(3,4,5-dimethylathiazol-2yl)-5-diphenyl-tetrazolium bromide (MTT), chloroquine (CQ), cycloheximide (CHX) and MG132 were purchased from Millipore Sigma. All other reagents were of the purest grade available.

HCT-116 (CCL-247) cells were purchased from American Type Culture Collection. Caco-2 cells were obtained from Korean Cell Line Bank. HCT-116 and Caco-2 cells were cultured in RPMI-1640 and Minimal Essential Medium (MEM) (Cytiva), respectively, supplemented with 10% fetal bovine serum (Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin (Cytiva).

Transfection was performed as previously described (16). The cells were cultured to ~80% confluency in a 35 mm dish and were transiently transfected using Lipofectamine^® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. Briefly, 500 µl of Opti-MEM I (Gibco; Thermo Fisher Scientific, Inc.), Lipofectamine (5 µl), DNA (2 µg), scrambled short interfering (si)RNA (1 µl) and siRNA (10 µl) were used for transfection. The mixtures were sequentially incubated at room temperature (RT) for 15 min, added to the cultured cells and then incubated at 37°C for 24 h. The Rab25 cDNA was subcloned into a pcDNA3 vector and an empty pcDNA3 vector was used as a negative control. RAS-V12 constructs were kindly provided by Dr A.R Moon (Duksung University, Seoul). Claudin-7 constructs were purchased from OriGene Technologies, Inc. siRNAs against Rab25 (SASI_Hs01_00216284), claudin-7 #1 (SASI_Hs01_00214821), claudin-7 #2 (SASI_Hs01_00214822) were purchased from Millipore Sigma. The sequence of siRNAs was as follows: Rab25, 5′-GAGCCAUCACCUCGGCGUA-3′ (sense) and 5′-UACGCCGAGGUGAUGGCUC-3′ (antisense); claudin-7 #1, 5′-CUAUGCGGGUGACAACAUC-3′ (sense) and 5′-GAUGUUGUCACCCGCAUAG-3′ (antisense); claudin-7 #2, 5′-CUGGUAUGGCCAUCAGAUU-3′ (sense) and 5′-AAUCUGAUGGCCAUACCAG-3′ (antisense). Scrambled siRNA as a negative control was obtained from Invitrogen (cat. no. 12935112; Invitrogen; Thermo Fisher Scientific, Inc.).

Cell viability was performed using the MTT assay as previously described (19,20). The cells (3×10⁴ cells/well) were cultured to ~80% confluency in a 24-well plate and were transfected for 24 h. Following two PBS washes, 0.5 mg/ml of MTT solution was added to the wells and incubated at 37°C for 2 h. Then, 1 ml of dimethyl sulfoxide (DMSO) was added to each well. Immediately after the purple formazan crystals dissolved, the solution was collected and pipetted into a 96-well plate. Optical density was measured using an ELISA plate reader (BioTek Instruments, Inc.) at 540 nm.

RT-qPCR was performed as previously described (21). The cells (1×10⁵ cells/well) were cultured to ~80% confluency in a 35 mm dish and were transfected for 24 h. Total RNA was extracted from the cells using TRIzol^® (Thermo Fisher Scientific, Inc.) and isolated by adding chloroform according to the manufacturer's protocols. Then, RNA was precipitated by mixing with isopropanol. The RNA pellet was solubilized in RNase-free water. The total RNA was reverse transcribed using dNTPs (Thermo Fisher Scientific, Inc.), oligo (dT) and M-MLV reverse transcriptase (Promega Corporation). The cDNA complex was synthesized using a TaKaRa PCR Thermal Cycler Dice^® (Takara Bio, Inc.) according to the manufacturer's protocols. The cDNA complex was amplified using an iQ5 Real-Time PCR Detection System (Bio-Rad Laboratories, Inc.) according to the iQ SYBR Green Supermix (Bio-Rad Laboratories, Inc.) protocols with the following primer sets: KRAS, 5′-TGTTCACAAAGGTTTTGTCTCC-3′ (forward) and 5′-CCTTATAATAGTTTCCATTGCCTTG-3′ (reverse); CLDN1, 5′-TTTACTCCTATGCCGGCGAC-3′ (forward) and 5′-GAGGATGCCAACCACCATCA-3′ (reverse); CLDN7, 5′-AGTTAGGAGCCTTGATGCCG-3′ (forward) and 5′-GCACAGGGAGTAGGATACGC-3′ (reverse); RAB25, 5′-CCATCACCTCGGCGTACTAT-3′ (forward) and 5′-TTTGTTACCCACGAGCATGA-3′ (reverse); and β-actin, 5′-AGAGCTACGAGCTGCCTGAC-3′ (forward) and 5′-AGCATCGTGTTGGCGTACAG-3′ (reverse). The mixture was initially denatured at 95°C for 3 min and then performed 40 cycles of denaturation at 95°C for 10 sec, annealing at 60°C for 10 sec, and extension at 72°C for 30 sec. The β-actin gene was used as a control for calculating the ΔCq value. The RT-qPCR data were analyzed using the 2^−ΔΔCq method (22). The results are from the experiment in triplicate of three independent experiments.

