JDP2 is directly regulated by ATF4 and modulates TRAIL sensitivity by suppressing the ATF4–DR5 axis

Here, we found transcriptional repressor JDP2 is transcriptionally induced by ATF4 via AARE in the JDP2 promoter. JDP2 functions as a negative feedback regulator of the ATF4 pathway and suppresses ATF4‐mediated death receptor DR5 expression. JDP2 suppresses TRAIL‐mediated apoptosis by inhibiting the ATF4‐DR5 axis.

Jun dimerization protein 2 (JDP2) is a bZip-type transcription factor, which acts as a repressor or activator of several cellular processes, including cell differentiation and chromatin remodeling. Previously, we found that a stress-responsive transcription factor, known as activating transcription factor 4 (ATF4), enhances JDP2 gene expression in human astrocytoma U373MG and cervical cancer HeLa cells; however, the role of JDP2 in the ATF4-mediated stress response remained unclear. Here, we reported that siRNA-mediated JDP2 knockdown enhances the expression of several ATF4 target genes, including ASNS, and death receptors 4 and 5 (DR4 and DR5) in HeLa cells. In addition, the results of a transient reporter assay indicate that JDP2 overexpression represses ER stress-mediated DR5 promoter activation suggesting that JDP2 negatively regulates ATF4-mediated gene expression. Curiously, knockdown of JDP2 increases the sensitivity of cells to TNF-related apoptosis-inducing ligand (TRAIL), which induces apoptosis in cancer cells through DR4 and DR5. These results indicate that JDP2 functions as a negative feedback regulator of the ATF4 pathway and contributes to TRAIL resistance in cancer cells.
ATF4 is a stress-responsive transcription factor that plays an important role in the integrated stress response [12]. ATF4 is activated in response to various stresses, such as endoplasmic reticulum (ER) stress and amino acid deficiency, and modulates the expression of target genes. ATF4 transactivates gene expression by binding to its recognition element, namely the amino acid response element (AARE) (the consensus sequence is TGATGnAAn, where n is any nucleotide) [12]. ATF4 upregulates several cytoprotective and stress response genes, such as those involved in amino acid metabolism and the unfolded protein response [12]. ATF4 also induces apoptosis by enhancing C/EBP homologous protein (CHOP) and death receptors 4 and 5 (DR4 and DR5) in response to stress [13,14]. CHOP activates proapoptotic factors, such as B-cell lymphoma-2 (Bcl-2)-like protein 11 and p53 upregulated modulator of apoptosis, but suppresses antiapoptotic factors, such as Bcl-2, thus leading to intrinsic mitochondria-dependent apoptosis [13]. DR4 and DR5 are receptors for TNF-related apoptosis-inducing ligand (TRAIL) and mediate TRAIL-dependent extrinsic apoptosis in target cells and the Fas/Fas ligand system [14,15]. TRAIL also induces TRAIL ligand-independent apoptosis during glucose starvation or ER stress [16,17]. TRAIL is mainly produced by immune cells as a membranebound protein or a secreted protein that leads to caspase-8-dependent apoptosis [18]. Considering that DR4 and DR5 are expressed in tumor cells but not in normal cells, TRAIL and the death receptor system play an important role in tumor cell elimination by regulating the immune system. However, many tumor cells escape TRAIL-mediated apoptosis by downregulating death receptors, overexpressing decoy receptors that lack a cytosolic domain, or overexpressing FLICE-like inhibitory protein (c-FLIP), which is a negative regulator of death receptor signaling [19]. Several transcription factors, including ATF4 and CHOP, are known to modulate DR4 and DR5 gene expression. However, the precise mechanism by which DR4 and DR5 are downregulated in tumor cells remains poorly understood.

Reverse transcription quantitative PCR (RT-qPCR)
Total RNAs were isolated using TRIzol reagent (Life Technologies) according to the manufacturer's instruction. A PrimeScript II RT Kit (Takara Bio, Otsu, Japan) was used to synthesize cDNA. SYBR Premix EX Taq II (Takara Bio) and a CFX96 thermal cycler (Bio-Rad, Hercules, CA, USA) were used for qPCR. The following primer pairs were used for the experiments: siRNA transfection U373MG or HeLa cells were seeded into 12-well plates (1 9 10 5 cells per well). The next day, the cells were transfected with siRNA by using Lipofectamine RNAiMax Reagent (Life Technologies). After a 24-h transfection, the medium was replaced with a fresh medium, and the cells were treated with the indicated drugs. The target sequence of human ATF4 and JDP2 are 5 0 -GCC TAG GTC TCT TAG ATG A-3 0 and 5 0 -GTG AGC TAG ATG AGG AAG A-3 0 , respectively. A negative control siRNA (cat. #1027310, Qiagen, Hilden, Germany) was used for negative control experiments.

