This article is mentioned in:

Introduction

Cigarette smoking is strongly associated with lung cancer incidence and mortality (1,2). As an important toxic component of cigarette smoke (3), benzo(a)pyrene (BaP) alone is sufficient to induce lung cancer (4, 5). However, BaP promotes lung cancer through different mechanisms (6–10), and one of these mechanisms is the induction of epithelial-mesenchymal transition (EMT) (10).

EMT is a process by which epithelial cells gain migratory and invasive characteristics to become motile mesenchymal stem cells. In physiological conditions, EMT is integral in embryo implantation, organ development and wound healing (11,12). However, under pathological conditions, EMT contributes to fibrosis and cancer progression (11,12). It is an important process in the development and progression of lung cancer induced by tobacco smoke (13), and it is an important cause of drug resistance and metastasis in lung cancer (14). In addition, the EMT secretory phenotype can also predict survival in patients with lung cancer (15).

Twist family BHLH transcription factor 1 (Twist1) is a basic helix-loop-helix domain-containing transcription factor that is encoded by the Twist1 gene (16), and is essential for embryonic differentiation during physiological conditions (16). Upregulated Twist1 expression is associated with the progression of lung cancer (17) and other types of cancer, such as gastric (18), breast (19), colorectal (20), endometrial (21) and prostate (22) cancer. In addition, Twist1 can induce EMT in cancer cells (23,24). Particularly, inhibiting the expression of Twist1 in lung cancer cells suppresses cell proliferation (17,25) and metastasis (25), activates oncogene-induced senescence (26) and suppresses the EMT process (25).

In the present study, we aimed to study the effects of BaP on migration ability, the EMT process and Twist1 expression in adenocarcinoma A549 cells. We found that BaP can enhance the migration of lung adenocarcinoma A549 cells through the promotion of EMT by upregulation of Twist1.

Materials and methods

Cell culture and treatments

Human A549 cells (Xiangya Cells Center, Central South University, Changsha, Hunan, China) were cultured in RPMI-1640 medium (HyClone, Logan, UT, USA) containing 10% fetal bovine serum (FBS) (Biological Industries, Beit Haemek, Israel) and 1% penicillin/streptomycin and incubated in a humidified incubator at 37°C with 5% CO2.

BaP (purity ≥96%; Sigma-Aldrich, St. Louis, MO, USA) was dissolved in dimethyl sulfoxide (DMSO) and added to the culture medium at a final concentration of 1 µM. Before proceeding with further experiments, A549 cells were treated with 1 µM BaP for 24, 48 and 72 h; 1, 2 and 4 weeks.

Wound healing assay

Wound healing assays were conducted to evaluate the lateral migration capacity of cells. After cells were seeded into 24-well plates and had grown into a monolayer, a ‘scratch’ was scraped in a straight line using a p200 pipet tip. The cells were washed 3 times with growth medium to smooth the edges of the scratch and remove debris, and the medium was replaced with fresh culture medium. The cells were then allowed to recover for 12 h. Before and after the 12 h incubation, the initial scratch and the recovery area, respectively, were photographed under a microscope. The wound closure rate was equal to the recovered distance divided by the original width of the scratch.

Transwell migration assay

Transwell plates (6.5 mm in diameter at the lower surface and 8-µm pore filters; Corning Costar, Cambridge, MA, USA) without Matrigel were used to detect the longitudinal migration ability of cells. Briefly, 5×104 cells in 200 µl of 0.1% FBS culture medium were seeded in the upper chamber and 800 µl of medium with 10% FBS was placed in the lower chamber. Following incubation for 24 h, a cotton swab was used to remove the cells adhering to the membrane of the upper chamber. The migrated cells on the lower side of the filter membrane surface were stained with 0.1% crystal violet, and the cells were counted under a light microscope.

