidizing equivalents in the ER, was shown to be induced by 125 or 625 lM DEHP in INS cells (Fig. 4B). In addition, mRNA level of Dnajc3 was decreased in 5 lM DEHP-exposed cells, but unaltered in 25, 125 or 625 lM  (Fig. 4B).

DEHP disturbs Ca 2+ homoeostasis in INS-1 cells
Because aberrant Ca 2+ homoeostasis has been shown to mediate ER stress and b-cell apoptosis, the effect of DEHP on Ca 2+ homoeostasis was evaluated with a fluorescent Ca 2+ indicator Fluo-3/AM. Figure 5A showed that DEHP exposure led to a significant and dose-dependent increase in [Ca 2+ ] i in INS-1 cells. Subsequently, cells were treated with Tg, which was able to dissipate ER Ca 2+ storage resulting in increased cytosolic free Ca 2+ which indirectly reflected the Ca 2+ level at ER [23]. As shown in Figure 5B, untreated control cells exhibited a sharp increase in [Ca 2+ ] i response to Tg stimulation, but Tg-driven increase in [Ca 2+ ] i was dose-dependently blunted when cells were exposed to DEHP. Likewise, exposure to 5-625 lM DEHP significantly reduced the mRNA levels of Serca2, which is essential for the maintenance of calcium movement across the cells. The expression of PMCA2 which encodes a major Ca 2+ extrusion pump involved in regulation of Ca 2+ signalling, were also reduced in cells exposed to 25, 125 or

Discussion
Exposure of humans to DEHP happens quite frequently and has been associated with the occurrence of type 2 diabetes. However, the underlying mechanisms mediating these adverse effects remain poorly understood. Considering that pancreatic b cells play a pivotal role in the maintenance of glucose homoeostasis, the present study was designed to investigate the cytotoxic effects of DEHP on the insulinoma-derived b cells INS-1, and explore the associated molecular mechanisms. INS-1 cells generated by Asfari et al. in 1992 display many important characteristics of the native b cells, including a high insulin content and responsiveness to glucose stimulation [24]. However, some discrepancies were observed among numerous laboratories passages and were able to secrete insulin in response to elevated glucose concentrations [25,26]. Unfortunately, INS-1E cells are not commercially available. Asfari et al. indicated that the amount of insulin in INS-1 cells remained unchanged over 80 passages. Moreover, INS-1 cells are highly susceptible to damage under oxidative stress because of their low expression of oxygen radical scavenging enzymes, mimicking the susceptibility of primary islets. Therefore, INS-1 cells have now been extensively used for the study of oxidative stress-mediated dysfunction and cytotoxicity [27]. In the present study, we demonstrated that DEHP exposure significantly and dosedependently decreased insulin secretion ability and induced early apoptosis in INS-1 cells. More importantly, we provided evidence that the adverse effects of DEHP on INS-1 cells were mediated by the activation of PERK-ATF4-CHOP ER stress signalling pathway and the dysregulations of Nrf2-dependent antioxidant protection.
DEHP is widely dispersed in the environment. As one example, a recent study of monitoring DEHP in 623 food samples found DEHP in almost all of the tested samples, including milk-based products, distilled liquor, wine, beverages, grains, meat, oil, biscuits (cookies) and canned food, with levels ranging from 0.02 to 2685 mg/kg [28]. Another representative survey of phthalates in U.S. food also indicated that DEHP concentrations were the highest of the phthalates measured in all foods (except beef) and pork contained the highest concentration of DEHP (mean 300 mg/kg; max 1158 mg/kg) [29]. In addition to the ingestion of contaminated food, individuals can be exposed to DEHP via inhalation of polluted air and dermal contact with tainted products. DEHP exposure in the general population from all sources is estimated to be 3-30 lg/kg/day [30]. Occupational exposure and specific medical treatments using PVC medical devices lead to DEHP exposure levels that are substantially higher than background levels. For example, the amount of DEHP in blood products stored in PVC bags ranged from 1.8 to 83.2 lg/ml (0.5-213.0 lM) and the amount would increase with increasing storage time [31]. Long-term haemodialysis resulted in the highest cumulative dose of DEHP (up to 2.2 mg/kg/day) in adults, and short-term blood transfusions to trauma patients also led to the highest acute DEHP exposure (up to 10 mg/kg/day) [32]. Moreover, almost all neonates in medical care have been shown to be highly exposed to DEHP. Maximum DEHP exposure of neonates exceeded the reference dose (20 lg/kg/ day) and the tolerable daily intake (50 lg/kg/day) by a factor of 100 [33]. DEHP absorbed in the body could be metabolized into MEHP; therefore, there were studies using MEHP to evaluate the effects of DEHP in vitro. Nevertheless, another study suggested that DEHP is active at the cellular level because it has some intrinsic activity in mediating cytotoxic effects. Moreover, cells have a capacity to convert of DEHP to its metabolites [34]. Studies in vitro also indicated that MEHP cannot pass the plasma membrane and enter into cells as readily as DEHP can [35]. Most importantly, it has been reported that 7.3% of DEHP is metabolized to MEHP and 25.8% of DEHP remains accumulated in body tissues after it entered into the human body [36]. In this study, we focused on the direct action of DEHP on INS-1 cells and we chose concentration ranges ranging from 5 to 625 lM based on experimental design of many in vitro studies [34,37,38].
