Effects of the anti-inflammatory drug celecoxib on cell death signaling in human colon cancer

The anti-inflammatory drug celecoxib, the only inhibitor of cyclooxygenase-2 (COX-2) with anticancer activity, is used to treat rheumatoid arthritis and can cause endoplasmic reticulum (ER) stress by inhibiting sarco/ER Ca2 +-ATPase activity in cancer cells. This study aimed to investigate the correlation between celecoxib-induced ER stress and the effects of celecoxib against cell death signaling. Treatment of human colon cancer HCT116 cells with celecoxib reduced their viability and resulted in a loss of mitochondrial membrane potential (ΔΨm\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta {\Psi }_{\mathrm{m}}$$\end{document}). Additionally, celecoxib treatment reduced the expression of genes involved in mitochondrial biogenesis and metabolism such as mitochondrial transcription factor A (TFAM) and uncoupling protein 2 (UCP2). Furthermore, celecoxib reduced transmembrane protein 117 (TMEM117), and RNAi-mediated knockdown of TMEM117 reduced TFAM and UCP2 expressions. These results suggest that celecoxib treatment results in the loss of ΔΨm\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Delta {\Psi }_{\mathrm{m}}$$\end{document} by reducing TMEM117 expression and provide insights for the development of novel drugs through TMEM117 expression.


Introduction
Non-steroidal anti-inflammatory drugs (NSAIDs) show antiinflammatory, antipyretic, and anti-pain activities by inhibiting cyclooxygenase (COX) activity, thereby reducing prostaglandin production. COX enzymes catalyze the initial step of prostaglandin synthesis and exist in three isoforms-COX-1, COX-2, and COX-3. COX-1 is constitutively expressed and performs homeostatic functions in cells. COX-2 expression is induced by inflammatory cytokines. Traditional NSAIDs inhibit both COX-1 and COX-2, whereas new NSAIDsalso called coxibs-selectively inhibit COX-2.
Celecoxib is a potent COX-2 selective inhibitor that is primarily used for the treatment of rheumatoid arthritis. It has been reported to exhibit anticancer activity. For example, it can prevent hypoxia-induced epithelial-mesenchymal transition (EMT) in colon cancer cells (Bocca et al. 2012), induce cell cycle arrest (Schiffmann et al. 2008), and inhibit proliferation and depolarize mitochondrial membrane potential ( ΔΨ m ) in human colorectal adenocarcinoma HT-29 cells (Srivastava et al. 2021). In addition, celecoxib reportedly reduces the incidence of colon polyps in patients with familial adenomatous polyposis (Steinbach et al. 2000).
Although several studies have reported the antitumor activity of celecoxib (Tołoczko-Iwaniuk et al. 2019), the underlying mechanism is still unclear.
It was postulated that the antitumor activity of celecoxib is independent of COX-2 inhibition, since methylcelecoxib, a close structural analogue of celecoxib lacking COX-2 inhibitory activity, was also found to reduce the viability of COX-2 negative human colon cancer HCT116 cells, whereas other coxibs such as valdecoxib did not affect cell viability (Schiffmann et al. 2008). Celecoxib was highly effective in inhibiting either basal or hypoxia-induced invasiveness in the COX-2 expression cell line, but not in the COX-2 negative cell line, suggesting that COX-2 is involved in the PI3K/ PTEN/Akt pathway, which is one of the most common pathways in human malignancy (Bocca et al. 2012). In addition to COX-2 inhibition, celecoxib reportedly shows other specific activities. For instance, celecoxib specifically inhibits sarco/ER Ca 2 + -ATPase (SERCA) (Johnson et al. 2002) and induces endoplasmic reticulum (ER) stress (Kardosh et al. 2008;Pyrko et al. 2007;Kim et al. 2007). The impairment of calcium homeostasis in ER results in accumulation of misfolded proteins in ER, which in turn causes ER stress and activates three membrane proteins-activating transcription factor 6 (ATF6), inositol-requiring enzyme 1 (IRE1), and PKR-like ER kinase (PERK)-to restore ER homeostasis. The cell defense mechanism against ER stress is called the unfolded protein response (UPR). UPR plays a role in cancer progression and can be targeted for cancer therapy (Riha et al. 2017). Although a moderate UPR due to limited nutrient and oxygen supplies to tumor cells can enhance tumor cell survival (Clarke et al. 2014), excess UPR induced by the anticancer drug thapsigargin, a non-competitive inhibitor of SERCA, inhibits cancer cell survival (Jaskulska et al. 2020).