Immunoblotting was performed as previously described (23). The cells were lysed using RIPA buffer (Millipore Sigma) containing a protease inhibitor cocktail (Roche Applied Science). Protein concentrations were measured using a bicinchoninic acid protein assay kit (Thermo Fisher Scientific, Inc.). Samples (30 µg) were separated by 8% [E-cadherin, phosphorylated (p-)EGFR, EGFR] or 12% (Rab25, Snail, Slug, Twist, Ras, β-actin, claudin-1, claudin-7, K-RAS) sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto PVDF membranes. After blocking with EzBlockChemi (ATTO Corporation) at RT for 30 min, the membrane was treated with primary antibody at 4°C overnight. The E-cadherin antibody (cat. no. 610182) was purchased from BD Biosciences. Antibodies for Rab25 (cat. no. 4314), Snail (cat. no. 3879), Slug (cat. no. 9585), Twist (cat. no. 46702), Ras (cat. no. 3965), phosphorylated (p-)EGFR (cat. no. v3777) and EGFR (cat. no. 2085) were obtained from Cell Signaling Technology, Inc. Antibodies for β-actin (cat. no. 47778), claudin-1 (cat. no. sc-166338) and claudin-7 (cat. no. sc-17670) were purchased from Santa Cruz Biotechnology, Inc. The K-Ras antibody (cat. no. ab180772) was purchased form Abcam. All primary antibodies were used at 1:1,000 dilution. Then, the membrane was incubated with the secondary antibody at RT for 2 h. Anti-rabbit (cat. no. 31463) and anti-mouse (cat. no. 31437) secondary antibodies were purchased from Thermo Fisher Scientific, Inc. (1:3,000), while an anti-goat (cat. no. sc-2354) secondary antibody was purchased from Santa Cruz Biotechnology, Inc. (1:3,000). The band was visualized using ECL reagents (cat. no. RPN2232; Amersham; Cytiva), and band densitometry was measured using ImageJ 1.53 version (National Institutes of Health).

Immunofluorescence was performed as previously described (24). The cells were fixed with cold methanol for 10 min and blocked with 1% bovine serum albumin (BSA; Rocky Mountain Biologicals, Inc.) solution. Briefly, antibodies of E-cadherin (cat. no. 610182; 1:500; BD Biosciences), Snail (cat. no. sc-271977; 1:500; Santa Cruz Biotechnology, Inc.) and claudin-7 (cat. no. ab207300; 1:500; Abcam) were used. The cells were reacted with Cy2-conjugated goat anti-mouse IgG (cat. no. 111-223-003; 1:500; Jackson ImmunoResearch Laboratories, Inc.) and Cy3-conjugated goat anti-rabbit IgG (cat. no. 111-156-003; 1:500; Jackson ImmunoResearch Laboratories, Inc.). The cell nuclei were stained with 4′,6′-diamidino-2-phenylindole dihydrochloride (cat. no. v62249; 1:1,000; Thermo Fisher Scientific, Inc.) at RT for 5 min. Fluorescence images were captured using confocal microscopy (LSM710 Carl Zeiss AG).

Wound healing assay was performed as previously reported (25). The cells were cultured to ~80% confluency in a 35 mm dish and were transfected at 37°C for 24 h. The cells were scraped using a 200 µl micropipette tip and placed in new serum-free medium. Images were captured immediately after scraping and after 24 and 36 h in the same locations. The wound closure rate was calculated by measuring the area of the wound at each time using ImageJ 1.53 version (National Institutes of Health), and the mean value of triple replicated experiments was shown.

