β-Carotene Inhibits Expression of Matrix Metalloproteinase-10 and Invasion in Helicobacter pylori-Infected Gastric Epithelial Cells

Matrix metalloproteinases (MMPs), key molecules of cancer invasion and metastasis, degrade the extracellular matrix and cell–cell adhesion molecules. MMP-10 plays a crucial role in Helicobacter pylori-induced cell-invasion. The mitogen-activated protein kinase (MAPK) signaling pathway, which activates activator protein-1 (AP-1), is known to mediate MMP expression. Infection with H. pylori, a Gram-negative bacterium, is associated with gastric cancer development. A toxic factor induced by H. pylori infection is reactive oxygen species (ROS), which activate MAPK signaling in gastric epithelial cells. Peroxisome proliferator-activated receptor γ (PPAR-γ) mediates the expression of antioxidant enzymes including catalase. β-Carotene, a red-orange pigment, exerts antioxidant and anti-inflammatory properties. We aimed to investigate whether β-carotene inhibits H. pylori-induced MMP expression and cell invasion in gastric epithelial AGS (gastric adenocarcinoma) cells. We found that H. pylori induced MMP-10 expression and increased cell invasion via the activation of MAPKs and AP-1 in gastric epithelial cells. Specific inhibitors of MAPKs suppressed H. pylori-induced MMP-10 expression, suggesting that H. pylori induces MMP-10 expression through MAPKs. β-Carotene inhibited the H. pylori-induced activation of MAPKs and AP-1, expression of MMP-10, and cell invasion. Additionally, it promoted the expression of PPAR-γ and catalase, which reduced ROS levels in H. pylori-infected cells. In conclusion, β-carotene exerts an inhibitory effect on MAPK-mediated MMP-10 expression and cell invasion by increasing PPAR-γ-mediated catalase expression and reducing ROS levels in H. pylori-infected gastric epithelial cells.


Introduction
Helicobacter pylori (H. pylori) is a Gram-negative bacterium that infects nearly half of the world's population. It is a human pathogen that causes stomach diseases such as gastric inflammation, ulcers, and gastric cancer [1]. H. pylori infection enhances the migration and invasion of gastric cells, which is closely associated with gastric cancer development [2]. Matrix metalloproteinases (MMPs) are zinc-dependent proteinases that can degrade the extracellular matrix (ECM) and cell-cell adhesion molecules [3]. They can promote cancer progression by increasing cancer-cell growth, migration, invasion, and metastasis. Among these, invasion is an important feature of malignant cancer progression [4]. H. pylori infection increases the expression and secretion of various MMPs, including MMP-1 [5,6], MMP-9 [7,8], MMP-7 [9], and MMP-10 [6,10], in the gastric epithelial cells or gastric cancer cells.
Among the MMPs, MMP-10 cleaves numerous ECM components, including fibronectin, proteoglycans, gelatins, and collagens [11]. Since MMPs are synthesized as inactive zymogens (proMMP) and subsequently activated by many factors to degrade the ECM, the activation of pro-MMP is linked to cancer development. MMP-10 cleaves pro-MMPs,