Transmembrane protein 117 (TMEM117) is associated with the ER stress-mediated mitochondrial apoptotic pathway (Tamaki et al. 2017). TMEM117 is localized in the ER and at the plasma membrane. It contains eight transmembrane domains, and an intracellular C-terminus that contains intrinsically disordered regions, and two extracellular N-glycosylated sites that contribute to its stability (Bürgi et al. 2016). Thapsigargin attenuates cell viability by reducing TMEM117 levels and inducing the depolarization of ΔΨ m (Tamaki et al. 2017). Mitochondria produce cellular energy and metabolites, alongside regulating apoptosis (Nunnari and Suomalainen 2012;Koh et al. 2021), and mitochondrial biogenesis is an important process for mitochondrial homeostasis. Additionally, mitochondria play an essential role in ER stress-mediated cell death; however, the signaling mechanisms remain to be elucidated. Mitochondrial transcription factor A (TFAM) is one of the most abundant mitochondria-localized DNA-binding proteins that control mitochondria DNA replication, transcription, and packaging (Koh et al. 2021;Dölle et al. 2016;Matsushima et al. 2010;Kang et al. 2018). TFAM is also involved in regulating reactive oxygen species (ROS) levels and mitochondrial uncoupling via regulation of ΔΨ m (Koh et al. 2019). The expression of TFAM is regulated by the binding of nuclear respiratory factor 1 (NRF1) to Tfam promoters (Virbasius and Scarpulla 1994).
Uncoupling proteins (UCPs) are a subfamily of the mitochondrial solute carrier family located in the inner mitochondrial membrane (IMM) and are essential regulators of ΔΨ m (Pan et al. 2018). UCP2 is widely distributed in tissues and regulates various processes such as ROS production, cancer, diabetes, and neuronal injury (Azzu and Brand 2010). Mitochondrial ROS production is related to ΔΨ m ; UCP2-mediated ROS attenuation is thought to occur via the dissipation of the mitochondrial proton motive force (Azzu and Brand 2010).
It is important to investigate the molecular mechanisms of celecoxib in relation to ER stress-mediated mitochondrial function to better understand the mechanisms underlying the antitumor activity of celecoxib, which is the aim of the current study. We demonstrated that celecoxib downregulates TMEM117 expression which could play a key role in the loss of ΔΨ m induced by the drug.

Cell culture and transfection
HCT116 (Public Health England Culture Collection, Salisbury, UK) and DLD-1 cells (Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer Tohoku University, Miyagi, Japan) were cultured in RPMI-1640 media (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 5% fetal bovine serum and 100 U/mL penicillin and 100 µg/mL streptomycin at 37 °C in a 5% CO 2 atmosphere. Cells less than 25 passages were used for experiments. Electroporation was performed using approximately 1 × 10 6 cells with 10 µg of siRNA using a NEPA21 Super Electroporator (Nepa Gene Co., Ltd., Chiba, Japan).

Quantitative reverse transcription-PCR (RT-qPCR)
Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer's instructions. ReverTra Ace™ qPCR RT Master Mix and THUNDERBIRD™ SYBR® qPCR Master Mix (TOY-OBO, Osaka, Japan) were used to synthesize and amplify cDNA from total RNA. RT-qPCR was performed using Rotor-Gene Q (QIAGEN, Venlo, Netherlands). PCR cycling conditions were as follows: denaturation for 1 min at 95 °C; followed by 40 or 45 cycles of a 3-step amplification method (denaturation for 15 s at 95 °C, annealing for 30 s at 61 °C, extension for 60 s at 72 °C). Gene expression was normalized to that of the acidic ribosomal phosphoprotein P0 (36B4) control. The standard curve method was used to evaluate the gene expression levels.

