Mitogen-Inducible Gene 6 Triggers Apoptosis and Exacerbates ER Stress-Induced β-Cell Death

Abstract

The increased insulin secretory burden placed on pancreatic β-cells during obesity and insulin resistance can ultimately lead to β-cell dysfunction and death and the development of type 2 diabetes. Mitogen-inducible gene 6 (Mig6) is a cellular stress-responsive protein that can negatively regulate the duration and intensity of epidermal growth factor receptor signaling and has been classically viewed as a molecular brake for proliferation. In this study, we used Mig6 heterozygous knockout mice (Mig6^+/−) to study the role of Mig6 in regulating β-cell proliferation and survival. Surprisingly, the proliferation rate of Mig6^+/− pancreatic islets was lower than wild-type islets despite having comparable β-cell mass and glucose tolerance. We thus speculated that Mig6 regulates cellular death. Using adenoviral vectors to overexpress or knockdown Mig6, we found that caspase 3 activation during apoptosis was dependent on the level of Mig6. Interestingly, Mig6 expression was induced during endoplasmic reticulum (ER) stress, and its protein levels were maintained throughout ER stress. Using polyribosomal profiling, we identified that Mig6 protein translation was maintained, whereas the global protein translation was inhibited during ER stress. In addition, Mig6 overexpression exacerbated ER stress-induced caspase 3 activation in vitro. In conclusion, Mig6 is transcriptionally up-regulated and resistant to global translational inhibition during stressed conditions in β-cells and mediates apoptosis in the form of caspase 3 activation. The sustained production of Mig6 protein exacerbates ER stress-induced β-cell death. Thus, preventing the induction, translation, and/or function of Mig6 is warranted for increasing β-cell survival.

Type 2 diabetes (T2D) is an endemic disease that greatly impacts the healthcare and financial systems in both developed and developing countries (1). T2D is characterized by tissue insulin resistance and pancreatic β-cell failure (2). During the development of T2D, insulin resistance is initially compensated for by increased insulin secretion by the β-cells and β-cell mass expansion. However, the body's insulin demands eventually exceed the β-cell secretory capacity, thus placing an insurmountable burden on the endoplasmic reticulum (ER); the unmitigated protein synthesis/folding stress in the ER of β-cells finally initiates an apoptotic response (3, 4).

The execution mechanism of ER stress-induced apoptosis remains an area of intensive study. Multiple signaling pathways tightly control a cell's life and death decisions. Although the activation of C/EBP homologous protein-10 (CHOP), also known as GADD153 and c-Jun N-terminal kinase (JNK) signaling pathways and their direct connections to the mitochondrial apoptotic program are considered to be the classical ER stress-induced cell death mechanism (5–7), the cross talk between the ER stress-responsive factors and canonical cell proliferation and survival signaling pathways is an emerging field. For example, Tribbles homolog 3, an inhibitor of the Akt, also known as protein kinase B signaling pathway, was recently found to respond to stress stimuli in β-cells. The up-regulation of TRIB3 (Tribbles homolog 3) leads to cell apoptosis by promoting translocation of BCL2-associated X protein to the mitochondria (8–10). Here, we identified mitogen-inducible gene 6 (Mig6) as an ER stress-responsive gene, which modulates β-cell apoptosis.

Mig6 was identified as a feedback inhibitor of epidermal growth factor receptor (EGFR) signaling (11, 12). By binding to EGFR, Mig6 controls the temporal and spatial continuity of EGFR signaling cascades (13–15). Interestingly, Mig6 expression is induced by many stress stimuli, including hypoxia, osmotic stress, mechanical strain, and lipopolysaccharide-induced infections in various cell types (16–20). Nevertheless, the role of Mig6 in response to ER stress in β-cells has not been investigated.

ER stress not only initiates adaptive responses by transcriptionally up-regulating stress-responsive proteins but also directly controls the protein translation machinery. Under normal physiological conditions, cap- and scanning-dependent translation is the default translation mechanism. In contrast, during stressed conditions, such as those that occur during the development of T2D, global protein translation is halted to conserve resources and mitigate the unfolded protein stress (21). Meanwhile, cells enhance cap-independent translation to generate proteins that are involved in stress alleviation or apoptosis induction. Intriguingly, Mig6 is resistant to global translation inhibition under hypoxic stress in a prostate cancer cell line (22). Thus, we speculate that Mig6 is a stress-responsive molecule that modulates cellular apoptosis in diverse disease states, including diabetes.