H. pylori Activates MAPKs and Expression of MMP-10 in AGS Cells
First, we wanted to investigate whether H. pylori induces mRNA and MMP-10 protein expression by real-time polymerase chain reaction (PCR) and Western blot analysis, respectively. AGS cells were infected with H. pylori at the indicated ratios. At 24 h, the MMP-10 mRNA was upregulated by H. pylori in a density-dependent manner ( Figure 1A). At a 50:1 bacteria/cell ratio, H. pylori increased the mRNA and protein levels of MMP-10 in a time-dependent manner. The maximum induction of MMP-10 in H. pylori-infected cells was observed at 24 h ( Figure 1B,C). To determine whether H. pylori activates the MAPK signaling pathway, phosphorylated and total forms of MAPKs were detected by Western blotting. H. pylori increased the levels of phosphorylated MAPKs (p-JNK1/2, p-p38, and p-ERK1/2) in AGS cells at 30 min, while the total levels were not changed ( Figure 1D). Levels of both p-JNK1/2 and p-38 steadily increased till 60 min but p-ERK1/2 decreased after 30 min. The expression of MMP-10 mRNA was analyzed by real-time PCR and normalized to β-actin mRNA. All data are shown as the mean ± standard error (S.E.) of three independent experiments. * p < 0.05 vs. none (cells without any treatment or infection). (C) Protein levels of MMP-10 were determined by Western blot analysis, using actin as the loading control. (D) Protein levels of phosphorylated or total form of JNK1/2, p38 and ERK1/2 were determined by Western blot analysis. Actin served as a loading control (left panel). Right panel: the densitometry data represent means ± S.E. from three immunoblots and are shown as relative density of phosphorylated protein band normalized to total form of protein level. * p < 0.05 vs. 0 min.

MAPK Inhibitors Prevent H. pylori-Induced Expression of MMP-10 in AGS Cells
To confirm the involvement of MAPKs in the H. pylori-induced expression of MMP-10, we used MAPK inhibitors SB600125 (JNK inhibitor), SB203580 (p38 inhibitor), and U0126 (ERK inhibitor). AGS cells were pretreated with SB600125 for 60 min, SB203580 for 60 min, and U0126 for 30 min, and infected with H. pylori for 24 h. All three MAPK inhibitors suppressed H. pylori-induced MMP-10 expression in AGS cells ( Figure 2). These results indicate that H. pylori induces MMP-10 expression through JNK, p38, and ERK signaling in AGS cells. The cells were pretreated with SB600125 (a JNK inhibitor, 20 µM) or SB203580 (a p38 inhibitor, 20 µM) for 60 min, or U0126 (an ERK inhibitor, 10 µM) for 30 min, and then infected with H. pylori for 24 h. MMP-10 levels were determined by Western blot analysis. Actin was used as a loading control.

β-Carotene
Inhibits H. pylori-Induced Activation of MAPKs and AP-1, and Expression of MMP-10 in AGS Cells Next, we examined the effect of β-carotene on the H. pylori-induced expression of MMP-10. AGS cells were infected with H. pylori in the presence or absence of β-carotene. β-Carotene inhibited H. pylori-induced mRNA and the protein expression of MMP-10 at 24 h ( Figure 3A,B). This prompted us to examine whether the inhibitory effect of β-carotene is mediated by suppressing MAPK activation. Levels of phosphorylated and total forms of MAPKs were determined in cells treated with or without β-carotene and infected with H. pylori. As shown in Figure 3C, β-carotene markedly reduced the phosphorylation of JNK1/2, p38, and ERK1/2 in a dose-dependent manner. Since MAPK induction is an upstream signaling event in AP-1-mediated MMP-10 expression, AP-1 DNA-binding activity was determined in the nuclear extracts of cells treated with or without β-carotene and infected with H. pylori ( Figure 3D). β-Carotene inhibited H. pylori-induced AP-1 activation in AGS cells These results suggest that β-carotene inhibits the activation of MAPKs and AP-1 and thus, MMP-10 expression in H. pylori-infected AGS cells.

β-Carotene Inhibits Cell Invasion Induced by H. pylori in AGS Cells
To determine whether β-carotene inhibits cell invasion in H. pylori-infected cells, AGS cells were infected with H. pylori in the presence or absence of β-carotene and analyzed using a cell-invasion assay. H. pylori significantly increased cell invasion, which was inhibited by β-carotene ( Figure 4). These results demonstrate that β-carotene inhibits cell invasion induced by H. pylori infection.

