Abstract Glycosylation is an important modification of membrane proteins that results in functional changes in many cellular activities, from cell-cell recognition to regulatory signaling. Fucosyltransferase 8 (FUT8) is the sole enzyme responsible for core fucosylation, and aberrant fucosylation by dysregulated expression of fucosyltransferases is responsible for the growth of various types of carcinomas. However, the function of FUT8 in the progress of osteosarcoma (OS) has not been reported. In this study, we found that FUT8 is expressed at lower levels in patients with OS and in human OS cell lines such as MNNG/HOS, U2OS, and 143B, suggesting that attenuated expression of FUT8 is involved in the growth and progression of OS. Mechanistically, FUT8 affects the survival strategy of OS by modifying core-fucosylation levels of TNF receptors (TNFRs). Lower fucosylation of TNFRs activates the non-canonical NF-κB signaling pathway, and in turn, decreases mitochondria-dependent apoptosis in OS cells. Together, our results point to FUT8 being a negative regulator of OS that enhances OS-cell apoptosis and suggests a novel therapeutic strategy for treating OS. Subject terms: Glycobiology, Bone cancer, Apoptosis Introduction Osteosarcoma (OS) is the most frequent and aggressive primary malignant neoplasm diagnosed in the skeletal system [[38]1]. Many patients have to face death threat after diagnosis due to its rapid progression, tendency to metastasize to the lungs, and high recurrence rate [[39]2]. Since the advent of a comprehensive therapeutic strategy in the 1970s, which includes neoadjuvant chemotherapy, surgery, and radiotherapy, the 5-year overall survival rate increased for localized OS [[40]1]. However, the survival rate has remained unchanged for decades, leaving 30–40% of patients who still cannot benefit from comprehensive therapy [[41]3]. Therefore, it is important to identify all the factors that contribute to OS progression. Identification of such factors could pave the way for the development of novel clinical treatments for OS. Glycosylation is one of the most important post-translational modifications of proteins in all cells [[42]4], as it plays an indispensable role in maintaining many biological functions of proteins. Aberrant glycosylation is closely linked to a variety of diseases, especially cancers, and is due to alternations in the expression of glycosyltransferase [[43]5]. Aberrant glycosylation affects well-known intrinsic traits of tumors, such as proliferation, metastasis, angiogenesis, and drug resistance, among others [[44]6–[45]9]. Fucosylation is a significant kind of glycosylation that is catalyzed by fucosyltransferases (FUTs) [[46]10]. FUTs can be divided into four types according to their catalytic products: FUT1-2, FUT3-7, FUT9-11, and FUT8 [[47]11, [48]12]. FUT8 exclusively catalyzes the formation of α-1,6 fucosylation—also known as core fucosylation—by connecting fucose to the inner GlcNAc residue of N-glycans [[49]11, [50]13]. FUT8 has been extensively investigated in various cancers. Recent studies indicate that FUT8 expression levels are highly correlated with cancer development, in particular, hepatocellular carcinoma [[51]14], melanoma [[52]15], breast cancer [[53]16], and gastric cancer [[54]17]. Whether FUT8 might also influence the progression of OS is unknown. Furthermore, it is unclear what the underlying mechanism could be. In the present study, we found significantly lower expression of FUT8 in human OS tissues compared to normal healthy bone tissue. We also found that the expression of FUT8 was negatively correlated with growth of human OS cell lines. In vitro and in vivo assays confirmed that the downregulation of FUT8 in OS cells promoted the progression of OS. Transcriptome sequencing analysis, cell biology experiments and signaling pathway validation provided strong evidence for a possible mechanism. When core fucosylation of TNF receptors was downregulated in OS cell lines due to knockdown of FUT8, the TNF/NF-κB2 signaling pathway was activated and mitochondria-dependent apoptosis