The CP1 polysaccharide was extracted from adlay seeds (C. lachryma-jobi L.) by decoction and alcohol precipitation as described in a previous study (36). The human A549 non-small cell lung cancer cell line was obtained from the Cell Bank of Type Culture Collection of Chinese Academy of Sciences (Shanghai, China) and cultured in RPMI-1640 medium (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) supplemented with L-glutamine (1 mM), 10% (v/v) heat-inactivated fetal bovine serum (FBS), penicillin (100 U/ml), and streptomycin (100 µg/ml) at 37°C in 5% (v/v) CO₂ incubator. In general, all experiments were conducted when cells reached 80–90% confluence. The cells were at <20 passages, remaining normal and with healthy cell morphology, and without mycoplasma contamination throughout the experiments.

The effect of CP1 on the viability of A549 cells was assessed by MTT assay (36). Briefly, exponentially growing cells in 96-well plates were treated with different concentrations (10–300 µg/ml) of CP1 in complete RPMI-1640 medium. Control cells were cultured in medium not containing CP1. MTT (20 µl, 5 mg/ml) was added following incubation of the cells for 24 and 48 h, and subsequently the cells were incubated for 4 h. The medium was then aspirated and 150 µl dimethyl sulfoxide (DMSO) was added into each well. The absorbance was measured at 570 nm using a 96-well microplate reader. All experiments were performed three times. The cell viability was calculated as follows: Ratio of cell viability (%) = (A-B / C-B) × 100; where A is the average optical density of CP1-treated cells, B is the average optical density of the control wells (culture medium without cells), and C is the average optical density of the negative control (culture medium containing DMSO and no CP1).

The cell scratch wound healing assay was performed based on the Yarrow method (37). The cells were seeded in 24-well plates for 24 h and cell density reached ~70–80% confluence as a monolayer. Gently and slowly, a scratch was made in the monolayer across the center of the well using a 1 µl pipette tip. While scratching across the surface of the well, the long-axial of the tip was always perpendicular to the bottom of the well. The resulting gap distance is therefore equal to the outer diameter of the end of the tip, and then the cells were washed with PBS buffer three times to remove cell debris. Scratch healing rate (%) = (0 h scratch width-12 or 24 h scratch width) / 0 h scratch width × 100 was photographically recorded and cell confluence area was measured to calculate the cell migration.

A Transwell migration chamber assay was performed to observe cell invasion, particularly the motility capability of tumor cell transmigration across Matrigel in vitro. For the assay, 1×10⁵ cells in FBS-free RPMI-1640 medium were plated in the top chamber of the Transwell insert with a Matrigel-coated polycarbonate membrane. RPMI-1640 medium with 10% FBS was added to the lower chamber as a chemoattractant. After incubation for 24 h (migration assay) or 36 h (invasion assay), cells on the lower surface of the membrane were fixed with 10% formalin and stained with 0.2% crystal violet. Cells that did not migrate through the pores were mechanically removed using a cotton swab (38). The images of migrated cells were acquired using an inverted light microscope at ×200 magnification. The number of invaded cells was counted from five or six randomly selected fields in a blind manner.

Total RNA was extracted from ~2×10⁶ cells for each test using an RNeasy Mini kit (Qiagen, Inc., Valencia, CA, USA) according to the manufacturer's protocol. The integrity of the total RNA was determined electrophoresis on 1% agarose gel. Reverse transcription (RT) was performed using 1 µl Ribolock™ RNase Inhibitor, 1 µl Oligo (dT) 18 primer, 2 µl 10 mM dNTP mix, 4 µl 5X RevertAid reaction buffer, 2 µl template RNA (100 ng/µl), 1 µl RevertAid™ reverse transcriptase (Thermo Fisher Scientific, Inc.) and nuclease-free water was added to a final volume of 20 µl. Reagents were mixed, and incubated at 42°C for 1 h, then at 70°C for 5 min to terminate the reaction. The cDNA was stored at −20°C. The polymerase chain reaction (PCR) product of S100A4 was detected using the primers listed in Table I, designed with Primer 3 software version 0.3.0 (frodo.wi.mit.edu). Primers were synthesized by Invitrogen (Thermo Fisher Scientific, Inc.). PCR of S100A4 was performed using 5 µl 10X Taq reaction buffer, 2 µl template cDNA, 1.5 µl primers each (forward and reverse), 1 µl dNTP mix (10 mM), 1 µl Taq DNA polymerase (Promega Corporation, Madison, WI, USA) and nuclease-free water to a final volume of 50 µl. The reaction was performed at 94°C for 30 sec, then 35 cycles of 94°C for 30 sec, 58°C for 30 sec and 72°C for 45 sec, with final extension at 72°C for 10 min, and then maintained at 4°C. The PCR fragments were visualized on a 1.2% agarose gel stained with ethidium bromide, semi-quantitatively analyzed using an ImageQuant LAS 4000 (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA), and sequenced by a commercial sequencing service company (Beijing Genomics Institute, Beijing, China) to identify it as S100A4.