Trypan blue staining
HCT116 or DLD-1 cells were seeded in 6-well plates at a density of 300,000 cells/well or 600,000 cells/well. After 24 h, cells were treated with 20, 40, 60, 80, or 100 µM celecoxib. After 6 h or 24 h, cells were counted using trypan blue staining. Trypan blue solution was added to the cells at a 1:1 ratio, and the cells were observed under a microscope (CKX41, Olympus) at a 100 × magnification.

WST-8 assay
Cell viability assay was performed as described previously (Tamaki et al. 2017). Briefly, HCT116 or DLD-1 cells were seeded in 96-well plates at a density of 10,000 cells/well or 20,000 cells/well. After 24 h, the cells were treated with 20, 40, 60, 80, or 100 µM celecoxib. After 6 h or 24 h, cells were incubated with WST-8 (Dojindo, Japan). The absorbance of WST-8 formazan was measured using a microplate reader (Bio-Rad, CA, USA) with a 450 nm filter.

Analysis of 1Ψ m
Flow cytometry experiments for ΔΨ m analysis were performed as described previously (Tamaki et al. 2017), with the following minor modification: HCT116 cells were seeded in 6-well plates at a density of 300,000 cells/well. After 24 h, the cells were treated with 40, 50, 60, or 70 µM celecoxib. After an additional 24 h, cells were incubated with 2 µM JC-1 (Life Technologies, Carlsbad, CA, USA), a cationic carbocyanine dye that exhibits potential-dependent accumulation in the mitochondria and is used as an indicator of ΔΨ m , in the growth media at 37 ℃ for 1 h. The cells were harvested, washed twice with 1 × PBS, and resuspended with 1 × PBS. JC-1 fluorescence (530 and 610 nm) was analyzed using FACSAria (Becton Dickinson, Franklin Lakes, NJ, USA) at a wavelength of 488 nm.

Western blotting
HCT116 cells were seeded in 6-well plates at a density of 300,000 cells/well. After 24 h, the cells were treated with 60 and 70 µM of celecoxib. After treatment, the cells were lysed in RIPA buffer (FUJIFILM Wako Pure Chemical Corporation, Tokyo, Japan) supplemented with 1% protease and 1% phosphatase inhibitor cocktails (Sigma-Aldrich, St. Louis, MO, USA). The lysates were separated via SDS-PAGE and transferred onto a nitrocellulose membrane. Thereafter, the membrane was blocked with 2% fish gelatin or 5% skimmed milk for 1 h at room temperature (23 ℃) and probed with the primary antibodies. The probed membrane was washed thrice with 0.1% Tween 20 in Tris-buffered saline (TBS) and then probed with anti-rabbit and anti-mouse secondary antibodies. The probed membrane was washed thrice with 0.1% Tween 20 in TBS; then, proteins were detected using Immu-noStarZeta (FUJIFILM Wako Pure Chemical Corporation).

Statistical analysis
Results are expressed as mean ± SEM. Statistical analysis was performed using Student's t -test to compare pairs of groups. Multiple comparison analysis followed by the Tukey-Kramer test was performed to compare more than two groups.