β-Carotene Upregulates PPAR-γ and Catalase and Reduces ROS Levels in H. pylori-Infected AGS Cells
Finally, we determined the antioxidant effect and the mechanisms underlying βcarotene in H. pylori-infected cells. Intracellular ROS levels and the expression of PPAR-γ and catalase were measured in cells treated with or without β-carotene (0.2 µM), and infected with H. pylori. As shown in Figure 5A, β-carotene reduced ROS levels, which were increased by H. pylori infection in AGS cells. Treatment with β-carotene (0.2 µM) increased the expression of PPAR-γ at 1 h, but increased the expression of catalase, a PPARγ target gene, till 3 h, in the absence of H. pylori ( Figure 5B). H. pylori infection reduced the expression of PPAR-γ, which was prevented by β-carotene ( Figure 5C). Additionally, H. pylori infection reduced the nuclear levels of PPAR-γ, which was restored by β-carotene treatment ( Figure 5D).
To scrutinize the involvement of PPAR-γ in the antioxidant mechanisms of β-carotene, H. pylori-infected cells were treated with the PPAR-γ antagonist GW9662 and β-carotene. Co-treatment with GW9662 reversed the effect of β-carotene on H. pylori-induced intracellular ROS production ( Figure 5E). These results indicate that β-carotene reduces ROS levels by activating PPAR-γ and inducing the expression of catalase in H. pylori-infected cells.