A549 cells were incubated to the logarithmic growth phase and 1×10⁶ cells were synchronized for another 10 h, then CP1 was added to a final concentration of 200 and 300 µg/ml, and incubated for 48 h. Cells were collected, washed twice with ice-cold PBS and lysed with lysis buffer [50 mM Tris, pH 7.4, 150 mM NaCl, 1 mM EDTA, 0.2 mM PMSF, 1.0% Triton X-100, protease inhibitor cocktail (Sigma-Aldrich; Thermo Fisher Scientific, Inc., Waltham, MA, USA)]. Lysates were incubated for 10 min on ice, sonicated and centrifuged for 15 min at 12,000 × g. Subsequently, protein concentrations were determined using the Bradford assay and the samples were boiled for 10 min. Equal amounts of protein (20 µg/lane) were separated by SDS-PAGE, transferred to nitrocellulose membranes and immunoblotted with a 1:1,000 dilution of a rabbit primary antibody against human S100A4 (catalog no. ab41532; Abcam, Cambridge, MA, USA) and 1:4,000 dilution of a rabbit primary antibody against human β-actin (catalog no. ab8227; Abcam) at 4°C overnight. The secondary antibody was a fluorescein-conjugated goat anti-rabbit IgG antibody (H+L; catalog no. 111-035-144; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA), diluted 1:5,000 in blocking solution for 1 h at room temperature. Immunoreactivity of PVDF membranes was visualized by scanning on LI-COR infrared laser imaging system (LI-COR, Inc., Lincoln, NE, USA). The values of the band density were normalized to β-actin using Multi-Gauge software version 2.0 (FujiFilm Corporation, Tokyo, Japan), therefore, the background was subtracted and only the non-saturated signals were quantified, resulting in a ratio which indicated the relative expression levels of the target protein for statistical analysis (n=3).

The protein structure of S100A4 used in the docking studies was obtained from the Protein Data Bank (http://www.rcsb.org/pdb/home/home.do cod3CGA) (21,22). All hydrogen atoms were added and the calcium ion, in each subunit, and an active site of a sphere was set around the following seven residues: Phe72, Tyr75, Phe78, Leu79, Met12, Val13 and Phe16, the coordinates of the sphere were 7.985, 6.865, −4.913. Iota-carrageenan (1CAR) was used as surrogate polysaccharide (37) for the simulation, it was energy minimized using ‘Powell algorithms’ method with a convergence gradient value of 0.05 kcal/(mol Å), 100 max interations were saved as mol2 format using the SYBYL-X 2.0 package (Tripos, Inc., St. Louis, MO, USA). Molecular docking was performed using GOLD 3.0.1 software (www.ccdc.cam.ac.uk/solutions/csd-discovery/components/gold/) that applied genetic algorithm. The number of generic algorithm runs was set to 10. Other parameters were the default.

The protein structure used in the docking studies was obtained from the Protein Data Bank (code 3ZWH). All hydrogen atoms were added, then the internal NMIIA peptide was removed. The polysaccharide in 1CAR was energy minimized using ‘Powell algorithms’ method with a convergence gradient value of 0.05 kcal/(mol Å), 100 max iterations, saved as mol2 format using SYBYL-X 2.0 package. Molecular docking was performed using GOLD 3.0.1 software that applied genetic algorithm, and the binding site was defined to encompass all atoms within a 10 Å sphere, whose origin (point 16.579, 1.084, 23.581) was located at the center of the residues at N-terminal of the NMIIA peptide (residues from Tyr1893 to Ala1907). The number of generic algorithm runs was set to 20. Other parameters were the default.

