Curcumin Inhibits Non-Small Cell Lung Cancer Cells Metastasis through the Adiponectin/NF-κb/MMPs Signaling Pathway

Adipose tissue is now considered as an endocrine organ involved in metabolic and inflammatory reactions. Adiponectin, a 244–amino acid peptide hormone, is associated with insulin resistance and carcinogenesis. Curcumin (diferuloylmethane) is the principal curcuminoid of the popular Indian spice, turmeric. Curcumin possesses antitumor effects, including the inhibition of neovascularization and regulation of cell cycle and apoptosis. However, the effects of adiponectin and curcumin on non-small cell lung cancer (NSCLC) remain unclear. In this study, we evaluated the expression of adiponectin in paired tumors and normal lung tissues from 77 patients with NSCLC using real-time polymerase chain reaction, western blotting, and immunohistochemistry. Kaplan–Meier survival analysis showed that patients with low adiponectin expression ratio (<1) had significantly longer survival time than those with high expression ratio (>1) (p = 0.015). Curcumin inhibited the migratory and invasive ability of A549 cells via the inhibition of adiponectin expression by blocking the adiponectin receptor 1. Curcumin treatment also inhibited the in vivo tumor growth of A549 cells and adiponectin expression. These results suggest that adiponectin can be a prognostic indicator of NSCLC. The effect of curcumin in decreasing the migratory and invasive ability of A549 cells by inhibiting adiponectin expression is probably mediated through NF-κB/MMP pathways. Curcumin could be an important potential adjuvant therapeutic agent for lung cancer in the future.


Introduction
Lung cancer remains the leading cause of cancer-related mortality worldwide and in Taiwan. Despite great advances in the understanding of lung carcinogenesis and in novel treatments in the past few decades, the overall 5-year survival rate remains poor. Innovative research effort

Demographic data and adiponectin expression in NSCLC patients
Of the 77 NSCLC patients in this study, 58 (75%) had histologically confirmed adenocarcinoma and 19 (25%) had squamous cell carcinoma. Their average age was 61.6 ± 10.3 years (range, 36-78 years). Adiponectin expression was not correlated with tumor (T), lymph nodes (N), and stages. NSCLC patients with metastasis had significantly higher adiponectin expression ratio ( Table 1). The Kaplan-Meier survival analysis showed that NSCLC patients with a low adiponectin expression ratio had a significantly longer survival time than those with a high adiponectin expression ratio (p = 0.015) (Fig 1). Multivariate-adjusted risk ratios were computed from Cox regression with the additional variables of sex (male vs. female), metastasis, tumor, lymph node involvement, and stages ( Table 2). The high adiponectin expression group had a 2.40-fold higher mortality risk (p = 0.04) than the low expression group.

Expression levels of adiponectin, adiponectin receptors, and MMPs in NSCLC patients
The expression levels of adiponectin, adiponectin receptors, and MMPs of NSCLC patients were analyzed (Fig 2). Immunohistochemical staining and western blotting of lung cancer   tissue showed an increased expression of adiponectin and adiponectin receptor 1 in the lung cancer tissue than in normal lung tissue (Fig 2A-2C). All MMPs were also increased in the lung cancer tissue than in the normal lung tissue (Fig 2D-2F).

Cytotoxic effect of curcumin on the expression of adiponectin and adiponectin receptors in A549 cells
The A549 cells were treated with different concentrations of curcumin. The MTT assay revealed that curcumin had a cytotoxic effect on the A549 cell lines at concentrations >75 μM (Fig 3A). Western blotting also revealed that the expression of adiponectin in A549 cells decreased significantly at a curcumin concentration of >50 μM (Fig 3B). The protein expression of adiponectin receptor 1 and adiponectin receptor 2 did not change with curcumin treatment even at high concentrations. The transfection of A549 cells with adiponectin vector or small interfering RNA (siRNA) plasmid successfully increased or silenced the expression of adiponectin, as demonstrated by polymerase chain reaction (PCR) and western blotting (Fig 3C and 3D). Effects of curcumin and adiponectin on the migratory and invasive ability of A549 cells Cellular migration was analyzed by wound scratch assay, while invasiveness was determined by Matrigel-coated Boyden chamber assay. Compared to the control group, the migratory and invasive ability of A549 cells was significantly inhibited with curcumin treatment in a dosedependent manner (p < 0.05) (Fig 4A and 4B). On the other hand, recombinant adiponectin increased the invasive and migratory ability of A549 cells in a dose-dependent manner (p < 0.05) (Fig 4C and 4D).