Discussion
Gastric cancer is the fourth leading cause of cancer. Infection with H. pylori is a significant risk factor for chronic gastritis, peptic ulcers, and gastric carcinogenesis. MMP expression is involved in the invasion and metastasis of malignant tumors, and is induced in H. pylori-infected gastric epithelial cells [5][6][7][8][9][10]. High levels of plasma MMPs were found in gastric cancer patients and, therefore, MMPs could serve as important diagnostic markers of gastric cancer [39]. MMP-10 is upregulated via ERK signaling in H. pylori-infected AGS cells [10]. Consistent with this study, we observed higher MMP-10 expression in AGS cells following H. pylori infection. In addition to ERK, JNK and p38 are activated, which induces MMP-10 expression in H. pylori-infected cells.
Cytokines, growth factors, and phorbol esters activate a group of protein kinases (MAPK kinase kinases (MAPKKKs)) that phosphorylate other kinases (MAPK kinases (MAPKKs)), which in turn are responsible for the phosphorylation and activation of MAPK. Upon activation by MAPKKs, MAPKs translocate to the nucleus to phosphorylate and activate various transcription factors [40,41]. Of particular relevance to MMP transcription, JNKs and ERKs phosphorylate and activate AP-1 family member c-Jun [42,43], which dimerizes with c-Fos to drive the transcription of multiple MMP genes. In the present study, we found that the H. pylori-induced activation of MAPKs mediates AP-1 activation and MMP-10 expression in AGS cells.
ROS have been implicated in the development of gastric cancer. The overproduction of ROS, which leads to the activation of pro-survival signaling pathways, including cancer cell proliferation and invasion, is induced by H. pylori infection [16]. Higher ROS levels activate MAPK signal transduction and the transcription factor, AP-1, in H. pylori-infected cells [18,19]. AP-1 transcription factor is important in regulating MMP family members and is sensitive to regulation by redox conditions [21]. Even though ROS generally increase the AP-1 level for the activation of MMP gene expression, ROS can also reduce the transcriptional activity of AP-1-through direct oxidation of cysteine residues contained within the DNA-binding domain [44]. Therefore, the cycling of cysteinyl residues is among the oxidant-dependent mechanisms that regulate the activity of AP-1. In fact, AP-1 DNAbinding activity increases after the reducing agent thioredoxin treatment in monocytes. When the conserved cysteine in the basic binding motif of both Fos and Jun is oxidized, the binding of these proteins to AP-1 sequences is diminished. On the other hand, if Fos/Jun heterodimers are bound to an AP-1 site, they cannot be oxidized [45]. In the present study, H. pylori infection increased both ROS levels and AP-1-DNA binding activity in gastric epithelial cells. Therefore, cysteine residues in the DNA-binding domain may not be affected nor oxidized by H. pylori infection at the infection ratio of 1:50 (cells: bacterium). On the other hand, Fos/Jun heterodimers may be bound directly to the AP-1 site by MAPK activation in H. pylori-infected cells. Further study should be performed to investigate whether a high infection ratio (more than 1:50, cells: bacterium) causes the oxidation of critical cysteine residues in the DNA-binding domain for AP-1 in gastric epithelial cells.
β-Carotene is the most frequently found carotenoid in the human diet. In the human body, β-carotene is absorbed, distributed, and metabolized by enzymatic and/or non-enzymatic oxidant cleavage into several metabolites. Despite the broadly accepted biological value of β-carotene, it is a double-edged sword, due to its potential antioxidant versus pro-oxidant behavior. Regarding the pharmacological role of β-carotene, a high concentration (50 and 100 µM) of β-carotene increases ROS levels and caspase-3 activity, which results in a reduction in DNA rapier protein Ku protein and apoptosis in gastric cancer AGS cells [46]. Furthermore, 100 µM of β-carotene induces apoptosis by increasing apoptotic protein p53 and decreasing anti-apoptotic Bcl-2 as well as nuclear ataxia-telangiectasia mutated (ATM), a sensor for DNA-damaging agents, in AGS cells [47]. Cui et al. [37] suggested that the chemopreventive activity of a high concentration of β-carotene (>20 µM) is associated with ROS production, which results in mitochondrial dysfunction, cytochrome c release, and apoptosis in human breast cancer MCF-7 cells. Palozza et al. [48] demonstrated that β-carotene (>10 µM) induced ROS production and apoptosis in human leukemic HL-60 cells. In addition, β-carotene (50-100 µM) increases ROS levels and inhibits cell growth with the expression of c-myc in colon adenocarcinoma WiDr cells. Thus, a high concentration of β-carotene may exert pro-oxidant activity, which contributes to anti-cancer effects.
However, our previous study showed that a low concentration (0.5 and 1 µM) of β-carotene reduced ROS levels and inhibited NF-B activation, NF-B-regulated expression of tumor necrosis factor receptor associated factor (TRAF), and hyper-proliferation in H. pylori-infected gastric epithelial cells [49]. A low concentration of β-carotene (0.15 and 0.3 µM) inhibited peroxynitrite anion-induced lipid peroxide production in rat brain synaptosomes [50]. Therefore, we used 0.1 and 0.2 µM of β-carotene to determine the antioxidant mechanism of β-carotene in relation to MMP expression in H. pylori-infected gastric epithelial cells in the present study.
In the present study, we found that β-carotene reduced ROS levels, which was increased by H. pylori infection. β-Carotene reduced cell invasion as well as MMP-10 expression in H. pylori-infected cells. The H. pylori-induced activation of MAPKs and AP-1 was inhibited by β-carotene. The JNK, p38, and ERK inhibitors suppressed H. pylori-induced MMP-10 expression. These results indicate that β-carotene inhibits H. pylori-induced invasion by downregulating the MAPK-mediated activation of AP-1, which in turn reduces MMP-10 expression in gastric epithelial cells.
For the antioxidant mechanism of β-carotene, we found that β-carotene induces the expression and nuclear translocation of PPAR-γ, which promotes catalase expression in H. pylori-infected AGS cells. These data demonstrate that β-carotene reduces ROS levels through the upregulation of the PPAR-γ-mediated expression of the antioxidant enzyme, catalase. Cui et al. [37] demonstrated that β-carotene upregulates PPAR-γ expression, but increases ROS production in MCF-7 breast cancer cells. Kaliappan et al. [56] demonstrated that dietary β carotene inhibits Angiotensin II-induced renal damage by repressing PPARγ and renin 1 in apo E knockout mice. Therefore, the effect of β-carotene on PPAR-γ expression and activation may be different depending on the tissues, stimuli, and environmental and genetic factors. Further studies should be performed to determine the interaction of PPAR-γ, ROS, and antioxidant enzyme expression on H. pylori-associated gastric cancer development.
Recently, Mao et al. [57] screened epidemiological studies and demonstrated that increased intake of phytochemicals in fruits and vegetables could reduce the risk of gastric cancer. They suggested that phytochemicals have the potential for the prevention and management of gastric cancer in humans. This study supports the present finding showing that dietary β-carotene, which is abundant in fruits and vegetables, may reduce H. pyloriassociated gastric cancer incidence by reducing the oxidative stress-mediated MMP-10 expression and invasive property of gastric epithelial cells.
In conclusion, β-carotene activates PPAR-γ and induces its downstream target, catalase, which leads to a reduction in ROS levels and the inhibition of JNK/p38/ERK signaling in H. pylori-infected gastric epithelial cells. Through this mechanism, β-carotene significantly suppressed H. pylori-induced MMP-10 expression and reduced the invasive phenotype of infected gastric epithelial cells.