The in vitro migration assay was performed in triplicates using a 48-well microchemotaxis chamber as previously described (21). Briefly, trypsinized cells were resuspended at a density of 2×10⁶ cells/ml in RPMI-1640. FBS medium (1%) was added to each well of the lower chamber. Type 1 collagen (Cell matrix Type I-P; Nitta Gelatin Inc.)-coated 8 µm (for HCT-116) or 12 µm (for Caco-2) pore polyvinyl pyrrolidine-free polycarbonate filters (Neuro Probe Inc.) were added for in vitro migration analysis. After incubation for 6 h at 37°C, invaded cells were fixed and stained with Diff-Quik reagents (Dade Behring Inc.). The average numbers of three random fields under the light microscope (magnification, ×200) of invasion filters were counted in each experiment.

The in vitro invasion assay was performed in triplicates using a 48-well microchemotaxis chamber as previously described (26). First, 1% FBS medium was added to each well of the lower chamber. Matrigel (Corning, Inc.), which contains extracellular matrix components, was used to laminate 8 µm (for HCT-116) or 12 µm (for Caco-2) pore polyvinyl pyrrolidine-free polycarbonate filters for in vitro invasion assay. Filters were laminated at RT for 1 h with Matrigel. The transfected cell suspension, 2×10⁶ cells/ml, was added to each well of the upper chamber. Following incubation for 16–18 h at 37°C, invaded cells were fixed and stained with Diff-Quik reagents (Dade Behring Inc.). The average numbers of three random fields under the light microscope (magnification, ×200) of invasion filters were counted in each experiment.

The 3D Matrigel invasion assay was performed as previously described (27–30). Cancer cells were labeled with DiI (Thermo Fisher Scientific, Inc.). The mixture was prepared by mixing of 20% type I collagen and Matrigel and solidify in the 3 µm pore size Transwell inserts (Corning, Inc.). A total of 5×10⁴ cells were mixed in 200 µl of medium supplemented with 0.2% FBS and plated on the gels. The 24-well plate was filled with culture medium or serum-free medium. After 3–5 days, the embedded gel was sectioned without fixation and the invaded cells were analyzed in five different fields using fluorescence confocal microscopy (magnification, ×100; LSM710; Carl Zeiss AG). In these images, the distance of invaded cells was measured from five different positions and calculated using the ZEN blue edition program 1.1.2.0 version (Carl Zeiss AG). The distance in µm was calculated as previously described (31).

The 3D cultures were observed as previously described (23). A total of 2×10⁴ cells were suspended in a 400 µl medium supplemented with 2% Matrigel and seeded over a layer of 100% Matrigel in an 8-well culture slide (cat. no. 345108, Corning, Inc.). Cells were grown for 5 days and the medium was changed every 2 days. Colony formation was monitored every day and examined using a light microscope (magnification, ×100).

The cells were cultured to ~80% confluency in a 35 mm dish and were transfected for 24 h. The supernatants were removed and RAS activation was determined using a RAS activation ELISA kit (cat. no. 17-497; Millipore Sigma) according to the manufacturer's instructions. The level of activated RAS was compared with that of vector transfectant as a control. The results represent triplicated experiments.

The cell lysates were prepared according to the manufacturer's recommendation (cat. no. BK008; Cytoskeleton, Inc.). The cells were cultured to ~80% confluency in a 60 mm dish and were transfected with the vectors at 37°C for 24 h. The cells were dissolved with the lysis buffer (cat. no. BK008; Cytoskeleton, Inc.) and centrifuged at 10,000 × g, 4°C for 1 min. The lysates (100 µg) were incubated with 30 µl of Raf-RBD beads at 4°C for 1 h and centrifuged at 5,000 × g, 4°C for 3 min. Then, the pellet was resuspended using 2X sample buffer (cat. no. BK008; Cytoskeleton, Inc.) and boiled at 95°C for 2 min. Samples were resolved using SDS-PAGE. Following resolution, procedures were the same as for immunoblotting. Antibodies for Ras (cat. no. 3965; Cell Signaling Technology, Inc.) K-Ras (cat. no. sc-30; Santa Cruz Biotechnology, Inc.), GST (cat. no. sc-138; Santa Cruz Biotechnology, Inc.) were used at 1:1,000 dilution.

Data were shown as the means ± standard deviation. Statistical analysis was assessed using the Student's t-test on the SigmaPlot software (Systat Software). Differences among three or more groups were estimated by one-way analysis of variance followed by Bonferroni multiple comparison tests. P<0.05 was considered to indicate a statistically significant difference.