Regulatory roles of adiponectin receptor-1 and adiponectin receptor 2 in migration and invasiveness of A549 cells
The A549 cells were transfected with adiponectin siRNA, adiponectin receptor 1 siRNA, and adiponectin receptor 2 siRNA plasmid. Plasmid transfection significantly reduced the expression levels of adiponectin, adiponectin receptor 1, and adiponectin receptor 2 in A549 cells, as  confirmed by western blotting (Fig 5A-5C). Silencing the adiponectin expression did not affect the expression of both adiponectin receptor 1 and receptor 2, but silencing the expression of adiponectin receptor 1 blocked the effect of adiponectin, which enhanced the migratory and invasive ability of A549 cells. However, silencing the expression of adiponectin receptor 2 did not antagonize the effect of adiponectin on the migration and invasiveness of A549 cells ( Fig  5D and 5E).

Co-immuno-precipitation (Co-IP) assay
This assay showed that adiponectin consisted of monomer, dimer, and multimer ( Fig 7A). Curcumin and silencing adiponectin transfection successfully decreased the expression of adiponectin monomer and dimer. Curcumin also decreased the p65 expression, as shown by the Co-IP and EMSA assays (Fig 7B and 7C). Enhanced adiponectin expression by transfection with adiponectin also increased NF-κB expression. Adiponectin can translocate to nucleus to bind with nuclear p65.

In vivo and in vitro effect of adiponectin on MMP expression
Transfection with adiponectin successfully increased the expression levels of MMP-2, -9, -1, and -14 in A549 cells. However, the expression of MMP-3 and -13 did not change with enhanced or silenced adiponectin expression (Fig 8).

In vivo effect of curcumin on tumor growth and expression of adiponectin and MMPs
Curcumin treatment significantly decreased the in vivo tumor growth and adiponectin expression compared with the control group (DMSO) (p < 0.05) (Fig 9A and 9B), but the expression of both adiponectin receptor 1 and receptor 2 did not change (Fig 9C). Curcumin treatment also significantly decreased the levels of all MMPs, except MMP-1 that remained uninfluenced, as compared with the control group (Fig 9D-9F).

Discussion
Chemotherapy remains the primary treatment for advanced lung cancer [26], although targeted therapy is more popular. Serious side effects of chemotherapy usually limit their clinical application. Consequently, clinicians are becoming increasingly interested in herbal medicine since the past few decades because of its safety and effectiveness as an adjuvant therapy. Curcumin, a phenolic compound derived from the plant Curcuma longa, has been used for centuries for its anti-inflammatory, chemopreventive, and potential chemotherapeutic properties. Till date, current evidence shows that curcumin can block cancer cell proliferation, transformation, and invasion [27]. This study showed that curcumin can inhibit the migratory and invasive capacity of A549 cells in vitro in a dose-dependent manner. Moreover, curcumin also significantly decreased the in vivo tumor growth. Tumor invasion and metastasis is a complicated multistep process that involves adherence, degradation of the surrounding matrix, migration, proliferation, and angiogenesis [28]. Curcumin possesses high natural antimetastatic ability. It can decrease the invasion of B16F-10 melanoma cells by inhibiting the expression of MMP-2 and MMP-9 and enhance the expression of antimetastatic proteins (E-cadherin and tissue inhibitors of MMP-2) [29,30]. In a xenograft tumor model of prostate or breast cancer, curcumin also showed high antimetastatic effect by suppressing the expression of MMP-9, NF-κB, and cyclooxygenase-2 [31,32]. In this study, curcumin can inhibit the migration and invasion of A549 in a dose-dependent manner in vitro. In our xenograft model with A549, curcumin can significantly decrease tumor growth and has a similar suppressing effect of the expression of MMPs, except for MMP-1 and -13.
Adipose tissue is considered as an endocrine system that can secrete a lot of hormones and cytokines, which are known as adipocytokines [33]. Adiponectin represents the most abundant adipose tissue protein with insulin-sensitizing, anti-inflammatory, and anti-atherogenic properties [34,35]. In addition, adiponectin may play a role in the development and progression of various types of malignancies [35]. The serum adiponectin level is inversely associated with the risk of obesity-related malignancies, including breast, endometrial, and prostate cancers, but it is positively correlated with gastric cancer and leukemia [35].
However, the prognostic role of circulatory adiponectin in lung cancer remains controversial [12,14]. In this report, adiponectin expression does not correlate to lung cancer TN and stages. Lung cancer patients with low adiponectin expression ratio showed longer survival time than those with high expression. Thus, adiponectin expression may be a prognostic factor of NSCLC.
In this study, we also demonstrated that NSCLC patients with metastasis have significantly higher adiponectin expression ratio than those without metastasis. The overexpression of adiponectin by transfection can increase the invasive and metastatic ability of A549 cells, perhaps through the activation of MMPs (1,9,14). Adiponectin can inhibit colon [36,37] and liver [37] cancer cell migration but promotes prostatic and endometrial carcinoma growth. Therefore, the role of adiponectin in carcinogenesis is variable in different cancer cell types.
Thus far, three types of adiponectin receptors have been identified: the two main receptors AdipoR1 and AdipoR2 and one receptor similar to the cadherin family. AdipoR1 and AdipoR2 are seven trans-membrane proteins that can mediate fatty acid oxidation and glucose uptake via adiponectin binding. Functional adiponectin receptors are normally expressed in lung epithelial cells. Using immunohistochemical staining, Jamshid et al. demonstrated that the expression of AdipoR1 and AdipoR2 is lower in NSCLC tissue than in normal lung tissue. Here, we demonstrated an increased AdipoR1 protein expression and immunohistochemical staining in NSCLC tissues. Silencing the AdipoR1 expression can block the effect of adiponectin on the migration and invasiveness of A549 cells. Blocking the AdipoR2 expression did not influence the migratory ability of A549 cells.
Curcumin has become an effective adjuvant agent in cancer therapy as it can modulate the activity of several transcription factors and their signaling pathways [38]. It can induce cellular apoptosis through the activation of the p38 pathway in lung cancer and ovarian tumor [39]. However, the regulatory role of curcumin in adiponectin expression remains unclear. We showed that curcumin inhibits adiponectin expression through the p38 and ERK pathways. Moreover, adiponectin can induce the activation of AKT pathway, which is also involved in the activation of NF-κB pathway in lung cancer.
In conclusion, adiponectin can be a novel potential prognostic marker of NSCLC. Except for classic chemotherapy, in vivo and in vitro results corroborate that curcumin may be used as an adjuvant agent in lung cancer therapy in the future. However, the poor absorption, metabolism, and bioavailability of curcumin may limit its clinical application. Several formulations have been prepared that include nanoparticles, liposomes, micelles, and phospholipid complexes, which could increase the bioavailability and promote longer circulation, better permeability, and resistance to metabolic processes of curcumin [40].

