CHOP deletion does not impact the development of diabetes but suppresses the early production of insulin autoantibody in the NOD mouse
- First Online:
- Cite this article as:
- Satoh, T., Abiru, N., Kobayashi, M. et al. Apoptosis (2011) 16: 438. doi:10.1007/s10495-011-0576-2
- 242 Views
C/EBP homologous protein (CHOP) has been proposed as a key transcription factor for endoplasmic reticulum (ER) stress-mediated β-cell death induced by inflammatory cytokines in vitro. However, the contribution of CHOP induction to the pathogenesis of type 1 diabetes is not yet clear. To evaluate the relevance of CHOP in the pathogenesis of type 1 diabetes in vivo, we generated CHOP-deficient non-obese diabetic (NOD.Chop−/−) mice. CHOP deficiency did not affect the development of insulitis and diabetes and apoptosis in β-cells. Interestingly, NOD.Chop−/− mice exhibited a delayed appearance of insulin autoantibodies compared to wild-type (wt) mice. Adoptive transfer with the diabetogenic, whole or CD8+-depleted splenocytes induced β-cell apoptosis and the rapid onset of diabetes in the irradiated NOD.Chop−/− recipients with similar kinetics as in wt mice. Expression of ER stress-associated genes was not significantly up-regulated in the islets from NOD.Chop−/− compared to those from wt mice or NOD-scid mice. These findings suggest that CHOP expression is independent of the development of insulitis and diabetes but might affect the early production of insulin autoantibodies in the NOD mouse.
KeywordsCHOPType 1 diabetesNon-obese diabetic mouseER stressAutoantibodyApoptosis
Type 1 diabetes is a chronic autoimmune disease resulting from specific destruction of the insulin-producing β-cells by an inflammatory reaction in and around the pancreatic islets . Two major molecular mechanisms of β-cell destruction via direct cell-mediated cytotoxicity, such as Fas/Fas Ligand (FasL)  or perforine/granzyme  and indirect cytokine-mediated cytotoxicity, have been proposed to be involved in type 1 diabetes [4–7]. Inflammatory cytokines including interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α) and interferon-γ (IFN-γ) are detected within the islets and the exposure to soluble mediators including cytokines, nitric oxide (NO), and oxygen free radicals probably causes β-cell death in the course of insulitis and type 1 diabetes [8–12].
Possible mechanisms of cytokine mediated β-cell death include (i) the activation of the stress-activated protein kinases such as cJUN NH2-terminal kinase (JNK), p38 mitogen-activated protein kinase (MAPK) and extracellular signal-regulated kinase, (ii) the release of death signals from mitochondria and (iii) triggering of endoplasmic reticulum (ER) stress [12, 13]. The ER stress response is triggered by the accumulation of unfolded proteins in the lumen of the ER. Upon the initiation of ER stress, the cells activate intracellular signaling pathways that transmit information from the ER to the cytoplasm and nucleus; this is known as the unfolded protein response (UPR) [14, 15]. The UPR promotes cell survival by adjusting the ER protein-folding capacity through translational attenuation and degradation of unfolded proteins, and up-regulation of the ER chaperones. But if homeostasis cannot be re-established, apoptosis is induced via transcriptional induction of C/EBP homologous protein (CHOP) , the activation of JNK [17, 18], and/or the activation of caspase-12 .
The β-cells of the pancreas have a highly developed ER to secrete insulin, and are susceptible to ER stress . The relevance of ER stress-mediated β-cell apoptosis to the pathogenesis of diabetes has been proven in several genetic syndromes [21–23] and in type 2 diabetes [24, 25]. NO-induced apoptosis in β-cells is shown to be mediated by the ER stress pathway. NO donors trigger an ER stress response induced by the depletion of ER Ca2+ content in β-cells leading to CHOP expression and apoptosis. CHOP-deficient islets have been shown to be much more resistant to NO-induced, ER stress-mediated β-cell apoptosis than wild-type islets [26, 27]. These observations suggest that the ER stress pathway via CHOP expression might be relevant for β-cell destruction in the pathogenesis of type 1 diabetes [16, 28]. A recent study, however, demonstrated that the attenuation of ER stress with siRNA-mediated CHOP reduction does not have an impact on cytokine-mediated β-cell death . The contribution of CHOP induction in cytokine-mediated β-cell apoptosis or in the pathogenesis of type 1 diabetes is still controversial.
This study was therefore conducted to investigate the role of CHOP in the pathogenesis of a spontaneous type 1 diabetes model studying the non-obese diabetic (NOD) mouse whose disease pathogenesis, specifically in relation to autoimmune-mediated β-cell destruction, is most likely similar to that in human type 1 diabetes.