During our characterization of the proliferation and survival of β-cells from mice lacking one Mig6 allele in the present study, we hypothesized that Mig6 is an inducer of apoptosis. In addition, we speculated that Mig6 would be induced during ER stress. Ultimately, we have identified a new role for Mig6 as a mediator of ER stress-induced apoptosis in pancreatic β-cells. Thus, Mig6 becomes a potential therapeutic target to preserve pancreatic β-cell mass and prevent the irreversible pathogenesis of diabetes.

Materials and Methods

Animals and pancreatic islets isolation

Animals were maintained and used according to protocols approved by the Indiana University School of Medicine Institutional Animal Care and Use Committee. C57Bl/6J mice lacking a Mig6 allele (Mig6^+/−) were a gift from Vande Woude (17) and were bred with wild type C57Bl/6J mice (Mig6^+/+). Mice were kept in a standard light-dark cycle with free access to a standard chow diet and water. Mouse islets were harvested as previously described (24) and cultured in 11.2 mm glucose RPMI 1640 medium (supplemented with 10% fetal bovine serum, 50 U/ml penicillin, and 50 μg/ml streptomycin).

Primary pancreatic islets were also collected from male Wistar rats weighing approximately 250 g (25, 26). After collagenase digestion, rat islets were hand picked and cultured in 5 mm glucose RPMI 1640 medium (supplemented with 10% fetal bovine serum, 50 U/ml penicillin, and 50 μg/ml streptomycin) overnight before use.

Intraperitoneal glucose tolerance test

Eight- to 10-wk-old Mig6^+/+ and Mig6^+/− mice were fasted for 5 h and weighed; blood samples were collected from a tail vein. Glucose (1.5 mg/g body weight) was injected ip into fasted mice. Blood samples were collected at the indicated time points after glucose injection, and blood glucose concentrations were measured using an AlphaTRAK glucometer (Abbott Laboratories, Abbott Park, IL). One day after the glucose tolerance test, the mice pancreata were fixed in Z-Fix (Anatech Ltd., Battle Creek, MI) and paraffin embedded for immunohistochemistry or immunofluorescence staining.

Immunohistochemistry and immunofluorescence staining

Mouse pancreatic sections (5 μm) were deparaffinized with xylene and rehydrated through a series of graded ethanol solutions. Afterward, antigen retrieval was performed by microwaving slides with antigen unmasking solution (Vector Laboratories, Burlingame, CA) followed by blocking (serum-free blocking reagent; Dako, Glostrup, Denmark) for 30 min. For immunohistochemistry staining, slides were incubated with guinea pig antiinsulin antibody [no. H-86 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), diluted 1:500 in antibody diluent from Dako] overnight at 4 C. The next day, immunodetection was performed with a peroxidase-conjugated antirabbit IgG antibody (ImmPRESS; Dako) and peroxidase substrate (VECTOR NovaRED; Vector Laboratories). Finally, slides were counterstained with hematoxylin and mounted. Digital images were acquired using an Axio-Observer Z1 inverted microscope (Zeiss, Oberkochen, Germany) equipped with an AxioCam color camera. The area of the insulin-positive cells (calculated using AxioVision software) was divided by the total pancreatic area to obtain the β-cell cross-sectional area as a percentage of total pancreatic area.

For immunofluorescence staining, slides were incubated with antiinsulin (1:250, no. 180067; Invitrogen, Carlsbad, CA), antiglucagon (1:1000, no. G2654; Sigma, St. Louis, MO), and antiphospho-histone H3 (1:1000, no. 06-570; Millipore, Bedford, MA) overnight at 4 C followed by incubation with Alexa Fluor 488- or Alexa Fluor 555-conjugated secondary antibodies (1:1000; Invitrogen) for 1 h at room temperature. Cells were counterstained with 4′,6-diamidino-2-phenylindole to visualize the nuclei and then imaged using an Axio-Observer Z1 inverted fluorescent microscope (Zeiss), equipped with an Orca ER CCD camera (Hamamatsu, Bridgewater, NJ).