Immunoblot analysis
Immunoblot analysis was performed as described before [9]. Briefly, an aliquot of nuclear extract was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and blotted with antibodies against ATF4 (CREB2) (C-20) and Lamin B (M-20) purchased from Santa Cruz Biotechnology (Dallas, TX, USA).

Plasmid construction
The JDP2 expression plasmid was constructed by PCR cloning. JDP2 cDNA was amplified using the primer pair and 5'-GGA ATT CTC ACT TCT TCT CGA GCT GCT C -3' and was digested with BamHI/EcoRI. The purified DNA fragment was subcloned into the BamHI/EcoRI sites of the pcDNA3-FLAGx3 vector [9]. To construct hJDP2-Luc, the human JDP2 promoter was amplified using the primer pair 5'-GCC AGA TCT GTC AGT GGG TGT GAA GCG CC-3' and 5'-GCC AAG CTT GCA GAA GGT GCG GGG GGA-3', and U373MG genomic DNA was used as a template. After BglII/StuI digestion, the DNA fragment was subcloned into the BglII/HindIII sites of the pGL3 vector (Promega, Madison, WI, USA). pJDP2-mt-Luc was generated using site-directed mutagenesis with the following primers: 5'-CTA TAT AGC CGG GGC GGG TGC AAC CCG TCC CGC-3' and 5'-GCG GGA CGG GTT GCA CCC GCC CCG GCT ATA TAG-3'. For the construction of pDR5-Luc, the human DR5 promoter was amplified using PCR primers (5'-AAA CAA ACC ACA GCC CGG GGC GCA-3' and 5'-CTG TCC CCG TTG TTC CAT GGC GGT-3') and subcloned into the SmaI/NcoI sites of the pGL3 vector.

Transient transfection and reporter assay
HeLa cells were seeded into a 24-well plate (5.0 9 10 4 cells per well) and incubated overnight. The next day, the cells were transfected with 200 ng of the luciferase reporter and 10 ng of pRL-TK or 2 ng of the pRL-EF Renilla luciferase coreporter plasmid and 200 ng of the effector plasmid by using the FuGENE transfection reagent (Promega). After a 4-h incubation, the medium was replaced with a fresh medium containing either DMSO or 2 lgÁmL À1 Tm. After a 20-h incubation, the cells were subjected to a luciferase assay. Luciferase activity was measured using Dual-Luciferase Assay Reagent (Promega) according to the manufacturer's protocol and was normalized to Renilla luciferase activity.

Cell viability assay
Cell viability was evaluated using a CCK-8 kit (Dojindo, Kumamoto, Japan) according to the manufacturer's protocol. Absorbance at 450 nm was measured using a microplate reader (Bio-Rad).

Statistical analysis
Student's t-test or one-way ANOVA with the Tukey-Kramer post hoc test was used to estimate the statistical significance of the differences between two or more groups. Differences between groups were considered statistically significant with P values < 0.05. All experiments were repeated at least three times, and the data are expressed as mean AE standard error (SE), and individual data points are shown as open circles.

