The B16-F0 mouse melanoma cell line (cat. no. JCRB0202) was purchased from Japanese Collection of Research Bioresources (JCRB) Cell Bank and maintained in Eagle's Minimum Essential Medium (EMEM; FUJIFILM Wako Pure Chemical Corporation) supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich; Merck KGaA), 100 U/ml penicillin and 100 µg/ml streptomycin (both from Nacalai Tesque, Inc.) under mycoplasma-free conditions. Clone M3 (Cloudman S91) melanoma cell line was obtained from European Collection of Authenticated Cell Cultures (ECACC) and cultured in Ham's F10 medium (FUJIFILM Wako Pure Chemical Corporation) supplemented with 15% FBS, 2 mM glutamine (Nacalai Tesque, Inc.), 100 U/ml penicillin and 100 µg/ml streptomycin. 293T and 293FT cells were obtained from Invitrogen; Thermo Fisher Scientific, Inc. HEK-Blue™ TGF-β cells were purchased from InvivoGen. 293T, 293FT and HEK-Blue™ TGF-β cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Nacalai Tesque, Inc.) supplemented with 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin. The cultured medium for 293FT cells was also supplemented with 1% non-essential amino acid solution (Nacalai Tesque, Inc.). All cell lines were cultured in a humidified incubator containing 5% CO₂ at 37°C.

TGF-β1 (PeproTech, Inc.), TGF-β2 (PeproTech, Inc.) and TGF-β3 (R&D Systems, Inc.) were used at concentrations of 1 ng/ml or 3 ng/ml depending on the experiment. SB431542 (FUJIFILM Wako Pure Chemical Corporation) was used at a concentration of 10 µM.

The correlation between the levels of expression of genes encoding all human TGF-β isoforms, TGFB1, TGFB2 or TGFB3 and overall survival of melanoma patients was performed using a public database, PrognoScan (http://dna00.bio.kyutech.ac.jp/PrognoScan/) which comprises multiple cancer microarray datasets (32). TGFB1, TGFB2 and TGFB3 were used as queries. The analysis was conducted by minimum P-value approach, which allowed grouping of patients into two groups based on the expression levels of TGFB1, TGFB2 and TGFB3 at all possible cutoffs (cutoff points providing the best minimum corrected P-value were 0.76 for TGFB1, 0.87 for TGFB2 and 0.66 for TGFB3 analyses, respectively). The analysis results in the present study were based on the evaluation of TGFB1, TGFB2 and TGFB3 expression levels and survival of melanoma patients whose data was combined in dataset: GSE19234 (33). The log-rank test was used for statistical analysis.

The effect of expression of Fc chimeric receptors on proliferation of B16 melanoma cells was evaluated by WST-1 assay. The B16 cells (3.5×10⁴) were seeded into a 12-well culture plate and incubated overnight at 37°C, 5% CO₂. The following day the medium was refreshed, and the cells were cultured for 72 h. The assay was conducted using the WST-1 reagent (Dojindo Molecular Technologies, Inc.) according to the manufacturer's protocol. The colorimetric changes in the substrate were measured with a microplate reader (Model 680; Bio-Rad Laboratories, Inc.) at 450 nm. To evaluate the effect of TGF-β isoforms on the proliferation of B16 melanoma cells, B16 cells (7.5×10⁴ cells/well) were seeded into 6-well culture plates and cultured overnight at 37°C in 5% CO₂. The following day, the medium was replaced with 1 ml of EMEM and cells were treated with each TGF-β (3 ng/ml) isoform or SB431542 (10 µM) for 72 h followed by direct cell counting with Bürker-Türk hemocytometer (cat. no. 03-303-1; Erma, Inc.). The experiment was performed in triplicate and repeated twice.

