A nonsteroidal anti-inflammatory drug, zaltoprofen, inhibits the growth of extraskeletal chondrosarcoma cells by inducing PPARγ, p21, p27, and p53

ABSTRACT Peroxisome proliferator-activated receptor gamma (PPARγ) is a nuclear receptor and master transcription factor of adipogenesis-related genes, and has been reported as an antitumor target for chondrosarcomas. Herein, we show that the nonsteroidal anti-inflammatory drug, zaltoprofen, induces the expression of PPARγ at the mRNA and protein levels, following the induction of PPARγ-activating factors, such as Krox20, C/EBPβ, and C/EBPα, in human extraskeletal chondrosarcoma H-EMC-SS cells. Upregulation of the cell cycle checkpoint proteins, p21, p27, and p53, was observed upon treatment of H-EMC-SS cells with zaltoprofen, which probably resulted in the inhibition of proliferation of these cells observed in vitro. Zaltoprofen treatment inhibited tumor growth, induced tumor cell apoptosis, and was well tolerated in a mouse model of extraskeletal myxoid chondrosarcoma. Our results provide mechanistic insights into the therapeutic effect of zaltoprofen that should promote further studies on the rational use of this drug for the effective treatment of sarcomas.


Introduction
Extraskeletal myxoid chondrosarcoma (EMC) is a soft tissue sarcoma that arises at various sites including the extremities [1]. Because radiation and chemotherapy are ineffective, wide resection is the only radical treatment available for EMC [1]. Although considered a slow-growing neoplasm, the recurrence and metastatic rates of EMC after radical resection (with values up to 50% for each) are higher than those for soft-tissue sarcomas [1,2].
We recently reported that peroxisome proliferator-activated receptor gamma (PPARγ), a ligandactivated nuclear receptor that controls the differentiation of adipocytes in normal tissues, could be a therapeutic target for inhibiting cell invasion and matrix metalloproteinase-2 (MMP2) expression in chondrosarcoma cells [3]. In addition, we demonstrated that zaltoprofen, a nonsteroidal antiinflammatory drug (NSAID), could effectively induce and activate PPARγ, and elicit antitumor effects in chondrosarcoma cells [3].
To better understand the therapeutic potential of PPARγ-targeted treatment using zaltoprofen, in the present study, we investigated the mechanism of zaltoprofen-mediated induction of PPARγ and its anti-proliferative effect in vitro and in vivo on human EMC cells (H-EMC-SS).

5-ethynyl-2-deoxyuridine (EdU) proliferation assay
To investigate the proliferation of cells after zaltoprofen treatment, EdU assay was performed as follows.

Animal experiments
To investigate the antiproliferation effect of zaltoprofen in vivo, four-week-old BALB/c female nude mice (Charles River Laboratories Japan, Kanagawa, Japan) were adapted to the laboratory conditions for a week before the experiment, and H-EMC-SS cells (4 × 10 6 /100 µL) suspended in Matrigel (BD Japan, Tokyo, Japan) were implanted subcutaneously on the flanks of mice. The mice were divided randomly into zaltoprofen (30 mg/ kg, oral gavage, twice a day, 42 days)-treated (n = 6) and vehicle (0.2% CMC sodium, 30 mL/kg, oral gavage twice a day for 42 days)-treated (n = 6) groups. Tumor diameters and body weights of mice were measured once a week using digital calipers and digital scales, respectively. Tumor volume (mm 3 ) was calculated using the following formula: tumor volume (mm 3 ) = largest diameter (mm) × smallest diameter (mm) × smallest diameter (mm) × 0.5 [4]. The animals were observed on a daily basis and humanely sacrificed if they met the following endpoint criteria: severe tumor burden (more than 20 mm in diameter), prostration, significant body weight loss, difficulty breathing, rotational motion and body temperature drop. All treatment groups were housed without cage separation to minimize potential confounders. After 42 days of treatment, the mice were sacrificed humanely, scanned by computed tomography, and tumors were resected for histopathological evaluation. The resected tumors were fixed in 10% formalin (Sigma-Aldrich Japan) and embedded in paraffin. Tumor sections were deparaffinized and rehydrated in xylene and ethanol, then stained with hematoxylin and eosin or immunostained with antibodies against Ki-67 (Abcam, Cambridge, UK), a proliferation marker, in combination with diaminobenzidine (Dako Japan, Kyoto, Japan). To observe apoptosis in the tumor sections, the terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) assay was performed using a cell death kit (Roche) according to the manufacturer's instructions. The TUNEL-stained area was measured and Ki-67-positive cells were counted using ImageJ software. To analyze the drug toxicity, kidney, intestine, and blood serum samples were collected from mice and evaluated pathologically and biochemically for side effects by an outsourced inspection facility (Safety Research Institute for Chemical Compounds, Sapporo, Japan).

