The Tumor Suppressor TGFBR3 Blocks Lymph Node Metastasis in Head and Neck Cancer

The TGF-β type III receptor (TGFBR3) is an essential constituent of the TGF-β signaling. In this study, we observed a down-regulation of TGFBR3 in oral cancer, a subtype of head and neck cancer (HNC), and patients with low TGFBR3 had poor clinical outcomes. Ectopic expression of TGFBR3 decreased migration and invasion of oral cancer cells and lymph node metastasis of tumors, whereas depletion of TGFBR3 had the opposite effect. In SMAD4-positive OC-2 oral cancer cells, TGFBR3-mediated suppression requires both of its cytoplasmic interacting partners ARRB2 and GIPC1. We demonstrated that TGFBR3 induces the abundance of secreted angiogenin (ANG), a known pro-angiogenic factor, and ANG is essential and sufficient to mediate TGFBR3-dependent inhibition of migration and invasion of oral cancer cells. Notably, in SMAD4-deficient CAL-27 oral cancer cells, only GIPC1 is essential for TGFBR3-induced suppressive activity. Accordingly, HNC patients with low expressions of both TGFBR3 and GIPC1 had the poorest overall survival. In summary, we conclude that TGFBR3 is as a tumor suppressor via SMAD4-dependent and -independent manner in both tumor and stromal cells during oral carcinogenesis. Our study should facilitate the possibility of using TGFBR3-mediated tumor suppression for HNC treatment.

we employed an antigen retrieval method to enhance the immunodetection before antibody incubation. Consecutive tissue sections from the same patient were individually incubated overnight with the indicated antibodies at 4°C followed by incubation with the secondary antibody. The immunocomplexes were detected by the Dako REAL™ EnVision™ Detection System, Peroxidase/DAB+ (Hamburg, Germany).

Isolation and cultivation of oral CAFs and NFs
Briefly, fresh tissues were washed several times with phosphate-buffered saline (PBS) and antibiotics. The tissues were sliced into small pieces (1 × 1 × 1 mm) and maintained in DMEM containing 20% FBS, glutamine (20 μg/ml), penicillin (100 U/ml), and streptomycin (100 μg/ml) at 37°C with 5% CO2. Following the cellular characterization of the identity of these isolated cells at passage 3, isolated NFs and CAFs at 5 to 7 passages were used for the subsequent studies.

RNA isolation and RT-qPCR
Total RNA was isolated by using TRIzol reagents from the indicated cells or snapfrozen tissues. One μg RNA was reverse-transcribed into cDNA using the High Capacity cDNA Reverse Transcription Kit. We amplified cDNA samples by using the Fast SYBR Green Master Mix and determined the cycle threshold (Ct), the fractional cycle number at which the amount of an amplified target reaching a fixed threshold.
The mRNA expression of the indicated genes in triplicates was calculated by using 2 -ΔCt (ΔCt = Ct target gene -Ct 28S rRNA ). The primers were listed in Table S1.

Nuclear and cytosol fractionation
Following the TGF-β stimulation (10 ng/mL) for one hour, the indicated cells were washed with phosphate-buffered saline and harvested in lysis buffer (10 mM Tris-HCI, pH 6.8, 10 mM NaC1, 3 mM MgC12, 0.05% NP-40, 1 mM EGTA, 1 mM Na3VO4, 50 mM NaF, and 1 nM okadaic acid) containing protease inhibitors. Following lysate centrifugation at 20,800 g for 15 min at 4 °C, the supernatant was collected for the cytosolic fraction. The pellet was subsequently washed with a wash buffer (10 mM PIPES, pH 6.8, 25 mM NaC1, 3 mM MgC12, 300 mM sucrose, 1 mM EGTA, 1 mM Na3VO4, and 50 mM NaF) followed by centrifugation at 2,700 g for 5 min at 4 °C. The pellet was resuspended in 100 μL of wash buffer and layered with a 1 ml sucrose buffer (1M Sucrose, 1 mM Na3VO4, and 50 mM NaF) followed by centrifugation at 2,700 g for 10 min at 4 °C. The resulting pellet was washed with the wash buffer and then extracted with an extraction buffer (20 mM HEPES, pH 7.9, 300 mM NaC1, 1.5 mM MgC12, 0.2 mM EDTA, 1 mM Na3VO4, 0.1 mM β-glycerophosphate, 50 mM NaF, and °C, the supernatant was stored as a nuclear extract for subsequent studies.