Materials and methods
Female NOD mice and NOD-scid mice were purchased from Clea Japan (Tokyo Japan) and CHOP knockout mice on C57Bl/6 background (B6.Chop−/−) were obtained by the method described previously [26, 30]. We generated NOD mice with the CHOP gene deleted by backcrossing with B6.Chop−/− mice to NOD mice. The targeted allele was introgressed into the NOD background using a marker-assisted “speed congenic” breeding approach wherein backcross segregants were fixed for homozygosity for NOD alleles at NOD/B6 polymorphic markers at known NOD diabetes susceptibility (idd) loci. Idd loci were fixed by backcross 4. Homozygous CHOP-deficient NOD background mice (NOD.Chop−/−) were produced by the intercrossing of heterozygotes at N10, and a permanent line of NOD.Chop−/−, wild-type mice (NOD.Chop+/+) and heterozygous mice (NOD.Chop+/−) were established at N10.
The mice were maintained in the Laboratory Animal Center for Biomedical Research at Nagasaki University under specific pathogen-free conditions. All animal experiments described in this study were approved by the institutional animal experimentation committee, and were conducted in accordance with the Guidelines for Animal Experimentation.
Monitoring blood glucose levels
Blood glucose levels were monitored weekly with a One-Touch Ultra (Johnson & Johnson, Tokyo, Japan). Mice with blood glucose levels above 250 mg/dl for two consecutive measurements were considered diabetic.
Pancreatic sections were histologically analyzed by fixing tissues in 10% formalin and staining the paraffin-embedded samples with hematoxylin and eosin. A minimum of 30 islets from each mouse were microscopically observed by two different observers for the presence of insulitis, and the levels of insulitis were scored according to the following criteria: 0, no lymphocyte infiltration; 1, islets with lymphocyte infiltration in less than 25% of the area; 2, 25–50% of the islet infiltrated; 3, 50–75% of the islet infiltrated; 4, more than 75% infiltrated or small retracted islets.
Measurement of insulin autoantibodies and serum IgG level
The mouse serum were collected and kept −20°C. The levels of insulin autoantibody (IAA) in serum were evaluated by using a 96-well filtration plate micro IAA assay as previously described . The index value of 0.01 was chosen as the cut-off limit of normal serum level of IAA in non-diabetic mouse strains. The level of serum IgG was quantified with Mouse IgG ELISA Quantitation Kit (Bethyl Laboratories, INC, TX, USA). The 96-well plates were precoated with 100 μl of 1:100 carbonate–bicarbonate buffer diluted primary antibody in kit at room temperature (RT) for 1 h, and then washed three times, followed by addition of 100 μl of 1:2 serially diluted 500 ng/ml standard IgG into seven wells and 1:100 diluted samples into other assigned wells respectively. The plates were incubated at RT for 1 h, and the wells were washed five times. Then 100 μl of 1:50,000 diluted horseradish peroxidase conjugated antibody was added into well, after 1 h incubation at RT, the samples were washed as previous described, color was developed with TMB and terminated by 2 M H2SO4. IgG concentration (μg/ml) was calculated according to the standard cure measured under 450 nm.
Adoptive transfer of diabetes
The 3 × 107 whole splenocytes or 2 × 107 splenocytes depleted CD8+ splenocytes with Auto-MACS (Milteny Biotech, Bergisch Gladbach, Germany) from newly diabetic female NOD mice were injected intraperitoneally into 8–12 week-old male recipient mice which were irradiated at 750 rad 4 h before the transfer .
The kidneys from mice treated with 1.0 mg/kg of tunicamycin (i.p.) and the pancreata were fixed in 10% formalin and embedded in paraffin. Tissue sections (5-μm thick) were sectioned and deparaffinized. The sections were then stained using the labeled streptavidin-biotin method (Histofine Staining Kit, Nichirei Co., Tokyo, Japan) with 10% rabbit serum, followed by incubation with mouse anti-insulin monoclonal antibody (Nichirei Co., Tokyo, Japan) and rabbit anti-Chop polyclonal antibody (Santa Cruz, CA, USA) (1:200 dilution).
Terminal deoxynucleotidyltransferase-mediated dUTP nick end-labeling (TUNEL) staining
For the confirmation of apoptosis by immunofluorescence analysis, TUNEL staining was employed to detect double-stranded DNA breaks. After the pancreatic tissues were fixed in 4% PFA 4°C for 15 min followed by immersion in PBS with 0.5% Tween 20 and 0.2% bovine serum albumin, we used Mebstain Apoptosis kit direct (MBL, Nagoya, Japan). After the pancreatic tissues were rinsed with distilled water, they were incubated with a 50 μl terminal deoxynucleotidyl transeferase solution at 37°C for 1 h. The stained pancreatic tissues were analyzed by confocal microscopy (LSM5, PASCAL; Carl Zeiss, Jena, Germany). After washing in PBS, the cells were mounted in Vectashield mounting medium (Vector Laboratories, Inc., Burlingame, CA, USA), and scanned by confocal microscopy. The number of apoptotic cells was counted in the areas of islets in four-nonconsecutive slices of pancreatic tissues derived from each mouse.
