The human cervical carcinoma HeLa and oral squamous carcinoma HSC4 cell lines were provided by the Institute of Biotechnology and Genetic Engineering, Chulalongkorn University (Bangkok, Thailand) and Associate Professor R. Surarit of Mahidol University (Bangkok, Thailand), respectively. The cell cultures were maintained in Dulbecco's modified Eagle's medium (DMEM; Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; Hyclone; GE Healthcare Life Sciences, Logan, UT, USA) and 1X antibiotic-antimycotic solution (Invitrogen; Thermo Fisher Scientific, Inc.) at 37°C in an atmosphere of 5% CO₂.

Total RNA was extracted using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.). A total of 1 µg RNA was treated with DNase (Invitrogen; Thermo Fisher Scientific, Inc.) and then the half amount of DNase-treated RNA was converted into complementary DNA (cDNA) using oligo-(dT)₁₈ and a RevertAid^™ First Strand cDNA Synthesis kit (Fermentas; Thermo Fisher Scientific, Inc.). The reverse transcription reaction was carried out at 42°C for 60 min. qPCR was then performed in an ABI 7500 Real-time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.). Each reaction mixture contained 5 µl cDNA (diluted 1:5), 10 µl 2X Maxima^™ SYBR-Green/ROX qPCR Master Mix (Fermentas; Thermo Fisher Scientific, Inc.) and 0.3 µM primer pairs for maspin or GAPDH (internal control). The primer pairs were as follows: Maspin forward, 5′-CGTAGAAAACTAATCAAGCGGCTCTAG-3′; maspin reverse, 5′-CCAATTCCTTTGCATAGGGTCTC-3′; GAPDH forward, 5′-CGTTGGGTGAAGGTCGGAGTCAAG-3′; and GAPDH reverse, 5′-GGCAACAATATCCACTTTACCAGA-3′. The thermocycling conditions were as follows: 95°C for 10 min; and 45 cycles of 95°C for 15 sec and 60°C for 60 sec. Relative maspin expression levels were normalized to GAPDH and calculated using the 2-^ΔΔCq method (19). All experiments were performed 3 times with 3 replicates per experiment.

The cytokines TGF-β1, TNF-α and IL-6 (Roche Diagnostics; Indianapolis, IN, USA) were added at 0.1, 1, or 10 ng/ml to cancer cells in serum-free DMEM and incubated at 37°C in an atmosphere of 5% CO₂ for 48 h or as indicated. Smad3 inhibitor (SIS3; Merck KGaA, Billerica, MA, USA; 1, 5 or 10 µM), 100 µM pyrrolidine dithiocarbamate (PDTC; Sigma-Aldrich; Merck KGaA, Darmstadt, Germany), 1 µM wortmannin (Sigma-Aldrich; Merck KGaA), 10 µM SP600125 (Sigma-Aldrich; Merck KGaA), 10 µM U0126 (Sigma-Aldrich; Merck KGaA) or 10 µM SB202190 (Merck KGaA) were incubated with confluent HeLa cells for 1 h at 37°C in a 5% CO₂ atmosphere. A total of 10 ng/ml TGF-β1 was then added and the cells were incubated for an additional 48 h.

