The human ovarian cancer cell line, RMG-I, which was originated from the tissues of human ovarian clear cell carcinoma, was donated by Professor M. Iwamori of Tokyo University of Japan. RMG-I-H cell line was established as previously reported (6,7). The RMG-I-C cells transfected with the vector alone were used as controls.

Recombinant human TGF-β1 was from peprotech. Anti-TGFβ RI, anti-TGFβ RII, anti-ERK1, anti-dually phosphorylated ERK (Thr202/Tyr204), anti-MEK-1/2, anti-p-MEK-1/2, anti-Akt, anti-p-Akt, anti-Smad7 antibody, horseradish peroxidase (HRP)-labeled secondary antibodies and protein A/G Plus-Agarose were from Santa Cruz Biotechnology. Anti-Smad2/3 and anti-p-Smad2/3 antibodies were from Abzoom. TRIzol, PrimeScript™ RT reagent kit, SYBR^® Premix Ex Taq™ and GAPDH primer (D3702) from Takara. Mouse anti-human LeY mAb was produced by immunization of female BALB/c weanling mice with LeY purified from a SK-LU-3 lung cancer cell line. Antigen preparation was mixed with complete Freunds adjuvant and mice were injected intraperitoneally with 0.2 ml of this preparation on Day 0 and intravenously on Day 21 (14). Harvested immune spleen cells were fused with myeloma cells 4 days after the second immunization to produced hybridomas as described previously (15). When the cells had reached 50% confluency in the majority of wells showing cell growth, the supernatants were collected and assayed by enzyme-linked immunosorbent assay for reaction with synthetic LeY and additionally with a panel of related A, B, H blood group antigens and Lewis antigens (Isosep AB, Tullinge, Sweden and Dextra Laboratories, Reading, UK) as described previously (16). The specific anti-LeY antibody was purified from culture supernatants using conventional ultrafiltration techniques. Protein concentration was determined spectrophotometrically by absorption at 280 nm. Immunoglobulin isotype was identified as of the IgM class by using a MonoAb-ID immunoassay kit (Invitrogen) and electrophoretic analysis (14). The mouse mAb was suspended in buffer containing 0.01 M phosphate-buffered saline, 0.1% sodium azide and 1% bovine serum albumin.

The method for cell culture has been described previously (6). For Western blot assays, subconfluent cell layers were rendered quiescent by serum starvation for 12–24 h. Cells were stimulated subsequently by addition of medium containing or lacking TGF-β1 (5 ng/ml) for the specified time period. For inhibition assay, LeY mAb (10 μg/ml) was added for different times (1, 10 and 30 min) before stimulation with TGF-β1.

Total RNA was extracted using trizol reagent. cDNA was synthesized using Takara PrimeScript™ RT reagent Kit. Primers used for amplification: TβRI (158 bp) forward, AGTGTTCTGGCTCCAAATGGTAGT; reverse, GGCCCATGGGTATTCCAGTAATC. TβRII (75 bp) forward, GCAGGTGGGAACTGCAAGAT; reverse, GAAGGACTCA ACATTCTCCAAATTC. Reaction conditions were 37°C for 15 min, 85°C for 5 sec, 4°C for 5 min. The real-time PCR reaction conditions were denature at 94°C for 20 sec, 45 cycles of 94°C for 20 sec and 60 or 58°C for 20 sec in a 20-μl reaction mixture containing SYBR^® Premix Ex Taq^TM (2X) 10 μl, forward primer (5 μmol/l) 1 μl, reverse primer (5 μmol/l) 1 μl, cDNA 2 μl, dH₂O 6 μl. GAPDH was used as the endogenous control. The Light Cycler PCR and detection system (Roche Diagnostics, Mannheim, Germany) was used for real-time PCR amplification and Ct value calculation. Once the amplification was completed, the melting curve was analyzed. The change of target gene expression level was calculated using the 2^−ΔΔCT method (17).

Cells were rinsed with PBS and 1% of Triton X-100 lysis buffer (20 mM Tris-HCl, pH 7.4, 10 mM EGTA, 10 mM MgCl₂, 1 mM benzamidine, 60 mM β-glycerophosphate, 1 mM Na₃VO₄, 20 mM NaF, 2 μg/ml aprotinin, 5 μg/ml leupeptin, 0.1 mM phenylmethylsulfonyl fluoride) was added. Then centrifuged, and the supernatants were collected. Protein content was measured using the protein assay BCA kit (Beyotime Biotechnology, China) and equal amounts of protein were loaded on SDS-PAGE gels. Subsequently, proteins were transferred to PVDF membranes (Millipore, Beaford, MA) and were probed with antibodies (1:1000). Immunoreactive bands were visualized by chemiluminescence (ECL; Pierce) using a secondary antibodies (1:8000).

Membrane proteins were extracted and concentrated with Mem-PER^® eukaryote membrane protein extraction kit and Pierce^® SDS-PAGE Sample Prep Kit (Pierce, Rockford, USA). Membrane proteins were then incubated at 4°C for 2 h with TβRI or TβRII antibody. The immune complexes were isolated by stirring the mixture at 4°C overnight with Protein A/G Plus-Agarose. Thereafter, the samples were loaded onto 10% SDS-PAGE for Western blotting with the procedure described above. The LeY antibody (1:2000) was used to detect the expression of LeY in TβRI and TβRII. TβRI and TβRII antibodies (1:1000) were used to detect the expression of TβRI and TβRII, respectively.