ATF4 regulates JDP2 expression
In a previous global transcriptome study, we found that JDP2 is induced by CA, which is a phytochemical that activates both Nrf2 and ATF4 in U373MG human astrocytoma cells [9]. However, whether ATF4 modulates JDP2 gene expression remains unknown. To investigate the role of ATF4 in the regulation of JDP2 gene expression, we used the ATF4 inducers Tm and Tg, which activate ATF4 via ER stress in a PERK-dependent manner. On the basis of RT-qPCR analysis, we found that both Tm and Tg significantly induced JDP2 gene expression in U373MG cells (Fig. 1A). The human JDP2 gene has at least four alternative first exons (Fig. 1B). To identify the JDP2 transcripts induced by ATF4 activation, we performed variant-specific RT-PCR by using the primer pairs shown in Fig. 1B. Given that variants 2 and 3 are highly expressed in human cells, we analyzed the expression of variants 2 and 3. Both Tm and Th specifically induced the transcript expression of JDP2 variant 3 (Fig. 1C). We also found that the Tm induction of JDP2 was completely abolished by ATF4 knockdown (Fig. 1D,E). Furthermore, both Tm and Th significantly induced JDP2 in human cervical cancer HeLa cells (Fig. 1F). These results indicate that ATF4 modulates JDP2 expression.
The JDP2 promoter is activated by ATF4 ATF4 regulates the expression of various target genes by binding to AAREs [12]. On the basis of in silico analysis, we found a consensus AARE sequence that is conserved between mammals immediately upstream of first exon of variant 3 of human JDP2 gene ( Fig. 2A). To confirm whether this AARE is responsible for the ATF4 induction of JDP2, we constructed a JDP2 gene promoter-luciferase plasmid and transfected it into HeLa cells with or without the ATF4 expression plasmid (Fig. 2B). Luciferase activity was significantly upregulated by ATF4 cotransfection (Fig. 2C). By contrast, no ATF4-mediated induction was observed in the AARE-mutated reporter construct (pJDP2-mt-Luc) (Fig. 2B,C). These results indicate that ATF4 regulates JDP2 expression in an AARE-dependent manner. knockdown experiments by using siRNA. We confirmed by RT-qPCR analysis that JDP2 siRNA transfection considerably reduced JDP2 expression in HeLa cells (Fig. 3A). ATF4 mRNA expression was significantly upregulated by JDP2 knockdown (Fig. 3B). Furthermore, ATF4 target genes, such as ASNS, DR5, and DR4, were also significantly upregulated by JDP2 knockdown (Figs. 3C-E). Other ATF4 target genes, including ATF3 and CHOP, were also substantially upregulated by JDP2 knockdown (data not shown). By contrast, the gene expression of c-FLIP, which inhibits DR5 and DR4 signaling, was not affected by JDP2 knockdown (Fig. 3F). In addition, ATF4 knockdown significantly reduced JDP2 and DR5 expression in HeLa cells (Fig. 3G,H). These results indicate that JDP2 negatively regulates the ATF4 pathway.

JDP2 negatively regulates TRAIL sensitivity
Given that DR5 plays an important role in TRAILmediated cancer cell apoptosis, we next investigated whether JDP2 affects TRAIL sensitivity of cancer cells. Although DR5 expression was also induced by JDP2 knockdown in U373MG cells (Fig. S1A,B), U373MG cells show resistance to TRAIL by overexpression of c-FLIP, that negatively regulates DR5 signaling (Fig. S1C) [20], we analyzed the role of JDP2 in the regulation of DR5 gene expression and TRAIL sensitivity in HeLa cells. ATF4 modulates DR5 gene expression by binding to the DR5 gene promoter [21]. To investigate whether JDP2 affects DR5 promoter activity, we performed reporter analysis by using a DR5 promoter-luciferase reporter construct (Fig. 4A). JDP2 overexpression decreased the activation of the DR5 promoter induced by Tm in a dose-dependent manner (Fig. 4B). Interestingly, although JDP2 knockdown did not decrease cell viability by itself, it significantly increased TRAIL sensitivity in HeLa cells (Fig. 4C,D). Although Tm induced endogenous DR5 expression, further up-regulation of DR5 expression by JDP2 knockdown was not observed, although ATF4 knockdown down-reduced DR5 expression (Fig. S1D,  E). On the other hand, JDP2 knockdown also increased TRAIL sensitivity of T98G human glioma cells (Fig. S2). These results indicate that JDP2 decreases TRAIL sensitivity, at least in part, by suppressing DR5 gene expression.