was blocked. Together, our findings highlight the important role of FUT8 in the oncogenesis of OS and suggest a new therapeutic strategy for inhibiting OS by changing its glycosylation profile, especially core fucosylation. Materials and methods Clinical samples All OS tissues and normal cancellous bone tissues were collected from 2014 to 2015 from the Department of Orthopaedic Oncology, Shanghai Jiao Tong University Affiliated Sixth People’s Hospital. All patients were pathologically diagnosed with osteosarcoma and underwent primary OS resection. Immediately after excision, the samples were frozen in liquid nitrogen to prevent protein and RNA degradation and then stored at −80 °C. Ethics approval was obtained from the Shanghai Sixth People’s Hospital Ethics Committee (YS-2016-064, 24 February 2016). Cell lines The human OS cell lines MNNG/HOS (short for MNNG in the following description), U2OS, and 143B and human osteoblast cell line hFOB1.19 were acquired from American Type Culture Collection (ATCC, USA). OS cell lines MNNG and 143B were cultured in DMEM (Corning, USA) at a 37 °C atmosphere containing 5% CO[2]. OS cell line U2OS was cultured in RPMI 1640 (Corning, USA) at a 37 °C atmosphere containing 5% CO[2]. Osteoblast cell line hFOB1.19 was cultured in D-MEM/F-12 (Gibco, USA) at a 33.5 °C atmosphere containing 5% CO[2]. All culture mediums were supplemented with 10% fetal bovine serum (Gibco, USA). Stable cell line construction The lentivirus shuttle plasmids containing full-length FUT8 (for overexpressing FUT8) and siRNA against FUT8 (for knockdown FUT8) (shFUT8-2 was selected according to efficiency) were co-transfected into HEK293T cells with lentivirus packing vectors, respectively. After 48 h, the supernatant of HEK293T, which containing lentivirus were collected, purified and performed titer determination. Then, MNNG cells were infected by the collected lentivirus at the MOI = 10.0. The positive cells were selected by puromycin at the concentration of 1.0 μg/ml after 72 h after infection. Finally, the stable cells overexpressing (MNNG-F8) and knock down FUT8 (MNNG-siF8) gene were verified by RT-qPCR and WB at both mRNA and protein levels. All the information of the vector and sequences of full-length FUT8 and siRNA against FUT8 has been provided in the Supplementary Table. [55]1. Total RNA extraction and Real-time Quantitative PCR Trizol reagent (Invitrogen, USA) was used to extract total RNA from clinical samples and cell lines. Reverse transcription was achieved by a RevertAid First Strand cDNA Synthesis Kit (Invitrogen, USA). Revsese transcription quantitative PCR (RT-qPCR) assays were carried out on an ABI Prism 7900HT real-time system (Applied Biosystems, USA) by applicating specifically primers and the SYBR gene PCR master mix (Invitrogen, USA). The 2^−ΔΔCt approach was used to calculate the relative mRNA expression of different genes. All primers are shown in Supplementary Table [56]2. Western blot and reagent RIPA solution (EpiZyme, PRC) containing proteinase inhibitor (Invitrogen, USA) was added to culture cells to gain lysis solution. Then the lysates were centrifuged (13000 g/15 min, 4 °C) and discarded precipitate. Equal amounts of collected total proteins were separated by SDS-PAGE and transferred to a PVDF membrane. Giving the membranes a dip in 5% milk at room temperature (RT) for 1 h. Then, proteins were detected by incubated membranes in primary antibodies solution at 4 °C overnight. After that, HRP-linked anti-IgG antibodies were employed as secondary antibodies. Primary antibodies were shown in Supplementary Table. [57]3. Real Time Cellular Analysis (RTCA) and colony formation assay The cell proliferation ability was measured by RTCA (ACEA Biosciences, USA) according to the manufacturer’s instructions. Particularly, first, 100 μl of culture medium was added to wells and incubation at 37 °C in a cell incubator for 1 h. Then the baseline value was measured by the incubated medium. Finally, the cells were seeded into wells, and make sure that there were 2.5 × 10^3 cells per well. Cell attachment