[³H]thymidine incorporation

Islet proliferation was assessed by measuring the incorporation of [³H]methyl-thymidine into genomic DNA (27). [³H]methyl-thymidine was added to groups of 100 islets at a final concentration of 1 μCi/ml medium for 16 h. Groups of 30 islets were picked in triplicate and washed twice with PBS. The DNA was precipitated in 500 μl of cold 10% tricholoroacetic acid and solubilized by the addition of 80 μl of 0.3 n NaOH. The amount of [³H]thymidine incorporated into DNA was measured by liquid scintillation counting and normalized by total cellular protein.

Cell culture and adenovirus transduction

INS-1-derived 832/13 rat insulinoma cells were cultured in complete RPMI 1640 medium containing 11.1 mm glucose supplemented with 10% fetal bovine serum, 10 mm HEPES, 2 mm l-glutamine, 1 mm sodium pyruvate, and 50 μm β-mercaptoethanol. For gene overexpression studies, cytomegalovirus (CMV) promoter-driven recombinant adenovirus containing the green fluorescent protein (GFP) was prepared as described (28, 29). The CMV promoter-driven Mig6 recombinant adenovirus was a gift from Kyriakis and Xu (30). Purified viruses were used to transduce cells at a concentration of 5 × 10¹¹ particles/ml medium for 24 h. The assays and drug treatments were performed 48 h after transduction. For gene knockdown experiments, adenoviruses containing a small interfering hairpin RNA sequence corresponding to rat Mig6 cDNA (5′-GATCCCCGGGACACTTTTACATTTGATTCAAGAGATCAAATGTAAAAGTGTCCCTTTTTGGAAA-3′) or a scrambled control RNA were constructed as previously described (31, 32); 832/13 cells were cultured with purified viruses for 24 h and cultured in virus-free complete RPMI 1640 medium for an additional 48 h before drug treatments.

Apoptosis assays

To induce apoptosis, cells were treated with etoposide (50 nm; BioVision, Mountain View, CA) for 4 h or thapsigargin (1 μm; Sigma) for up to 6 h. Cells were then lysed with radioimmunoprecipitation assay buffer (Santa Cruz Biotechnology, Inc.). Caspase 3 enzyme activity in the cell lysate was determined using a Caspase 3 Fluorometric Substrate Assay (Upstate, Lake Placid, NY) and a SpectraMax M5 microplate reader (Molecular Devices, Sunnyvale, CA). Protein concentration was measured using the BCA Protein Assay kit (Pierce, Rockford, IL) to normalize caspase 3 activity.

To measure cell death, 832/13 cells were grown on chamber slides, transduced with adenoviruses carrying CMV promoter-driven GFP (cmvGFP) or CMV promoter-driven Mig6 (cmvMig6) and then treated with dimethylsulfoxide (DMSO) or 1 μm thapsigargin for 6 h. Cells were incubated with propidium iodide (PI) and visualized on a fluorescent microscope. Cell death was reported as the percentage of PI and GFP copositive cells per GFP positive cells.

Pharmacological inhibitors

Inhibition of MAPK pathways by pharmacological agents was achieved via pretreating 832/13 cells with 10 μm the MAPK kinase inhibitor UO126, the p38 inhibitor PD169316, or the JNK inhibitor SP600125 (Sigma) for 1 h. Cells were then treated with thapsigargin.

Immunoblot analysis

Cells were lysed in radioimmunoprecipitation assay reagent supplemented with protease inhibitors (Santa Cruz Biotechnology, Inc.) and phosphatase inhibitor cocktails (Sigma). Lysates were resolved on a 10% NuPAGE Bis-Tris gel (Invitrogen), transferred to an Immobilon-FL Transfer Membrane (Millipore), and incubated with antibodies to caspase 3 (1:1000, no. 9662; Cell Signaling, Beverly, MA), phospho-eukaryotic translation initiation factor 2 α (eIF2α) (1:1000, no. 3398; Cell Signaling), eIF2α (1:1000, no. 5324; Cell Signaling), Mig6 (1:1000, no. 2440; Cell Signaling), pancreas and duodenal homeobox 1 (Pdx-1) (1:5000, no. 07-696; Millipore), CHOP (1:200, no. 7351; Santa Cruz Biotechnology, Inc.), actin (1:5000, no.691002; MP Biomedicals, Solon, OH), or γ-tubulin (1:2000, no. T6557; Sigma). Subsequently, membranes were incubated with IRDye 800 or 700 fluorophore-labeled secondary antibodies from LI-COR (Lincoln, NE). Protein bands were visualized using the Odyssey System (LI-COR) and quantified with ImageJ software (National Institutes of Health, Bethesda, MD).