Fc chimeric receptors were generated by transfection of 293T cells with respective Fc chimeric receptor expression vectors. The construction of expression vectors was performed as previously described (31). Briefly, to express Control-Fc protein, the Fc region of human IgG fused to the interleukin-2 signal peptide was inserted into pENTR201 vector (Invitrogen; Thermo Fisher Scientific, Inc.). For the expression of TβRII-Fc chimeric receptor the extracellular domain (ECD) of TβRII-Fc corresponding to the 184 amino acids (ECD_1-184) was fused with the Fc region of human IgG and inserted into pENTR201 (Invitrogen; Thermo Fisher Scientific, Inc.). To express TβRI-TβRII-Fc chimeric receptor the ECD of TβRI corresponding to the 128 amino acids (ECD_1-128) was fused with ECD of TβRII-Fc lacking signal peptide (ECD_23-184) followed by the addition of the Fc region of human IgG and inserted into pENTR201 (Invitrogen; Thermo Fisher Scientific, Inc.). The Gateway Technology (Invitrogen; Thermo Fisher Scientific, Inc.) was used to transfer Control-Fc, TβRII-Fc and TβRI-TβRII-Fc cDNAs into pCSII-EF-RfA (a gift from Dr Hiroyuki Miyoshi, Keio University, deceased) to generate lentiviral expression vectors; pCSII-EF-RfA-Control-Fc, pCSII-EF-RfA-TβRII-Fc and pCSII-EF-RfA-TβRI-TβRII-Fc. 293T cells (9.0×10⁶) were seeded into 10-cm tissue culture dishes and cultured overnight at 37°C and 5% CO₂, followed by transfection with pCSII-EF-RfA vectors (20 µg/dish) expressing each chimeric receptor pCSII-EF-RfA-Control-Fc, pCSII-EF-RfA-TβRII-Fc and pCSII-EF-RfA-TβRI-TβRII-Fc, using Lipofectamine 2000 Transfection Reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's protocol. A total of 4 h post-transfection, the medium was replaced with serum-free Opti-MEM (Gibco; Thermo Fisher Scientific, Inc.) and the cells were incubated for 48 h to allow accumulation of secreted Fc chimeric receptors in culture medium. The accumulation of soluble chimeric receptors in the conditioned media was evaluated by immunoblotting with rabbit polyclonal anti-human IgG-Fc fragment (1:5,000; A80-105; Bethyl Laboratories, Inc.) as described in the Immunoblot analysis section. The concentration of each Fc chimeric receptor was assessed by enzyme-linked immunosorbent assay (ELISA) with the Human IgG ELISA Quantitation Set (E80-104; Bethyl Laboratories, Inc.). The collected conditioned media were aliquoted and stored at −80°C until use.

Total RNA was extracted from B16 and Clone M3 cells. The extraction was performed using Sepasol (R)-RNA I Super G (Nacalai Tesque, Inc.) and reverse-transcribed to cDNA with ReverTra Ace qPCR-RT Master Mix (Toyobo Life Science) according to the manufacturer's protocol. Quantitative PCR analysis was performed using SYBR Green (Roche Applied Science) on a Step One Plus Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific, Inc.) under the following cycling conditions: 95°C, 10 min, followed by 40 cycles at 95°C for 15 sec and 60°C for 30 sec with a final incubation at 95°C for 5 sec. The relative standard curve method was used to determine the relative expression of target genes (34). All expression data were normalized to the expression of β-actin. The genes and corresponding primer sequences are listed in Table SI.

B16, Clone M3 and 293T cells were lysed using radioimmunoprecipitation assay buffer (RIPA Lysis Buffer System; Santa Cruz Biotechnology, Inc.) in the presence of a protease inhibitor (Sigma-Aldrich; Merck KGaA), followed by repeated freeze and thawing. Cell lysates were cleared by centrifugation at 16,400 × g for 30 min at 4°C, and the supernatants were collected. The protein concentration in obtained lysates was determined using a BCA Protein Assay Kit (Thermo Fisher Scientific, Inc.).

