NR4A1 counteracts JNK activation incurred by ER stress or ROS in pancreatic β‐cells for protection

Abstract Sustained hyperglycaemia and hyperlipidaemia incur endoplasmic reticulum stress (ER stress) and reactive oxygen species (ROS) overproduction in pancreatic β‐cells. ER stress or ROS causes c‐Jun N‐terminal kinase (JNK) activation, and the activated JNK triggers apoptosis in different cells. Nuclear receptor subfamily 4 group A member 1 (NR4A1) is an inducible multi‐stress response factor. The aim of this study was to explore the role of NR4A1 in counteracting JNK activation induced by ER stress or ROS and the related mechanism. qPCR, Western blotting, dual‐luciferase reporter and ChIP assays were applied to detect gene expression or regulation by NR4A1. Immunofluorescence was used to detect a specific protein expression in β‐cells. Our data showed that NR4A1 reduced the phosphorylated JNK (p‐JNK) in MIN6 cells encountering ER stress or ROS and reduced MKK4 protein in a proteasome‐dependent manner. We found that NR4A1 increased the expression of cbl‐b (an E3 ligase); knocking down cbl‐b expression increased MKK4 and p‐JNK levels under ER stress or ROS conditions. We elucidated that NR4A1 enhanced the transactivation of cbl‐b promoter by physical association. We further confirmed that cbl‐b expression in β‐cells was reduced in NR4A1‐knockout mice compared with WT mice. NR4A1 down‐regulates JNK activation by ER stress or ROS in β‐cells via enhancing cbl‐b expression.

mechanisms by which pancreatic β-cells counteracts the apoptotic effect from ER stress or ROS may provide clues for β-cell protection.
NR4A1 is an inducible multi-stress response factor. [7][8][9] It belongs to the NR4A family, and its molecular structure includes a transcriptional activation domain, a DNA-binding domain and a putative ligand-binding domain. 10 So far, no nature ligand has been found in this family. Therefore, NR4A1 belongs to orphan nuclear receptors. 11 NR4A1 acts as a transcriptional factor to enhance the expression of some genes without ligand binding. 12,13 NR4A1 participates in various physiological activities, including cell proliferation, differentiation and apoptosis. 14-18 ER stress or ROS-inducing NR4A1 expression in pancreatic β-cells has been documented. 4,5 Controversy still exists regarding the role of NR4A1 in apoptosis. Some studies showed that it had anti-apoptotic effects, 15,[19][20][21][22][23][24] whereas others showed it was pro-apoptotic. 16,18,[25][26][27] A possible explanation for the above controversial conclusions may be accounted by the cellular location of NR4A1. When NR4A1 remains in the nucleus as a transcriptional factor, it enhances the expression of some anti-apoptotic factors, such as WT1 and survivin 4,5,24 ; if NR4A1 is translocated into the cytoplasm, it associates with bcl2 and results in cell apoptosis. 27 JNKs, also known as stress-activated protein kinases, form an important subgroup of the mitogen-activated protein kinase (MAPK) superfamily and have three isoforms (JNK1, JNK2 and JNK3). Upon activation by upstream kinases, the activated JNK has diverse cellular functions, including cell proliferation and differentiation under physiological conditions. The role of activated JNK on pro-apoptosis has been described as in cell type-dependent manner. 28 Cellular stresses, such as ER stress and oxidative stress (ROS), incur high level or sustained JNK activation, which further triggers apoptotic pathways in different cells. [29][30][31][32][33][34][35][36][37][38][39] Our preliminary data showed that either TG (thapsigargin, an ER stress inducer) or H 2 O 2 (hydrogen peroxide, a kind of ROS) increased the accumulation of phosphorylated JNK in pancreatic β-cells, whereas overexpressing NR4A1 was able to reduce JNK phosphorylation. NR4A1 is a transcription factor. How can a transcript factor regulate JNK phosphorylation? To answer this question, we used pancreatic mouse insulinoma 6 cells (MIN6 cells) and NR4A1 KO mice to explore the underlying mechanism.