Oxidative stress is widely implicated as a pathogenetic mechanism of DEHP toxicity. Numerous studies have confirmed that DEHP induces reproductive toxicity mainly by inducing ROS production and by disrupting the activity of antioxidant enzymes [16,17,34]. In hepatocytes, DEHP has also been reported to induce apoptosis via the activation of the ERK/NF-jB signalling pathway, in which intracellular Ca 2+ and ROS act as pivotal mediators [37]. Consistent with these studies, our study revealed that ROS generation was significantly increased in INS-1 cells exposed to 25, 125 or 625 lM DEHP. In addition, we found that DEHP at these concentrations decreased the amount of Nrf2 in both the nucleus and the cytosol, with a corresponding decrease in the transcription of the genes encoding antioxidant enzymes, such as Hmox-1, Cat, Gclc, Gclm, Nqo1 and GPx. Nrf2 is one of the most important cellular defence mechanisms that can neutralize ROS and detoxify harmful chemicals to limit oxidative damage and maintain cellular redox homoeostasis [39,40]. Blunting the Nrf2-mediated antioxidant response would result in irreversible oxidative cellular damage. Indeed, insulin secretion was decreased and cellular apoptosis was demonstrated by positive Annexin-PI staining and caspase-3 activation in a dose-dependent manner in DEHPexposed INS-1 cells. Surprisingly, 5 lM DEHP decreased GSIS in INS-1 cells, but did not increase ROS production, implying that low and high DEHP doses might work through different mechanisms to impair insulin secretion. Under lower dose DEHP stimulation, the nuclear translocation and accumulation of Nrf2 and increase transcription of Nrf2 target genes were exhibited, indicating that INS-1 cells are able to trigger a defence mechanism to counteract low-level DEHP toxicity. Although the activation of the Nrf2-mediated antioxidant response has the potential to detoxify ROS and alleviate oxidative damage, it can also block normal glucose-dependent ROS signals involving in insulin secretion [41]. In this situation, Nrf2 activation might become the primary cause of decreased GSIS observed in the cells exposed to 5 lM DEHP. Collectively, these data indicated a critical role for the Nrf2-mediated antioxidant response in DEHP-induced b-cell dysfunction and death.
Endoplasmic reticulum is one of major sites in cells for protein synthesis and ER stress is currently considered a crucial event that drives cell apoptosis. b cells possess an extremely well-developed ER because of their functions in secreting insulin and glycoproteins; they are susceptible to changes in ER homoeostasis. ROS is reported to be an important inducer of ER stress, and ER stress in turn exacerbates the accumulation of excess ROS, ultimately entering a vicious circle and leading to cell dysfunction and apoptosis. In this study, we first found that exposure to DEHP significantly and dose-dependently induced ROS generation in INS-1 cells. Therefore, whether exposure to DEHP would trigger the ER stress response was analysed in subsequent experiments. It is known that the ER proteins PERK, ATF6 and IRE-1 are maintained in an inactive state by binding with ER-localized chaperones GRP78 and GRP94. ER stress can cause activation of PERK, ATF6 and IRE-1, as well as their downstream signalling pathways. This study focused only on the PERK-eIF2a signalling pathway because the PERK branch is the only pathway in ER stress that has been shown to induce an antioxidant stress response [42,43]. As expected, we found that a 24-hr DEHP treatment elicited ER stress in INS-1 cells, as demonstrated by the up-regulation of the major ER-localized chaperones GRP78 and GRP94, as well as an increase in PERK phosphorylation concomitant with the stimulation of eIF2a phosphorylation. Simultaneous to the ER stress response, there was a significant decrease in insulin content and secretion in INS-1 cells after exposure to DEHP, demonstrating that ER stress partially contributed to b-cell dysfunction.