Denatured cell lysates (30 µg of total proteins in Figs. 5A and S5B; 20 µg of proteins in Figs. 3D, 4C, S3B and S4B; or 10 µg of total protein in Fig. S1) were separated on 10.5% or 12% SDS-PAGE gel depending on the experiment, followed by transfer onto PVDF membranes (Merck KGaA). The membranes were then blocked with 3% bovine serum albumin (BSA; FUJIFILM Wako Pure Chemical Corporation) for 30 min at room temperature and incubated with appropriate primary antibodies diluted in 3% BSA (Nacalai Tesque, Inc.): Rabbit monoclonal anti-αSMA (1:1,000; product no. 19245; Cell Signaling Technology, Inc.), rabbit polyclonal anti-TAGLN/Transgelin (SM22α; 1:2,000; product code ab14106; Abcam), rabbit polyclonal anti-human IgG-Fc fragment (1:5,000), and rabbit polyclonal anti-α-tubulin (1:10,000; product code ab4074; Abcam) overnight at 4°C. The membranes were then incubated with goat anti-rabbit IgG HRP-linked antibody (1:5,000; product no. 7074S; Cell Signaling Technology, Inc.) for 1 h at room temperature. The target proteins were detected using an Enhanced Chemiluminescence Kit (ECL detection reagent; Cytiva) and visualized with a Fusion Solo S Imaging System (SOLO.6S.EDGE; Vilber Lourmat).

B16 and Clone M3 cells (3.5×10⁴ cells/well) were seeded on cover glasses placed into 12-well tissue culture plates and treated with TGF-β1, -β2, and -β3 in the presence or absence of Fc chimeric receptors for 72 h, at 37°C in 5% CO₂. The cells were then fixed with methanol/acetone (1:1) for 20 sec on ice and blocked in PBS containing 1% BSA (FUJIFILM Wako Pure Chemical Corporation) for 30 min at room temperature and incubated with primary antibodies diluted in Blocking One buffer: Rabbit polyclonal anti-TAGLN/Transgelin (1:1,000), mouse monoclonal anti-actin, αSMA-Cy3™ (1:1,000; cat. no. C6198-2ML; Sigma-Aldrich; Merck KGaA) overnight at 4°C. To visualize SM22α and nuclei, samples were incubated for 1 h at room temperature with a mixture of donkey polyclonal anti-rabbit IgG (H+L) Alexa Fluor 488-conjugated secondary antibodies (1:1,000 in Blocking One buffer; cat. no. A-21206; Thermo Fisher Scientific, Inc.) and 500 ng/ml Hoechst33342 (Cell Signaling Technology, Inc.) for nuclear staining. The samples were then embedded in Fluoromount-G mounting medium (Cosmo Bio Co., Ltd.). Images were captured under a fluorescence microscope (BZ-X710; Keyence Corporation).

B16 or Clone M3 cells (3.5×10⁴ cells/well) were seeded into tissue culture plates (12-well plate or 6-well plate, depending on the experiment) and cultured overnight at 37°C in 5% CO₂. The next day, the medium was replaced with 1 ml of serum-free Opti-MEM. The Fc chimeric receptor/ligand complexes were generated by mixing the conditioned medium from the 293T cells expressing Fc chimeric receptors containing the equal amount of Fc chimeric receptors (600 ng of Fc chimeric receptors in 500 µl of Opti-MEM) with TGF-β1, -β2, or -β3 at the concentration of 3 ng/ml. Samples were incubated for 2 h at 37°C to allow formation of the Fc chimeric receptor/ligand complexes and added to the B16 or Clone M3 cells. The cells were then incubated at 37°C and 5% CO₂ with Fc chimeric receptor/ligand complexes for 4 or 72 h (depending on the experiment) and subjected to gene expression analysis by RT-qPCR, immunocytochemistry, or immunoblotting. Cells treated with each TGF-β ligand in serum-free Opti-MEM were used as controls for upregulation of the TGF-β signal while samples treated with the mixture of SB431542, a TβRI kinase inhibitor, and TGF-β in serum-free Opti-MEM were used as controls for inhibition of the TGF-β signal.