Quantitative PCR analysis

Total RNA was isolated from 832/13 cells or islets using the RNeasy Mini or Micro kit, respectively (QIAGEN, Valencia, CA). Reverse transcription was completed with a High Capacity cDNA Reverse Transcription kit (Applied Biosystems, Foster City, CA). The threshold cycle methodology was used to calculate the relative quantities of the mRNA products of Mig6, Pdx1, cyclinD1, and ATF4 (TaqMan; Applied Biosystems). Samples were corrected for total RNA using Gapdh (TaqMan; Applied Biosystems). The PCRs were performed in triplicate for each sample from at least three independent experiments.

Polyribosome analysis

Polyribosome analysis was performed as described previously (33–35). Briefly, 832/13 cells were cultured with or without 1 μm thapsigargin for 6 h. Ten minutes before harvesting, cells were treated with 50 μg/ml cycloheximide to block translation (Sigma). Cells were washed and lysed in a buffer of 20 mm (pH 7.5) Tris-HCl, 100 mm NaCl, 5 mm MgCl₂, and 0.4% Nonidet P-40 supplemented with 50 μg/ml cycloheximide. Cell lysates were homogenized by passing through a 25-gauge needle and incubated on ice for 10 min. The cell lysate was precleared by centrifugation (9000 × g for 10 min at 4 C) and transferred onto a 10–50% sucrose gradient solution made with the above lysis buffer. A portion of unfractionated lysate was saved and used as the “input” to determine steady-state mRNA levels of Mig6, ATF4, and cyclinD1. The gradients were then ultracentrifuged in a Beckman SW-41Ti rotor for 2 h at 35,000 rpm at 4 C (Beckman Coulter, Inc., Brea, CA). Ten gradients were fractionated using a Biocomp Gradient Station and collected by a Fraction Collector (Bio-Rad, Hercules, CA). Meanwhile, absorbance of RNA at 254 nm was recorded using an ECONO UV Monitor (Bio-Rad). RNA was isolated as described above using the RNeasy Mini kit (QIAGEN) from each fraction. The samples from two adjacent fractions were pooled and subjected to quantitative (q) RT-PCR analysis to determine Mig6, ATF4, and cyclinD1 transcripts levels.

Statistical analysis

Statistical differences were detected using ANOVA; differences were considered significant when P < 0.05. Differences within ANOVA were determined using Tukey's post hoc tests. Data are reported as means ± sem.

Results

Mig6 heterozygous knockout mice have similar glucose tolerance and pancreatic β-cell area but decreased islet proliferation

Previous reports showed that deletion of Mig6 in mice leads to hyperactivation of EGFR signaling pathways and the development of tumors in the skin, lungs, gall bladder, and bile duct (36, 37). Mice with both Mig6 alleles nullified also had increased embryonic lethality (17), and in our breeding facility, we rarely obtain Mig6 homozygote knockout mice (three out of 234 pups; 1.3% observed from heterozygous matings). We, therefore, used Mig6 heterozygous knockout mice (Mig6^+/−) and their wild-type control littermates (Mig6^+/+) to study the function of Mig6 in regulating glucose homeostasis and the proliferation of β-cells in vivo. Intraperitoneal glucose tolerance tests were performed on 8- to 10-wk-old mice after a 5-h fast. Blood glucose concentrations in Mig6^+/+ and Mig6^+/− mice were not different at fasting or throughout the glucose tolerance test (Fig. 1A). We then harvested the pancreata from the same animals to study the islet morphology and β-cell cross-sectional area. Similarly, Mig6^+/+ and Mig6^+/− showed comparable β-cell cross-sectional area and islet morphology (Fig. 1, B and C). Surprisingly, islet proliferation of Mig6^+/− mice was lower than Mig6^+/+ mice, as measured by the numbers of β-cells positive for phosphorylated histone H3 in vivo (Fig. 1D) and tritiated-thymidine incorporation in isolated islets in vitro (Fig. 1E).

Fig. 1.