Cells were fixed with 4% paraformaldehyde. After blocking with normal goat serum, cells were incubated with LeY and TβRI (or TβRII) antibodies (1:100) for 1 h at RT. Cells were then incubated with goat anti-mouse tetramethylrhodamine isothiocyanate (TRITC) conjugated antibody and goat anti-rabbit fluorescein isothiocyanate (FITC) labeled antibody (1:200) (Zhongshan Biotech, Beijing, China) for 1 h at RT in dark. 4,6-Diamidino-2-phenylindole (DAPI) was used to stain the nuclei at RT for 1 min. Stained slide was observed with a laser confocal microscope (C1-SI; Nikon, Tokyo, Japan). Data were collected using a computer and the digital images were generated.

The SPSS 12.0 statistical analysis software was used, while the analysis of variance was employed. p<0.05 was regarded as with statistical significance.

Cells were subjected to real-time PCR analyses to assess the mRNA levels of TβRI and TβRII. The results show that the TβRII mRNA levels in RMG-I-H cells were 1.69- and 1.74-fold (>1.3-fold) higher than that in the RMG-I and RMG-I-C cells, respectively. However, TβRI mRNA level in RMG-I-H cells was 1.03- and 1.00-fold compared to that in the RMG-I and RMG-I-C cells, respectively (Fig. 1).

To estimate the expression of the LeY oligosaccharide in TGF-β receptors, we performed a series of immunoprecipitation experiments to determine whether LeY mAb would bind to membrane extracts precipitated by TβRI and TβRII antibody. After SDS-PAGE, followed by immunoblotting, anti-LeY mAb stained the 53/70 kDa protein bands precipitated by TβRI and TβRII antibody, respectively. The results showed that both TβRI and TβRII contained LeY structures, and TβRI and TβRII showed absolute or relative increase in the content of LeY in the RMG-I-H cells, respectively (Fig. 2).

Immunoprecipitation assay showed that TβRI and TβRII contain LeY structure, we verified the spatial orientation of TGF-β receptors and LeY by immunofluorescence double staining. TβRI and TβRII were labeled with FITC and LeY antigen was labeled with TRITC. Images were scanned using a confocal microscope in serial Z-sections and then overlaid. The results clearly showed that TβRI and TβRII in green fluorescence were mainly localized on the cell membrane with a small amount localized in the cytoplasm (Fig. 3A and E); LeY antigen in red fluorescence was mainly on the cell membrane (Fig. 3B and F). As shown in the merged figures, spatial co-localization of TGF-β receptors and LeY exhibited yellow fluorescence (Fig. 3D and H, white arrow).

To further characterize the effect of LeY on TGF-β/Smad pathway, Western blot analysis was performed to analyze the expression of Smad2/3, p-Smad2/3 and Smad7 (inhibitory Smad), which are key members of TGF-β1/Smad, in cells before and after α1,2-FT gene transfection after TGF-β1 stimulation (5 ng/ml). The results showed that total protein levels of Smad2/3 did not change significantly (p>0.05), however, p-Smad2/3 level was significantly decreased in RMG-I-H cells compared to RMG-I and RMG-I-C cells. The expression of Smad7 was up-regulated in RMG-I-H cells (Fig. 4). The results indicated that LeY down-regulated the activation of TGF-β1/Smad pathway.

Given that ERK/MAPK and PI3K are two important pathways that can also be activated by TGF-β in some cell types (18–21), we first analyzed the expression of ERK1/2 and Akt in TGF-β1 stimulated RMG-I-H cells following serum starvation by Western blot analysis to determine whether TGF-β can activate ERK/MAPK and PI3K in RMG-I-H cells. As shown in Fig. 5, expression of EKR1/2 and Akt did not change (p>0.05), however, p-EKR1/2, p-Akt increased over time, suggesting TGF-β1 indeed activate the ERK and PI3K pathways in RMG-I-H cells. Then we compared the expression of ERK1/2, MEK1/2 and Akt in cells before and after α1,2-FT gene transfection, the results showed that the total protein levels of ERK1/2, MEK1/2 and Akt did not change significantly among cells (p>0.05), but the p-ERK1/2, p-MEK1/2 and p-Akt levels were significantly increased in RMG-I-H cells than those in RMG-I and RMG-I-C cells (Fig. 6). These results demonstrate that LeY up-regulated the activation of TGF-β1 mediated ERK and PI3K pathways.

The expression of phosphorylated ERK1/2, Smad2/3 and Akt in RMG-I-H cells at different time-points (1, 10 and 30 min) after treatment with LeY mAb (10 μg/ml) was analyzed by Western blotting. The results show that the expression levels of p-Smad2/3, p-ERK1/2 and p-Akt decreased over the time of antibody treatment (Fig. 7), indicating LeY antigen involvement in the TGF-β1-dependent Smad, ERK and PI3K pathways.