The lentiviral particles were produced as previously described (31). Briefly, the 293FT (8.0×10⁶) cells were co-transfected with 5.5 µg of expression plasmids (pCSII-EF-RfA-Control-Fc, pCSII-EF-RfA-TβRII-Fc, and pCSII-EF-RfA-TβRI-TβRII-Fc) and packaging plasmids pCMV–VSV-G-RSV-Rev (3.25 µg; RIKEN BioResource Center) and pCAG-HIVgp (3.25 µg; RIKEN BioResource Center) using Lipofectamine 2000 Transfection Reagent (Invitrogen; Thermo Fisher Scientific, Inc.) in 5 ml Opti-MEM supplemented with 10% FBS. The control lentiviral particles expressing green fluorescent protein (GFP) were prepared by transfecting 293T cells with 5.5 µg of pCSII-EF-RfA-GFP and packaging plasmids pCMV–VSV-G-RSV-Rev (3.25 µg) and pCAG-HIVgp (3.25 µg). A total of 24 h post-transfection, the transfection medium was refreshed with 7.5 ml Opti-MEM, 10% FBS, and the recombinant lentiviruses were harvested 48 h later. The conditioned media containing viral particles were collected by centrifugation at 4°C for 5 min at 1,700 × g, and incubated at 4°C for 7 days on the rotary shaker with Lenti-X Concentrator (Takara Bio, Inc.). The viral particles were then centrifuged at 4°C for 45 min, at 1,500 × g, and resuspended in 140 µl of Opti-MEM. The 70-µl of concentrated lentiviral particles were used to infect B16 melanoma cells (3×10⁵ cells/well in 12-well tissue culture plates). Transduction efficiency was evaluated using lentiviral particles expressing green GFP. The successful expression was examined by immunoblotting using rabbit polyclonal anti-human IgG-Fc antibody as described in the Immunoblot analysis section. The second generation of transduced B16 cells was used for further experiments.

The ability of Fc chimeric receptors, secreted by transduced B16 cells, to trap TGF-β ligands was examined using the HEK-Blue TGF-β reporter system. The B16 cells (1×10⁶ cells) expressing each of the Fc chimeric receptors were seeded into 10 cm tissue culture plates and incubated overnight at 37°C in 5% CO₂. The following day, the medium was replaced with 5 ml of serum-free Opti-MEM and the cells were incubated for 48 h to allow the accumulation of Fc chimeric receptors in the culture supernatant. The conditioned media were collected and stored at −80°C until use.

HEK-Blue TGF-β reporter cells (1.0×10⁵) were seeded into 96-well plates and incubated overnight at 37°C in 5% CO₂. The following day, the medium was replaced with 200 µl of serum-free DMEM, and the cells were incubated for 3 h. The B16 cell-derived conditioned medium containing Control-Fc, TβRII-Fc, or TβRI-TβRII-Fc chimeric proteins, was mixed with TGF-β ligands (1 ng/ml) and incubated for 2 h at 37°C to allow the formation of Fc chimeric receptor/ligand complexes. Next, Fc chimeric receptor/ligand complexes were added to the HEK-Blue TGF-β reporter cells, followed by incubation for 24 h, at 37°C. The activation of TGF-β/Smad signals was detected using QUANTI-Blue substrate (InvivoGen) following incubation for 30 min at 37°C. The colorimetric change of the substrate by the secreted alkaline phosphatase (SEAP) was quantified at 640 nm using a microplate reader (Bio-Rad Laboratories, Inc.).

A total of 62 female C57/BL6 mice (5–6 weeks old; average weight, 14–19 g) were purchased from Japan SLC, Inc. All animal experimental protocols were approved (approval no. R-02-017-1) by the Animal Experiment Committee of the Graduate School of Dentistry, Osaka University (Osaka, Japan). The mice were kept under a temperature of 23–24°C with 40–60% humidity and a 12-h light/dark cycle. Mice were provided with access to food and water ad libitum throughout the experiment. A total of 20 mice was used for injection of B16-Control-Fc cells, 21 for B16-TβRII-Fc cells, and 21 for B16-TβRI-TβRII-Fc cells. All animals underwent general anesthesia with mixture of medetomidine (0.3 mg/kg; Nippon Zenyaku Kogyo, Co., Ltd.) midazolam (4 mg/kg; Astellas Pharma, Inc.) and butorphanol (5 mg/kg; Meiji Seika Kaisha, Ltd.) by intraperitoneal administration (35,36). The B16 cells (5.0×10⁵) expressing the Fc chimeric receptors, B16-Control-Fc, B16-TβRII-Fc, B16-TβRI-TβRII-Fc were suspended in 50 µl serum-free EMEM and subcutaneously injected into left flank region. Mice that did not develop any palpable tumor or did not survive until the endpoint of the experiment were excluded from the analysis. Tumor growth was monitored for 26 days. The tumor volume was measured twice per week and estimated using the following equation: Tumor volume (mm³) = [length (mm) × width (mm)²]/2. The size of developed tumors was selected as the humane endpoint. Mice were sacrificed when the size of the largest primary tumors started to exceed the 1,000 mm³. As large melanoma tumors often develop necrotic changes that would likely affect the experimental outcome, a total of 32 mice bearing primary tumors derived from Control-Fc (n=11), TβRII-Fc (n=9), and TβRI-TβRII-Fc (n=11) were thus euthanized on day 26 by intraperitoneal injection of the mixture of medetomidine (3 mg/kg), midazolam (40 mg/kg) and butorphanol (50 mg/kg).