Statistical analyses
Easy R (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a modified version of R commander (The R Foundation for Statistical Computing, Vienna, Austria), which includes statistical functions for biostatistics, was used for all statistical analyses [5]. Data were compared using Student's t-test or a one-way analysis of variance with Tukey's post-hoc test. Each experiment was repeated at least three times. All data are expressed as means ± standard error of the mean. All p-values are two-sided.

Induction of PPARγ by zaltoprofen
Zaltoprofen treatment upregulated the mRNA and protein levels of PPARγ in H-EMC-SS cells as evidenced by qRT-PCR and western blot analyses, respectively. This demonstrated that zaltoprofen can induce PPARγ in H-EMC-SS cells (Figure 1(a,  b)). Next, we investigated the induction of Krox20 and C/EBPs, which regulate PPARγ in normal adipose tissue, in zaltoprofen (400 µmol/L)-treated H-EMC-SS cells. Krox20 was significantly upregulated from 3 to 24 h after the initiation of treatment, followed by a significant upregulation of C/EBPβ starting 12 h and of C/EBPα at 24 h (Figure 1(c)), demonstrating the signal pathway of PPARγ induction by zaltoprofen in H-EMC-SS cells.

Inhibition of cell proliferation by zaltoprofen
The inhibitory effect of zaltoprofen on the proliferation of H-EMC-SS cells in vitro was examined by the WST-8 and EdU assays. After treatment of H-EMC-SS cells with zaltoprofen (0-400 µmol/L) for 96 h, their growth was significantly reduced in concentration-and time-dependent manners (Figure 2(a)). To investigate PPARγ-dependency, cell growth was analyzed after downregulation of PPARγ using shRNA. The knockdown of PPARγ was confirmed using qRT-PCR (Figure 2(b)). The relative expression of PPARγ mRNA was significantly decreased in H-EMC-SS shPPARγ cells compared to that in H-EMC-SS shNT cells (p < 0.001). The growth of H-EMC-SS shPPARγ cells was significantly increased compared to that of H-EMC-

Inhibitory effects of zaltoprofen on xenograft tumors in mice
The effect of zaltoprofen on xenograft tumors and its safety were tested using a xenograft mouse model of subcutaneously implanted H-EMC-SS cells. We determined the effect of 30 mg/kg zaltoprofen, which is equivalent to 400 µmol/L in vitro (the effective concentration). Zaltoprofen treatment significantly inhibited tumor growth at all time points compared with that in the vehicle group ( Figure 5 (a), p = 0.032). Computed tomography images of tumors and photographs of resected tumors from zaltoprofen-treated mice at the treatment end point showed decreased tumor size and increased calcification inside the tumor (Figures 5(b) and 6(a, b)).
Histopathological analyses were conducted on the excised tumors. Hematoxylin and eosinstained sections showed that tumors in the vehicle group consisted of viable highly dense tumor cells, whereas zaltoprofen-treated tumors had scattered scar areas (Figure 6c   The body weights in both vehicle-and zaltoprofen-treated mice post-administration were higher than the pre-administration weights (p = 0.005 and p = 0.014, respectively; Figure 8, suggesting natural body weight increase over 6 weeks. Generally, long-term administration of NSAIDs can cause kidney or gastrointestinal mucosa injury. However, there were no gastrointestinal damages evident upon gross observation, no abnormal renal function as evident from the absence of abnormalities in the blood levels of urea nitrogen (BUN) and serum creatinine (Table 1), and no pathological abnormalities based on renal autopsy results in both groups. In addition, no hepatotoxicity was detected in the analysis of serum total bilirubin (T-Bil) and alanine aminotransferase (ALT) in either of the groups (Table 1). There were no other observable side effects or deaths in either of the groups.