Effects of celecoxib treatment on cell viability
It was reported that HCT116 cells were COX-2 negative in human colon cancer (Schiffmann et al. 2008). We performed RT-qPCR to verify the COX-2 expression levels of the HCT116 cells used in this study. A good amplification was obtained using human kidney cDNA as a positive template input, whereas no amplification was observed on HCT116 cells. The expression of 36B4, a housekeeping gene, was confirmed on both human kidney cDNA and HCT116 cells. These results suggested that COX-2 was not expressed in the HCT116 cells used in this study (Fig. 1a). Next, to determine whether celecoxib affects the viability of HCT116 cells, we performed trypan blue staining. After treatment with trypan blue solution, the live cells remained unstained, whereas the nuclei of dead cells appeared blue. The number of live cells treated with celecoxib for 6 h rapidly decreased at concentrations higher than 80 µM in trypan blue staining. We also measured cell viability using the WST-8 assay, which decreased in a dose-dependent manner (Fig. 1b). However, this result was not consistent with trypan blue staining since trypan blue solution cannot enter cells unless the cell membrane is compromised. Thus, we tested the cells that were treated with celecoxib for 24 h, and the result was consistent with the WST-8 assay (Fig. 1c). The IC 50 of celecoxib in HCT116 cells was approximately 50 µM. Next, we measured the viability of DLD-1 cells treated with celecoxib to confirm the data using other colon cancer cell lines. The result was similar to that obtained using HCT116 cells (Fig. 1d-e).

Effect of celecoxib on the mitochondrial membrane potential and biogenesis
Next, we investigated whether celecoxib treatment affected ΔΨ m . Cells treated with 100 µM mitochondrial-uncoupling reagent CCCP for 5 min were used as a positive control. Loss of ΔΨ m was analyzed by a decrease in the JC-1 red/ green fluorescence ratio. As shown in Fig. 2a, a ΔΨ m -sensitive color shift was observed in the presence of CCCP and celecoxib. In Fig. 1c, the number of live cells decreased rapidly after celecoxib treatment between 40 and 80 µM. Thus, we performed flow cytometry assay for mitochondrial membrane potential using this range of celecoxib concentration. The treatment of HCT116 cells with 40, 50, 60, and 70 µM celecoxib resulted in ΔΨ m loss in 7.6, 11.3, 35.8, and 55.0% of cells, respectively, whereas ΔΨ m loss occurred in only 6.4% of control cells. Beyond 80 µM of celecoxib treatment, ΔΨ m could not be detected since most of the cells were dead (data not shown).
Next, we examined the expression of key genes involved in mitochondrial biogenesis. The impairment of mitochondrial biogenesis leads to mitochondrial dysfunction, which can be detected as loss of ΔΨ m . SIRT1 plays an important role in deacetylation of proliferator activator receptor gamma-coactivator 1 α (PGC-1α ), which is essential for TFAM expression in skeletal muscle (Baar et al. 2002;Botta et al. 2013). In addition, the PGC-1α -NRF1-TFAM pathway regulates mitochondrial biogenesis in neuronal cells in Alzheimer's disease (Sheng et al. 2012). Thus, we treated cells with 70 µM celecoxib for 24 h, which led to a decrease in more than 50% of the number of live cells and ΔΨ m (Figs. 1c and 2). Additionally, we performed RT-qPCR to measure the expression of genes involved in mitochondrial biogenesis and metabolism. The expression of CHOP, which was considered a positive control (Tsutsumi et al. 2006), increased (p < 0.05, Fig. 3a). SIRT1 expression reduced by 0.85-fold (p < 0.05, Fig. 3b), NRF1 expression decreased by 0.7-fold (p < 0.05, Fig. 3c), and TFAM expression reduced by approximately 50% (p < 0.01, Fig. 3d). Since UCP2 is involved in the maintenance of ΔΨ m (Ding et al. 2019), we also measured UCP2 expression and found that it reduced by 0.3-fold (p < 0.01, Fig. 3e). Mitochondria-associated ER membrane (MAM) proteins are involved in calcium homeostasis, apoptosis, and mitochondrial functions. The expression of MFN2, an MAM and mitochondrial fusion protein, reduced by 0.6-fold (p < 0.01, Fig. 3f). The expression of Sig1R, a chaperone that regulates Ca 2 + signaling (Hayashi and Su 2007), reduced by 0.35-fold (p < 0.01, Fig. 3g). A decrease in expression of these genes in the background of celecoxib treatment was observed in DLD-1 cells (Fig. 4). These results suggest that the reduction in ΔΨ m by celecoxib treatment is partly caused by the downregulation of genes involved in mitochondrial biogenesis and metabolism.