| Cell culture and reagents
MIN6 cells were obtained from ATCC and were cultured under the same conditions as described in the reference. 4 NR4A1 overexpression cell lines (designated as OV cells) and control cell lines (designated as NC cells) were generated and maintained as described previously. 4 cbl-b knockdown cells (designated as KD-cbl-b cells) or the control cell clone (designated as CON-cbl-b cells) were cultured in the same medium as NC/OV cells plus G418 at 1 mg/
The genotypes of the new-born mice were verified with PCR according to the manual from Cyagen Biosciences. Mouse islets were isolated with collagenase P (purchased from Roche Applied Science) as described previously. 40

| Lentiviral infection and stable cell line selection
Lentivirus encoding full-length NR4A1 and control lentivirus were constructed and generated by GenePharma. MIN6 cells were infected with recombinant lentivirus encoding NR4A1 or control lentiviral vector, and stable cells were selected under puromycin selection.
Lentivirus encoding shRNA targeting to cbl-b and control lentivirus were constructed in Obio Technology, and the targeting sequence was 5′-AACACAGACGCCATGATTTGC-3′. 41 OV cells were infected with cbl-b shRNA lentivirus or control scramble shRNA lentivirus, and the stable cells were selected under G418 drug pressure.

| Plasmid construction
Mice genomic DNA was extracted as previously described. 42 Mice cDNA library was obtained from WT C57BL/6J mouse liver with ReverTra Ace qPCR RT Kit (Toyobo). We designed a pair of primers with HindIII and XbaI restriction sites at 5′ and 3′ to amplify the cDNA of MKK4 and cloned the PCR product into pFLAG-CMV™-2 vector (Sigma). The sequence information of the primer is listed below: Primer R: 5′-GGGTCTAGATCAGTCGACATACATGGGCGAGC-3′.
We designed four pairs of primers with Kpnl and XhoI restriction sites at 5′ and 3′, respectively, to amplify four different lengths of the cbl-b promoter (1987, 1440, 981 and 485 bp) from mice genomic DNA. And then, the PCR products were cloned into pGL3-basic luciferase reporter vector (Promega). All constructed plasmids were sequenced by Sangon Biotech Co., Ltd., to verify these sequences were completely aligned with the original DNA sequences from NCBI.

| Quantitative real-time PCR assay
Total RNA was extracted and purified from cultured cells by using RNAiso Plus (Takara) and from mouse islets using a RNeasy Mini Kit (Qiagen). qPCR was performed as previously described. 4 According to the Cq value of the target gene obtained from the qPCR machine, the relative expression multiple of the target gene was calculated with a relative quantitative method for 2 -ΔΔCq , and the numerical value was further corrected by using the housekeeping gene 18s rRNA as the internal control. All experiments were repeated three times, and each sample was assayed in triplicate. The primers used for PCR are listed in Table 1.

| Western blot analysis
Protein samples were obtained, and Western blot analysis was performed according to the methods as previously described. 43 The antibodies used in the experiment are as follows: rabbit antibodies,

| Dual-luciferase reporter assays
Plasmids with different lengths of cbl-b promoter were co-transfected into cells with TK (thymidine kinase promoter-Renilla luciferase reporter) plasmid, or pGL3-basic vector was co-transfected into cells with TK plasmid as control. The luciferase activity was measured by a dual-luciferase reporter assay kit (Vazyme Biotech Co., Ltd) after 48 hours post-transfection.

| Chromatin immunoprecipitation (ChIP) analysis
MIN6 cells were infected with adenovirus encoding NR4A1-HA or control adenovirus. After 48 hours post-infection, the infected cells were applied for ChIP assay. ChIP analysis was accomplished with a ChIP Assay Kit (Beyotime) as previously described. 4 The specific anti-HA monoclonal antibody (12CA5) was applied to pull down the DNA fragments associated with HA-tagged NR4A1.

| Immunofluorescence
The pancreatic tissues from WT mice or NR4A1 KO mice were fixed with paraformaldehyde (4%). The primary anti-insulin mice antibody (Santa Cruz Biotechnology, Inc) and the primary anti-cbl-b rabbit antibody (ProteinTech Group, Inc) were applied for immuno-staining. The overlapped colour (yellow) of insulin (red) with cbl-b (green) in islets of pancreatic tissue indicated cbl-b expression in pancreatic β-cells.

| Statistical analyses
Data were expressed as the mean ± SD. Statistical analysis was done with t-test or analysis of variance (ANOVA) by using GraphPad Prism 5.0. A value of P < .05 (*) was considered a statistically significant difference.