Western Blot analysis
Cells were lysed in the lysis buffer (50 mM HEPES pH 7.4, 150 mM NaCl, 1% Triton X-100, 10% glycerol, 1 mM EGTA, 1 mM EDTA, 10 mM sodium pyrophosphate, 100 mM sodium fluoride, 0.2 mM sodium orthovanadate) with a protease inhibitor cocktail (Biotool, Houston,TX, USA). Following centrifugation at 13,000 rpm for 15 minutes to remove cell debris, we measured the protein concentration by Bio-Rad Bradford Protein Assays (Hercules, CA, USA). Equal amounts of total protein were subjected to SDS-PAGE, followed by Western blots probed with the indicated antibodies, and detected by Chemiluminescence Reagent. Densitometry was used to quantify the expression of the indicated protein.

Cell proliferation assay
Two different assays, cell enumeration and OD492 measurement by MTS kits, were used for measuring cell proliferation. For cell enumeration, the indicated cells were seeded in triplicate at 10-20% confluence in 24-well plates. Cells were harvested for viable cell count by trypan blue exclusion on a daily basis for four days after seeding. For OD492 measurement, the indicated cells were seeded in quadruplicate in 96-well plates and subjected to growth in a CO2 incubator for two days prior to the use of MTS kits. This experiment was independently repeated three times. Data are mean ± SD.

Wound repair assay
Culture dishes (35 mm), coated overnight with type I collagen (5 μg/mL) from rat tails, were seeded with 90% confluent cell density in the growth medium. After 16 hours, cells were treated with mitomycin C for 24 hours. Cell monolayers were wounded by scraping with a pipette tip and incubated at 37°C with the CM if needed. Cell migration was monitored and photographed at the indicated time post-wounding. The mean distance of ten wound widths along the wound before and after the migration was calculated. The migration rate was the cell migration distance per hour and expressed as Mean ± SD. This experiment was independently repeated three times.

Invasion assay
Invasion assays were performed in 24-well Transwell units with 8-µm-pore polycarbonate membranes. The indicated cells (3x10 5 cells per well for cancer cells or 3x10 4 cells per well for CAFs) in 250 μL of the starvation medium were added in duplicate onto upper chambers, precoated with 1 mg/mL Matrigel for cancer cells or 2 mg/mL collagen for CAFs. Lower chambers were filled with 500 μL growth medium or CM. After 24-h (for cancer cell) or 48-h (for CAFs) incubation, cells that remained attached to the upper side of the filter were removed with cotton swabs. Cells that had migrated through the membrane to the lower surface were stained with Giemsa solution and counted in five random fields under a light microscope at 100X or 200X magnification. Each experiment was repeated three times, and results were expressed as mean ± SD.  Figure S1. The decrease of TGFBR3 protein expression in oral cancer relative to adjacent normal tissue. Following IHC staining, the stained tissue image of another representative oral cancer was taken under 400 x magnification. Scale bar, 100 µm.

Figure S5. Characterization of adjacent normal fibroblasts (NFs) and cancerassociated fibroblasts (CAFs) isolated from human oral cancer tissue samples. a A
representative image of cell morphology of NFs and CAFs. b Western blot analysis was used to analyze the expression of Pan-CK (an epithelial cell marker), vimentin (a fibroblast marker), FSP-1 (a fibroblast marker), and α-SMA (a CAF marker). All the uncropped blots with molecular weight markers for Figure S5b are shown in Figure   S20.   Figure S21. Figure S8. The deregulation of ANG mRNA has no impact on TCGA-HNC patient clinical outcomes. a An Oncomine analysis of ANG mRNA expression in 3 HNC patient cohorts. We used box-plot diagrams to compare the mRNA levels of ANG in normal tissues with those in tumor tissues using Oncomine datasets. b The overall survival rates of HNC patients (N=497) were analyzed using the Kaplan-Meier curve and log-rank test based on high (> median) and low (< median) mRNA levels for ANG or TGFBR3/ANG from the TCGA cohort.