Transcriptional regulation of cyclooxygenase-2 in response to proteasome inhibitors involves reactive oxygen species-mediated signaling pathway and recruitment of CCAAT/enhancer-binding protein delta and CREB-binding protein

Abstract

Inhibition of ubiquitin-proteasome pathway has been shown to be a promising strategy for the treatment of inflammation and cancer. Here, we show that proteasome inhibitors MG132, PSI-1, and lactacystin induce COX-2 expression via enhancing gene transcription rather than preventing protein degradation in the human alveolar NCI-H292 and A549, and gastric AGS epithelial cells. NF-IL6 and CRE, but not NF-kappaB elements on the COX-2 promoter were involved in the gene transcription event. The binding of CCAAT/enhancer binding protein (C/EBP)beta and C/EBPdelta to the CRE and NF-IL6 elements, as well as the recruitment of CBP and the enhancement of histone H3 and H4 acetylation on the COX-2 promoter was enhanced by MG132. However, it did not affect the total protein levels of C/EBPbeta and C/EBPdelta. MG132-induced DNA-binding activity of C/EBPdelta, but not C/EBPbeta was regulated by p38, PI3K, Src, and protein kinase C. Small interfering RNA of C/EBPdelta suppressed COX-2 expression, further strengthening the role of C/EBPdelta in COX-2 gene transcription. In addition, the generation of intracellular reactive oxygen species (ROS) in response to MG132 contributed to the activation of MAPKs and Akt. These findings reveal that the induction of COX-2 transcription induced by proteasome inhibitors requires ROS-dependent protein kinases activation and the subsequent recruitments of C/EBPdelta and CBP.

Figure 1.

Proteasome inhibitors stimulate COX-2 protein expression via enhancing gene transcription in NCI-H292, A549, and AGS epithelial cells. (A) Cells were treated with various concentrations of MG132, PSI-1, or lactacystin for 16 h, and then whole-cell lysates were prepared and subjected to Western blotting using antibodies specific for COX-2 or actin as described in Materials and Methods. (B) NCI-H292 cells were treated with 25 μM MG132 for the indicated time, and then whole-cell lysates were prepared and subjected to Western blotting using antibodies specific for COX-2 or actin, or mRNA was harvested and RT-PCR was performed to quantify COX-2 mRNA. Fold of induction of COX-2 protein or mRNA expression normalized with actin or β-actin was quantified using ImageQuant. (C and D) Cells were treated with MG132, PSI-1, or lactacystin for 16 h. Whole-cell lysates were immunoprecipitated with anti-COX-2 or anti-ubiquitin antibody, and then the precipitates were subjected to SDS-PAGE and Western blotting using anti-COX-2 (C) or anti-ubiquitin (D) antibody. Results are representative of three independent experiments.

Figure 2.

Involvements of NF-IL-6 and CRE in the MG132-induced COX-2 promoter activity in NCI-H292 cells. (A) Cells were transfected with –327/+59, KBM, ILM, or CRM luciferase expression vector, and then treated with 25 μM MG132 for 6 h. Luciferase activity was measured as described in Materials and Methods. The data have been normalized with β-galactosidase activity and expressed as the mean ± SEM of three independent experiments performed in triplicate. * p < 0.05; compared with –327/+59. (B-E) Cells were treated with 25 μM MG132 for 30 min, 1 h, 2 h, or 24 h, and then nuclear extracts were prepared and DNA-protein binding activity measured by EMSA using NF-IL6 (B and D) and CRE (C and E) oligonucleotides probe as described in Materials and Methods. One hundred-fold molar excess of unlabeled NF-IL6, CRE, or NF-κB oligonucleotides was used as competitor (D and E). Results are representative of three independent experiments.

Figure 3.

Identification of transcription factors binding to the NF-IL6 and CRE sites. NCI-H292 cells were incubated with 25 μM MG132 for 1 h. Nuclear extracts were prepared and DNA-protein binding activity measured by EMSA using NF-IL6 (A) and CRE (B) oligonucleotides probe. Supershift assays were performed using 2 μg of the indicated antibodies as described in Materials and Methods. Results are representative of three independent experiments.

Figure 4.

Kinetics of MG132-induced translocation of C/EBPβ and C/EBPδ and phosphorylation of C/EBPβ and their binding to CRE and NF-IL6 sites. (A) NCI-H292 cells were treated with 25 μM MG132 for 10 min, 30 min, 1 h, or 2 h, and then whole-cell lysates and nuclear extracts were prepared and subjected to Western blotting using antibodies against C/EBPβ, C/EBPδ, and phospho-C/EBPβ (Thr-235). (B and C) Cells were treated with 25 μM MG132 for the indicated time, and then nuclear extracts were prepared and mixed with biotinylated COX-2 promoter probe containing CRE (B) or NF-IL6 (C) sites or without probe (Control) and streptavidin-agarose beads. C/EBPβ or C/EBPδ in the complex was detected by Western blotting. Results are representative of three independent experiments.

Figure 5.

MG132 induced recruitment of C/EBPβ, C/EBPδ, and CBP to the COX-2 promoter and acetylation of histone H3 and H4 in vivo. (A) Schematic illustration of various regulatory elements on the COX-2 promoter (–397 to +7). (B–D) NCI-H292 cells were treated with 25 μM MG132 for the indicated time or 10 ng/ml TNF-α for 60 min, and then ChIP assays were performed using anti-C/EBPβ, anti-C/EBPδ (B), anti-CBP (C), anti-acetyl H3, or anti-acetyl H4 (D) antibody or without antibody (Control) to assay the precipitated COX-2 promoter regions (–397 to –119 or –109 to +7) as described in Materials and Methods. One percent of the chromatin was assayed to verify equal loading (Input). Results are representative of three independent experiments.