Subsequent to the aforementioned treatments, cells were washed twice with cold PBS, lysed with a Mammalian Protein Extraction buffer (GE Healthcare Life Sciences, Piscataway, NJ, USA) containing a protease and phosphatase inhibitor cocktail (Roche Diagnostics), and centrifuged at 14,000 × g for 10 min at 4°C. A total of 10 µg of protein, measured using a Bio-Rad™ Protein Assay (Bio-Rad Laboratories, Inc., Hercules, CA, USA), were separated by 10% SDS-PAGE, then transferred onto a nitrocellulose membrane. The blot was incubated with 5% non-fat dried milk in Tris buffer saline (TBS) with 0.01% Tween-20. Membranes were then treated with 0.5 µg/ml mouse monoclonal anti-human maspin (cat. no. 554292; BD Biosciences, San Jose, CA, USA), rabbit monoclonal anti-Smad2 (cat. no. #3122; Cell Signaling Technology, Inc., Danvers, MA, USA; dilution 1:1,000), rabbit monoclonal anti-phospho-Smad2 (cat. no. #8828; Cell Signaling Technology, Inc.; dilution, 1:2,000), mouse polyclonal anti-p38 mitogen-activated protein kinase (cat. no. #9212; Cell Signaling Technology, Inc.; dilution, 1:2,000), or mouse polyclonal anti-phospho-p38 mitogen-activated protein kinase (p38 MAPK; cat. no. #9211; Cell Signaling Technology, Inc.; dilution, 1:2,000) primary antibodies. Subsequently, the membranes were incubated with horseradish peroxidase-conjugated goat anti-mouse (cat. no. 1706516; dilution, 1:7,500; Bio-Rad Laboratories, Inc.) or goat anti-rabbit immunoglobulin G (cat. no. 1706515; dilution, 1:7500; Bio-Rad Laboratories, Inc.) secondary antibodies, and immunoreactive protein bands were visualized using Western Lightning^® Plus-ECL substrate (PerkinElmer, Inc., Waltham, MA, USA). For normalization of protein loading, the blots were then stripped and re-probed with primary rabbit polyclonal anti-human actin antibodies (cat. no. #A2066; Sigma-Aldrich; Merck KGaA; dilution, 1:1,000) as described previously. All experiments were performed 3 times with 3 replicates per experiment.

A full-length maspin promoter in a luciferase reporter plasmid (pLightSwitch; Active Motif, Carlsbad, CA, USA) was used to generate two 5′ truncated maspin promoters (−600 maspin promoter and −300 maspin promoter) in a pLightSwitch luciferase reporter plasmid by a PCR-based method (20). Point mutations in two p53-binding sites, p53 I and p53 II, in the maspin promoter region were constructed using the QuickChange II Site-Directed Mutagenesis kit (Stratagene, La Jolla, CA, USA) and a pLightSwitch plasmid containing −300 maspin promoter as the template. A pair of primers, 5′-GCCCCTTCCTGCCCtatatacaacGAGGCCTTTTGGAAGC-3′ and 5′-GCTTCCAAAAGGCCTCgttgtatataGGGCAGGAAGGGGC-3′ (mutated nucleotides illustrated in lowercase letters), were used to introduce sequence substitution at the p53 I binding site. The primers for mutation of p53 II binding site were 5′-GCCGAGAGGATTGCCGTAatataGTCTGTACGTATGCATG-3′ and 5′-CATGCATACGTACAGACtatatTACGGCAATCCTCTCGGC-3′. The correct changes to the mutated sequences were confirmed by DNA sequencing at First BASE Laboratories (Selangor, Malaysia). Each construct was transfected into the HeLa cells using TurboFect^™ in vitro Transfection Reagent (Fermentas; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. After 4 h transfection, the medium was replaced with serum free-DMEM and the cells were incubated for a further 2 h. The cells were then incubated with 10 ng/ml TGF-β1, or without TGF-β1 as control, at 37°C in a 5% CO₂ atmosphere for 48 h. Luciferase activity was determined by adding LightSwitch Luciferase Assay Reagent^™ (Active Motif) according to manufacturer's protocol, and measuring the luminescence in a Synergy^™ HT Multi-Detection microplate reader (BioTek Instruments, Inc., Winooski, VT, USA). The results were recorded as the fold of the induction of the reporter plasmid in the presence of TGF-β1, subsequent to normalization with the transfection controls without TGF-β1 treatment.

The in vitro invasion assay was performed using polycarbonate 8 µM-pore Multiscreen MIC 96-well pre-coated with 5 µg/ml Matrigel (BD Biosciences). A total of 5×104 HeLa cells/well were pre-treated for 24 h with cytokines in serum-free DMEM and then transferred to the upper wells of the filter plate. DMEM culture medium containing 10% FBS was added to the lower chambers. After 48 h incubation at 37°C in a 5% CO₂ incubator, the number of invaded cells in the bottom wells were measured using a CyQUANT^® Cell Proliferation Assay kit (Molecular Probes; Thermo Fisher Scientific, Inc.); fluorescence (excitation at 450 nm and emission at 530 nm) was measured using a Biotech K-40 spectrofluorometer (BioTek Instruments). Cells in triplicate wells without Transwell inserts served as controls for cell proliferation and/or death during the incubation period. Cell invasion was calculated from the fluorescence values as follows: Cell invasion=Relative fluorescence value of invaded cells/total cells plated in upper chamber without Transwell insert.