Discussion
We found that zaltoprofen induced the expression of PPARγ in human H-EMC-SS cells in a concentration-dependent manner by inducing KROX20, C/EBPβ, and C/EBPα in that order ( Figure 1). Zaltoprofen (400 µmol/L) treatment also induced expression of the cell cycle checkpoint proteins, p21, p27, and p53, and significantly inhibited the proliferation of H-EMC-SS cells (Figures 2-4). PPARγ knockdown did not completely cancel the antitumor effect of zaltoprofen (Figures 2(c) and 3), suggesting that zaltoprofen affects cell proliferation in both PPARγ-dependent and -independent manners. The antiproliferative effects of zaltoprofen on H-EMC-SS cells were also demonstrated in vivo for the first time (Figures 5-7) without observable side effects (Table 1 and Figure 8).
PPARγ expression has been reported to correlate with malignancy and prognosis in several cancer types [12]. Downregulation of PPARγ stimulates cell proliferation in cancer cells such as esophageal carcinoma or hepatocellular carcinoma cells, suggesting that PPARγ possesses a possible protective role in cancer development [13,14]. In the present study, the knockdown of PPARγ by shRNA in H-EMC-SS cells significantly increased cell viability and proliferation, suggesting that PPARγ is involved in the process of malignancy in EMC. However, on the contrary,   downregulation of PPARγ has been reported to inhibit cell proliferation in prostate and bladder cancers [15,16]. Further studies are required to elucidate the exact role of PPARγ in the oncogenesis and malignancy of cancer including EMC. C/EBPs form a family of six homologous transcription factors that are involved in regulating cellular proliferation and differentiation, metabolism, and immunity in various organs [17,18]. C/ EBPβ and C/EBPδ are induced early during adipogenesis and are involved in the induction of PPARγ [19,20]. C/EBPα is induced at a later stage of adipogenesis, which is almost the same stage at which PPARγ is induced, and is activated in mature adipocytes [21,22]. C/EBPα promotes adipogenesis both with and without PPARγ induction [21,22]. Krox20, also referred to as early growth response protein 2, a transcription regulatory factor, is an upstream regulator of C/EBPs and one of the earliest factors to be induced during adipogenesis [23,24]. In normal tissues such as fat tissue, some extrinsic and intrinsic factors such as dexamethasone, insulin, or 3-isobutyl-1-methylxanthine, a phosphodiesterase inhibitor which stimulates intracellular cyclic adenosine monophosphate (cAMP), have been reported to induce PPARγ via inducing Krox20 and C/EBP β and α [25]. We believe that similar signal pathways may be responsible for PPARγ induction by zaltoprofen in H-EMC-SS cells. Herein, we demonstrate that, in H-EMC-SS cells, zaltoprofen induces Krox20 at the earliest stage, followed by C/EBPβ and C/EBPα, but not C/EBPδ. This pathway could be one of the mechanisms for the induction of PPARγ by zaltoprofen in H-EMC-SS cells. However, more detailed analysis to understand the relationship among these factors, including their protein expression, is needed to clarify how zaltoprofen up-regulates these proteins in tumor cells.
It has been reported that Krox20 could directly affect the apoptotic pathway mediated by p53 [26]. The expression of endogenous Krox20 is significantly low in hepatocellular cancers, and the proliferation of cancer cells is suppressed by Krox20 induction [26]. C/EBP α can directly inhibit cyclin-dependent kinase (CDK) 2/CDK 4 activity, and its reduced expression is correlated with poor survival in hepatocellular cancer patients [17].
Biological properties of C/EBP β resemble those of C/EBP α [17]. C/EBP β has been reported to be an independent predictor of overall survival in breast cancer patients [27]. Whether zaltoprofen exerts its antitumor effect not only via PPARγ but also via C/EBPs and Krox20, or how these factors interact is a topic for future studies.