Role of TMEM117 in cell death signaling by celecoxib
A previous study demonstrated that TMEM117 knockdown results in the loss of ΔΨ m (Tamaki et al. 2017). Therefore, we investigated whether celecoxib treatment affects TMEM117 expression. We observed that TMEM117 expression remained unchanged 6 h after celecoxib treatment. However, TMEM117 expression reduced 24 h after celecoxib treatment (Fig. 5a). These results suggest that Fig. 1 Celecoxib treatment reduced the viability of HCT116 and DLD-1 cells. a The gene expression levels of COX-2 in human kidney and HCT116 cells were measured using quantitative RT-qPCR. COX-2 expression was normalized to that of the acidic ribosomal phosphoprotein P0 (36B4) control. HCT116 (b, c) and DLD-1 cells (d, e) were treated with 20, 40, 60, 80, or 100 µM celecoxib for 6 h (b, d) or 24 h (c, e). Cell viability was measured using trypan blue staining (•) and WST-8 assay (○); the number of live cells and cell viability were represented in percentages, with no drug treatment (DMSO) condition set at 100% viability. Data are expressed as mean ± SEM (n = 3 for trypan blue staining; n = 6 for WST-8 assay)   celecoxib treatment reduces TMEM117 expression, resulting in a loss of ΔΨ m . Next, we examined the expression of ER stress response proteins and apoptosis signaling factors. ER stress activates three membrane proteins-ATF6, IRE1, and PERK-in the early stage of the ER stress response. We observed that PERK was activated 1 h after celecoxib treatment (Fig. 5b). Then, we examined other ER stress response proteins, which Upregulation of the spliced X-box binding protein 1 (XBP1), which is activated by IRE1 autophosphorylation, and CHOP, a transcription factor related to ER stress-induced apoptosis and a downstream molecule of PERK signaling, was observed.
After 24 h of celecoxib treatment, caspase-3 and caspase-8 were activated and BCL-2-a negative regulator of apoptosis-was downregulated (Fig. 5b). These results indicate that TMEM117 plays a role as a transducer of cell death signaling from CHOP to the caspase signaling cascade.
To determine whether TMEM117 expression is associated with caspase activation, HCT116 cells were preincubated with 10 µM caspase inhibitor Z-VAD-FMK and then treated with 70 µM celecoxib for 24 h. Consistent with the results of a previous study (Pyrko et al. 2007), depletion of BCL-2 and accumulation of cleaved caspase-3 were observed in cells treated with celecoxib, and this accumulation reduced upon treatment of cells with Z-VAD-FMK (Fig. 6). Conversely, TMEM117 expression was not affected by treatment with Z-VAD-FMK, suggesting that TMEM117 regulation in celecoxib-induced cell death was not related to caspase activation (Fig. 6).

Effect of TMEM117 on the regulation of genes involved in mitochondrial biogenesis
Next, we examined whether TMEM117 regulates the expression of genes involved in mitochondrial biogenesis c d e f and metabolism. To examine the downstream effects of TMEM117 signaling, we performed TMEM117 gene silencing. We used two siRNAs against TMEM117 (TMEM117psi1 and TMEM117_2) in a manner similar to that used in our previous study (Tamaki et al. 2017). Both the siRNAs reduced TMEM117 expression (Fig. 7a). We observed that the expression of MFN2 and Sig1R was reduced upon silencing TMEM117 using either siRNA (Fig. 7b-c), suggesting a specific effect of silencing. Then, we used the mixture of TMEM117psi1 and TMEM117_2 on TMEM117 silencing. Upon TMEM117 knockdown (Fig. 8a), the expression of CHOP increased (p < 0.05, Fig. 8b), consistent with the findings of our previous study (Tamaki et al. 2017). When the expression of genes in the background of transfection with control siRNA was set at 100%, the expression of SIRT1, NRF1, and TFAM was found to be reduced by 30% (Fig. 8c-e; c, p < 0.05; d, e, p < 0.01). The expression of UCP2 reduced by approximately 70% (p < 0.01, Fig. 8f), whereas that of MFN2 and Sig1R reduced by approximately 40% (p < 0.01, Fig. 8g-h). These results suggest that TMEM117 is involved in mitochondrial biogenesis. Furthermore, we examined the effects of celecoxib in the background of TMEM117 silencing. In the background of TMEM117 knockdown and celecoxib treatment, the cell viability and the expression of genes involved in mitochondrial biogenesis decreased, suggesting that celecoxib treatment enhances the reduced mitochondrial biogenesis through TMEM117 silencing (Fig. 9).