Reverse transcription PCR (RT-PCR)

Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA, USA), and cDNA was synthesized using a PrimeScript First Strand cDNA Synthesis kit (Takara Biotech, Kusatsu, Shiga, Japan) according to the manufacturer's instructions. The following are the primers used for PCR amplification: Twist1 forward, 5′-CTCAGCTACGCCTTCTCGGT-3′ and reverse, 5′-AGTCCATAGTGATGCCTTTC-3′; N-cadherin forward, 5′-ATCCAGACCGACCCAAACAG-3 and reverse, 5′-GGAATTCCATTGTCAGAAGC-3; vimentin forward, 5′-CCAGGCAAAGCAGGAGTC-3 and reverse, 5′-CGAAGGTGACGAGCCATT-3; E-cadherin forward, 5′-CTGAGAACGAGGCTAACGTCG-3 and reverse, 5′-AGTGTAGGATGTGATTTCCTG-3; β-actin forward, 5′-AGAGCTACGAGCTGCCTGAC-3 and reverse, 5′-AGCACTGTGTTGGCGTACAG-3.

Western blotting

Whole cell protein was prepared using ice-cold RIPA buffer (Millipore, Temecula, CA, USA) mixed with a protease inhibitor cocktail. The protein concentrations were determined after cell lysates were centrifuged at 12,000 × g at 4°C for 20 min to clear the cell debris. Equal amounts of protein (50 µg) from each sample were loaded/lane and separated on 10% SDS-PAGE, followed by transfer to polyvinylidene difluoride (PVDF) membranes (Millipore) via electroblotting. Then, 5% non-fat milk was used to block the membranes for 2 h. Primary antibodies, diluted according to the manufacturer's recommendations, were incubated with the membranes overnight at 4°C. The antibodies used included antibodies against Twist1 (Santa Cruz, Heidelberg, Germany), N-cadherin (BD Biosciences, San Jose, CA, USA), vimentin (Santa Cruz) and E-cadherin (BD Biosciences), and β-actin (Cell Signaling Technology, Beverly, MA, USA) served as the loading control. The membranes were washed 3 times (10 min/time) in Tris-buffered saline with Tween-20 (TBST) buffer, followed by incubation with horseradish peroxidase-conjugated secondary antibodies for 2 h at room temperature. The protein bands were detected using an enhanced chemiluminescence (ECL) reagent kit (Thermo Scientific Pierce, Rockford, IL, USA).

Construction of plasmid-expressing short hairpin RNA targeting Twist1

A short hairpin RNA sequence (GCTGAGC AAGATTCAGACCCTTTCAAGAGAAGGGTCTGAATCT TGCTCAGC) targeting Twist1 (Twist1-shRNA) was inserted into a pGCsilencer™ U6/Neo/GFP/RNAi plasmid (GeneChem, Shanghai, China). A negative control shRNA (TTCTCCGAAC GTGTCACGT) vector was also constructed using a pGCsilencer™ U6/Neo/GFP/RNAi plasmid to verify the sequence specificity of Twist1-shRNA. The cells were transfected with plasmids using Lipofectamine™ 2000 reagent (Invitrogen). The transfection efficiency was observed under a fluorescence microscope after transfection for 24 h. The knockdown efficiency of Twist1-shRNA was detected with western blotting after transfection for 48 h.

Statistical analysis

Statistical analyses were conducted using SPSS 18.0 software (SPSS, Inc., Chicago, IL, USA) with a Windows operating system. Repeated measurement data were presented as the mean ± standard deviation (SD). When two sets of data were compared, differences were calculated using Student's t-test (two-tailed). Statistical significance was defined as p<0.05.

Results

BaP promotes the migration of A549 cells

To study the effect of BaP on the migration of A549 cells, we first treated A549 cells with 1 µM BaP for different lengths of time (24, 48 and 72 h; 1, 2 and 4 weeks). Thereafter, we used a wound healing assay to study the lateral migration ability of BaP-treated A549 cells. As shown in Fig. 1, after the cells in the wound healing assay were allowed to recover for 12 h, BaP enhanced the lateral migration ability of A549 cells when the treatment time was at least 48 h. Furthermore, we used Transwell migration assays to research the longitudinal migration ability of BaP-treated A549 cells. As shown in Fig. 2, after the Transwell migration assay was performed for 24 h, BaP enhanced the longitudinal migration ability of A549 cells when the intervention time was at least 48 h, similar to the enhancement effect observed on lateral migration. In addition, we found that the effect of BaP on the migration of A549 cells was gradually enhanced with prolonged treatment time (Figs. 1 and 2).