Cell Line and Culture Conditions
Human gastric epithelial AGS cells were purchased from the American Type Culture Collection (Rockville, MD, USA). The cells were cultured in complete medium consisting of Roswell Park Memorial Institute (RPMI) 1640 medium (GIBCO, Grand Island, NY, USA) with 10% fetal bovine serum (FBS), 2 mM glutamine, 100 units/mL penicillin, and 100 µg/mL streptomycin (Sigma-Aldrich, St. Louis, MO, USA). The cells were cultured at 37 • C and maintained in a humidified atmosphere containing 5% CO 2 .

Bacterial Strain and Growth Conditions
All experiments were performed with H. pylori, cag A positive strain 60190 (ATCC 49503). Chocolate agar plates (Becton Dickinson Microbiology Systems, Cockeysvile, MD, USA) were used to grow the bacteria at 37 • C under microaerophilic conditions using an anaerobic chamber (BBL Campy Pouch ® System, Becton Dickinson Microbiology Systems, Franklin Lakes, NJ, USA).

Infection of AGS Cells with H. pylori
AGS cells were seeded and cultivated overnight to reach 80% confluency. H. pylori were harvested from the chocolate agar plates and resuspended in antibiotic-free RPMI 1640 medium supplemented with 10% FBS. Subsequently, AGS cells were infected with H. pylori at multiplicity-of-infection of 50:1.

Experimental Protocols
To investigate the effect of β-carotene, AGS cells (1.5 × 10 5 /2 mL, 7.0 × 10 5 /10 mL) were pretreated with β-carotene (0.1 or 0.2 µM) for 2 h and then infected with H. pylori for 30 min (for ROS levels, PPAR-γ DNA binding activity, protein levels of PPAR-γ and catalase), 1 h (for protein expression of JNK/p38/ERK and DNA binding activity of AP-1), and 24 h (for mRNA expression and protein levels of MMP-10, and invasion assay). To determine the involvement of MAPKs, the MAPK inhibitors SB203580 and SP600125 were added to the culture medium 1 h before infection with H. pylori. In the case of U0126, the cells were pretreated with U0126 for 30 min before infection with H. pylori.

Preparation of Cell Extracts
The cells were collected by treatment with trypsin-ethylenediaminetetraacetic acid (EDTA), followed by centrifugation at 1000× g for 5 min. The cell pellets were resuspended in lysis buffer, including 10 mM Tris pH 7.4, 15 mM NaCl, 1% NP-40, and protease inhibitor cocktail (Complete; Roche, Mannheim, Germany), and lysed by passing the cells through a 1 mL syringe several times. The resulting mixture was incubated on ice for 30 min and centrifuged at 13,000× g for 15 min. The supernatants were collected and used as wholecell extracts. The cytosolic and nuclear extracts were prepared using a NE-PER ® nuclear and cytoplasmic extraction kit (Thermo Fisher, Waltham, MA, USA) according to the manufacturer's instructions. In brief, cells were resuspended in the cytoplasmic extraction reagent containing protease inhibitors and vortexed for 15 s, followed by centrifugation at 13,000× g for 10 min. The supernatants were used as cytosolic extracts. The nuclear pellet was resuspended in nuclear extraction reagent on ice, and then centrifuged at 13,000× g for 10 min. The supernatants were collected and used as nuclear extracts. The specificity of the nuclear extracts was confirmed by the predominant presence of lamin B1 in the nuclear fraction. A Bradford assay (Bio-Rad Laboratories, Hercules, CA, USA) was used to determine the protein concentrations.