SPSS 19.0 software (IBM SPSS, Armonk, NY, USA) was used for statistic analysis. All experimental results were expressed as the mean ± standard deviation, and were analyzed by one-way analysis of variance with Dunnett's multiple comparison tests. P<0.05 was considered to indicate a statistically significant difference.

MTT assays demonstrated that A549 cell proliferation was significantly inhibited by CP in a time- and concentration-dependent manner. Fig. 1 presents the viability of cells treated with CP at various concentrations for 24 and 48 h. The cell viability was reduced 84.11 and 76.33% of the control following treatment with 200 and 300 µg/ml CP, respectively, for 24 h, and was 83.88 and 71.23% following treatment with 200 and 300 µg/ml CP, respectively, for 48 h.

The effects of CP1 on cell migration were determined using a cell scratch wound healing assay. The cells gradually grow back to confluence over scratch wound and were almost healed in the control group after 24 h. However, in the CP treatment groups wound healing was significantly reduced compared with the control (P<0.01), as shown my measuring the cell free area. Additionally, 300 µg/ml CP exerted a more potent effect on cell migration compared with 200 µg/ml CP following scratches. This indicated that CP1 inhibited cell mobility in a dose-dependent manner (Fig. 2).

Aggressive cancer cells can penetrate surrounding tissues or different organs by epithelial invasion, which directly and indirectly reflects clinical progression. The results of Fig. 3 demonstrated that, compared with untreated cells, as the CP1 concentration increases, the number of cells penetrating through the Matrigel was significantly reduced in a dose-dependent manner (P<0.01).

Different concentrations of CP1 were incubated with A549 cells for 24 h. The relative S100A4 mRNA level was significantly decreased compared with the control in a dose-dependent manner (P<0.01), which demonstrated that CP1 inhibited the expression of the S100A4 gene (Fig. 4). Western blot analysis was performed to detect the expression of S100A4 protein, which demonstrated that treatment with CP1 polysaccharides of 24 h inhibited the S100A4 protein expression in a dose-dependent manner (P<0.01; Fig. 5).

Each S100A4 subunit contains two calcium-binding motifan N-terminal pseudo-EF hand and a C-terminal canonical EF-hand. The polysaccharide analog did not target the dimerization site as it was demonstrated that 1CAR interacted between the two helices from each subunit of S100 homodimers, stabilized by noncovalent that form an X-type four-helix bundle dimerization motif. With full flexibility to be as close as possible to a natural conformation of S100A4 (Protein Bank Database, 3CGA), 1CAR GOLD docking occurred in the dimerization site at the turning point of the amino acids Phe72, Tyr75, Phe78 and Leu79 of the H4 helix with Met12, Val13 and Phe16 of the H1 helix of the 2nd S100A4 dimer protein (Fig. 6). The 1CAR docking pocket was more clearly demonstrated with the molecular surface (Fig. 6B) compared with the helix (Fig. 6A). The three strongest bonds between S100A4 and 1CAR were scored at 20.26, 17.12 and 9.92, respectively, which indicate a low affinity, and weak interaction with the S100A4 dimer, however affinity may be markedly effected by in the situation of the electric charge, for instance when cell senescence, over oxidation or carcinogenesis occur (37).

The original setting (point 16.579, 1.084, 23.581) was located at the center of the residues at the N-terminal of the NMIIA peptide (residues from Tyr1893 to Ala1907), the number of generic algorithm runs was set to 20. Other parameters were the default. 1CAR was targeted to S100A4 Asp-51 Arg-49 Gly-47 epitopes, revealing the potential for an intermolecular hydrogen bond by computational chemistry. The in vivo structure of S100A4 is associated with NMIIA protein. The most rigid portion of the structure is the N-terminal half of the S100A4 chains (residues 2–44) with no marked differences between subunit A and B. These residues are partially involved in subunit-subunit interactions and in forming the surface of the dimer on the diagonally opposite side to the NMIIA peptide tail binding interface (Fig. 7).