Figure 1. BaP (1 µM) stimulates A549 cell migration, shown by a wound healing assay. After intervention with BaP for different lengths of time (24, 48 and 72 h; 1, 2 and 4 weeks), wound healing assays were performed with a 12-h recovery time. Graphics present the wound closure rate of cells among different treatment groups, and the results are presented as the mean ± SD of 3 independent experiments; *p<0.05, **p<0.01. BaP, benzo(a)pyrene.

Figure 2. BaP (1 µM) stimulates A549 cell migration, as revealed by a Transwell migration assay. After intervention with BaP for different lengths of time (24, 48 and 72 h; 1, 2 and 4 weeks), Transwell migration assays were performed for 24 h (the microscopy fields are shown at a magnification of ×200). Graphics display the number of cells/field between different treatment groups, and the results are presented as the mean ± SD of 3 independent experiments; *p<0.05, **p<0.01. BaP, benzo(a)pyrene.

Effect of BaP treatment on the expression of Twist1, N-cadherin, vimentin and E-cadherin in A549 cells

As shown in Figs. 3 and 4, after A549 cells were exposed to 1 µM BaP for different durations of time (24, 48 and 72 h; 1, 2 or 4 weeks), both the mRNA (Fig. 3) and protein expression (Fig. 4) of Twist1 were gradually increased with prolonged intervention times. Accordingly, the expression of N-cadherin and vimentin were also gradually increased with prolonged BaP intervention (Figs. 3 and 4). However, conversely, the expression of E-cadherin gradually decreased as the BaP intervention time was extended (Figs. 3 and 4).

Figure 3. Effect of BaP on the mRNA expression of Twist1, N-cadherin, vimentin and E-cadherin in A549 cells. (A) After A549 cells were exposed to 1 µM BaP for 24, 48 and 72 h; 1, 2 or 4 weeks, the mRNA levels were determined using RT-PCR. (B) The statistical results are presented as the mean ± SD of 3 independent experiments; *p<0.05, **p<0.01. BaP, benzo(a)pyrene; Twist1, Twist family BHLH transcription factor 1.

Figure 4. Effect of BaP on the protein expression of Twist1, N-cadherin, vimentin and E-cadherin in A549 cells. (A) After A549 cells were exposed to 1 µM BaP for 24, 48 and 72 h; 1, 2 or 4 weeks, the protein levels were determined with western blotting. (B) The statistical results are presented as the mean ± SD of 3 independent experiments; *p<0.05, **p<0.01. BaP, benzo(a)pyrene.

Downregulation of Twist1 inhibits the migration ability of A549BaP-4w cells

Compared with A549 cells without BaP intervention, Twist1 was highly expressed in A549 cells treated with BaP for 4 weeks (A549BaP-4w) (Figs. 3 and 4), and the migration capacity of A549BaP-4w cells was significantly enhanced (Figs. 1 and 2). We hypothesized that downregulation of Twist1 expression may decrease the migration of A549BaP-4w cells. Thus, we applied wound healing and Transwell migration assays to assess A549BaP-4w cells that were transfected with recombinant plasmid containing short hairpin RNA (Twist1-shRNA). The results of both the wound healing (Fig. 5A) and Transwell migration assay (Fig. 5B) revealed that downregulation of Twist1 inhibited the migration ability of A549BaP-4w cells.

