The ubiquitin-like protein FAT10 enhances the autophagy-mediated degradation of ZO-1 by stabilizing ATG3 to promote the lung metastasis of colon cancer Authors: Zijing Li, Hengqing Zhu, Dongnian Du, Dandan Zhang, Xuzhe Yu, Fei Xie, Miao Ye, Qing Wang, Rong Wan, Jin Ge, Jianghua Shao Published: 2026 ABSTRACT PurposeThe ubiquitin-like modifier HLA-F adjacent transcript 10 (FAT10) directs substrates to the 26 S proteasome, but its role in autophagic protein degradation remains unclear.MethodsFAT10 expression was analyzed by qRT-PCR and western blotting in colon cancer (CC) tissues and cells. Bioinformatics revealed FAT10-associated biological processes in CC. In vitro and in vivo assays examined CC cell invasion and metastasis. Autophagy was assessed by western blotting, mRFP-GFP-LC3 reporter (tfLC3), and electron microscopy. In vitro ubiquitination assays measured ubiquitination of zona occludens 1 (ZO-1) and autophagy-related gene 3 (ATG3). liquid chromatography‒tandem mass spectrometry (LC-MS/MS) identified FAT10-interacting proteins. Co-IP and GST pull-down confirmed FAT10–ATG3 binding.ResultsFAT10 was upregulated in CC tissues and cells. Bioinformatic analyses revealed that FAT10 expression was correlated with extracellular matrix binding and cell migration in CC. FAT10 overexpression enhanced CC invasion and metastasis. FAT10 promoted autophagic degradation of ZO-1, a key component of tight junctions, facilitating invasion and migration. Mechanistically, FAT10 stabilized ATG3 by competing with ubiquitin for binding, inhibiting ATG3 ubiquitination and activating autophagy, leading to ZO-1 degradation. The ATG3 inhibitor compound 189 reversed FAT10-induced ZO-1 loss and suppressed lung metastasis.ConclusionOur findings demonstrate a novel pathway by which FAT10 activates autophagy by stabilizing ATG3, leading to ZO-1 degradation and promoting CC lung metastasis. This study provides new insights into the role of FAT10 in regulating protein homeostasis and offers a potential strategy for targeting the FAT10-ATG3-autophagy axis in the treatment of CC metastasis.Supplementary InformationThe online version contains supplementary material available at 10.1007/s13402-026-01220-7. Introduction Colon cancer (CC), the fifth most commonly diagnosed cancer worldwide, accounts for approximately 1.2 million new cases and results in nearly 600,000 deaths annually [1]. Distant metastasis is the main cause of death in CC patients, and the metastatic foci often invade the liver and lung [2]. Currently, the clinical outcomes of patients with liver metastases significantly improve after active local treatment (such as surgery) combined with systemic treatment [3]. However, when CC patients develop lung metastasis, they often lose the opportunity for radical local treatment, leading to a disease that is difficult to control, continuous progression and a poor prognosis [4]. Therefore, an in-depth exploration of the mechanism of lung metastasis of CC can provide a basis for developing targeted therapeutic strategies, and improving the quality of life and prolonging the survival of patients. Human leukocyte antigen F locus adjacent transcript 10 (FAT10), a ubiquitin-like protein, has two core structures similar to ubiquitin; and is involved in various cellular processes including the immune response, cell cycle, apoptosis, and chromatin instability [5–9]. In the past decade, the role of FAT10 in promoting cancer progression has received widespread attention. Research has confirmed that high expression of FAT10 promotes the invasion and migration of breast cancer cell through the stabilization of ZEB2 [10]. FAT10 promotes migration and invasion of non-small-cell lung cancer cell [11]. FAT10 promotes the invasion and metastasis of hepatocellular carcinoma by modifying β-catenin degradation [12]. In addition, silencing FAT10 could inhibit metastasis of osteosarcoma [13]. Although, studies have reported reporting that FAT10 can promote lung metastasis in patients with CC [14], the specific mechanism remains unclear. Tight junctions (TJs)are a complex composed of multiple proteins which located at the top of epithelial and endothelial cells and TJs mainly function as barriers to maintain cellular connections [15]. In recent years, studies have confirmed that TJs play important roles in tumor invasion and migration [16]. Zona occludens 1 (ZO-1), the first discovered tight junction protein, acts mainly as a scaffold to recruit other tight junction proteins, thus promoting the formation of TJs [17–20]. Studies have shown that ZO-1 expression is decreased in various tumors, including CC, and that decreased ZO-1 promotes the invasion and migration of cancers and lung metastasis [21–28]. As a cellular degradation pathway, autophagy has been shown to regulate ZO-1 protein expression, but whether autophagy promotes or inhibits ZO-1 degradation in different cells remains contradictory [29–34]. However, whether autophagy promotes or inhibits the degradation of ZO-1 remains unclear in CC cells. Therefore, clarifying the regulatory effect of autophagy on ZO-1 and its mechanism in CC cells will provide new ideas for improving the comprehensive treatment strategies for CC metastasis. In this study, our results revealed that FAT10 promotes the lung metastasis of CC. Mechanically, FAT10 overexpression antagonizes the ubiquitination of autophagy-related gene 3 (ATG3) to stabilize ATG3 expression and promote the autophagic degradation of ZO-1. Finally, we found that compound 189, an autophagy inhibitor, can block the increased autophagic degradation of ZO-1 mediated by FAT10 overexpression and thereby inhibit the lung metastasis of CC. Materials and methods Datasets and data preprocessing The GEO dataset was downloaded from the GEO database (http://www.ncbi.nlm.nih.gov/geo). Relevant data or figures from The Cancer Genome Atlas Program database (TCGA; http://cancergenome.nih.gov) were downloaded and analyzed using Gene Expression Profiling Interacting Analysis (GEPIA; http://gepia.cancer-pku.cn). GO analysis was performed using the GO enrichment analysis at the Gene Ontology Consortium homepage (http://geneontology.org/). Gene set enrichment analysis (GSEA) was performed using GSEA 3.0 (http://www.broadinstitute.org/gsea/). Data analysis was performed using R software (http://www.r-project.org/, version 3.1.3). Gene set variation analysis (GSVA) was performed using the GSVA package (v1.36.3). Patients and tumor specimens All procedures conformed to the principles set forth by the Declaration of Helsinki. Colon cancer tissue and normal colon tissue samples were obtained from 50 patients with CC who were admitted to the Department of General Surgery, the Second Affiliated Hospital of Nanchang University. Fresh tumor tissues were collected at the time of surgery, validated by pathological diagnosis, immediately snap-frozen in liquid nitrogen, and stored at − 80 °C for further analysis. The protocols for the collection of human tissue in the study were approved by the Ethics and Research Committees of the Second Affiliated Hospital of Nanchang University. Cell cultures The following cell lines were purchased from the National Collection of Authenticated Cell Cultures (Shanghai, China): normal human colon cells (NCM460); the human CC lines HCT116, COLO 205, LoVo, SW620, SW480, and HT-29; and embryonic kidney HEK293T cells. All the cell lines were authenticated by the Cell Bank using short tandem repeat profiling, and a mycoplasma detection kit (Shanghai GeneChem Co., Ltd.) was used to confirm the absence of mycoplasma contamination. All the cell lines were routinely maintained in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, Grand Island, NY, USA) at 37 °C in a humidified atmosphere containing 5% CO2. Plasmids and lentivirus Lentiviral vectors for silencing and overexpressing FAT10 (shFAT10 and Flag-FAT10), ATG3-overexpressing and shATG3-expressing plasmids, and ZO-1-overexpressing and shZO-1-expressing plasmids were purchased from GeneChem (Shanghai, China). The target sites of shRNA are listed in Table S11. Lentiviral vectors for silencing and overexpressing FAT10 (shFAT10 and Flag-FAT10), ATG3-overexpressing and shATG3-expressing plasmids, and ZO-1-overexpressing and shZO-1-expressing plasmids were purchased from GeneChem (Shanghai, China). The target sites of shRNA are listed in Table S1. Generation of stably transfected cells One day before transfection, 1 × 10⁵ SW620 and HCT116 cells were seeded into 6-well plates. Cells were seeded to 70% confluence 12 h before infection, and the cell culture medium was replaced with medium containing the indicated lentivirus. After 12 h of infection, the medium was replaced with fresh medium, and 48 h later, the infected cells were selected using 10 µg/mL puromycin (InvivoGen, San Diego, CA, USA). To select stably transfected cells, the cells were cultured in high-glucose DMEM supplemented with 5 µg/mL puromycin for 4 w. Clones demonstrating puromycin resistance were selected and expanded. Quantitative real-time PCR (qRT-PCR) Total RNA was extracted from cells after treatment using TRIzol (Invitrogen, Carlsbad, CA, USA) and purified using TB Green Premix Ex Taq II (Takara Bio Inc. Kusatsu, Shiga, Japan)). cDNA was reverse transcribed using a PrimeScript RT reagent Kit (Takara Bio, Inc., Kusatsu, Shiga, Japan). The experiments were performed according to the instructions of the kit. The sequences of primers used in this experiment are summarized in Supplementary Table 1. Label-free liquid chromatography‒tandem mass spectrometry (LC‒MS/MS) analysis Fresh SW620 cell samples from the negative control and FAT10-knockdown groups were collected, placed in 15 mL centrifuge tubes, and then immediately stored on dry ice. Protein expression was analyzed using label-free LC‒MS/MS (Personalbio, Shanghai, China). Western blot For western blotting, total protein was extracted from cells or tumor specimens, and equal amounts of protein were separated via sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS‒PAGE). Proteins were transferred to a nitrocellulose filter (Millipore, Bedford, MA, USA) by electroblotting. The membranes were then incubated with primary antibodies according to the manufacturer’s instructions. After a night of incubation at an appropriate concentration of 4 °C, the membranes were then washed with 1× TBST three times and incubated with the appropriate secondary antibodies. Afterward, specific antibody interactions were visualized by enhanced chemiluminescence (Thermo, Waltham, MA, USA). The intensity of each band was measured using Quantity-One software (Bio-Rad, Hercules, CA, USA). Western blot analysis was performed using the following antibodies: anti-FAT10 (Proteintech, 13003-2-AP, 1:500), anti-ATG3 (Proteintech, 11262-2-AP, 1:1000), anti-ZO-1 (Proteintech, 21773-1-AP, 1:1000), anti-LC3B (Abcam, ab192890, 1:1000), anti-ubiquitin (Santa Cruz, sc-8017, 1:200), anti-p62/SQSTM1 (Proteintech, 18420-1-AP, 1:1000), anti-K48-linkage Specific Polyubiquitin(cell signaling technology, 12805), anti-K63-linkage Specific Polyubiquitin(cell signaling technology, 5621)and tubulin (Proteintech, 66031-1-Ig, 1:1000). Co-IP The cells were lysed in cell lysis buffer for western blotting and IP (Beyotime Biotechnology, 20 mM Tris, pH 8.0, 150 mM NaCl, 5 mM EDTA, and 1% Triton X-100) and centrifuged at 12,000 rpm for 15 min to remove debris. In total, 60 µL of cleared lysates was used as the input group, and the remaining lysates were subjected to immunoprecipitation with antibodies. Rolling was performed at 4 °C overnight. Afterward, protein A/G PLUSAgarose (Santa Cruz Biotechnology Cat# sc-2003; RRID: AB_10201400) was added, and the plate was allowed to rest for 6 h. Washing was performed three times with wash buffer. Afterward, the appropriate Bis–Tris gel was used for SDS‒PAGE and western blot analysis. Glutathione transferase (GST) pull-down assay A Pierce™ GST Protein Interaction Pull-Down Kit (Thermo Scientific) was used to perform the GST pull-down. The purified proteins were incubated with glutathione agarose beads overnight at 4 °C in buffer. After they were washed at least three times with the same buffer, the proteins were eluted with 4X SDS‒PAGE loading buffer (Takara) and boiled. After separation by SDS‒PAGE, Western blotting was performed as described above. Immunohistochemical (IHC) staining The tissues were fixed in 10% formalin, embedded in paraffin, and sectioned. The tissue sections were then dewaxed. Some of the dewaxed sections were stained with hematoxylin‒eosin (H&E) to detect morphological changes, and some were blocked with 5% goat serum at room temperature for 30 min after antigen repair. The sections were subsequently incubated overnight at 4 °C with the corresponding primary antibody. The sections were labeled with an EnVision HRP kit (Solarbio) for 30 min at room temperature, incubated with DAB substrate solution (Solarbio), and counterstained with Mayer’s hematoxylin (Solarbio). All the sections were observed and photographed with a light microscope and a DP70 CCD system (Olympus Corp.). The staining intensity and percentage of positive cells were scored semiquantitatively by two pathologists who were unaware of the clinical parameters. Transwell assays Migration and invasion assays were performed using a Transwell system (8-µm pore size; BD Biosciences, NJ, USA). In brief, 5 × 104 cells resuspended in DMEM were seeded into the upper chambers, which were either uncoated (for the migration assay) or coated with Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) (for the invasion assay). The lower chambers were filled with medium containing 15% FBS. After incubation for the indicated periods, the upper chambers were removed, and the cells on the lower surface of the membrane were fixed, stained with 0.1% crystal violet, and photographed. Cells were manually counted in five random visual fields per chamber. Electron microscopy Cells were fixed with 2.5% glutaraldehyde (Solarbio, Beijing, China) for 4 h at 4 °C, washed with phosphate-buffered saline (PBS), and postfixed with 1% OsO4 buffer for 2 h at 4 °C. The cells were then washed and dehydrated in a graded series of ethanol solutions and embedded in Epon812 epoxy resin. Ultrathin (90 nm) sections were collected on copper grids, double-stained with 1% uranyl acetate and 0.2% lead citrate, and examined using a JEOL-1200EX transmission electron microscope (Beijing, China). In vivo xenograft mouse study Nude mice were injected with 1 × 106 colon cancer cells into the lateral vein of the tail. The mice were sacrificed after 30 days. Lung tissues were collected for metastatic foci evaluation and a standard histopathological study. For the ATG12-ATG3 inhibitor 1 (MedChemEpress, 333351-38-5) treatment, the same number of CC cells were injected into the tail veins of the nude mice. ATG12-ATG3 inhibitor 1 (80 mg/kg/per day) was initiated on the same day. The mice were sacrificed after 30 days. Lung tissues were collected for metastatic foci evaluation and a standard histopathological study. All animal experiments were approved by the animal research committee in the Laboratory Animal Science Center of Nanchang University. Statistical analysis All the data are expressed as the means ± standard deviations. Differences between two groups were analyzed by Student’s t test, and differences among more than two groups were evaluated by one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test. The Kaplan‒Meier method was used to generate survival curves, and the log-rank test was used to determine the significance of differences in survival. P values less than 0.05 were considered to indicate statistical significance. Statistical analyses were conducted using GraphPad Prism software 9.0. Results FAT10 is highly expressed in colon cancer An analysis of CC samples from GSE223119 in the GEO database showed that FAT10 was significantly overexpressed in CC tissues compared with normal tissues, and the paired analysis also revealed higher