Mig6^+/+ and Mig6^+/− mice have same similar glucose tolerance and islets morphology but different islets proliferation rates. Eight- to 10-wk-old Mig6^+/+ and Mig6^+/− mice (n = 6–8 per group) were submitted to (A) ip glucose tolerance testing and pancreata were assessed for (B) β-cell cross-sectional area and (C) islet morphology (insulin, red; glucagon, green; and nuclei, blue). D, Relative β-cells proliferation rates were determined in vivo by counting phospho-histone H3 (green)- and insulin (red)-stained cells normalized to total islet numbers. E, Islets proliferation rates were measured in vitro by a [³H]thymidine incorporation assay (islets from same genotype were pooled, n = 3). *, P < 0.05 between Mig6^+/+ and Mig6^+/−. DAPI, 4′,6-diamidino-2-phenylindole. IPGTT, IP glucose tolerance test; pH3, phospho-histone H3.

Mig6 regulates caspase 3-mediated β-cell apoptosis

Because the regulation of β-cell mass involves the balance between processes that both increase and decrease the numbers of β-cells (38), we hypothesized that Mig6^+/− mice islets must have decreased apoptosis, because they have comparable β-cell area and decreased β-cell proliferation. To study the role of Mig6 in regulating β-cell apoptosis, we first employed an adenovirus carrying a small hairpin RNA against rat Mig6 mRNA to knockdown Mig6 in rat insulinoma INS-1-derived 832/13 cells (75% knockdown efficiency) (Supplemental Fig. 1A, published on The Endocrine Society's Journals Online web site at http://mend.endojournals.org). After the knockdown, cell apoptosis was induced by etoposide (a topoisomerase toxin that causes DNA double strands break). The activation of cellular apoptosis was measured by the amount of cleaved caspase 3 by Western blotting (Fig. 2, A and B) and a caspase 3 activity assay (Fig. 2C). Indeed, knocking down Mig6 protected β-cells from etoposide-induced apoptosis as indicated by decreased caspase 3 activity after treatment with etoposide (Fig. 2C). In addition, adenoviral-mediated overexpression of Mig6 (Supplemental Fig. 1B) significantly exacerbated etoposide-induced apoptosis (Fig. 2, D and E). These results suggested that Mig6 regulates caspase 3-mediated β-cell apoptosis.

Fig. 2.

Knockdown of Mig6 attenuates etoposide-induced apoptosis, whereas overexpression of Mig6 exacerbates apoptosis. INS-1-derived 832/13 cells were transduced with adenoviruses producing either a scrambled control RNA (siCon) or a small interfering hairpin RNA sequence against Mig6 (siMig6). After 4 h of 50 nm etoposide treatment, immunoblotting was performed to determine the cleaved caspase 3 levels as an indicator of apoptosis (A and B), or cell lysates were used for a caspase 3 activity assay (C). INS-1 cells were transduced with adenoviruses carrying cmvGFP or cmvMig6. After treating with 50 nm etoposide (Etop), cell lysates were collected for immunoblotting (D and E). Data are presented as representative immunoblots and means ± sem; n = 3–4. *, P < 0.05 between etoposide-treated siCon and siMig6; #, P < 0.05 between etoposide-treated cmvGFP and cmvMig6.

Mig6 exacerbates ER stress

Because chronic hyperglycemia and hyperlipidemia in diabetes disrupt pancreatic β-cell ER homeostasis and lead to β-cell death (39), we further investigated the apoptosis-regulating function of Mig6 during ER stress. To induce ER stress, we used a pharmacological ER calcium channel blocker thapsigargin to the compromise protein folding capacity of the ER. As hypothesized, overexpression of Mig6 in 832/13 cells remarkably exacerbated thapsigargin-induced β-cell apoptosis (Fig. 3, A and B). The phosphorylation of eIF2α and the induced CHOP protein expression served as hallmarks of the unfolded protein response (UPR) in an ER stress environment (Fig. 3, A, C, and D). Interestingly, overexpression of Mig6 did not influence the magnitude and induction of the UPR, suggesting that the apoptosis-promoting mechanism of Mig6 is downstream and/or independent of the UPR pathways.

Fig. 3.

Overexpression of Mig6 exacerbates apoptosis. INS-1-derived 832/13 cells were transduced with adenoviruses carrying cmvGFP or cmvMig6. After 48 h, cells were treated with 1 μm thapsigargin (Tg) for 0, 4, or 6 h. A, Cell lysates were collected for immunoblotting to determine cleaved caspase 3 (B), phosphorylated eIF2α (C), and CHOP (D) levels. Data are presented as representative immunoblots and means ± sem; n = 3–4. *. P < 0.05 between cmvGFP and cmvMig6 groups at both 4 and 6 h of thapsigargin treatment. E, Cell death was measured by PI staining in 832/13 cells transduced with adenoviruses carrying cmvGFP or cmvMig6 and treated with DMSO or 1 μm thapsigargin for 6 h. Data represent means ± sem; n = 5. *, P < 0.05 between cmvGFP and cmvMig6 groups.