Tumor sample collection
The NSCLC and corresponding normal tissues were collected from 77 patients who underwent surgical resection between 2004 and 2011 at the Division of Thoracic Surgery, Department of Surgery, Kaohsiung Medical University Hospital. The tissue samples were immediately placed in liquid nitrogen for shipment to the laboratory and then stored in −80°C freezers until RNA isolation and protein extraction. Complete staging procedures, including chest radiography, bronchoscopy, computed tomography of the brain and thoracic cavity, sonography, and bone scintigraphy, were performed to precisely determine the tumor characteristics according to the TNM International Staging System for lung cancer [41]. All the patients were followed up until March 2012. Details of their demographic and survival data were updated.

Ethics statement
The koahsiung medical university hospital's Institutional Review Board for Research approved the study protocol (KMUH-IRB-990357). All of the participants provided written informed consent.

RNA extraction and real-time PCR
Total RNA was isolated from frozen lung tumor tissues of NSCLC patients and corresponding normal adjacent lung tissues. RNA isolated from normal lung tissue, lung tumor tissue, and lung cancer cells was analyzed using real-time PCR, which was performed as described previously [42]. The following primers for real-time PCR were designed using the Primer Express software (RealQuant, Roche, Mannheim, Germany) using published sequences: human adiponectin sense primer: 5 0 -GGT GAT GGC AGA GAT GGC AC-3 0 and adiponectin antisense primer: 5 0 -GCC TTG TCC TTC TTG AAG AG-3 0 and human GAPDH sense primer: 5 0 -AGC CAC ATC GCT CAG ACA-3 0 and GAPDH antisense primer: 5 0 -GCC CAA TAC GAC CAA ATC C-3 0 .
Fluorescence data were acquired at the end of the extension. Melt analysis was performed for all products to determine the specificity of the amplification. In addition, the PCR products were electrophoresed on 1% agarose gels to confirm the presence of correct band sizes. The relative expression was calculated as a ratio of the expression in the lung cancer tissue compared to the expression in the normal adjacent tissue (tumor lesion/normal tissue >1, high expression, and <1, low expression) [42][43][44].