From the results of WST-8 assay and EdU assay in the present study, cell proliferation of H-EMC-SS cells were significantly inhibited by zaltoprofen treatment, suggesting that zaltoprofen could inhibit the cell cycle. In addition, zaltoprofen treatment up-regulated p21, p27, and p53 in a PPARγdependent manner demonstrated by western blot analysis, which could result in the suppression of proliferation in H-EMC-SS cells. The cyclindependent kinase inhibitors, p21 and p27, are essential tumor suppressor proteins that play significant roles in inhibiting the growth of cancer cells [28]. It has been reported that activation of PPARγ induces cell cycle arrest through the upregulation of p21 and p27 in various cancer cells, such as hepatic cell carcinoma [29][30][31], melanoma [32], and pancreatic carcinoma [33]. The tumor suppressor protein, p53, also regulates the cell cycle by repairing DNA, inducing apoptosis, and inducing p21 and p27; thus, p21, p27, and p53 interact in a complex manner to suppress the cell cycle [34]. PPARγ has been reported to directly target p53, contributing to cell-cycle arrest and/or apoptosis of cancer cells [35,36]. Nuclear factor κB (NFκB), which is a modulator of p53, can be an intermediary between p53 and PPARγ; upregulation of PPARγ downregulates NFκB, resulting in the upregulation of p53 [37]. Whether the upregulation of p53 by zaltoprofen is modulated by PPARγ directly or indirectly through the other factor, as well as the relationship between PPARγ and p53 needs to be further analyzed.
In the present study, p21, p27, and p53 were induced in H-EMC-SS cells upon zaltoprofen treatment in a PPARγ-dependent manner and inhibited their proliferation. Similarly, it has been reported that a PPARγ agonist, ciglitazone induced cell-cycle arrest in bladder cancer via upregulation of p21, p27, and p53 [38]. More detailed analysis is needed to clarify the interaction between zaltoprofen-induced PPARγ expression and p21, p27, p53 expression and examination of the involvement of other mechanism in tumor cells. Flow cytometric analysis is required to elucidate the detailed antiproliferative effects of zaltoprofen on the cell cycle. These include identifying which phases of the cell cycle are involved in zaltoprofen-treated cells. In addition, the expression of the other cell-cycle related proteins that correspond to p21 and p27 such as CDK 2, CDK 4/6, and cyclin A, D and E should be studied in future. Whether zaltoprofen induces p21 or p27 in a p53dependent or -independent manner, and PPARγindependent antitumor effects of zaltoprofen, as evident in the WST-8 and EdU assays in the present study, require confirmation in future studies. A recent study has demonstrated that flurbiprofen and zaltoprofen platinum complexes exerted remarkable DNA damage and apoptosis with upregulation of p53 in cancer cells both in vitro and in vivo [39]. In this series, inflammation in tumor tissues is restrained by the suppression of inflammasomes including cyclooxygenase (COX)-2, matrix metalloproteinase (MMP)-9, NLR family pyrin domain containing 3 (NLRP3) and caspase-1 which have been reported to promote cancer development [39]. These inflammasomes may have also influenced the zaltoprofen-treated tumors in the present study.
In the present study, we demonstrated the in vivo antiproliferative effect of zaltoprofen on musculoskeletal tumor cells for the first time. The downregulation of Ki-67, a proliferation marker expressed during the active cell-cycle phases, in zaltoprofen-treated tumors indicates inhibition of the cell cycle in vivo by zaltoprofen, consistent with the in vitro results. There were more TUNEL-stained cells in zaltoprofen-treated tumors, indicating that zaltoprofen could induce apoptosis of tumor cells. The good tolerability of zaltoprofen, which was revealed by natural increase in body weight over 6 weeks and the absence of major complications including body weight loss, nephrotoxicity, and hepatotoxicity, supports further analysis on the clinical efficacy of this drug against musculoskeletal tumors. Further studies are required to determine the efficacy of zaltoprofen on normal cells and to evaluate its anti-tumorigenic effect and cell toxicity.
In conclusion, we demonstrated that zaltoprofen induces Krox20, C/EBPβ, C/EBPα, and PPARγ, and subsequently p21, p27, and p53 in H-EMC-SS cells, and contributes to the inhibition of cell proliferation. The antiproliferative effect of zaltoprofen was also demonstrated in vivo with good tolerability.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This work was supported by JSPS KAKENHI grant number [JP17K16682, JP 20K22802].