Statistical analysis was carried out by EZR software (37). Results are presented as the mean ± standard deviation (SD) or standard error (SE). Each experiment was performed in triplicate and repeated twice. Comparisons of quantitative data were conducted using one-way ANOVA with post hoc Tukey's test or Mann-Whitney U test with post hoc Bonferroni test depending on experiment. P<0.05 was considered to indicate a statistically significant difference.

TGF-β has been revealed to promote invasiveness and progression of melanoma (13,16). Previous studies revealed that melanoma cells expressed all three TGF-β isoforms (12) and that an elevated level of TGF-β in melanoma patients was associated with metastatic outcomes (38). In addition, meta-analysis using a public database, PrognoScan (http://dna00.bio.kyutech.ac.jp/PrognoScan/) (32) and dataset: GSE19234 (33), revealed that high expression levels of TGF-β2, but not that of TGF-β1 or TGF-β3, were associated with overall survival of melanoma patients (Fig. 1). Therefore, it was examined whether TβRI-TβRII-Fc chimeric receptor could be applied in the melanoma model. In our study, B16 melanoma cells were used, in which the EMT program can be activated in response to TGF-β. 293T cells were transfected with the vectors expressing Control-Fc, TβRII-Fc, or TβRI-TβRII-Fc chimeric receptors and the accumulation of chimeric proteins in the conditioned media was confirmed by immunoblotting (Fig. S1). Such conditioned media were then used to examine the effect of soluble chimeric receptors on the activation of TGF-β signals. B16 cells were incubated in conditioned media of 293T cells expressing Control-Fc, TβRII-Fc, or TβRI-TβRII-Fc chimeric receptors in an absence or presence of TGF-β1, -β2, or -β3, respectively, followed by an analysis of the expression of genes directly responding to TGF-β, TMEPAI, and PAI-1 by RT-qPCR. Incubation of B16 melanoma cells with any of the TGF-β isoforms upregulated the expression of both TMEPAI (Fig. 2A) and PAI-1 (Fig. 2B). As anticipated, SB431542, a TβRI kinase inhibitor, reduced the expression of both direct target genes to the background level (Fig. 2). Incubation of B16 melanoma cells with Control-Fc protein did not reduce the expression of TMEPAI, and PAI-1 induced by TGF-βs. Reduced expression of both genes in the presence of TβRII-Fc chimeric receptor was only observed when cells were treated with TGF-β1 and TGF-β3, but not by TGF-β2 (Fig. 2). Conversely, TβRI-TβRII-Fc chimeric receptor significantly inhibited the expression levels of TMEPAI and PAI-1 induced by all TGF-β isoforms indicating that TβRI-TβRII-Fc chimeric receptor effectively interfered with TGF-β signals also in the melanoma model. To generalize our findings, the same set of experiments were performed with another melanoma cell line, Clone M3. Clone M3 cells responded to all TGF-β isoforms as indicated by upregulated expression of TMEPAI (Fig. S2A) and PAI-1 (Fig. S2B). In addition, incubation with TβRI-TβRII-Fc decreased the expression of TMEPAI and PAI-1 induced by all TGF-βs when compared with the expression of both genes detected in the Clone M3 cells incubated with Control-Fc protein indicating that TβRI-TβRII-Fc chimeric receptor suppressed TGF-β signals in multiple types of melanoma cells.