Previously, we generated NR4A1 overexpression cells (OV cells) from MIN6 cells and the control cells (NC cells). The mRNA and
protein expression levels of NR4A1 were confirmed in OV cells and NC cells ( Figure 1A-C). Figure 1 showed that the phosphorylation level of JNK was increased in both OV cells and NC cells after they were treated with TG or H 2 O 2 , but overall the JNK phosphorylation level was lower in OV cells compared with that in NC cells ( Figure 1D,E,H,I), whereas the total JNK level was close between the two cells ( Figure 1F,G,J,K). Overexpressing NR4A1 turned down the JNK phosphorylation in MIN6 cells rather than total JNK level.
F I G U R E 1 NR4A1 down-regulates JNK phosphorylation induced by TG or H 2 O 2 in MIN6 cells. A-C, the relative NR4A1 mRNA and protein levels were determined by qPCR and Western blot in both OV (NR4A1 overexpression) and NC (control) cells. D-G, Western blotting exhibited the protein levels of p-JNK and JNK in response to 100 μmol/L H 2 O 2 at various time-points in both OV and NC cells. H-K, Western blotting exhibited the protein levels of p-JNK and JNK in response to 0.5 μmol/L TG at various time-points in both OV and NC cells. Densitometric analyses of the blots were presented as histogram, and these data represented the means of three independent experiments; *P < .05, **P < .01 and ***P < .001 vs ns; error bars indicate SD

| NR4A1 down-regulates MKK4 protein level in MIN6 cells
To explore the mechanism by which NR4A1 attenuates JNK phosphorylation induced by ER stress or ROS, we searched online and found that there were two upstream kinase molecules specific for JNK, namely MKK4 and MKK7. The mRNA or protein level of MKK7 did not change between OV and NC cells (Figure 2A-C), but the protein level of MKK4 in OV cells was markedly reduced compared with NC cells (Figure 2A,B), whereas the mRNA level of MKK4 did not change ( Figure 2C). and MKK7 phosphorylation levels. By comparison, the induced fold of p-MKK4 was higher than that of p-MMK7 ( Figure 2D-E).
The OV cells and NC cells were treated with TG or H 2 O 2 at different time-points, both MKK4 and MKK7 were examined in these samples, and we found that the MKK4 protein was maintained at a lower level in OV cells compared with that in NC cells during these series of treatments ( Figure 2F-I).

| NR4A1 increases MKK4 ubiquitination and degradation in MIN6 cells
We tested whether overexpressing NR4A1 was able to enhance MKK4 ubiquitination level. In Figure 3, equal numbers of OV cells  Figure 3A showed that more ubiquitin modified exogenous Flag-MKK4 in OV cells compared with that in NC cells. From the data above, it was reasonable to conclude that more overexpressed exogenous Flag-MKK4 in OV cells was ubiquitinated compared with that in NC cells.
To further confirm that MKK4 was reduced by ubiquitination-based proteasome degradation, MG132 was applied to figure out the dynamics of MKK4 in OV cells. As expected, MG132 was

| NR4A1 enhances the expression of cbl-b in MIN6 cells
Proteasome-based protein degradation depends on the abundance of the specific E3 ligase. Each E3 ligase has a specific target or substrate to be modified with ubiquitin. It was reported that NR4A1 was able to enhance the expression of three possible E3 ligases, which were TRAF2, 19 Smurf1 44 and cbl-b (casitas B-lineage lymphoma b). 41 Our data showed that OV cells had higher mRNA and protein levels of cbl-b ( Figure 4A-C), whereas the mRNA levels of TRAF2 or Smurf1 did not increase in OV cells compared with NC cells ( Figure 4A). During TG or H 2 O 2 treatment for different time-points, OV cells always maintained a higher level of cbl-b protein expression compared with NC cells (Figure 4D-G).