expression of FAT10 in CC tissues than in corresponding adjacent tissues (Fig. 1A). An analysis of data from TCGA database confirmed that FAT10 was significantly more highly expressed in CC than in normal control tissues (Fig. 1B-D). Furthermore, an examination of the mRNA expression of FAT10 in 50 pairs of CC tissues using quantitative reverse transcription polymerase chain reaction (qRT‒PCR) revealed that FAT10 was upregulated in CC tissues, and western blotting showed that the FAT10 protein level was increased in CC tissues compared with the corresponding adjacent tissues (Fig. 1E-G). Immunohistochemical (IHC) also showed that the expression of the FAT10 protein was significantly increased in CC tissues (Fig. 1H). Moreover, qRT‒PCR and western blotting revealed that FAT10 mRNA and protein levels were higher in the five CC cell lines tested (HT-29, SW620, LoVo, COLO 205, HCT116 and SW480) than in normal colon cells (NCM460) (Fig. 1I). These results indicated that FAT10 is highly expressed in both CC tissues and cell lines. Fig. 1FAT10 is overexpressed in colon cancer tissues and cells. (A) Expression of FAT10 in CC in the GEO cohort (GSE223119). (B) Volcano plot showing the marker mRNAs of colon cancer cells in the TCGA cohort. C, D. Comparison of FAT10 expression between normal tissues and tumor tissues in the TCGA-COAD cohort. E-G. qRT–PCR and western blot analysis of FAT10 mRNA and protein levels in CC tissues and adjacent nontumor tissues (n = 50). The mRNA levels of FAT10 were normalized to those of GAPDH. H. Representative IHC staining and quantitative analysis of FAT10 protein expression in 50 pairs of CC tissues and adjacent non-tumor tissues (n = 50). Scale bars = 100 μm. I. Western blot analysis and quantitative analysis of FAT10 protein expression in normal human colon NCM460 cells and six tested CC cell lines (HCT116, COLO 205, LoVo, SW620, SW480, and HT-29). The data are presented as the means ± SDs. *p < 0.05 and ***p < 0.001 FAT10 is overexpressed in colon cancer tissues and cells. (A) Expression of FAT10 in CC in the GEO cohort (GSE223119). (B) Volcano plot showing the marker mRNAs of colon cancer cells in the TCGA cohort. C, D. Comparison of FAT10 expression between normal tissues and tumor tissues in the TCGA-COAD cohort. E-G. qRT–PCR and western blot analysis of FAT10 mRNA and protein levels in CC tissues and adjacent nontumor tissues (n = 50). The mRNA levels of FAT10 were normalized to those of GAPDH. H. Representative IHC staining and quantitative analysis of FAT10 protein expression in 50 pairs of CC tissues and adjacent non-tumor tissues (n = 50). Scale bars = 100 μm. I. Western blot analysis and quantitative analysis of FAT10 protein expression in normal human colon NCM460 cells and six tested CC cell lines (HCT116, COLO 205, LoVo, SW620, SW480, and HT-29). The data are presented as the means ± SDs. *p < 0.05 and ***p < 0.001 FAT10 overexpression promotes the lung metastasis of CC cells The results of GSEA and GO enrichment analyses of data from colon cancer patients from TCGA database indicated close relationships between FAT10 expression and the extracellular matrix binding and cell migration (Fig. 2A, B). Furthermore, our results showed that FAT10 promoted the invasion and migration of SW620 and HCT116 cells. qRT‒PCR and western blotting revealed that mRNA and protein expression were increased in FAT10-overexpressing CC cells compared with the control group. (Fig. 2C, Fig. S1A). Moreover, transwell assays and real-time cellular analysis (RTCA) revealed increases in the invasion and migration of CC cells overexpressing FAT10 (Fig. 2D, E, Fig. S1B, C). Conversely, when FAT10 was knocked down, the invasion and migration of CC cells were inhibited, as confirmed by transwell assays and RTCA (Fig. 2F-H, Fig. S1D-F) These results indicated that FAT10 could increase the invasion and migration of CC cells in vitro. Fig. 2FAT10 overexpression promoted the invasion and metastasis of colon cancer cells in vitro and in vivo. A, B. GSEA and GO was performed on genes related to FAT10 expression in the TCGA-COAD cohort. The bar chart (A) and bubble chart (B) indicate enrichment pathways associated with FAT10. C. Western blotting and qRT‒PCR were used to detect the expression of FAT10 in HCT116 cells from different treatment groups(n = 4). The mRNA levels of FAT10 were normalized to those of GAPDH. D, E. Transwell (D) and RTCA (E) assays were used to detect the invasion and migration capacities of HCT116 cells. Scale bars = 150 μm. F. Western blotting and qRT‒PCR were used to detect the expression of FAT10 in SW620 cells from different treatment groups(n = 4). The mRNA levels of FAT10 were normalized to those of GAPDH. G, H. Transwell (G) and RTCA (H) assays were used to measure the invasion and migration capacities of SW620 cells. Scale bars = 150 μm. I, J. Representative bioluminescent images and quantification analysis (I) of lung tumors in nude mice intravenously injected with FAT10-overexpressing HCT116 cells and H&E-stained images and quantification analysis (J) of metastatic nodules in the lungs (n = 10 mice per group). K, L. Representative bioluminescent images and quantification analysis of lung tumors (K) in nude mice intravenously injected with FAT10 knockdown SW620 cells, H&E-stained images and quantification analysis (L) of metastatic nodules in the lungs. (n = 10 mice per group). Scale bars = 100 μm. Data represent the means ± SDs. **p < 0.01 and ***p < 0.001 FAT10 overexpression promoted the invasion and metastasis of colon cancer cells in vitro and in vivo. A, B. GSEA and GO was performed on genes related to FAT10 expression in the TCGA-COAD cohort. The bar chart (A) and bubble chart (B) indicate enrichment pathways associated with FAT10. C. Western blotting and qRT‒PCR were used to detect the expression of FAT10 in HCT116 cells from different treatment groups(n = 4). The mRNA levels of FAT10 were normalized to those of GAPDH. D, E. Transwell (D) and RTCA (E) assays were used to detect the invasion and migration capacities of HCT116 cells. Scale bars = 150 μm. F. Western blotting and qRT‒PCR were used to detect the expression of FAT10 in SW620 cells from different treatment groups(n = 4). The mRNA levels of FAT10 were normalized to those of GAPDH. G, H. Transwell (G) and RTCA (H) assays were used to measure the invasion and migration capacities of SW620 cells. Scale bars = 150 μm. I, J. Representative bioluminescent images and quantification analysis (I) of lung tumors in nude mice intravenously injected with FAT10-overexpressing HCT116 cells and H&E-stained images and quantification analysis (J) of metastatic nodules in the lungs (n = 10 mice per group). K, L. Representative bioluminescent images and quantification analysis of lung tumors (K) in nude mice intravenously injected with FAT10 knockdown SW620 cells, H&E-stained images and quantification analysis (L) of metastatic nodules in the lungs. (n = 10 mice per group). Scale bars = 100 μm. Data represent the means ± SDs. **p < 0.01 and ***p < 0.001 Furthermore, we wanted to investigate whether FAT10 can promote the lung metastasis of CC. Lung metastasis models were established in BALB/c nude mice via a tail vein injection of Flag-vector-HCT116 and Flag-FAT10-HCT116 cells. After four weeks, the images acquired with an in vivo imaging system (IVIS) showed a significant increase in the fluorescence intensity in the Flag-FAT10-HCT116 group compared with that in the Flag-vector-HCT116 group (Fig. 2I). Hematoxylin and eosin (H&E) staining of lung tissues confirmed a significant increase in the number of metastatic lung nodules in the Flag-FAT10-HCT116 group (Fig. 2J). Conversely, FAT10 knockdown in SW620 cells inhibited lung metastasis (Fig. 2K, L). Together, our results confirmed that FAT10 promoted the invasion and migration of CC cells and promoted lung metastasis of CC. FAT10 promotes CC cell invasion and migration by decreasing the expression of the ZO-1 protein Next, we explored the potential mechanisms by which FAT10 