Melanoma cells can activate the EMT program in response to TGF-β. Previous studies revealed that melanoma cells treated with TGF-β upregulated the expression of mesenchymal markers (14,17). Therefore, it was examined whether similar changes can be observed in B16 melanoma cells. B16 melanoma cells were cultured for 72 h in the absence or presence of each TGF-β isoform or SB431542, and the expression of various mesenchymal markers was determined using RT-qPCR. The treatment with any of TGF-β isoform resulted in upregulated expression of all mesenchymal markers; SM22α (Fig. 3A), αSMA (Fig. 3B) and fibronectin (Fig. 3C), while SB431542 did not exhibit any effect (Fig. 3A-C). These results were also confirmed at the protein level using immunoblotting and immunocytochemical analysis. A significant increase was observed in the band intensity corresponding to each mesenchymal marker, SM22α and αSMA (Fig. 3D), as well as an increase in a fluorescent signal related to the presence of SM22α-positive and αSMA-positive cells upon TGF-β treatment (Fig. 3E), indicating that B16 melanoma cells activated TGF-β-dependent EMT. The activation of the EMT program was also confirmed using Clone M3 cells. Treatment of the Clone M3 cells with any of the three TGF-β isoforms resulted in upregulated expression of SM22α as revealed by RT-qPCR (Fig. S3A), immunoblotting (Fig. S3B), and immunocytochemical analysis (Fig. S3C), supporting the findings that treatment with TGF-β upregulated the expression of SM22α in multiple types of melanoma cells.

Our data indicated that the EMT program was activated upon treatment with any of the three TGF-β isoforms. In addition, as shown in Fig. 1B, high TGFB2 expression was a poor prognostic factor in overall survival in melanoma patients, indicating that inhibiting the EMT program may have beneficial effects on melanoma treatment. Therefore, in the following experiment, the effect of TβRI-TβRII-Fc chimeric receptor on activation of the EMT program was examined in B16 melanoma cells. The B16 melanoma cells were treated without or with TGF-β1, -β2 or -β3 in the presence of conditioned medium derived from 293T cells expressing Control-Fc, TβRII-Fc, or TβRI-TβRII-Fc chimeric receptors for 72 h, followed by RT-qPCR analysis for the expression of mesenchymal markers, SM22α (Fig. 4A) and αSMA (Fig. 4B). The expression levels of both mesenchymal markers were upregulated when B16 melanoma cells were incubated in the conditioned medium containing Control-Fc protein. A significant suppression of SM22α (Fig. 4A) and αSMA (Fig. 4B) expression was observed in the cells stimulated with TGF-β1 or TGF-β3 in the presence of TβRII-Fc or TβRI-TβRII-Fc chimeric receptors. Conversely, the induction of the EMT program by TGF-β2 was inhibited only in the presence of TβRI-TβRII-Fc chimeric receptor (Fig. 4A and B), indicating that TβRI-TβRII-Fc chimeric receptor could modulate the response to TGF-β2 in the melanoma model. The aforementioned findings were also confirmed at the protein level using immunoblotting (Fig. 4C) and immunocytochemical analyses (Fig. 4D and E). In agreement with the RT-qPCR results, changes in the intensity of bands corresponding to upregulated expression of SM22α and αSMA proteins were observed (Fig. 4C) in response to all TGF-β isoforms, in the absence or presence of Fc chimeric receptors, as well as the number of SM22α-positive and αSMA-positive cells (Fig. 4D and E, respectively). Consistent with the RT-qPCR results, effective inhibition of the TGF-β2-induced EMT program was observed only in the presence of TβRI-TβRII-Fc chimeric receptor while Control-Fc and TβRII-Fc did not demonstrate such an effect (Fig. 4C-E). The suppressive effect of TβRI-TβRII-Fc chimeric receptor on EMT-associated changes in SM22α expression was also examined at both RNA and protein levels in Clone M3 cells. As anticipated, the expression of SM22α induced by any of the three TGF-β isoforms was inhibited only by TβRI-TβRII-Fc chimeric receptor (Fig. S4), indicating that TβRI-TβRII-Fc chimeric receptor could be potentially used for targeting all TGF-β isoforms present in the TME of melanoma tumors.