All experiments were performed at least 3 times with 3–5 replicates per experiment. The differences in the mean values among the groups were determined by one-way analysis of variance, and differences between individual by a Dunnett post-hoc test using SPSS v.18.0 software (SPSS Inc., Chicago, IL, USA). Data are presented as the mean ± standard deviation. P<0.05 was considered to indicate a statistically significant difference.

The effects of TGF-β1, TNF-α and IL-6 on maspin gene expression were evaluated in two human cancer cell lines: HeLa cervical carcinoma cells and HSC4 oral squamous cell carcinoma cells. TGF-β1 treatment at a concentration of 10 ng/ml induced an increase in the expression levels of maspin in HeLa and HSC4 cells (Fig. 1A). By contrast, 10 ng/ml IL-6 exhibited no effect, and 10 ng/ml TNF-α demonstrated a slight inhibitory effect on maspin levels in the two cell lines. In the HeLa cells, 0.1–10 ng/ml TGF-β1 treatment increased maspin gene expression levels at 24 and 48 h. in a dose-dependent fashion compared with control cells, and this effect was demonstrated at the mRNA (Fig. 1B) and protein levels (Fig. 1C). Similar effects were observed with HSC4 cells (data not shown).

As expected, TGF-β1-treated HeLa cells produced phospho-Smad2, and the presence of SIS3, a specific inhibitor of Smad3 (21), inhibited the induction of maspin gene expression in a dose-dependent manner (Fig. 2A). It is notable that the inhibition of TGF-β1 signaling via the Smad pathway by SIS3 did not interfere with the non-Smad pathway, as demonstrated by the presence of phospho-p38 MAPK (Fig. 2A). In order to determine whether the non-Smad signaling pathway was also activated by TGF-β1, a series of kinase inhibitors, including PDTC (a nuclear factor kB inhibitor), wortmannin (a PI3K inhibitor), SP600125 (a JNK inhibitor), U0126 (a MEK1/2 inhibitor) and SB202190 (a p38 MAPK inhibitor), were employed. U0126 and SB202190 were effective in inhibiting the TGF-β1-induced maspin gene expression, whereas PDTC, wortmannin or SP600125 were not (Fig. 2B). Inhibition by U0126 of MEK1/2, an upstream activator of p38 MAPK, resulted in an absence of p38 MAPK phosphorylation, whereas SB202190 exhibited no effect on p38 MAPK phosphorylation, as the inhibitor only binds to the ATP pocket of p38 and blocks its intrinsic ATPase activity (22). Additionally, TGF-β1-dependent Smad2 phosphorylation was not affected by the presence of these kinase inhibitors (Fig. 2B). Taken together, these data suggest that Smad and non-Smad signaling pathways are involved in TGF-β1-induced maspin gene expression in HeLa cells. Notably, the blockage of TGF-β1-induced maspin expression was achieved by inhibiting one pathway without affecting the activation of the other.

The maspin promoter region contains two p53-binding sites, p53 I and p53 II, which are located within the first upstream 300 nucleotides (nt) (Fig. 3A). Using luciferase reporter plasmid constructs, which contain point mutations in maspin promoter p53-binding sites that have been previously demonstrated to abolish p53 protein binding (18), transfection of HeLa cells and subsequent addition of TGF-β1 revealed that mutations in either the p53 I or p53 II binding sites significantly diminished the ability of TGF-β1 to induce luciferase activity compared with control cells transfected with a construct containing the full-length maspin promoter sequence (nt −872 to +193) (Fig. 3B). Therefore, the two p53-binding sites in the maspin promoter are necessary and probably sufficient in HeLa cells for TGF-β1-induced maspin gene expression, as the luciferase expression plasmid containing 5′ truncated maspin promoter element (from nt −600 or −300 to −872) expressed luciferase activity comparable with the cells containing the full length maspin promoter sequence (Fig. 3B).

The biological activities of maspin involve the inhibition of carcinoma cell migration and invasion (23). As incubation with TGF-β1 upregulated the expression of maspin in HeLa cells, whether this inhibits the invasive ability of HeLa cells was examined using a Matrigel invasion assay. Notably, 10 ng/ml TGF-β1 significantly increased the invasiveness of HeLa cells compared with the untreated control (Fig. 4).