Figure 5. Downregulation of Twist1 in A549BaP-4w cells inhibits the migration ability. After downregulation of Twist1 using Twist1-shRNA, (A) a wound healing assay was performed with 12 h of recovery and (B) a Transwell migration assay was performed at 24 h (the microscopic fields are shown at a magnification of ×200). The graphics display the (A) wound closure rate and (B) the number of cells/field between the different groups. The statistical results are presented as the mean ± SD of 3 independent experiments; *p<0.05, **p<0.01. A549BaP-4w, A549 cell model incubated with 1 µM BaP for 4 weeks, resulting in the upregulation of the expression of Twist1; control, A549BaP-4w cell group without plasmid transfection; blank control, A549BaP-4w cell group transfected with blank plasmid; negative control, A549BaP-4w cell group transfected with negative plasmid; Twist1-shRNA, A549BaP-4w cell group transfected with recombinant plasmid containing short hairpin RNA to downregulate Twist1 gene expression. BaP, benzo(a)pyrene; Twist1, Twist family BHLH transcription factor 1.

Decreased expression of Twist1 in A549BaP-4w cells results in downregulation of N-cadherin and vimentin and upregulation of E-cadherin

As shown in Fig. 6, the expression of Twist1 in A549BaP-4w cells was obviously downregulated after transfection with Twist1-shRNA. Accordingly, the expression of N-cadherin and vimentin was also markedly decreased after A549BaP-4w cells were transfected with Twist1-shRNA (Fig. 6A and B). However, decreased expression of Twist1 in A549BaP-4w cells resulted in the upregulation of E-cadherin expression (Fig. 6A and B).

Figure 6. Decreased expression of Twist1 in A549BaP-4w cells results in the downregulation of N-cadherin and vimentin and upregulation of E-cadherin. (A) The mRNA levels and (B) protein expression of Twist1, N-cadherin, vimentin and E-cadherin were determined with RT-PCR and western blotting, respectively. The results are grouped by A549BaP-4w cells with different transfection statuses (control, A549BaP-4w cell group without plasmid transfection; blank control, A549BaP-4w cell group transfected with blank plasmid; negative control, A549BaP-4w cell group transfected with negative plasmid; Twist1-shRNA, A549BaP-4w cell group transfected with recombinant plasmid containing short hairpin RNA to downregulate Twist1 gene expression). BaP, benzo(a)pyrene; Twist1, Twist family BHLH transcription factor 1.

BaP induces EMT in A549 cells by upregulating Twist1

Under a microscope, epithelial cells are characterized by a flat and polygonal shape, and mesenchymal cells are characterized by a relatively small cell body that is long and thin. As shown in Fig. 7A, after 4 weeks of 1 µM BaP intervention, most of the of A549BaP-4w cells transformed from cells with epithelial characteristics to cells with mesenchymal characteristics. However, after transfection of A549BaP-4w cells with Twist1-shRNA (Fig. 7B) to downregulate the expression of Twist1 (Fig. 7C), the A549BaP-4w cells were transformed from cells with mesenchymal characteristics back into cells with epithelial characteristics (Fig. 7D). Consequently, we concluded that BaP induced EMT in A549 cells by upregulating Twist1.

Figure 7. (A) Morphological observation reveals that most of the A549BaP-4w cells were transformed from cells with epithelial characteristics to cells with mesenchymal characteristics after 4 weeks of 1 µM BaP intervention. (B) After transfection with recombinant plasmid, (C) containing short hairpin RNA to downregulate Twist1 gene expression, (D) A549BaP-4w cells were transformed into cells with epithelial characteristics from cells with mesenchymal characteristics. Blank control, A549BaP-4w cell group transfected with blank plasmid; negative control, A549BaP-4w cell group transfected with negative plasmid; Twist1-shRNA, A549BaP-4w cell group transfected with recombinant plasmid containing short hairpin RNA to downregulate Twist1 gene expression. BaP, benzo(a)pyrene; Twist1, Twist family BHLH transcription factor 1.

Discussion

In the present study, we first found that BaP promotes the migration of lung adenocarcinoma A549 cells in a time-dependent manner. Increased expression of Twist1 in A549 cells resulted in upregulation of N-cadherin and vimentin and downregulation of E-cadherin. Then, after the expression of Twist1 was knocked down in A549 cells that were treated with BaP for 4 weeks (A549BaP-4w), the resulting decrease in the expression of Twist1 inhibited the migration capacity of A549BaP-4w cells, downregulated the expression of N-cadherin and vimentin and upregulated the expression of E-cadherin. Finally, along with the morphological results observed under the microscope, we concluded that BaP enhanced the migration of lung adenocarcinoma A549 cells through the promotion of EMT by upregulation of Twist1.