To verify that caspase 3 cleavage was indicative of apoptosis and ultimately cell death, PI staining was performed on cells transduced with control or Mig6-overexpressing adenoviruses. Because both adenoviruses express GFP, cell death was measured in only those cells successfully transduced by adenoviruses (i.e. GFP positive cells). As indicated by PI staining, Mig6 overexpression induced cell death in cells treated with DMSO or thapsigargin (Fig. 3E).

Mig6 mRNA expression is induced and stabilized by ER stress

The transcription of Mig6 was induced by ER stress in a time-dependent manner in 832/13 cells (Fig. 4A), suggesting that Mig6 is a bona fide stress-responsive protein (16, 40, 41). Importantly, Mig6 gene expression was also induced in primary rat islets exposed to thapsigargin for 6, 16, and 24 h (Fig. 4B). In contrast, the mRNA level of pdx-1 was unchanged in thapsigargin-treated 832/13 cells (Fig. 4A) and notably was decreased in thapsigargin-treated rat islets at all time points studied (Fig. 4B). Despite the elevated Mig6 mRNA level during ER stress, Mig6 protein expression was unchanged (Fig. 4, C and D). In contrast, Pdx-1 protein level decreased in the stressed environment.

Fig. 4.

Thapsigargin-induced Mig6 mRNA expression. A, INS-1-derived 832/13 cells were treated with 1 μm thapsigargin (Tg) for the indicated times. Isolated mRNA was subjected to qRT-PCR analysis with Mig6 and Pdx-1 primers and probes. B, Isolated rat islets were treated with 1 μm thapsigargin for the indicated times and collected for qRT-PCR analysis. C and D, INS-1-derived 832/13 cells were treated with 1 μm thapsigargin for 4 and 6 h. Cell lysates were collected to determine Mig6 protein level. E, INS-1-derived 832/13 cells were pretreated with 1 μm thapsigargin for 2 h, followed by 5 μg/ml actinomycin D (ActD) treatment for 0, 20, 40, 60, 120 min, and subjected to qRT-PCR analysis. Data are presented as means ± sem; n = 3–4. *, P < 0.05 between the indicated groups.

Mig6 is categorized as an immediate early gene, which is usually induced by p38 and JNK/stress-activated protein kinases (SAPKs) (5). We thus investigated the participation of three different MAPKs in regulating Mig6 transcription. Pharmacological MAPK kinase (UO126), p38 (PD169316), or JNK (SP600125) inhibitors did not block the thapsigargin-induced Mig6 mRNA expression (Supplemental Fig. 2A). However, the combination of JNK and p38 inhibitors hindered the ER stress-stimulated Mig6 expression by 40% (Supplemental Fig. 2B).

Sequence analysis of the Mig6 mRNA reveals putative adenylate-uridylate-rich elements (AREs) in the 3′-untranslated region (42). ARE-containing mRNAs are translated into proteins that normally control cell survival and are usually labile (43). In response to stimuli, ARE-containing mRNAs associate with different ARE-binding proteins, thus yielding increased or decrease stability. We analyzed the half-life of Mig6 during ER stress using actinomycin D (an antibiotic that causes a transcription initiation blockade). We discovered that Mig6 mRNA was stabilized by ER stress (Fig. 4E).