promotes CC cell invasion and migration. We knocked down FAT10 in SW620 cells and detected differentially expressed proteins in shNC-SW620 and shFAT0-SW620 cells using LC‒MS. Compare to shNC group, there are 726 down-regulated proteins and 612 up-regulated proteins in shFAT10 SW620 cells (Fig. 3A). GSVA analysis of the differentially expressed proteins showed that FAT10 was significantly negatively related to apical junctions (Fig. 3B). Furthermore, among the genes related to apical junctions, ZO-1 exhibited the most significant change in expression (Fig. 3C). Therefore, we speculated that FAT10 may promote the invasion and migration of CC cells by regulating ZO-1 expression. To confirm this hypothesis, we first investigated the association between FAT10 and ZO-1 in CC cells and tissues. Compared with the level in normal NCM460 colon cells, the expression of ZO-1 in CC cells (HT-29, SW620, LoVo, COLO 205, HCT116 and SW480) was decreased, and the expression of FAT10 was significantly negatively correlated with ZO-1 expression (Fig. 3D). The results of IHC staining showed that ZO-1 was expressed at lower levels in CC tissues than in adjacent tissues and that the expression of ZO-1 was significantly negatively correlated with FAT10 expression in CC tissues. (Fig. 3E). These results indicated a negative correlation between the expression of the FAT10 and ZO-1 proteins in CC cells and tissues. Fig. 3FAT10 promotes invasion and migration of CC cells by downregulating ZO-1 expression. (A) Mass spectrometry was used to detect protein expression profiles in the indicated two groups of cells (n = 3). (B) Based on the mass spectrum data, GSVA enrichment analysis revealed pathways closely related to FAT10. (C) Differential gene expression analysis of 13 genes in the “apical junction” signature. Genes are ranked on the basis of fold change (-log10 scaled). The blue and red colors represent low and high expression, respectively. (D) Western blot analysis of FAT10 and ZO-1 protein expression in normal human colon NCM460 cells and six tested CC cell lines (HCT116, COLO 205, LoVo, SW620, SW480, and HT-29). Scatter plots showing a negative correlation between the protein expression of FAT10 and that of ZO-1 in CC cells. (E) Representative IHC staining of FAT10 and ZO-1 protein expression in 50 pairs of CC tissues and adjacent nontumor tissues (n = 50). Scale bars = 100 μm. Scatter plots showing a negative correlation between the protein expression of FAT10 and that of ZO-1 in CC cells. (F) HCT116 cells were transfected with a FAT10 overexpression lentivirus, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10 and ZO-1 were normalized to those of GAPDH. (G) SW620 cells were transfected with shRNA against FAT10, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10 and ZO-1 were normalized to those of GAPDH. (H) Western blotting was used to analyze the protein levels of FAT10 and ZO-1 in HCT116 cells subjected to the indicated treatments. I, J. Transwell (I) and RTCA assays (J) were used to determine the invasion and migration capacities of HCT116 cells. Scale bars = 150 μm. K. Western blotting was used to analyze the protein levels of FAT10 and ZO-1 in SW620 cells subjected to the indicated treatments. L, M. Transwell (L) and RTCA assays (M) were used to measure the invasion and migration capacities of HCT116 cells. Scale bars = 150 μm. The data are presented as the means ± SDs. ns, p > 0.05; **p < 0.01; ***p < 0.001 FAT10 promotes invasion and migration of CC cells by downregulating ZO-1 expression. (A) Mass spectrometry was used to detect protein expression profiles in the indicated two groups of cells (n = 3). (B) Based on the mass spectrum data, GSVA enrichment analysis revealed pathways closely related to FAT10. (C) Differential gene expression analysis of 13 genes in the “apical junction” signature. Genes are ranked on the basis of fold change (-log10 scaled). The blue and red colors represent low and high expression, respectively. (D) Western blot analysis of FAT10 and ZO-1 protein expression in normal human colon NCM460 cells and six tested CC cell lines (HCT116, COLO 205, LoVo, SW620, SW480, and HT-29). Scatter plots showing a negative correlation between the protein expression of FAT10 and that of ZO-1 in CC cells. (E) Representative IHC staining of FAT10 and ZO-1 protein expression in 50 pairs of CC tissues and adjacent nontumor tissues (n = 50). Scale bars = 100 μm. Scatter plots showing a negative correlation between the protein expression of FAT10 and that of ZO-1 in CC cells. (F) HCT116 cells were transfected with a FAT10 overexpression lentivirus, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10 and ZO-1 were normalized to those of GAPDH. (G) SW620 cells were transfected with shRNA against FAT10, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10 and ZO-1 were normalized to those of GAPDH. (H) Western blotting was used to analyze the protein levels of FAT10 and ZO-1 in HCT116 cells subjected to the indicated treatments. I, J. Transwell (I) and RTCA assays (J) were used to determine the invasion and migration capacities of HCT116 cells. Scale bars = 150 μm. K. Western blotting was used to analyze the protein levels of FAT10 and ZO-1 in SW620 cells subjected to the indicated treatments. L, M. Transwell (L) and RTCA assays (M) were used to measure the invasion and migration capacities of HCT116 cells. Scale bars = 150 μm. The data are presented as the means ± SDs. ns, p > 0.05; **p < 0.01; ***p < 0.001 Furthermore, we altered the expression of FAT10 in SW620 and HCT116 cells and detected the expression of ZO-1. qRT‒PCR revealed that FAT10 overexpression did not affect ZO-1 mRNA expression, but the western blotting results showed that the upregulation of FAT10 expression decreased the expression of the ZO-1 protein in HCT116 and SW620 cells (Fig. 3F, Fig. S2A). In addition, qRT‒PCR and western blotting results revealed that FAT10 knockdown did not affect ZO-1 mRNA expression but did increase the expression of the ZO-1 protein in SW620 and HCT116 cells (Fig. 3G, Fig. S2B). Furthermore, rescue experiments indicated that FAT10 overexpression decreased ZO-1 expression, whereas ZO-1 upregulation attenuated the decrease in ZO-1 expression caused by FAT10 overexpression in HCT116 cells (Fig. 3H). transwell and RTCA assays showed that FAT10 overexpression significantly increased the invasion and migration of HCT116 cells, whereas ZO-1 upregulation reversed the increases in migration and invasion caused by FAT10 overexpression (Fig. 3I, J). Conversely, FAT10 knockdown upregulated ZO-1 expression and inhibited the invasion and migration of SW620 cells, whereas the downregulation of ZO-1 reversed these changes (Fig. 3K-M). Overall, these results indicated that FAT10 overexpression inhibited ZO-1 protein expression to promote the invasion and migration of CC cells. FAT10 promotes ZO-1 degradation through the autophagy in CC cells Next, we explored the specific mechanism by which FAT10 regulates the expression of the ZO-1 protein in CC cells. Co-IP showed that FAT10 and ZO-1 did not directly bind to each other in HCT116 and SW620 cells (Fig. S3A, B). Studies have indicated that ZO-1 can be degraded by both the ubiquitin‒proteasome system (UPS) and autophagy pathways [28, 30]. Therefore, we wanted to investigate whether FAT10 affected ZO-1 protein expression through the UPS or autophagy. Our results first revealed that ZO-1 can be degraded by the UPS in CC cells. Western blotting showed that when the proteasome was blocked with MG132 in HCT116 cells, the expression of the ZO-1 protein gradually increased (Fig. S3C). Ubiquitination experiments revealed that increase in ubiquitin (Ub) levels increased the level of ZO-1 ubiquitination and decreased ZO-1 protein expression in HCT116 cells (Fig. S3D-E). Further results revealed that FAT10 didn’t affect the degradation of ZO-1 by the UPS in HCT116 cells (Fig. S3F). In addition, our results confirmed that ZO-1 was degraded by autophagy but not by chaperone-mediated