Prior to the present study, Yoshino et al (10) reported that BaP can induce the EMT of lung cancer cells. In addition, D'Angelo et al (23) and Yoon et al (24) found that Twist1 can induce the EMT of cancer cells. We therefore hypothesized that Twist1 may be the target of BaP-treated lung adenocarcinoma A549 cells, and we thus demonstrated this hypothesis in the present study. However, as the mechanism of cigarette smoking-induced lung cancer is very complex, we surmised that Twist1 cannot be the only target of BaP in lung cancer cells. More studies on the mechanism of cigarette smoking/BaP-induced lung cancer need to be conducted in the future.

Yoshino et al (10) exposed A549 cells to 1 µM BaP for a long period of time (24 weeks) and used a gene chip to conduct a microarray analysis. They reported that the mRNA expression of Twist was upregulated and the mRNA expression of E-cadherin was downregulated. However, they did not observe morphological changes under these experimental conditions (10). In the present study, by treating A549 cells with 1 µM BaP for different lengths of time (24, 48 and 72 h; 1, 2 and 4 weeks), we found that the expression of Twist1, N-cadherin and vimentin were increased at both the mRNA and protein levels, while the expression of E-cadherin was decreased. Furthermore, we discovered through observation under a microscope that most of the A549 cells were transformed from cells with epithelial characteristics to cells with mesenchymal characteristics when treated with BaP for 4 weeks. Through a comparative analysis, we hypothesized that a relatively short period of BaP intervention can promote EMT in A549 cells, whereas a long duration of intervention may reverse this effect. However, the related mechanisms of this phenomenon remain to be clarified.

Wang et al (27) incubated A549 cells with 8 µM BaP for 24 h and reported that BaP promoted A549 cell migration. They found that the mRNA and protein expression of Twist were upregulated in A549 cells with treatment with 8 µM BaP for 24, 48 and 72 h (27). In addition, when the expression of Twist was knocked down in A549 cells, the migration capacity was blocked by intervention with 8 µM BaP for 24 h (27). In the present study, after treating A549 cells of 1 µM BaP for at least 48 h, we obtained results similar to those of Wang et al, specifically, migration promotion and Twist1 upregulation. Similarly, we also revealed that downregulation of Twist1 can inhibit the migration capacity of A549BaP-4w cells, indicating that Twist1 indeed plays an important role in the progression of smoke-induced lung cancer.

Generally, N-cadherin and vimentin are considered to be mesenchymal markers (28–30) while E-cadherin is regarded as an epithelial marker (28,29,31). In the present study, notably, we found that the expression trend of Twist1 was consistent with that of N-cadherin and vimentin but contrary to that of E-cadherin. Furthermore, we also noted a similar situation in many previous cancer studies (32–35). Therefore, we speculate that Twist1 is a new potential mesenchymal biomarker in the process of cancer progression. However, future studies are warranted to confirm this speculation.

Previous studies have shown that Twist1 is an oncogene (36–38). It promotes proliferation (17,39–41) and inhibits apoptosis (42) of cancer cells. Furthermore, emerging evidence suggests that Twist1 significantly enhances EMT-associated cell migration and invasion to promote cancer metastasis (16). It also plays a role in chemotherapeutic resistance (16,43). Therefore, Twist1 is considered to be a potential therapeutic target for cancer (16,44,45). Since lung cancer is reported to be the leading cause of cancer-related deaths (46), and cigarette smoking is the main factor leading to lung cancer (47), the development of therapeutic strategies targeting Twist1 is particularly important in the treatment of lung cancer.

In conclusion, the results from the present study indicate that BaP enhances EMT-associated migration of lung adenocarcinoma A549 cells by upregulating Twist1.

Acknowledgements

The present study was supported by the National Natural Science Foundation of China (grant no. 81572284) and the Important Research and Development Plan of Hunan Provincial Science and Technology Department (grant no. 2015SK20662).

References