As our in vitro data revealed effective inhibition of the EMT program, examination of the effect of TβRI-TβRII-Fc chimeric receptor on melanoma tumor growth in vivo was performed. Therefore, B16 melanoma cells expressing Control-Fc, TβRII-Fc, and TβRI-TβRII-Fc chimeric proteins were established by infecting B16 melanoma cells with lentiviral vectors (Fig. S5). As revealed in Fig. 5A, all Fc chimeric receptors were expressed in B16 cells (Fig. 5A; cell lysate) and secreted into the culture media (Fig. 5A; conditioned medium); however, the amount of accumulated TβRI-TβRII-Fc chimeric receptor was lower when compared with the secreted amount of Control-Fc or TβRII-Fc chimeric receptors.

The anti-proliferative effect of TGF-β on normal epithelial cells has been previously reported (39). Moreover, in the early stage of melanoma progression, TGF-β has been revealed to inhibit cell growth (15). In agreement with these previous findings, B16 melanoma cells incubated in the presence of TGF-βs for 72 h demonstrated decreased proliferation when compared with the non-treated control cells, independently of the TGF-β isoform used (Fig. S6). The proliferation of B16 melanoma cells in the presence of SB431542, a TβRI kinase inhibitor, did not differ from the proliferation of control cells (Fig. S6). As B16 melanoma cells have been revealed to secrete active TGF-βs (40), there was a possibility that the expression of Fc chimeric receptors could alter their extracellular environment and affect cell proliferation. Therefore, the proliferation of B16 cells expressing each Fc chimeric receptor was examined and it was revealed that there was not any difference in the proliferation exhibited by B16 cells expressing TβRII-Fc and TβRI-TβRII-Fc chimeric receptors when compared with B16 cells expressing Control-Fc protein (Fig. 5B).

Next, the effect of TβRII-Fc and TβRI-TβRII-Fc chimeric receptors accumulated in conditioned media of B16 cells on TGF-β signaling was examined using HEK-Blue TGF-β reporter cells. HEK-Blue cells were cultured in the conditioned medium of B16 cells expressing Control-Fc, TβRII-Fc, or TβRI-TβRII-Fc chimeric receptors in the absence or presence of TGF-β1, -β2, or -β3. As anticipated, stimulation of HEK-Blue cells with any TGF-β isoform in the presence of conditioned medium from B16 cells expressing Control-Fc protein resulted in upregulation of TGF-β signals (Fig. 5C). Conditioned medium derived from B16 cells expressing TβRII-Fc chimeric receptor significantly inhibited signals induced by TGF-β1 or TGF-β3 and had no effect on signals induced by TGF-β2 (Fig. 5C). Complete inhibition of TGF-β signals was observed only in the presence of conditioned medium derived from B16 cells expressing TβRI-TβRII-Fc chimeric receptor (Fig. 5C), indicating that TβRI-TβRII-Fc could trap all TGF-β isoforms.

Finally, the effect of Fc chimeric receptors on melanoma tumor growth was examined in vivo. The B16 cells expressing Control-Fc, TβRII-Fc, or TβRI-TβRII-Fc chimeric receptors were subcutaneously inoculated in the left flank of C57/BL6 mice. The tumor growth and body weight were then monitored for 26 days. Expression of TβRI-TβRII-Fc chimeric receptor inhibited B16 melanoma tumor growth in vivo (Fig. 6A) when compared with Control-Fc. Moreover, the size of tumors originating from B16 cells expressing TβRI-TβRII-Fc chimeric receptor was significantly smaller than that developed from B16 cells expressing Control-Fc protein (Fig. 6B) indicating that it could effectively trap all TGF-β isoforms residing in the TME. Of note, no significant difference was observed between Control-Fc and TβRII-Fc or TβRII-Fc and TβRI-TβRII-Fc groups. In addition, no significant differences in body weight were observed between the three experimental groups (Fig. 6C).