In order to determine the role of Rab25 in colon cancer cell invasion, the cells were transfected with a control vector or Rab25 in colon cancer Caco-2 and HCT-116 cells. MTT analysis showed little difference in the viability between vector and Rab25 transfectants (Fig. S1). In addition, ectopic expression of Rab25 (Fig. S2) significantly attenuated the invasiveness (Fig. 1A) and the wound closure rates (Figs. 1B and S3) of these colon cancer cells. Rab25 markedly reduced HCT-116 cancer cell invasion (Fig. 1C). Furthermore, it was observed that Rab25 dramatically reduces the colony size of colon cancer cells on the 3D Matrigel system (Fig. 1D). To confirm the Rab25-induced inhibition of colon cancer cell invasion, the cells were transfected with Rab25 siRNA and it was noted that silencing of Rab25 (Fig. S4) significantly induced colon cancer cell invasion (Fig. 1E). Therefore, these data clearly indicated that Ra25 suppresses colon cancer cell invasion and migration.

Given that an EMT transcription factor, Snail, is closely associated with Rab25-induced cancer cell invasiveness of various types of cancer (17), it was determined whether Rab25 regulated the expression of EMT factors. Immunoblotting data showed that Rab25 upregulated E-cadherin expression, while Snail expression was reduced by Rab25 (Fig. 2A). Consistently, immunofluorescence results demonstrated the upregulation of E-cadherin and downregulation of Snail expression by Rab25 in colon cancer cells (Fig. 2B). In contrast, silencing of Rab25 expression reduced and increased E-cadherin and Snail expression, respectively (Fig. S5). Furthermore, ectopic expression of Snail recovered colon cancer cell invasion repressed by Rab25 (Fig. 2C). Therefore, these data implied that Rab25 inhibited colon cancer cell EMT through the downregulation of Snail expression, leading to suppression of colon cancer cell invasion.

Previous data showed that EGFR is a critical upstream governor of Snail in Rab25-induced cancer cell invasion (17). In addition, EGFR has been closely associated with Rab25-induced cancer progression (17,32). Therefore, the involvement of EGFR in Rab25-induced suppression of colon cancer cell invasion was analyzed. Intriguingly, Rab25 profoundly reduced the levels of p-EGFR expression (Fig. 3A and B), suggesting that Rab25 inactivated EGFR and subsequent Snail expression for colon cancer cell invasion. Since one of downstream EGFR effectors is Ras, whether Rab25-induced suppression of EGFR activity influences Ras activity was determined. It was observed that Rab25 reduced Ras activity in colon cancer cells (Fig. 3C and D). In addition, ectopic expression of constitutively active KRAS (KRAS-V12) rescued Rab25-induced HCT-116 cell invasion (Fig. 3E), thus consolidating the involvement of Ras in Rab25-induced suppression of colon cancer cell invasion. Overall, these data indicated that Rab25 inhibits the EGFR/Ras/Snail signaling cascade to attenuate colon cancer cell invasion.

As claudin-7 induces Rab25 expression and suppresses colon cancer cell growth in a Rab25-dependent manner (16), the present study investigated whether Rab25 induces claudin-7 expression and consequently suppresses colon cancer cell invasion. Notably, Rab25 markedly increased claudin-7 expression (Fig. 4A and B). However, the present study did not observe Rab25 inducing claudin-7 transcript (Fig. 4C). Instead, Rab25 maintained the level of claudin-7 expression in the presence of CHX (Fig. 4D). Furthermore, treatment of the cells with pharmacological inhibitors of proteasome (MG-132; Fig. 4E) and lysosome (CQ; Fig. 4F) increased Rab25-induced claudin-7 expression. Collectively, these results suggested that Rab25 induces claudin-7 expression by stabilizing its protein.

Next, the role of claudin-7 in Rab25-induced suppression of colon cancer cell invasion was explored. Ectopic expression of claudin-7 reduced the expression of p-EGFR and Snail (Fig. 5A). In addition, claudin-7 upregulated E-cadherin expression (Fig. S6) and attenuated colon cancer cell invasion, which was reversed by enforced Snail expression (Fig. 5B). Consistently, silencing of claudin-7 expression increased colon cancer cell invasion (Fig. 5C) and migration (Fig. 5D). Furthermore, claudin-7 siRNA recovered Rab25-induced EGFR inactivation and suppression of Snail expression (Fig. 5E). In addition, it was observed that claudin-7 significantly reduced Ras activity in colon cancer cells (Fig. 5F) and that ectopic expression of constitutively active KRAS (KRAS-V12) rescued claudin-7-induced suppression of HCT-116 cell invasion (Fig. 5G). Therefore, these data implied that claudin-7 mediated Rab25-induced attenuation of colon cancer cell invasion.