| Cbl-b expression negatively correlates with MKK4 protein level and p-JNK level upon H 2 O 2 or TG treatment in NR4A1-overexpression cells
To further confirm that cbl-b modulates MKK4 protein expression in MIN6 cells, we applied RNA interference scheme to knock down cbl-b expression in OV cells. Figure

| NR4A1 enhances the transactivation of cbl-b promoter via physical association
It was reported that NR4A1 enhanced the cbl-b promoter transactivation in human cells. 41 We tested whether NR4A1 directly regulates cbl-b transactivation in mouse cells regarding that humans and mice have some differences in promoter sequences.
The promoter sequence of cbl-b between −14 and −2008 has four putative NR4A1 binding sites as shown in Figure 6A. We cloned four cbl-b luciferase reporters with different DNA lengths as shown in Figure 6B. We transfected the four cbl-b luciferase reporter plasmids into OV and NC cells, the dual-luciferase assay results showed that To verify whether NR4A1 associates with cbl-b promoter, we infected MIN6 cells with adenovirus encoding NR4A1-HA, and after the infection, the cells were applied for ChIP assay. The ChIP products obtained were applied for PCR examination. As shown in Figure 6D

| Confirmation of the impact of NR4A1 on cbl-b expression in NR4A1-knockout mice
To confirm the conclusion that NR4A1 enhances the expression of cbl-b in vivo, we tested the expression of cbl-b in NR4A1-KO mice. We purified the islets from WT mice or NR4A1-KO mice, and further extracted the RNA and protein for real-time PCR and Western blotting.
The mRNA expression of cbl-b in NR4A1-KO mice was much lower than that in WT mice ( Figure 7A), whereas the cbl-b protein level in NR4A1-KO mice was also lower than that in WT mice ( Figure 7B).
The immuno-staining data showed that the colocalization of cbl-b with insulin was dramatically decreased in the islets of pancreatic tissue from NR4A1-KO mice compared with that from WT mice ( Figure 7C), which indicated that cbl-b protein expression in pancreatic β-cells was significantly reduced in NR4A1-KO mice. Figure 7D is a graphic model for the mechanism that NR4A1 protects β-cells from ER stress or ROS-induced JNK phosphorylation. In In short, NR4A1 reduces p-JNK level by indirectly down-regulating the availability of its upstream MKK4, therefore, reduces the possibility of pancreatic β-cell apoptosis.