Mig6 translation is maintained during ER stress

During ER stress, general protein biosynthesis is decreased through a phospho-eIF2α-dependent manner as a protective mechanism to conserve cellular energy and prevent protein-folding overload (44). However, some UPR adapter proteins are translated continuously during ER stress. Most of the alternatively translated mRNAs contain upstream open reading frames or internal ribosomal entry sites (IRESs) (45). Not surprisingly, the 5′-untranslated region of Mig6 mRNA is rich in GC content (42), which might form secondary structures and possibly serve as an IRES. To test whether Mig6 is alternatively translated during ER stress, we employed polyribosome analysis to examine the polyribosome association pattern of Mig6 mRNA. After stimulating 832/13 cells with thapsigargin (or DMSO as an unstressed control), sucrose gradient layered-cell lysates were fractionated, and the isolated mRNAs from the fractions were analyzed by qRT-PCR. Upon thapsigargin treatment, the overall polyribosome level was significantly reduced, whereas monoribosome levels were robustly elevated (Fig. 5A). The Mig6 mRNAs distribution in the ribosome profiles showed the same occupancy in control and stressed conditions (Fig. 5B and Table 1). Under normal conditions, 76% of Mig6 mRNAs were associated with translating polysomes (which was defined as fractions 6–10). During ER stress, 60% of Mig6 mRNA remained associated with polysomes (the difference between control and thapsigargin-mediated polysomal occupancy was not statistically significant). In this report, ATF4 served as positive control (Fig. 5C and Table 1), because ATF4 mRNA associates with translating ribosomes under stress conditions (46). On the other hand, cylinD1 was used to demonstrate general protein translation blockade (47). The polysome-associated cyclinD1 mRNA level shifted from 87% in control group to 57% in ER stress conditions (Fig. 5D and Table 1). These findings suggested that Mig6 translation was maintained, whereas global translation was attenuated during ER stress.

Fig. 5.

Translation (polyribosome association) of Mig6 is maintained during ER stress. A, INS-1-derived 832/13 cells were treated with DMSO or 1 μm thapsigargin (Tg) for 6 h. Cell lysates were separated on a sucrose gradient. Ten fractions were collected while absorbance at 254 nm was continuously monitored to indicate the 40S ribosome subunits, 60S ribosome subunits, and 80S monosomes and polysomes. Samples from adjacent two fractions were pooled and subjected to RT-PCR analysis. Fractions one to four represent the monosomes, whereas fractions 5–10 represent the polysomes. Mig6 (B), ATF4 (C), and cyclinD1 (D) mRNA in the fractions were presented as percentage of unfractionated inputs. Data are presented as means ± sem; n = 4.

Table 1.

Association of mRNAs to the translating polysomes (as a percentage of total mRNA)

	Mig6	ATF4	Cyclin D1
DMSO	0.76 ± 0.07	0.44 ± 0.00	0.87 ± 0.02
Tg	0.60 ± 0.03	0.65 ± 0.04a	0.57 ± 0.01a

Tg, Thapsigargin. Data are presented as means ± sem; n = 4.

^a P < 0.05 vs. DMSO.

Discussion

The hyperglucolipidemic milieu present during the pathogenesis of diabetes compromises the integrity of the ER in pancreatic β-cells. The perturbed ER eventually initiates a series of molecular pathways to trigger cell death. Our work presented here highlights a new mechanism linking ER stress to β-cell apoptosis through an adaptor protein, Mig6. During our characterization of Mig6^+/− mice, we speculated that Mig6 could also be a regulator of β-cell death. Consequently, we demonstrated that Mig6 exacerbates ER stress-induced cell apoptosis. Interestingly, Mig6 transcript was both induced and stabilized by ER stress. Finally, we showed that the translation of Mig6 was preserved during ER stress, whereas global protein translation was attenuated.

Mig6 was initially characterized as an immediate early gene induced by glucocorticoids and hormones (42, 48). More recently, Mig6 was identified as a negative regulator of EGFR in multiple tissues (37, 49, 50). Much of the research regarding Mig6 has focused on its role as an antiproliferative tumor suppressor in various tissues (12, 13, 36, 37). On a molecular level, Mig6 inhibits EGFR activation through a two-tiered mechanism: suppressing EGFR kinase and directing EGFR degradation (15). Correspondingly, a loss of Mig6 is associated with murine and human tumorigenesis (36, 37, 52, 53). Nevertheless, increasing evidence suggests that Mig6 responds to diverse pathophysiological stimuli, including mechanical stress, tissue injury, and hypoxia (16, 18, 54). However, the detailed stress-sensing mechanism and the consequences of Mig6 induction in different tissues remains to be fully characterized. Our study suggests that Mig6 regulates pancreatic β-cell apoptosis during ER stress. This finding not only provides a new drug target to prevent ER stress-associated β-cell death in diabetes but also indicates a possible role of Mig6 as a proapoptotic molecule in treating cancers.