Real-time quantitative PCR
Islets were isolated from 20-week-old female mice by pancreatic digestion with collagenase (Wako Pure Chemical Industries, Ltd., Japan). Islets were purified by Histopaque (Sigma-Aldrich) density gradient centrifugation. Isolated islets were stored at −80°C until use. Total RNA was extracted from the isolated islets and kidneys. cDNA synthesis was done using primers with SuperScript III First-strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA, USA). cDNAs were used as templates in SYBER green real-time PCR assays on a LightCycler (Roche Diagnostics). The primers used in the PCR reaction for inositol requiring 1 (IRE1), PKR-like ER kinase (PERK), JNK1, JNK2, and Caspase-12 were obtained from SA Biosciences. The primers for CHOP and Binding immunoglobulin protein (Bip) were synthesized as described in the previous study . Sample data were analyzed according to the comparative cycle threshold method and were normalized by stable reference genes of GAPDH and 18S selected by geNorm VBA applet among three house keeping genes including β-actin, GAPDH and 18S. All results were expressed as a percent of the value in control extracts.
Group differences were analyzed by the Mann–Whitney U test, χ2 test or Student’s t test and differences between Kaplan–Meier survival curves were estimated by the log-rank test with the use of Dr SPSS II for Windows software (SPSS Inc., Chicago, IL, USA). P values less than 0.05 was considered statistically significant. Insulitis levels were analyzed by Ridit analysis, and levels of t higher than 1.96 or lower than −1.96 were considered statistically significant.
CHOP deletion did not affect the spontaneous development of diabetes and insulitis
CHOP deletion suppressed the expression of insulin autoantibody at 8 weeks of age
Levels and frequencies of IAA
Age in weeks
Mean levels of IAA index (×10−3)
1.7 ± 0.3
41.4 ± 17.7
142.8 ± 70.4
146.7 ± 81.9
1.8 ± 0.2
4.0 ± 1.1*
70.4 ± 40.3
146.7 ± 81.9
Mean levels of IgG (μg/ml)
13.2 ± 0.6
14.9 ± 0.5
13.2 ± 0.5
11.8 ± 0.3
11.8 ± 0.4
13.4 ± 0.7
12.0 ± 0.6
11.7 ± 0.5
Frequencies of IAA positive/total (%)
CHOP deletion did not affect diabetes in an adoptive transfer study with whole or CD8+-depleted splenocytes
CHOP deletion did not affect apoptosis in islet cells
Expression of ER stress genes in the islets from pre-diabetic mice
Both type 1 diabetes and type 2 diabetes are characterized by deficits in β-cell mass and increased β-cell apoptosis [38–43]. It has been suggested that ER stress induced β-cell apoptosis may be a common cause in type 1 and type 2 diabetes [26, 27, 44, 45]. In type 1 diabetes, exposure to cytokines and NO during insulitis inhibits expression of the sarcoendoplasmic reticulum pump Ca2+ ATPase (SERCA) 2 in β-cells and deplete endoplasmic reticulum calcium stores, leading to ER stress and apoptosis together with increased levels of ER stress-associated molecules such as IRE-1α, CHOP and ATF-4 . The pathogenesis of type 2 diabetes is characterized by both insulin resistance and impaired insulin secretion due to decreased β-cell function and β-cell mass [46, 47]. High levels of free fatty acids (FFAs) palmitate triggers ER stress and induces β-cell apoptosis together with increased levels of ER stress-associated molecules [44, 45]. Thus, cytokines and palmitate trigger ER stress in β-cells evidenced by increased expression of ER stress-associated molecules including CHOP. A recent study, however, demonstrated that siRNA-mediated CHOP reduction as well as PBA which acts as a chemical chaperone attenuate palmitate-mediated ER stress and β-cell death but not cytokine-mediated ones . Furthermore, CHOP deletion in several mouse models of type 2 diabetes also demonstrated improved glycemic control and the maintenance of β-cell mass via the UPR and oxidative stress response [48, 49].
In the present study, we investigated the impact of CHOP deletion in the NOD mouse, a model of type 1 diabetes, in order to determine whether ER stress-mediated β-cell death via the CHOP is involved in the pathogenesis of type 1 diabetes. We found that CHOP deletion does not affect the development of insulitis and diabetes. We also demonstrated that CHOP expression is not increased in islets from pre-diabetic NOD mice compared to those from NOD-scid mice. A previous study has shown that CHOP expression is frequently observed in islets from obese type 2 diabetes patients but not from type 1 diabetes patients . These findings suggest that CHOP is fundamental factor that links ER stress and apoptosis in β-cells in type 2 diabetes but not type 1 diabetes.