| D ISCUSS I ON
Hyperglycaemia and hyperlipidaemia result in ER stress and/or increased ROS in pancreatic β-cells, and cause β cells damage and apoptosis. [1][2][3][4][5][6] Pancreatic β-cell loss plays an essential role in the pathogenesis of type 2 diabetes. Our pancreatic β-cells are always challenged with hyperglycaemia and/or hyperlipidaemia after meal, but they are always able to survive over the subsequent cellular stress by an unknown "protective network" with some protective molecules. If the "protective network" is weak, consequently, the pancreatic β-cells will be damaged or even undergo apoptosis. To search the molecules for pancreatic β-cell protection under cellular stress is important for type 2 diabetes prevention.
It was reported that ER stress or ROS resulted in increased JNK activation or phosphorylation in various cells [31][32][33]35 and activated JNK further triggered the signalling pathways to apoptosis. [36][37][38][39] We confirmed that MIN6 cells were treated with TG or H 2 O 2 resulted in increased JNK phosphorylation. Some researcher applied JNK phosphorylation inhibitor which partially reversed the ER stress or ROS-induced apoptosis in Bcr/Abl-positive chronic myelogenous leukaemia cells, 45 which indicated that JNK phosphorylation was critical for ER stress or ROS-induced apoptosis. 46 As for the mechanisms of ROS or ER stress induced JNK activation, many research articles reported that ROS induced ER stress, which has been popularly accepted. 47 Previously, we found that the pancreatic cells from NR4A1 KO mice islets were more sensitive to TG or H 2 O 2 -induced apoptosis compared with that from WT mice. 4,5 We observed that high-fat diet feeding resulted in increased 8-OHdG (a senescence marker) level and apoptosis rate in pancreatic cells from NR4A1 KO mice compared with that from WT mice. 5 It is known that pancreatic cell senescence or/and pancreatic cell loss results in diabetes. We We previously found that NR4A1 was inducible upon treat- is closely related to their ubiquitination level and proteasome degradation, which lies in the protein level of substrate-specific E3 ligase.
We verified that the loss of MKK4 protein by NR4A1 overexpression was proteasome-dependent by using a specific drug MG132.
Thus, we searched the literatures and found that the NR4A1 was able to enhance the expression of three E3 ligases. 19,40,41 However, we only found cbl-b mRNA level was increased in NR4A1 overexpression MIN6 cells. We further checked the protein expression Cbl-b deficiency could exaggerate HFD-induced insulin resistance through saturated FA-mediated macrophage activation. 58 We first discovered that cbl-b was able to degrade MKK4 through the ubiquitination-proteasome pathway, and thus to attenuate JNK phosphorylation in response to ER stress or ROS.
We further studied the possible mechanism(s) by which NR4A1 may up-regulate cbl-b expression. It was reported that NR4A1 enhanced cbl-b promoter transactivation in human cells. To verify whether NR4A1 directly enhances cbl-b transactivation in mice β-cells, we exploited luciferase assay and ChIP assay. We further found NR4A1 was able to enhance the transactivation activity of F I G U R E 7 Confirmation of NR4A1 impacting cbl-b expression in NR4A1-KO mice. A, the relative mRNA levels of cbl-b were determined by real-time quantitative PCR in wild-type mice and NR4A1 KO mice. B, the relative protein levels of cbl-b and NR4A1 in pancreatic islets from WT or NR4A1 KO mice were determined by Western blotting. C, the colocalization of cbl-b (green) with insulin (red) was dramatically decreased in the islets of pancreatic tissue from NR4A1-KO mice compared with that from WT mice. The optical magnification of the image was 300×, and the larger square on the left bottom of each image was the enlarged image of the smaller section of the islet (the smaller square indicated within the islet). D, A graphic model for the mechanism that NR4A1 protects β-cells from JNK phosphorylation induced by ER stress or ROS. The data represented the means of three independent experiments; ***P < .001 vs ns; error bars indicate SD cbl-b promoter and NR4A1 was able to physically associate with cbl-b at some putative targeting sequence. What we detected is the binding sites of NR4A1 in cbl-b promoter regulatory element in mouse cells were not the same as people found in human cells as reported.
In addition to our in vitro study, we further found the mRNA level or protein level of cbl-b was markedly reduced in the islets from NR4A1-KO mice compared with that from WT mice, which was further confirmed that NR4A1 was a positive factor for cbl-b expression in mice pancreatic islets. As shown in insulin immuno-staining data, the majority of cells (about 85%-90%) in pancreatic islets In summary, our data demonstrated that NR4A1 could repress cellular ER stress or ROS-induced JNK activation via enhancing cbl-b expression to degrade MKK4 in pancreatic β-cells. Namely, NR4A1 down-regulates p-JNK level through NR4A1/cbl-b↑/MKK4↓/p-JNK↓ pathway. Our study indicates NR4A1 might be a major player in the protective network of pancreatic β-cells, and our finding might provide clues for β-cell protection and diabetes prevention.

ACK N OWLED G EM ENTS
Thankfulness. We are much thankful to Dr Guang-wei Wei for his suggestions and for his help with the high-quality animal facility.

CO N FLI C T O F I NTE R E S T
No potential conflict of interest relevant to this article exists.