Real-Time PCR Analysis
Total RNA was isolated using TRI reagent (Molecular Research Center, Inc., Cincinnati, OH, USA). The isolated RNA was reverse transcribed into cDNA using random hexamers and MuLV reverse transcriptase (Promega, Madison, WI, USA) at 23 • C for 10 min, 37 • C for 60 min, and 95 • C for 5 min. For real-time PCR, the cDNA was amplified with specific primers for human MMP-10 and β-actin. For the MMP-10 PCR product, 5 -CATTCCTTGTGCTGTTGTGTC-3 (forward primer) and 5 -TGTCTAGCTTCCCGTTCACC-3 (reverse primer) were used. The sequence of β-actin primers was the following: 5 -ACCAACTGGGACGACATGGAG-3 (forward primer) and 5 -GTGAGGATCTTCATGAGG TAGTC-3 (reverse primer). The cDNA was amplified by 45 cycles including denaturation at 95 • C for 30 s, annealing at 55 • C for 30 s, and extension at 72 • C for 30 s. During the first cycle, the 95 • C step was extended to 3 min. Amplification of the β-actin gene was performed in the same reaction and served as the reference gene.

Invasion Assay
Invasion of AGS cells infected with H. pylori was evaluated in Matrigel-coated 24-well invasion chambers (#354480, Corning Life Sciences, Teterboro, NJ, USA). The upper and lower compartments were separated by Matrigel-coated filters with 8 µm pore size. AGS cells (4.0 × 10 4 cells/well) were seeded on the reconstituted basement membrane in RPMI 1640 supplemented with 0.01% FBS. Following infection with H. pylori, the cells were incubated for 24 h in the upper chamber with or without β-carotene. Cells which crossed the filter and attached to the lower surface of the Matrigel-coated membrane (invasive cells) were washed with PBS and fixed in 4% formaldehyde for 30 min at room temperature. The cells were rinsed twice with PBS and then stained using 4 ,6-diamidino-2-phenylindole (DAPI) for 30 min. The number of cells that had migrated to the lower surface was counted in twelve random fields using a laser scanning confocal microscope (Zeiss LSM 880, Carl Zeiss Inc., Thornwood, NY, USA) (10×).

Measurement of Intracellular ROS Levels
For the measurement of intracellular ROS, cells were treated with 10 µg/mL of dichlorofluorescein diacetate (DCF-DA; Sigma-Aldrich, St. Louis, MO, USA) and incubated at 37 • C for 30 min. DCF fluorescence (excitation 495 nm and emission 535 nm) was measured using a Victor5 multi-label counter (PerkinElmer Life and Analytical Sciences, Boston, MA, USA). ROS levels were quantified based on the relative increase in fluorescence.
Protein level was compared to that of the loading control actin or total form of MAPK (ERK1/2, JNK1/2, p-38). Intensity of each protein band was densitometrically quantified by using the software Image J (National Institutes of Health, USA). The densitometry data represent means ± S.E. from three immunoblots and are shown as relative density of protein band normalized to actin or total form of MAPK.

Statistical Analysis
One-way analysis of variance followed by the Newman-Keuls post-hoc test was used for statistical analysis. All data are reported as the mean ± S.E. of three independent experiments. For each experiment, the number of each group was 4 (n = 4 per each group). A p-value less than 0.05 was considered statistically significant.

Data Availability Statement:
The data used to support the findings of this study are included within the article.

Conflicts of Interest:
The authors declare no conflict of interest.