Discussion
Herein, we investigated whether celecoxib treatment affects the expression of ER stress response proteins and apoptosis signaling factors. Celecoxib treatment activated PERK, XBP1, caspase-3, and caspase-8; upregulated CHOP; and downregulated BCL-2. These results are consistent with those of previous studies (Tsutsumi et al. 2006(Tsutsumi et al. , 2004Park et al. 2018Park et al. , 2015Cha et al. 2014;Huang and Sinicrope 2010). Moreover, celecoxib induced loss of ΔΨ m and a reduction in the expression of genes involved in mitochondrial biogenesis. Compared to selective COX-2 inhibition (Riendeau et al. 2001), the effect was observed relatively in higher concentration. These results provide evidence that celecoxib affects cell death via its effects on mitochondria. ER stress signals are transmitted to mitochondria through specific sites in the ER called MAMs, where the ER membrane is in close contact with mitochondria both physically and physiologically. ER stress affects ER calcium homeostasis by altering SERCA and inositol 1, 4, 5-triphosphate receptor (IP 3 R) activities, which in turn causes pathophysiological calcium flow from IP 3 R in MAMs to the mitochondrial matrix through voltage-dependent anion channel 1 and Celecoxib treatment caused ER stress-induced HCT116 cell death. HCT116 cells were treated with 60 or 70 µM celecoxib at 1, 2, 3, 6, and 24 h. DMSO was used as control. a TMEM117 protein expression at various time points was measured using western blotting. b ER stress pathway protein expression at various time points was measured using western blotting mitochondria Ca 2+ uniporter located in the outer mitochondrial membrane (OMM) and IMM, respectively. Under ER stress, Sig1R, a stabilizer of IP 3 R, is upregulated by ATF4, a target gene of PERK (Mitsuda et al. 2011). In addition, Sig1R binds to and stabilizes IRE1 at MAMs (Mori et al. 2013). Thus, the ER stress signal is reinforced at MAM sites and transferred to mitochondria. Accordingly, celecoxib can elicit these ER stress responses (Johnson et al. 2002;Kardosh et al. 2008;Pyrko et al. 2007;Kim et al. 2007); we observed the activation of PERK and IRE1 signaling after celecoxib treatment. The loss of ΔΨ m and downregulation of gene expression (as shown in Fig. 3) in mitochondria after celecoxib treatment suggest that the ER stress signal transfer was not only caused via physiological signaling such as Ca 2+ loading but also possibly through changes in the expression of mitochondrial proteins involved in the cell death signaling, although the latter requires experimental verification. For example, MFN2 is a mitochondrial fusion protein located in both the OMM and MAM and its deficiency activates the UPR and increases ROS production and mitochondrial calcium overload by PERK activation (Muñoz et al. 2013). Moreover, PERK localizes at MAMs and promotes the rapid transfer of ROS across the membrane to induce apoptosis; this activity occurs in parallel to the conventional PERK-ATF4-CHOP pathway (Verfaillie et al. 2012). In addition, the decrease in ΔΨ m and TFAM expression levels in MFN2-depleted mouse embryonic fibroblast cells were reported to be greater than that in MFN2-positive Fig. 9 The expression of genes involved in mitochondrial biogenesis with celecoxib treatment was decreased by TMEM117 silencing. HCT116 cells were transfected with 10 µg control siRNA (Control) or mixture of 5 µg each of TMEM117psi1 and TMEM117_2 (siTMEM117).