autophagy (CMA). We used the autophagy agonist EBSS, the autophagy antagonist Chloroquine (CQ) and Bafilomycin A1(BafA1) to treat SW620 and HCT116 cells, and western blotting results showed that EBSS treatment significantly downregulated ZO-1 protein expression, whereas CQ and BafA1 treatment upregulated ZO-1 protein expression in CC cells (Fig. 4A, Fig. S3G, H). Furthermore, treatment with the autophagy-specific agonist rapamycin promoted the degradation of ZO-1, whereas treatment with the specific CMA agonist AR7 had no significant effect on CC cells (Fig. 4B, Fig. S3H). To further confirm autophagic degradation of ZO-1, Co-IP and confocal microscopy of immunofluorescence staining were used to verify the combination of ZO-1 and LC3 in HCT116 cells. The results of CO-IP showed that ZO-1 interacted with LC3 in HCT116 cells (Fig. S3J). Confocal microscopy of immunofluorescence show shat the interaction between ZO-1 and LC3 was significantly increased by EBSS treatment; inversely, he interaction between ZO-1 and LC3 was significantly inhibited by CQ treatment s (Fig. S3K, L). Fig. 4FAT10 promotes autophagic degradation of ZO-1. (A) SW620 cells were treated with EBSS and for the indicated times, and the cell lysates were subjected to western blotting. (B) SW620 cells were treated with rapamycin and AR7 for the indicated times, and the cell lysates were subjected to western blotting. (C) Western blotting analysis of FAT10 and LC3B-II protein expression in HCT116 cells transfected with Flag-vector and Flag-FAT10. The LC3BII: I levels were normalized to those of the Flag-vector. (D) Immunofluorescence images of HCT116 cells transfected with Flag-vector/Flag-FAT10 and GFP-RFP-LC3. (E) Representative electron micrographs of autophagic vesicles in HCT116 cells transfected with Flag-vector and Flag-FAT10. (F) Western blot analysis of FAT10 and LC3-II protein expression in SW620 cells transfected with shNC and shFAT10. The LC3BII: I levels were normalized to those of tubulin. (G) Immunofluorescence images of SW620 cells transfected with shNC/shFAT10 and GFP-RFP-LC3. (H) Representative electron micrographs of autophagic vesicles in SW620 cells transfected with shNC and shFAT10. (I) Determination of the protein levels of FAT10, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to tubulin). (J) Determination of the protein levels of FAT10, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to tubulin). K. Determination of the protein levels of FAT10, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to tubulin). The data are presented as the means ± SDs. **p < 0.01; ***p < 0.001 FAT10 promotes autophagic degradation of ZO-1. (A) SW620 cells were treated with EBSS and for the indicated times, and the cell lysates were subjected to western blotting. (B) SW620 cells were treated with rapamycin and AR7 for the indicated times, and the cell lysates were subjected to western blotting. (C) Western blotting analysis of FAT10 and LC3B-II protein expression in HCT116 cells transfected with Flag-vector and Flag-FAT10. The LC3BII: I levels were normalized to those of the Flag-vector. (D) Immunofluorescence images of HCT116 cells transfected with Flag-vector/Flag-FAT10 and GFP-RFP-LC3. (E) Representative electron micrographs of autophagic vesicles in HCT116 cells transfected with Flag-vector and Flag-FAT10. (F) Western blot analysis of FAT10 and LC3-II protein expression in SW620 cells transfected with shNC and shFAT10. The LC3BII: I levels were normalized to those of tubulin. (G) Immunofluorescence images of SW620 cells transfected with shNC/shFAT10 and GFP-RFP-LC3. (H) Representative electron micrographs of autophagic vesicles in SW620 cells transfected with shNC and shFAT10. (I) Determination of the protein levels of FAT10, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to tubulin). (J) Determination of the protein levels of FAT10, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to tubulin). K. Determination of the protein levels of FAT10, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to tubulin). The data are presented as the means ± SDs. **p < 0.01; ***p < 0.001 Next, we investigated whether FAT10 affects ZO-1 degradation by influencing autophagy in CC cells. We first examined whether FAT10 affects autophagy in CC cells. Western blotting showed that the LC3B-I/LC3B-II ratio was increased in FAT10 overexpression HCT116 cells (Fig. 4C). The results of the analysis of autophagy flux using an mRFP-GFP-LC3 reporter (tfLC3) showed that autophagic flux was also increased in FAT10 overexpression HCT116 cells (Fig. 4D, Fig. S3M) Transmission electron microscopy (TEM) also revealed that autophagic activity in HCT116 cells was significantly increased following FAT10 overexpression (Fig. 4E). In contrast, FAT10 knockdown significantly inhibited autophagy in SW620 cells, as detected by western blotting, tfLC3 and TEM (Fig. 4F-H, Fig. S3N). These results revealed that FAT10 overexpression can promote autophagy in CC cells. Next, we aim to investigate whether FAT10 can affect the autophagic degradation of ZO-1 in CC cells. Western blotting showed that the ZO-1 and p62 protein expression levels decreased, and the transformation of LC3B-II increased in FAT10-overexpression CC cells, whereas treatment with CQ or 3-MA reversed these changes induced by FAT10 overexpression (Fig. 4I, J). Moreover, autophagy agonists reversed increased expression of ZO-1 and p62 caused by FAT10 knockdown in CC cells. Western blotting showed that FAT10 knockdown decreased the transformation of LC3B-II and increased ZO-1 and p62 expression, but rapamycin reversed the decreases in autophagy levels and the increase in ZO-1 expression caused by FAT10 knockdown in CC cells (Fig. 4K). Together, our data confirmed that FAT10 promotes ZO-1 degradation via the autophagy in CC cells. FAT10 promotes the autophagic degradation of ZO-1 by regulating ATG3 in CC cells Next, we aimed to investigate why FAT10 can promote the autophagic degradation of ZO-1 in CC cells. IP‒LC‒MS was performed in SW620 cells, and the results showed that ATG3, a key autophagy-related gene [35], is a binding protein of FAT10 (Fig. 5A). Molecular docking analyses with the ZDOCK server (http://zdock.umassmed.edu/) revealed multiple possible binding sites for FAT10 and ATG3 (Fig. 5B). Furthermore, we detected the expression of the FAT10, ATG3 and ZO-1 proteins in 50 postoperative CC tissues and the corresponding adjacent tissues from clinical patients. Western blotting showed that compared with the adjacent tissues, the protein expression of FAT10 and ATG3 was increased, while the protein expression of ZO-1 was decreased in CC tissues (Fig. 5C). Furthermore, the correlation analysis revealed that ZO-1 expression was negatively correlated with FAT10 and ATG3 expression, whereas ATG3 expression was positively correlated with FAT10 expression in CC tissues (Fig. 5D). In addition, qRT‒PCR and western blotting results revealed that upregulating FAT10 expression could increase the expression of ATG3 and decrease the expression of ZO-1 without altering the mRNA levels of ATG3 and ZO-1 in HCT116 and SW620 cells (Fig. 5E, Fig. S4A). Conversely, qRT‒PCR and western blotting results showed that FAT10 knockdown could inhibit the expression of ATG3 and increase the expression of ZO-1 without altering the mRNA levels of ATG3 and ZO-1 in CC cells (Fig. 5G, Fig. S4B). Based on these results, we speculated FAT10 can regulate ATG3 expression and thereby affect the autophagic degradation of ZO-1. Fig. 5FAT10 inhibits the expression of ZO-1 by increasing ATG3 expression to promote the invasion of colon cancer cells. (A) Table listing the FAT10-interacting proteins identified by coimmunoprecipitation mass spectrometry. (B) Putative docking models between FAT10 and ATG3. C, D. Determination of FAT10, ATG3, ZO-1 protein levels in paired CC tissues and adjacent nontumor tissues by western blot analysis (n = 50). Scatter plots revealed a negative correlation between FAT10, ATG3 and ZO-1 and a positive