Immuno-histochemical staining
Tumor specimens were dissected from human lung tissue, fixed in 4% buffered formalin solution overnight, embedded in paraffin, and cut into 5-μm thickness sections. The paraffin sections were de-paraffinized with xylene and stained with anti-human adiponectin (GeneTex, Irvine, CA, USA), adiponectin receptor 1 (AdipoR1), and adiponectin receptor 2 (AdipoR2) (Santa Cruz, CA, USA). After washing with phosphate-buffered saline (PBS), the sections were incubated with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. For color reactions, diaminobenzidine (DAB) was used and counterstained with hematoxylin. For negative controls, the antibody was replaced with control IgG.
Silencing the expression of adiponectin, adiponectin receptor 1, and adiponectin receptor 2 in A549 cells siRNA, a specific double-stranded 21-nucleotide RNA sequence homologous to the target gene, was used to silence the expression of AdipoR1, and AdipoR2. Human adiponectin siRNA sense primer: 5 0 -GUG UGG GAU UGG AGA CUU ATT-3 0 and antisense primer: The silencing effect of siRNA transfection was assessed by real-time PCR and immunoblotting. Briefly, the cells were grown in a six well plate and transiently transfected with 20 nM siRNA using 8μL siPORT Amine (Ambion, Austin, TX, USA) for a total transfection volume of 0.5 mL. After incubation at 37°C, under 5% CO 2 for 5 h, 1.5 mL of the normal growth medium was added, and the cells were incubated for 48 h.

MTT assay for cell viability
MTT assays were performed as described previously [42].

In vitro migration/invasion analyses
The migratory ability of the cells was assayed in a monolayer denudation assay, as described previously [42]. The confluent cells were wounded by scraping with a 100-μL pipette tip, which denuded a strip of the monolayer that was 300 μm in diameter. We washed the culture twice with PBS and added 5% FBS as a medium. After 24 hours, cells will migrated into the denuded area and were photographed. The areas were analyzed using the Wimasis WimScratch software, a new generation web-based image tool for cell migration analysis. Edge detection techniques can easily recognize the leading edge and the gap area. Users can upload the images and analysis will start automatically. Cellular invasion was quantified using a modified Matrigel Boyden chamber assay. The Bio-Coat Matrigel invasion chamber (BD Biosciences, Bedford, MA) was used according to the manufacturer's instructions. Cells (4 × 10 4 ) in serum-free media were seeded onto Matrigelcoated filters. In the lower chambers, 5% FBS was added as a chemoattractant. After incubation for 24 h, the membrane was washed briefly with PBS and fixed with 4% paraformaldehyde. The upper side of the membrane was wiped gently with a cotton ball. The membrane was then stained using hematoxylin and removed. The magnitude of cell migration was evaluated by counting the migrated cells in 6 random clones under high power (×100) microscopic fields.

Co-immunoprecipitation assay
To determine the protein-protein interactions between adiponectin and p65 (NF-κB subunit), the cells were harvested using immunoprecipitation lysis buffer (GeneTex). Nuclear proteins (800 μg) were cleared by incubating with protein A/G agarose beads (30 μL/tube), under rotating conditions, for 1 h at 4°C and then washed. The supernatants were collected and incubated with 2 μg of anti-adiponectin, anti-p65, or control rabbit IgG antibodies for 4 h. Protein A/G agarose beads (50 μL/tube) were then added to each tube, followed by overnight incubation at 4°C. The supernatants were removed by centrifuging at 12,000 ×g for 10 min and disrupted by boiling in 1% SDS. The immune complexes were analyzed by immunoblotting with anti-adiponectin and anti-p65 antibodies.

Mouse model
The animal use protocol was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Kaohsiung Medical University (IACUC 99067). Eight-week-old BC-17 SCID mice were purchased from Lasco, Taiwan. The effect of curcumin in attenuating NSCLC cell (A549) growth was assessed using a subcutaneous (s.c.) xenograft tumor model. Mice were subcutaneously inoculated in the right flank with 5 × 10 7 /ml cells, with one tumor site per rat. When the tumor growth reached 100 mm 3 , the mice received intraperitoneal injection (IP) of curcumin (45 mg/kg) or vehicle (DMSO) five times each week.
After two weeks of treatment with the above-mentioned agents, the tumors were isolated and the proteins were harvested and stored at −80°C. Some tumor tissues were collected in a tube containing 4% paraformaldehyde/PBS and stored at 4°C for 12 h. After washing three times with PBS, the tissues were embedded in paraffin and cut into 5-μm sections. Hematoxylin and eosin (H&E) and immunohistochemical staining were performed to examine the morphology and expression of adiponectin and MMP proteins in tumor tissues.

Nuclear extract preparation and EMSA
Nuclear protein extracts were prepared as previously described and analyzed using EMSA [42,45].

Statistical analyses
Statistical differences in the expression ratio of adiponectin and clinical parameters were tested using the student's t-test or one-way analysis of variance (ANOVA). Survival curves were drawn using the Kaplan-Meier analysis. The results were expressed as mean ± SEM. Data were analyzed using the JMP software (SAS, JMP, Version 8.0, Cary, NC). The chi-square test was used for statistical analysis. One-way ANOVA was used to assess the differences among the groups. Statistical significance was set at p < 0.05.