After 24 h (a) or 48 h (b-e), the cells were treated with 40 µM (a) or 50 µM (b-e) celecoxib (DMSO was used as control) for 24 h. a Cell viability was measured using WST-8 assay; the cell viability was represented in percentages, with no drug treatment (DMSO) condition transfected with control siRNA set at 100% viability. Data are expressed as mean ± SEM (n = 6). b-e RT-qPCR of TMEM117, NRF1, TFAM, and UCP2. Data are expressed as mean ± SEM (n = 3). Statistical analysis was performed using a one-way ANOVA followed by the Tukey-Kramer test. Different letters up the bars indicated significant differences (p < 0.05), and vice versa cells (Kawalec et al. 2015). The inhibition of mitochondrial fusion induces apoptosis (Karbowski et al. 2004). Since celecoxib reduces MFN2 expression (Fig. 3), we speculate that it induces apoptosis by regulating mitochondrial fusion. Although TMEM117 is localized in the ER (Bürgi et al. 2016), it remains unknown whether TMEM117 is an MAM-related protein. TMEM117 can downregulate Sig1R and MFN2, but further studies are required to investigate the interaction between MAM-related proteins and TMEM117. Interestingly, celecoxib treatment affected mitochondrial biogenesis homeostasis. Sirt1 is an NAD + -dependent protein deacetylase and functions in energy metabolism, stress response, and cell survival (Tang 2016). Sirt1 also plays an important role in mitochondrial function. Specifically, Sirt1 interacts with PGC-1α to enhance mitochondrial biogenesis (Tang 2016). We observed the downregulation of sirt1 by celecoxib treatment, which suggested that celecoxib affects mitochondrial homeostasis through the regulation of sirt1 expression. NRF1 is a transcription factor that regulates the expression of genes involved in mitochondrial biogenesis and components of the respiratory chain, and TFAM, which is a nuclear gene. TFAM translocates to the mitochondrion where it activates mitochondrial DNA transcription. We showed that NRF1 and TFAM expressions reduced upon celecoxib treatment. These results support the hypothesis that the disruption of mitochondrial function is caused by celecoxib-induced ER stress (Kardosh et al. 2008;Pyrko et al. 2007;Kim et al. 2007). Since TFAM also regulates Serca2 expression (Watanabe et al. 2011), the depletion of Ca 2 + in ER by celecoxib could interfere with mitochondrial Ca 2 + regulation, leading to loss of ΔΨ m (Koh et al. 2019;Xie et al. 2016). Celecoxib inhibits SERCA activity (Johnson et al. 2002), suggesting that the downregulation of TFAM by celecoxib may abolish cellular Ca 2 + homeostasis and celecoxib may cause apoptosis by inducing mitochondrial dysfunction. ATF4 is translated under ER stress and regulates the NRF1-TFAM pathway, which induces mitochondrial dysfunction in alcoholic liver disease (Hao et al. 2021). Furthermore, UCP2 expression was suppressed by celecoxib treatment. UCP2 is reportedly involved in ΔΨ m (Ding et al. 2019) and ROS production (Galley 2010), and TMEM117 knockdown leads to ΔΨ m loss and an increase in ROS levels (Tamaki et al. 2017), suggesting that the loss of ΔΨ m induced by celecoxib is caused by reduced UCP2 levels via TMEM117 suppression (Galley 2010). The loss of ΔΨ m induced by celecoxib may involve the regulation of TFAM and UCP2 expression by TMEM117. However, the role of TMEM117 in mitochondrial dysfunction is still unknown. Further studies are required to investigate whether the expression of genes involved in mitochondrial biogenesis and celecoxib-induced loss of ΔΨ m could be rescued by TMEM117 overexpression. Interestingly, TFAM expression is reportedly altered in and associated with breast and lung cancer; additionally, polymorphisms in TFAM are reportedly associated with a variety of cancers (Xie et al. 2016;Golubickaite et al. 2021aGolubickaite et al. , 2021bGao et al. 2016;Granados et al. 2017;Guo et al. 2011;Hu et al. 2020;Peng et al. 2013). UCP2 expression is also involved in cancer (Pons et al. 2015;Derdák et al. 2006;Derdak et al. 2008). Therefore, the anticancer activity of celecoxib could be attributed to its effects on TFAM and UCP2 expressions.