correlation between FAT10 and ATG3 at the protein level in CC tissues. E, F. SW620 cells were transfected with a FAT10 overexpression lentivirus, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10, ZO-1 and ATG3 were normalized to those of GAPDH. G. SW620 cells were transfected with shRNA against FAT10, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10, ZO-1 and ATG3 were normalized to those of GAPDH. H, I. Determination of the protein levels of ATG3, ZO-1, LC3B-II and p62 by western blotting in SW620 cells after the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to that of tubulin). J, K. Determination of the protein levels of FAT10, ATG3, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to that of tubulin). The data are presented as the means ± SDs. **p < 0.01; ***p < 0.001 FAT10 inhibits the expression of ZO-1 by increasing ATG3 expression to promote the invasion of colon cancer cells. (A) Table listing the FAT10-interacting proteins identified by coimmunoprecipitation mass spectrometry. (B) Putative docking models between FAT10 and ATG3. C, D. Determination of FAT10, ATG3, ZO-1 protein levels in paired CC tissues and adjacent nontumor tissues by western blot analysis (n = 50). Scatter plots revealed a negative correlation between FAT10, ATG3 and ZO-1 and a positive correlation between FAT10 and ATG3 at the protein level in CC tissues. E, F. SW620 cells were transfected with a FAT10 overexpression lentivirus, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10, ZO-1 and ATG3 were normalized to those of GAPDH. G. SW620 cells were transfected with shRNA against FAT10, and the cell lysates were subjected to qRT‒PCR and western blotting (n = 4). The mRNA levels of FAT10, ZO-1 and ATG3 were normalized to those of GAPDH. H, I. Determination of the protein levels of ATG3, ZO-1, LC3B-II and p62 by western blotting in SW620 cells after the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to that of tubulin). J, K. Determination of the protein levels of FAT10, ATG3, ZO-1, LC3B-II and p62 by western blotting in SW620 and HCT116 cells subjected to the indicated treatments. The number below the band represents the relative ratio of LC3B-II/I (normalized to that of tubulin). The data are presented as the means ± SDs. **p < 0.01; ***p < 0.001 To confirm this hypothesis, our results first found that ATG3 affected the autophagic degradation of ZO-1 in CC cells. Western blotting results showed that ATG3 overexpression reduced ZO-1 and p62 expression and promoted the transformation of LC3B-II, whereas CQ treatment reversed the reduction in ZO-1 expression and the increase in LC3B-II expression caused by ATG3 overexpression in CC cells (Fig. 5H Fig. S4C). Conversely, ATG3 knockdown increased the expression of ZO-1 and p62 and inhibited the transformation of LC3B-II, whereas rapamycin reversed the increase in ZO-1 expression and the decrease in LC3B-II expression caused by ATG3 knockdown in CC cells (Fig. 5I Fig. S4D). Furthermore, rescue experiments indicated that FAT10 overexpression decreased ZO-1 and p62 expression and activated autophagy, whereas ATG3 knockdown attenuated the decrease in ZO-1 expression and increase in autophagy caused by FAT10 overexpression in CC cells (Fig. 5J). In contrast, ATG3 overexpression reversed the FAT10 knockdown-induced increase in ZO-1 and p62 expression and inhibition of autophagy in CC cells (Fig. 5K). These results confirmed that FAT10 can promote the autophagic degradation of ZO-1 through the regulation of ATG3 expression in CC cells. FAT10 stabilizes ATG3 expression by antagonizing its ubiquitination Next, we aimed to explore the mechanisms by which FAT10 regulates ATG3 protein expression. Researchers have reported that ATG3 is degraded by the UPS [36]. The results of our previous LC‒MS and ZDOCK analyses revealed there was a binding between FAT10 and ATG3. In addition, our studies indicated that FAT10 stabilizes its substrates by antagonizing their ubiquitination [12, 37]. Therefore, we speculated that FAT10 could regulate ATG3 protein expression by antagonizing its ubiquitination. To verify this hypothesis, we first confirmed that FAT10 binds to the lysine site in the C-terminal catalytic domain (aa 131–341) of ATG3 in SW620 and HCT116 cells. Co-IP and confocal microscopy of immunofluorescence staining showed that FAT10 interacted with ATG3 in CC cells (Fig. 6A, B, Fig. S5A, B). Moreover, the co-IP results revealed that FAT10 could bind to the ATG3 C-terminal catalytic domain in CC cells (Fig. 6D, Fig. S5C). We subsequently transfected a plasmid encoding mutant ATG3 (HA-ATG3 Mut) with all the lysines mutated to arginines in the C-terminal catalytic domain into CC cells, and the co-IP results indicated that the binding ability of the mutant ATG3 to FAT10 was lost in CC cells (Fig. 6F, Fig. S5D). Furthermore, we found that FAT10 can regulate the ATG3 protein expression by UPS in SW620 and HCT116 cells. Western blotting revealed that MG132 treatment abolished the change in ATG3 protein expression mediated by changes in FAT10 expression in CC cells (Fig. 6G, Fig. S5E). Furthermore, an analysis of the dynamics of degradation showed that the half-life of exogenously expressed ATG3 was significantly increased in FAT10-overexpressing CC cells but was significantly reduced in FAT10-knockdown CC cells compared with the control cells (Fig. 6H-J, Fig. S5F-G). In addition, in vivo ubiquitination assays revealed that FAT10 overexpression markedly decreased ATG3 ubiquitination levels and increased ATG3 protein expression in CC cells; in contrast, FAT10 knockdown increased ATG3 ubiquitination levels and decreased ATG3 protein expression (Fig. 6K, Fig. S5H). To further determine whether FAT10 ubiquitinates ATG3 via K48- or K63-linked ubiquitination, we determined K48-linked or K63-linked ubiquitination levels. The results demonstrated K48-linked ubiquitination of ATG3 mediated by FAT10 (Fig. S5I), rather than K63-linked ubiquitination (Fig. S5J) in CC cells. Fig. 6FAT10 Stabilizes ATG3 in SW620 cells. A. Co-IP of ATG3 and FAT10 in SW620 cells. B. Representative confocal immunofluorescence images of ATG3 (red) colocalized with FAT10 (green) in SW620 cells. C. Schematic of ATG3, which has three functional domains. D. SW620 cells were cotransfected with HA-ATG3 domains 1 and 2, HA-ATG3 and Flag-FAT10. The cells were lysed for immunoprecipitation using anti-Flag beads to detect HA binding. (E) Schematic distribution of the 9 lysine residues in domain 2 of ATG3. (F) SW620 cells were transfected with the indicated constructs. Cells were lysed for Co-IP using anti-Flag beads to detect HA binding. (G) Determination of the protein levels of ATG3 and FAT10 by western blotting in SW620 cells subjected to the indicated treatments. H-J. The cells were exposed to 20 µg/ml CHX for the indicated times, and the degradation of ATG3 was detected by western blotting. K. SW620 cells were transfected with the indicated plasmids. The ubiquitination of ATG3 was detected. L, M. The binding of ATG3 during the course of the competition was analyzed by a GST pulldown experiment. N. SW620 cells were transfected with increasing amounts of Flag-FAT10 plasmid and HA-ATG3 (WT) plasmid. The cells were lysed for immunoprecipitation using anti-Ub and anti-FAT10 beads to detect ATG3 binding. O. SW620 cells were transfected with the indicated plasmids. The cells in each group were treated with MG132. The ubiquitination of ATG3 was detected. P. SW620 cells were transfected with increasing amounts of the Flag-FAT10 plasmid and HA-ATG3 (Mut) plasmid. The cells were lysed for immunoprecipitation using anti-Ub and anti-FAT10 beads to detect ATG3 binding. Q. SW620 cells were transfected with the indicated plasmids. The cells in each group were treated with MG132. The ubiquitination of ATG3 was detected. The data are presented as the means ± SDs. ***p < 0.001 FAT10 Stabilizes ATG3 in SW620 cells. A. Co-IP of ATG3 and FAT10 in SW620 cells. B. Representative confocal immunofluorescence images of ATG3 (red) colocalized