PERK is able to co-ordinate the convergence of ER stress with oxidative stress signalling, resulting in Nrf2 and ATF4 activation [44]. However, the data presented in this study suggested that DEHP-mediated activation of PERK demonstrated no obvious effect on the nuclear translocation of Nrf2 and the subsequent induction of target genes induction in INS-1 cells. Both nuclear and cytoplasmic Nrf2 were decreased in INS-1 cells when they were exposed to 25, 125 or 625 lM DEHP, despite that phosphorylation of PERK was increased in these cells. The decline of Nrf2 levels and activity could be consequences of the large amounts of ROS-mediated cellular damage. On the other hand, PERK-dependent phosphorylation of eIF2a unexpectedly enhanced the transcription and translation levels of ATF4 and CHOP in INS-1 cells after DEHP treatment. CHOP is considered to be a crucial regulator of ER stress-related apoptosis signalling [45]. CHOP mediates cell apoptosis through the induction of genes such as Gadd34 and Ero1a [46]. Ero1a contributes significantly to the accumulation and production of ROS in ER stressed cells [43], and so the observations made here suggested that 125 and 625 lM DEHP could facilitate ROS formation inside the ER by activating Ero1a, finally leading to apoptosis in INS-1 cells. Of note, activation of Ero1aderived ROS is mainly confined to the ER, while the Nrf-2-mediated antioxidant response is triggered by ROS accumulation in the cytosol [47], which could partially explain why higher doses of DEHP did not activate Nrf2 in INS-1 cells. Although an up-regulation of Gadd34 expression was also observed in INS-1 cells exposed to high levels of DEHP, the negative feed-back control of Gadd34-mediated p-eIF2a dephosphorylation [48] was not observed, suggesting that DEHPexposed INS-1 cells might possess a specific resistance to p-eIF2a dephoshporylation, but this remained to be elucidated in the future studies. In addition to Gadd34 and Ero1a, higher level DEHP increased transcription of Trib3, a novel ER stress-inducible gene and a target of CHOP/ATF4 involved in ATF4/CHOP-mediated apoptosis as a second messenger during ER stress response [49,50]. The balance between pro-apoptotic and anti-apoptotic processes determines the fate of b cells. Increased apoptosis would result in a reduction in b-cell number and loss of mass, finally leading to a deterioration of key b-cell functions, such as GSIS. Therefore, we suggested that DEHP-mediated activation of some apoptotic signals, such as increases in the expression of Trib3 in b cells, might specifically blunt insulin signalling and inhibit the GSIS response in INS-1 cells.
Endoplasmic reticulum is a major intracellular source for Ca 2+ . The maintenance of Ca 2+ homoeostasis plays a key role in several aspects of b-cell physiology, including insulin production and secretion and the maintenance of ER function. The ER stress response can be activated in conditions of ER Ca 2+ depletion. After establishing that DEHP triggered the ER stress response in INS-1 cells, this study subsequently investigated the possibility that DEHP would perturb the normal regulation of ER Ca 2+ homoeostasis. Our data showed that DEHP significantly decreased ER Ca 2+ storage and increased the intracellular free Ca 2+ ([Ca 2+ ] i ), paralleling the activation of ER stress. Elevated [Ca 2+ ] i observed in DEHP-exposed cells was likely a result of  enhancing Ca 2+ leakage from the ER. In this study, we showed that DEHP exposure significantly decreased the expression of Serca2b, one of the predominant SERCA isoforms expressed in the pancreatic islet, which was consistent with the data from the studies of rodent diabetic islets [51,52] and human islets isolated from cadaveric T2DM diabetic donors [53]. Because Serca2b is mainly responsible for removing Ca 2+ from the cytosol back into ER Ca 2+ storage [54], DEHP-induced down-regulation of Serca2b expression could result in an unbalance between the continuous leakage of Ca 2+ from the ER and the SERCA pump mediated reuptake of Ca 2+ , ultimately causing ER Ca 2+ depletion and activation of the ER stress. In addition, PMCA2 transcription levels were decreased by the treatment with 25, 125 or 625 lM DEHP, whereas the amount of Ncx1 was increased with 625 lM DEHP. PCAM and NCX are responsible for Ca 2+ extrusion and clearance from the b cells [55,56]. The changes in PMCA2 and Ncx1 levels would result in an inability to transport the increased Ca 2+ load, the progressive intracellular Ca 2+ toxicity, and ultimately, activation of apoptosis in INS-1 cells after higher doses of DEHP.
In conclusion, the present study provided the first evidence that oxidative stress and ER stress response were able to be induced by DEHP in INS-1 cells, where they may interrelate and form a 'stress loop', finally contributing to the abrogated GSIS, reduced insulin content and the activation of apoptosis. DEHP is an important environmental risk factor for type 2 diabetes. However, we acknowledge that there are some limitations of this study: there are difficulties in extrapolating the result obtained in vitro to that in vivo; therefore, this study cannot be taken as clear evidence for concern regarding real-life human exposure. To address this question, further analysis in animal models are necessary to better understand the cellular and molecular mechanisms of DEHP on pancreatic b cells on the basis of the observed effects of this study.