In the early stage of the ER stress response, three ER membrane proteins-ATF6, IRE1, and PERK-are activated and upregulate the expression of their target genes. One of these target genes is CHOP, a major transcription factor involved in ER stress-induced apoptosis and triggers the caspase cascade in the middle stage of the ER stress response. Severe ER stress can induce cellular apoptosis through the extrinsic or intrinsic pathway (Ghobrial et al. 2005). The extrinsic pathway is regulated by the death receptor and activates several caspases, including caspase-8. The intrinsic pathway is triggered by cytochrome c release and caspase-9 activation. Moreover, the intrinsic pathway suppresses the expression of the anti-apoptotic protein BCL-2, which is localized at the ER membrane and regulates Ca 2 + homeostasis in the ER. During the last stage of the ER stress response, caspase-3 is activated by both pathways to induce apoptosis. Celecoxib activates ER stress sensors (Tsutsumi et al. 2006) and increases the expression of BiP, an ATF6regulated chaperone protein in HepG2 cells (Maeng et al. 2017). Celecoxib treatment also upregulates CHOP and activates caspases (Tsutsumi et al. 2006(Tsutsumi et al. , 2004Park et al. 2018Park et al. , 2015Cha et al. 2014;Huang and Sinicrope 2010). Furthermore, celecoxib downregulates TMEM117 expression, which suggests that its regulation is related to celecoxibinduced apoptosis. Although celecoxib-induced cell death could be apoptosis due to caspase activation, no evidence has been reported regarding the effects of celecoxib on other cell death pathways, such as necroptosis and ferroptosis; further studies are required to explore this aspect. Celecoxib reduces the viability of HCT116 cells (Schiffmann et al. 2008) due to the ER stress response (Park et al. 2018(Park et al. , 2015Cha et al. 2014;Huang and Sinicrope 2010;Tsutsumi et al. 2004). Our results suggest that TMEM117 downregulation plays a role in the celecoxib-induced ER stress response; however, the interaction between TMEM117 and ER stress response proteins remains unclear.
In summary, the results of our study showed that celecoxib can induce loss of ΔΨ m and downregulate TMEM117. Celecoxib also downregulates mitochondrial biogenesis and metabolic genes such as TFAM and UCP2, and their expressions are suppressed by TMEM117 silencing. Therefore, TMEM117 might play a role in the celecoxib-induced ER stress-induced apoptosis by disrupting mitochondrial biogenesis and metabolism. Although the expression of ER stress markers is increased by celecoxib treatment , the effect of celecoxib treatment on the expression of mitochondrial biogenesis genes has not been studied. Herein, we elucidated that celecoxib disrupts mitochondrial homeostasis and ER homeostasis. In addition, it was reported that 4-phenyl butyric acid (4-PBA), an ER stress inhibitor, can prevent celecoxib-induced apoptosis and reverse the suppression of MFN2 expression by 3-chloro-1, 2-propanediol Zhong et al. 2021). Further studies assessing the effects of ER stress inhibitors such as 4-PBA on the other genes involved in mitochondrial biogenesis are required to conclude that the inhibition of celecoxibinduced ER stress can restore mitochondrial biogenesis. Celecoxib is an anti-inflammatory drug that also exhibits anticancer activity. However, a high dose of celecoxib is required when it is used as an anticancer drug in comparison to when it is used as an anti-inflammatory drug. Our results provide insights for the creation of novel drugs to reduce the side effects of celecoxib, such as gastrointestinal injury, cardiovascular events, hepatotoxicity, and nephrotoxicity (Xiao et al. 2022), through TMEM117 expression.