with FAT10 (green) in SW620 cells. C. Schematic of ATG3, which has three functional domains. D. SW620 cells were cotransfected with HA-ATG3 domains 1 and 2, HA-ATG3 and Flag-FAT10. The cells were lysed for immunoprecipitation using anti-Flag beads to detect HA binding. (E) Schematic distribution of the 9 lysine residues in domain 2 of ATG3. (F) SW620 cells were transfected with the indicated constructs. Cells were lysed for Co-IP using anti-Flag beads to detect HA binding. (G) Determination of the protein levels of ATG3 and FAT10 by western blotting in SW620 cells subjected to the indicated treatments. H-J. The cells were exposed to 20 µg/ml CHX for the indicated times, and the degradation of ATG3 was detected by western blotting. K. SW620 cells were transfected with the indicated plasmids. The ubiquitination of ATG3 was detected. L, M. The binding of ATG3 during the course of the competition was analyzed by a GST pulldown experiment. N. SW620 cells were transfected with increasing amounts of Flag-FAT10 plasmid and HA-ATG3 (WT) plasmid. The cells were lysed for immunoprecipitation using anti-Ub and anti-FAT10 beads to detect ATG3 binding. O. SW620 cells were transfected with the indicated plasmids. The cells in each group were treated with MG132. The ubiquitination of ATG3 was detected. P. SW620 cells were transfected with increasing amounts of the Flag-FAT10 plasmid and HA-ATG3 (Mut) plasmid. The cells were lysed for immunoprecipitation using anti-Ub and anti-FAT10 beads to detect ATG3 binding. Q. SW620 cells were transfected with the indicated plasmids. The cells in each group were treated with MG132. The ubiquitination of ATG3 was detected. The data are presented as the means ± SDs. ***p < 0.001 Finally, we confirmed that FAT10 could compete with Ub for binding to ATG3, thereby antagonizing its ubiquitination to stabilize its protein expression. The results of the GST pulldown experiment showed that the addition of increasing amounts of FAT10 resulted in a gradual increase in the levels of the FAT10–ATG3 complex and a gradual decrease in the levels of the Ub–ATG3 complex; conversely, as the amount of Ub gradually increased, the level of the Ub–ATG3 complex gradually increased, and the level of the FAT10–ATG3 complex gradually decreased (Fig. 6L, M). Second, FAT10 competed with Ub to bind ATG3 in vivo, thereby inhibiting its polyubiquitination. Co-IP revealed that as the amount of the FAT10 plasmid added increased, the level of FAT10–ATG3 complexes gradually increased, and the level of Ub–ATG3 complexes gradually decreased in SW620 and HCT116 cells (Fig. 6N, Fig. S5L). In vivo ubiquitination assays showed that with increasing FAT10 expression, ATG3 K48-linked polyubiquitination gradually decreased in SW620 and HCT116 cells (Fig. 6O, Fig. S5K, M, N). However, as the amount of the FAT10 plasmid added increased, the level of FAT10–ATG3 mutant complexes was undetectable, and neither the level of Ub–ATG3 mutant complexes nor the level of ATG3 polyubiquitination changed with increasing FAT10 expression in SW620 and HCT116 cells (Fig. 6P, Q, Fig. S5O, P). Overall, the above results confirm that FAT10 stabilizes the expression of ATG3 by competing with Ub to bind to ATG3, reducing the level of ATG3 ubiquitination and thereby stabilizing ATG3 in CC cells. Compound 189 mitigates the FAT10-mediated autophagy enhancement and downregulation of ZO-1 thereby inhibiting the invasion and metastasis of CC Based on the importance of ATG3 in the ability of FAT10 to promote the autophagic degradation of ZO-1 in CC cells, we investigated whether inhibiting ATG3 could alleviate the autophagic degradation of ZO-1 mediated by FAT10 overexpression, thereby inhibiting the lung metastasis of CC cells. To this end, we used compound 189, an inhibitor of the ATG3–ATG12 interaction [38], to inhibit ATG3-mediated autophagy in CC cells. Western blotting showed that treatment with compound 189 reversed the increased transformation of LC3B-II and decreased ZO-1 and p62 protein expression caused by FAT10 overexpression but did not affect ATG3 expression in CC cells (Fig. 7A, Fig. S6A). Moreover, the results of the transwell and RTCA assays indicated that treatment with compound 189 alleviated the increases in the migration and invasion induced by FAT10 overexpression in CC cells (Fig. 7B, C, Fig. S6B, C). The results from these in vitro experiments confirmed that compound 189 reversed FAT10-induced ATG3-mediated autophagy, leading to the upregulation of ZO-1 expression and impeding the migration and invasion of CC cells. Fig. 7Compound 189 inhibits FAT10 overexpression-induced decreases in ZO-1 expression and colon cancer lung metastasis. A. Western blotting was used to detect the protein expressions of FAT10, ATG3, ZO-1, LC3B-II and p62 in SW620 cells with the indicated treatments (n = 4). The number below the band represents the relative ratio of LC3B-II/I (normalized to that of tubulin). B, C. Transwell (B) and RTCA (C) assays were used to measure the invasion and migration capacities of SW620 cells. Scale bars = 150 μm. D. Schematic of the lung metastasis of CC cells in a nude mouse model and evaluation of the effect of compound 189 in mice injected with Flag-FAT10 SW620/HCT116 cells. E, F. Representative bioluminescent images and quantification analysis (E) of lung tumors in nude mice (n = 10 mice per group). G, H. H&E-stained images and quantification analysis of metastatic nodules in the lungs (n = 10 mice per group). Scale bars = 100 μm. I. Western blot analysis was used to measure FAT10, ATG3 and ZO-1 protein expression levels in metastatic pulmonary tumors (n = 4). J. Survival analysis of mice with the indicated treatments (n = 10 mice per group). The data are presented as the means ± SDs. **p < 0.01 and ***p < 0.001 Compound 189 inhibits FAT10 overexpression-induced decreases in ZO-1 expression and colon cancer lung metastasis. A. Western blotting was used to detect the protein expressions of FAT10, ATG3, ZO-1, LC3B-II and p62 in SW620 cells with the indicated treatments (n = 4). The number below the band represents the relative ratio of LC3B-II/I (normalized to that of tubulin). B, C. Transwell (B) and RTCA (C) assays were used to measure the invasion and migration capacities of SW620 cells. Scale bars = 150 μm. D. Schematic of the lung metastasis of CC cells in a nude mouse model and evaluation of the effect of compound 189 in mice injected with Flag-FAT10 SW620/HCT116 cells. E, F. Representative bioluminescent images and quantification analysis (E) of lung tumors in nude mice (n = 10 mice per group). G, H. H&E-stained images and quantification analysis of metastatic nodules in the lungs (n = 10 mice per group). Scale bars = 100 μm. I. Western blot analysis was used to measure FAT10, ATG3 and ZO-1 protein expression levels in metastatic pulmonary tumors (n = 4). J. Survival analysis of mice with the indicated treatments (n = 10 mice per group). The data are presented as the means ± SDs. **p < 0.01 and ***p < 0.001 Furthermore, the findings from our in vivo experiments confirmed that compound 189 alleviated FAT10-induced ATG3-mediated autophagy, leading to the upregulation of ZO-1 expression and impeding the lung metastasis of CC cells. First, we injected Flag-vector or Flag-FAT10-CC cells into BALB/c nude mice through the tail vein. One week later, we treated the FAT10-overexpressing group with compound 189 or normal saline (Fig. 7D). Four weeks after the injection, IVIS imaging and HE staining revealed that compared with those in the control group, the fluorescence intensity in the lung and the number of lung metastatic nodules in the treatment group were significantly increased; however, compared with those in the control group, both the fluorescence intensity in the lung tissue and the number of lung metastatic nodules in the compound 189-treated group were significantly decreased (Fig. 7E-H, Fig. S6D, E). Western blotting of lung metastases showed that treatment with compound 189 attenuated the decrease in ZO-1 