Figure 6.

Effects of ROS scavengers on MG132-induced COX-2 expression and ROS production. (A) Cells were treated with 25 μM MG132, PSI-1, or 200 μM H₂O₂ for 16 h in the presence or absence of pretreatment with 20 mM NAC or GSH for 30 min. Whole-cell lysates were prepared and subjected to Western blotting using antibody specific for COX-2 or actin. (B) Cells were loaded with 10 μM DCF for 1 h and then replaced with fresh medium followed by incubation with 25 μM MG132 or 200 μM H₂ O₂ for 1 h in the present or absence of 20 mM NAC or GSH. Fluorescent images were observed and photographed with a fluorescence microscope as described in Materials and Methods. Results are representative of three independent experiments. (C) The fluorescence intensity were determined with ImageQuant software and represented as means of spots in three independent fields. * p < 0.05; compared with MG132 alone.

Figure 7.

Effects of ROS scavengers on MG132-induced MAPKs and Akt activations. Cells were treated with 25 μM MG132 for the indicated time (A), or pretreated with 20 mM NAC or GSH for 30 min before incubation with 25 μM MG132 for 30 min (B). Whole-cell lysates were prepared and subjected to Western blotting using antibodies specific for phosphorylated and unphosphorylated form of ERK, p38, JNK, or Akt. The immunoblots shown are representative of three independent experiments.

Figure 8.

The effects of various inhibitors on the MG132-induced COX-2 expression and promoter activity. (A) NCI-H292 cells were pretreated with 10 μM PP2, 10 μM LY294002 (LY), 1 μM Ro 31-8220 (Ro), 200 ng/ml actinomycin D (Act D), 500 nM cycloheximide (CHX), 30 μM PD98059 (PD), 30 μM SB203580 (SB), or 25 μM SP600125 (SP) for 30 min before incubation with 25 μM MG132 for 16 h. Whole-cell lysates were prepared and subjected to Western blotting using antibodies specific for COX-2 and actin. Fold of induction of COX-2 protein expression normalized with actin was quantified using ImageQuant. (B) Cells transfected with pGS459 luciferase expression vector were pretreated with 30 μM PD98059, 30 μM SB203580, 10 μM PP2, 10 μM LY294002, or 1 μM Ro 31-8220 for 30 min before incubation with 25 μM MG132 for 6 h, and then luciferase activity was measured. The data were normalized with β-galactosidase activity and expressed as the mean ± SEM of three independent experiments performed in triplicate. * p < 0.05; compared with MG132 alone. (C) Cells were pretreated with various inhibitors as described in B for 30 min before incubation with 25 μM MG132 for 2 h, and then mRNA was harvested and RT-PCR was performed to quantify COX-2 mRNA. Results are representative of three independent experiments.

Figure 9.

The effects of various kinase inhibitors on the MG132-induced translocation of C/EBPβ and C/EBPδ and their binding to the CRE and NF-IL6 sites. NCI-H292 cells were pretreated with 30 μM PD98059 (PD), 30 μM SB203580 (SB), 10 μM PP2, 10 μM LY294002 (LY), or 1 μM Ro 31-8220 (Ro) for 30 min before incubation with 25 μM MG132 for 1 h, and then nuclear extracts were prepared and subjected to Western blotting using antibodies specific for C/EBPβ, C/EBPδ, or phosphorylated form of C/EBPβ (A) or mixed with biotinylated COX-2 promoter probe containing CRE (B) or NF-IL6 (C) sites or without probe (Control) and streptavidin-agarose beads. C/EBPβ or C/EBPδ in the complex was detected by Western blotting. In D, after cells were treated, ChIP assays were performed using anti-C/EBPβ, or anti-C/EBPδ antibody or without antibody (Control) to assay the precipitated COX-2 promoter regions (–397 to –119 or –109 to +7) as described in Materials and Methods. One percent of the chromatin was assayed to verify equal loading (Input). (E) NCI-H292 cells were transfected with pSilencer 3.1 plasmids expressing scrambled (–), or 2 or 4 μg C/EBPδ siRNA. After 48 h, cells were incubated with 25 μM MG132 for 16 h, and then whole-cell lysates were prepared and subjected to Western blot using antibodies specific for anti-C/EBPβ, anti-C/EBPδ, and anti-COX-2. Results are representative of three independent experiments.

Figure 10.

The effects of various kinase inhibitors on MG132-induced CRE promoter activity and CRE- or NF-IL6-specific DNA-protein complex formation. (A and B) NCI-H292 cells transfected with CRE luciferase expression vector were pretreated with 30 μM PD98059, 30 μM SB203580, 10 μM PP2, 10 μM LY294002, or 1 μM Ro 31-8220 for 30 min (A) or cotransfected with the empty vector or the DN mutants of p38 (DN), Akt (DN), or c-Src (DN; B) before incubation with 25 μM MG132 for 6 h, and then luciferase activity was measured. The data were normalized with β-galactosidase activity and expressed as the mean ± SEM of three independent experiments performed in triplicate. * p < 0.05 compared with MG132 alone. (C and D) Cells were pretreated with various inhibitors as described in (A) before incubation with 25 μM MG132 for 1 h. Nuclear extracts were prepared and the CRE (B) or NF-IL6 (C) DNA-protein complex formation was measured by EMSA. Results are representative of three independent experiments.