It is possible that other mechanisms of β-cell death such as direct cell-mediated cytotoxicity induced by activated CD8+ T cells (CTLs) may compensate for the lack of CHOP-mediated cell death in the NOD.Chop−/− mice, resulting in the similarity of the kinetics of diabetes development with the kinetics in wild-type mice. However, our adoptive transfer study with CD8+-depleted splenocytes confirmed that CHOP deletion does not have an influence on the development of diabetes, even without CTL-mediated cytotoxicity.
We also investigated the mRNA expressions of JNK [17, 18] and caspase-12 [19, 51] in order to rule out the possibility of the involvement of other apoptosis pathways mediated by ER stress. We did not find any evidence of enhancement of the expression of ER stress response genes or ER stress-associated apoptosis genes in the NOD.Chop−/− mice or the wild-type mice in comparison to control mice. Our results indicate that ER stress response in the islets is not a major mechanism of cytokine induced β-cell death in the pathogenesis of type 1 diabetes. It has been proposed that possible other pathways contributing to the cytokine induced β-cell death include an activation of stress-activated kinases and the release of mitochondrial death signals . JNK group of MAPK is activated by IL-1β in β-cells and inhibitors of JNK prevent cytokine-induced apoptosis in insulin-producing cells [52, 53]. The JNK1 is encoded by the Mapk8 gene and Mapk8−/− mice exhibit profound defects in insulin resistance and obesity caused by eating high fat diet, suggesting that JNK1 is implicated in type 2 diabetes [54, 55]. In contrast, JNK2 encoded by the Mapk9 gene play an important role for type 1 diabetes since Mapk9 deletion suppress the development of insulitis and diabetes in the NOD background mice. However the β-cells derived from Mapk9−/− NOD mice and wild type NOD mice are found to be equally sensitive to cytokine induced apoptosis in vitro , suggesting that JNK2 is independent of cytokine induced β-cell death. We investigated the mRNA expression of both JNK1, JNK2 in the islets of NOD.Chop+/+, NOD.Chop−/− and NOD-scid mice but no significant differences were observed among the three groups.
Mitochondria are key organelles for β-cell function but also play an important role in triggering apoptosis. Members of the Bcl-2 protein family regulate the mitochondrial response to pro-apoptotic signals  and prevent release of cytochrome c which execute dell death. It has been reported that overexpression of Bcl-2 partially protects cytokine-induced β-cell death in mouse  and human . However, overexpression of Bcl-2 in vivo does not prevent spontaneous diabetes in NOD mice .
Thus, proposed pathways contributing to cytokine-induced β-cell death could not fully explain the pathogenesis of type 1 diabetes in vivo. Alternative approaches with engineered NOD background mouse are need to be investigate to clarify the relevance of cytokine-induced β-cell death in type 1 diabetes.
We found that the expression of IAA was suppressed in NOD.Chop−/− mice at 8 weeks of age. NOD mice usually express IAA as early as 4 weeks of age, and the early appearance of IAA at 4–8 weeks is strongly associated with the subsequent early development of diabetes . Since the development of diabetes was not suppressed in NOD.Chop−/− mice, CHOP might be relevant for the production of autoantibody independent of the development of autoimmune diabetes. A previous study demonstrated that the ER stress response via the transcription factor X-box binding protein 1 (XBP-1) is required for the terminal differentiation of B cells to plasma cells . The expression of CHOP as well as XBP-1 has been shown to be rapidly induced in the B cells after stimulation with IL-4 . CHOP deficient mice exhibit lower levels of serum IgM secretion after lipopolysaccharide (LPS) stimulation independent of the differentiation of plasma cells . Although the spontaneous IAA in NOD mice is predominantly IgG, CHOP deficiency did not affect on the total serum IgG levels in this study. CHOP might play a crucial role in the early production of antigen-specific autoantibody such as IAA in the NOD mouse.
In this study, we demonstrated here that a lack of CHOP does not affect the natural course of diabetes development in NOD mice. Our results, taken together with these previous findings, suggest that the mechanisms of β-cell death could be fundamentally different between type 1 diabetes and type 2 diabetes, and different therapeutic approaches will be required to prevent β-cell death and reverse diabetes in each types of diabetes .
We would like to thank M. Motomura and Y. Kataoka for their technical assistance. This study was supported by research grants from the Japan Society for the Promotion of Science (#21591143, #22930032, #21790874, #22790865).