expression and increase in autophagy induced by FAT10 overexpression but did not affect the expression of ATG3 (Fig. 7I, Fig. S6F). Moreover, compared with that of the control group, the survival period of mice subjected to a tail vein injection of CC cells overexpressing FAT10 was significantly shorter, while treatment with compound 189 effectively improved the adverse prognosis caused by FAT10 overexpression (Fig. 7J, Fig. S6G). Together, our results indicated that compound 189 can attenuate the FAT10-mediated increase in autophagy and subsequently increase the expression of ZO-1 to inhibit the invasion and metastasis of CC cells both in vitro and in vivo. Discussion Lung metastasis is a key determinant of a poor prognosis for CC patients [1]. In this study, we observed high FAT10 expressing in CC tissues and cells and found that its overexpression promoted the lung metastasis of CC. FAT10 promotes the autophagic degradation of ZO-1 by stabilizing ATG3, ultimately facilitating the invasion and migration of CC cells. Finally, our in vivo experiments confirmed that compound 189 can block the lung metastasis of CC caused by FAT10 overexpression. Taken together, our results confirmed that FAT10 promoted the lung metastasis of CC, and we elucidated the underlying mechanism (Fig. 8). Autophagy plays an important role in regulating ZO-1 protein expression. However, its effect on ZO-1 degradation appears to differ depending on the cell type [29–34]. Some studies have indicated that autophagy can promote the degradation of ZO-1. For instance, under starvation stress, VDR deficiency inhibits LC3B transformation, thereby blocking autophagic flux, which leads to reduced autophagic degradation of ZO-1 and increased expression of ZO-1 in podocytes [19]. Studies have reported that LC3B can recognize ZO-1 through the LIR motif and that increased binding between LC3B and ZO-1 can increase the autophagic degradation of ZO-1 in Caco-2 cells [32]. Moreover, autophagy promotes ZO-1 degradation in the human cutaneous squamous cell carcinoma cell line COLO-16 [31]. In contrast, autophagy has been reported to inhibit the degradation of ZO-1. Autophagy activation can restore ZO-1 expression in stressed ileal epithelial cells [34]. Increased autophagy could also upregulate ZO-1 protein expression, thereby preserving the airway epithelial barrier function of BEAS-2B epithelial cells [30]. In this study, the autophagy activator rapamycin increased LC3B transformation and reduce the expression of p62 thereby promoting the autophagic degradation of ZO-1 in CC cells. Conversely, the autophagy inhibitors CQ and 3-MA suppressed LC3B transformation, consequently inhibiting the autophagic degradation of ZO-1 in CC cells. These results indicated that autophagy can promote the degradation of ZO-1 in CC cells. Ubiquitin-like proteins (UBLs) are crucial for maintaining cellular homeostasis and play an important role in regulating substrate protein expression [39]. UBLs regulate substrate protein expression through the UPS and/or autophagy pathways. For instance, interferon-stimulated gene 15 (ISG15) can promote PD-L1 degradation via the UPS [40]. Meanwhile, ISG15 can regulate autophagy-mediated RIG-I degradation [41]. Moreover, studies have reported that ubiquitin fold modifier-1 (UFM1) can promote the degradation of many substrates by the UPS [42]. In addition, UFM1 promotes the degradation of CYB5R3 and RPL26 via autophagy [43, 44]. As an important UBL, research on FAT10 has focused mainly on its role in regulating protein expression through its effects on the UPS [12, 37, 45–48]. In a previous study, we confirmed that FAT10 competes with SUMO1 for the K734 modification site of SIRT1, further reducing LC3 deacetylation and suppressing autophagy in cardiomyocyte [49]. However, whether FAT10 can promote autophagy, thereby promoting protein degradation, remains unknown. In this study, we confirmed that, like other UBLs, FAT10 promotes autophagy, thereby affecting protein degradation. Our findings showed that increased FAT10 expression promotes LC3B transformation, increases autophagic flux, and increases the formation of autophagosomes. Furthermore, FAT10 overexpression can increase autophagy and thereby promote the autophagic degradation of ZO-1. Conversely, FAT10 knockdown inhibited autophagy and the autophagic degradation of ZO-1. This discovery expands the understanding of the mechanism by which FAT10 regulates protein expression and deepens the understanding of the role of FAT10 in maintaining cellular homeostasis. ATG3 is a 36 kDa protein that comprises two domains: an N-terminal domain and a C-terminal catalytic domain [50]. ATG3 acts as an E2 enzyme in membrane extension to regulate the transformation of LC3B-I to LC3B-II [51]. Studies reported ATG3 plays a significant role in the onset and progression of various diseases [50, 52]. ATG3 expression is regulated by multiple factors. For instance, SP100-AS1 can stabilize ATG3 expression by inhibiting its degradation via the UPS, thereby promoting radiation resistance in CRC patients [36]. FOXO3 can upregulate ATG3, ultimately initiating autophagy in granulosa cells [53]. The lncRNA nuclear enriched abundant transcript 1 (NEAT1) can upregulate ATG3 expression to promote autophagy in HCC [54]. In addition, VDR deficiency downregulates ATG3 expression, thus impairing autophagy flux in podocytes [19]. Based on the important function of ATG3 in autophagy, a search for new factors that regulate ATG3 is particularly important. In our study, we identified FAT10 as a new regulator of ATG3. Our results showed that FAT10 can stabilize ATG3 expression by competing with Ub to bind ATG3, thus inhibiting the ubiquitination of ATG3. Therefore, our findings expand the regulatory network of autophagy and provide new targets for intervening in autophagy. The prognosis for patients with lung metastatic cancer is extremely poor [55, 56]. Chemotherapy is the main treatment for lung metastasis, but due to tumor heterogeneity and drug resistance, the effect of chemotherapy is limited [57–59]. Therefore, the development of new therapeutic agents for the clinical treatment of lung metastatic cancer is urgently needed. In recent years, autophagy-targeted therapy has shown promising prospects in treating lung metastatic cancer [60]. For instance, QU-NPs inhibited the lung metastasis of breast cancer by promoting autophagy [61]. Epimedokoreanin B blocks autophagic flux to inhibit the migration of triple-negative breast cancer cells [62]. LyP-1-trifluoperazine (TFP) inhibits the lung metastasis of melanoma by disrupting autophagy flux [63]. Classic autophagy inhibitors such as CQ and hydroxychloroquine (HCQ) suppress lung metastasis by inhibiting autophagy [64]. At present, research on autophagy-targeting strategies for the lung metastases of CC is still lacking. Our study demonstrated that compound 189 antagonizes the FAT10-mediated upregulation of ATG3 expression, thereby counteracting autophagy activation in CC cells. This process leads to increased expression of ZO-1 and ultimately inhibits CC cell invasion and migration. Furthermore, in vivo experiments revealed that compound 189 can inhibit the lung metastasis of FAT10-overexpressing CC and prolong survival. Conclusion We elucidated the role of the FAT10–ATG3–ZO-1 axis in the invasion and metastasis of CC cells and a novel mechanism by which FAT10 promotes the invasion and metastasis of CC cells. Importantly, we discovered that compound 189 can inhibit the FAT10-mediated autophagic degradation of ZO-1, thereby inhibiting the lung metastasis of CC. These findings may provide a basis for the development of new therapeutic strategies to prevent the lung metastasis of CC. Fig. 8The proposed model by which FAT10 promotes invasion and metastasis of CC cells via the ATG3-ZO-1 axis The proposed model by which FAT10 promotes invasion and metastasis of CC cells via the ATG3-ZO-1 axis Electronic Supplementary Material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3