Because Mig6 is an inhibitor of EGFR, we hypothesized that haploinsufficiency of Mig6 would result in increased pancreatic β-cell mass. However, Mig6^+/− mice had the same glucose tolerance and β-cell mass compared with Mig6^+/+ mice under normal chow diet-fed conditions. Surprisingly, islet proliferation from Mig6^+/− mice was lower than that in Mig6^+/+ mice. How islets from Mig6^+/− mice respond to a challenge that requires an expansion of β-cell mass, such as after a partial pancreatectomy or during high-fat feeding, remains to be determined.

Because β-cell mass is regulated by the balance between cell proliferation and death, we speculate that Mig6^+/− islets have lower levels of apoptosis. Conversely, we hypothesized that elevated levels of Mig6 could induce cell death. To address this hypothesis, we used the INS-1-derived rat insulinoma cell line 832/13 to identify whether or not Mig6 regulates apoptosis. We observed that although Mig6 overexpression exacerbated β-cell apoptosis and cell death through caspase 3-mediated pathways, silencing Mig6 mitigated apoptosis. During chemical-induced ER stress, Mig6 overexpression also led to increased β-cell apoptosis and death. Our results suggest that Mig6 most likely regulates β-cell death during ER stress independent of the canonical CHOP-mediated pathway. Instead, Mig6 possibly compromises traditional growth factor receptor-mediated cell survival signals, such as Akt, Erk, and their downstream signaling molecules (18, 56).

Intriguingly, ER stress induced Mig6 mRNA expression, partially through the activation of p38 and SAPK pathways. Being a classical immediate early gene, the Mig6 mRNA contains AREs, which are targets for fast degradation (Mig6 t_1/2 = 34.64 min) (see Fig. 4E). However, Mig6 mRNAs were, in fact, stabilized under ER stress conditions (Mig6 t_1/2 > 120 min) (see Fig. 4E). Typically, translating mRNAs are degraded rapidly during stress conditions. By clearing the transcription and translation load, the ER is freed to correct unfolded/misfolded proteins (57). For example, proinsulin mRNAs are unstable in stressed β-cells (58, 59). Yet our work presented here suggested an enhanced stability of Mig6 mRNA during ER stress. We speculate that β-cells conserve Mig6 transcripts during stress conditions should the need to initiate the apoptotic program arise.

In adaptation to ER stress, global protein synthesis is inhibited to prevent the accumulation of misfolded proteins in the ER lumen. The profound protein translation attenuation is correlated with the activation of protein kinase-like endoplasmic reticulum kinase and the phosphorylation of eIF2α. Phosphorylated eIF2α binds to eIF2B, whereby it interferes with the assembly of the translation initiation complex and, thus, cap-dependent translation initiation stalls (60). However, the genes essential for alleviating the stress or triggering apoptosis escape this translational blockade. Many of the preferentially translated genes contain upstream open reading frames or IRESs in the 5′-untranslated regions, which allows for cap-independent translation initiation to occur during cellular stress (23, 33, 46, 51, 61). Mig6 should be added to the list of genes that evade complete translational blockade during stress as its 5′-untranslated region has secondary structures that possibly serve as the IRES (48). From our observation, Mig6 protein level was maintained during ER stress, whereas other protein levels were decreased. The polyribosome analysis in 832/13 cells revealed that Mig6 mRNAs remained associated with actively translating polyribosomes during ER stress. In agreement with our work, Thomas and Johannes (22) reported that Mig6 mRNAs continue to associate with polyribosomes during hypoxic stress. On the other hand, mRNAs of the classically induced ER stress gene ATF4 were shifted from untranslating monoribosomes to translating polyribosomes upon ER stress in β-cells. In contrast, cyclinD1 mRNAs were redistributed from polyribosomes to monoribosomes in the stressed conditions. Indeed, translation inhibition in the UPR is not limited to cyclin D1. However, the drastic depression of cyclin D1 translation might slow down cell cycle progression, thus permitting β-cells to either reestablish ER homeostasis or proceed to apoptosis (55).

This study demonstrates that Mig6 is a stress-responsive protein mediating β-cell death. Mig6 transcripts are induced, stabilized, and remain translated during ER stress. Additionally, Mig6 regulates β-cell apoptosis through a caspase 3-dependent pathway. Further studies must be undertaken to understand on the molecular level how Mig6 bridges growth factor signaling and ER stress-induced apoptosis. This work highlights that Mig6 should be considered as a target for alleviating ER stress and β-cell death for the prevention and treatment of diabetes.