Discussion
In this study, we revealed for the first time that JDP2 decreases TRAIL sensitivity via the repression of an ATF4-DR5 axis. DR5 gene regulation by ATF4 and its downstream targets ATF3 and CHOP is well established. ATF4 enhances the DR5-dependent apoptosis of HeLa cells during glucose deprivation [14]. CHOP induces DR5 expression in response to Tg or fenretinide, which is a semisynthetic retinoid [22,23]. Edagawa et al. [24] demonstrated that zerumbone and celecoxib activate ATF4 and downstream targets ATF3 and CHOP to induce DR5 expression in human colorectal cancer cells.
In this study, we found ATF4/AARE-mediated JDP2 gene regulation (Figs 1 and 2). As shown in Fig. 2A, AARE in JDP2 gene promoter are well conserved among mammals, suggesting its importance in JDP2 gene regulation. We demonstrated that JDP2 knockdown substantially upregulated DR5, ATF3, CHOP, and ATF4 expression (Fig. 3B,D, and data not shown) and that JDP2 negatively regulated ATF4mediated DR5 gene promoter activity (Fig. 4B), thus indicating that JDP2 is a negative feedback regulator of the ATF4-DR5 axis. JDP2 was reported to recruit multiple HDAC proteins to the ATF3 promoter and suppress ATF3 expression in HEK293Ts and MEFs [3]. Furthermore, JDP2 inhibits ATF4-mediated CHOP induction in HeLa cells [21]. Therefore, it is likely that JDP2 downregulates DR5 expression via the suppression of the ATF4-ATF3 and ATF4-CHOP axes. However, the mechanisms by which JDP2 suppresses ATF4 activity remain unknown. ATF4 interacts with multiple AP-1 family proteins, including ATF3 and CHOP [25]. Among them, JDP2 is more similar to ATF3 in the bZip domain at the amino acid sequence level, which is responsible for dimer formation. Our preliminary study revealed that JDP2 interacts with ATF4 in vitro (data not shown), thus suggesting that JDP2 may suppress ATF4 activity by heterodimerizing with ATF4. Since Tm induction of DR5 expression was not further enhanced by JDP2 knockdown (Fig. S2), JDP2-mediated DR5 gene regulation mechanism may be mediated by more complicated process. Further studies are needed to clarify this point. In addition to ATF4, ATF3, and CHOP, DR5 gene expression is regulated by multiple transcription factors, including AP-1, FOXO3a, p53, and NF-jB [26]. Given that JDP2 was first reported as an AP-1 repressor, it is possible that JDP2 also inhibits AP-1 activity to modulate DR5 expression. Interestingly, Zou et al. [27] reported that c-Jun N-terminal kinase (JNK) upregulates DR5 expression via AP-1-mediated CHOP induction. JNK is reported to phosphorylate JDP2 at Thr148 for proteasomal degradation [28], thus suggesting that JNK may regulate DR5 expression by modulating not only AP-1 but also JDP2 activities.
Although TRAIL selectively induces apoptosis in cancer cells, it is well known that several cancer cells develop TRAIL resistance via several mechanisms [19].
One mechanism is the downregulation of DR5 or DR4 by somatic mutation or epigenetic modification. Given that JDP2 is known as an epigenetic modulator, it may contribute to TRAIL resistance in cancer cells by suppressing DR4 and DR5 expression. However, it remains controversial whether JDP2 is a tumor suppressor or oncogenic factor. The inhibitory effect of JDP2 on oncogenic AP-1 suggests that JDP2 may be a tumor suppressor. JDP2 inhibits the transformation of NIH3T3 cells by TPA, which activates AP-1 [29]. Yuanhong et al. [30] reported that JDP2 downregulation is associated with metastasis in pancreatic cancer patients. It was also reported that JDP2 inhibits the epithelial-mesenchymal transition of pancreatic cancer BxPC3 cells [31]. By contrast, the present study suggests that JDP2 negatively regulates TRAIL/DR5 signaling and promotes tumor cell survival. It was reported that JDP2 overexpression enhances diethylnitrosamine-induced liver cancer in mice [32]. The JDP2 gene locus is activated by retroviral insertion in mouse T-cell lymphomas [33]. Interestingly, JDP2 expression is suppressed by tumor suppressor p53, and this finding suggests the tumor promotion activity of JDP2 [34]. However, it is well established that many stressresponsive transcription factors, such as ATF4, exhibit protective or antitumor activity in normal cells but support tumor cell survival under severe stress conditions [35]. This suggests that the function of JDP2 in tumor formation may change depending on physiological and pathological situations.

Conclusions
In conclusion, this study shows that JDP2 functions as a negative feedback regulator of the ATF4 pathway and modulates TRAIL sensitivity in cancer cells. Further studies are needed to confirm whether JDP2 is a potential therapeutic target for TRAIL-mediated cancer immunotherapy. manuscript. All authors discussed the results and commented on the manuscript.

Data accessibility
The additional data used to arrive at these conclusions can be obtained from the corresponding author on reasonable request.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article.