and cell proliferation were continuously recorded for 6 h and 168 h, respectively. Cell index was used to representing cell attachment and cell proliferation [[58]18]. For colony formation ability detection, cells were incubated in a 6-wells plate for 2 weeks at a density of 1 × 10^3 cells per wells. After being fixed with 4% paraformaldehyde (PFA), the cells were stained with crystal violet for 30 min. The colonies which contain over 50 cells were recorded and counted by image J software. Subcutaneous tumor model Four- to five-week-old female nude mice were raised in the Laboratory Animal Research Centre of Shanghai Sixth People’s Hospital. All experiments were approved by the Animal Research Committee of Shanghai Sixth People’s Hospital. For the subcutaneous xenograft tumor model, ten mice were randomly assigned to two groups (N = 5) and anesthetized by intraperitoneal injection of 1% pentobarbital sodium at a dose of 0.01 ml/g. Then, the cells were harvested from cell culture flasks and resuspended in PBS to a final concentration of 5 × 10^6/ml. After that, 200 μl cell suspension containing 1 × 10^6 cells were injected into the nude mouse flank [[59]19]. Researchers blinded to group allocation measured the tumors with callipers until the longest diameter of the largest tumor reached 200 mm, and all tumors were extracted via surgery. The volume of tumors was calculated as the length (mm) × width (mm)^2. Cell apoptosis analysis Cells were cultured in a 6-wells plate. After being treated with 0.05 μM staurosporine (STS) (Sigma, USA) for 24 h [[60]20], cells were washed with a cold Phosphate Buffered Saline (PBS) solution followed by digested with trypsin. Digested cells were processed by the Annexin V-PI kit (Beyotime, PRC) according to the manufacturer’s instructions. The result was measured by flow cytometry. Mitochondrial membrane potential assay Cells were cultured in a 6-wells plate. After being treated with 0.05 μM STS for 24 h, cells were washed with a cold PBS solution. Cells were processed by mitochondrial membrane potential assay kit (Beyotime, PRC) according to the manufacturer’s instructions. The images were taken by DM6B fluorescence microscope (Leica, BRD). RNA-seq and analysis Total RNA was extracted from MNNG-siF8, MNNG-F8, and control cell lines by Trizol reagent (Invitrogen, USA). The RNA-Seq libraries were synthesized by the TruSeq™ RNA Sample Preparation Kit (Illumina, USA) following standard Illumina guidelines. After purification, the quantification and validation of the libraries were performed by Qubit® 2.0 Fluorometer (Life Technologies, USA) and Agilent 2100 bioanalyzer (Agilent Technologies, USA), and finally, the library was sequenced by Illumina NovaSeq 6000 (Illumina, USA). The library construction and sequencing were performed by Sinotech Genomics Co., Ltd (Shanghai, PRC). The selection criteria of differential expressed genes (DEG) was a less than five percentage false discovery rate (FDR) and changed expression higher than 1.5 or lower than 0.67 folds. All cell lines were tested three times. The raw RNA-seq data was uploaded to NCBI SRA database. The SRA accession number: PRJNA772908. Cell nucleus isolation For detection of the nuclear translocation of p52, the nucleus was isolated for western blot analysis. Cells were cultured in a 100 mm dish, and nuclei were extracted with a Nuclei EZ Prep kit (Sigma–Aldrich, USA). Briefly, the cells were cultured in a 100 mm diameter cell culture dish. When the cells grew to a density of 80–90% (approximately 1 × 10^7 cells), they were harvested and lysed by adding 4 ml of ice-cold Nuclei EZ lysis buffer to the cell dish. Then, the lysates were centrifuged at 500 g for 5 min at 4 °C, and the nuclei were resuspended and washed in 4 ml of ice-cold Nuclei EZ lysis buffer. Centrifugation was performed again at 500 g for 5 min at 4 °C, and the nuclei pellet was resuspended in 200 µl of ice-cold Nuclei EZ storage buffer and stored at −70 °C. Then, the expression of p52 was detected in both the nucleus and cytoplasm. PCNA and GAPDH were used as internal references.