Fig 2: TRIM25 mediates LC3 polyubiquitination and degradation.a Representative images of immunofluorescence showing the subcellular localization of DSS1 (magenta), LC3 (yellow), and PSMD3 (azure) in ACHN cells (n = 3 independent experiments). Nuclei: DAPI (blue). Scale bar: 5 μm. b LC-MS/MS analysis of LC3 interactomes identifies E3 ligases potentially associated with LC3 (n = 3 independent experiments; Proteome Discoverer 2.4, Thermo Scientific; False discovery rate <0.01). c The mRNA expression of TRIM25 in ccRCC patients from the TCGA-KIRC dataset (n = 595 distinct samples, tumor vs. normal, two-tailed Welch’s test). Boxplot: Center line = median; box = 25th to 75th percentiles; whiskers = minima to maxima. d Representative images of immunohistochemistry analysis showing TRIM25 protein levels upregulated in ccRCC tissues vs. normal (n = 66 distinct pairs; scale bar: 50 μm; error bar: mean (centre) ± SD; two-tailed paired t-test). e, f Immunoblotting showing TRIM25 knockdown increases LC3 levels, while TRIM25 overexpression reduces LC3 levels (n = 3 independent experiments, error bar: mean ± SD, two-tailed Welch’s t-test). g Co-IP analysis in Caki-1 cells showing the interaction between endogenous TRIM25 and LC3 (n = 3 independent experiments). h GST-pulldown analysis for TRIM25-LC3B interaction using 50 ng recombinantly expressed human full-length LC3B and 1 mg GST-TRIM25 (n = 3 independent experiments; Glutathione agarose beads: 50 µL). Asterisk: GST-TRIM25. i Immunoblotting of HEK293T cells transfected with HA-Ub, pcDNA3.1, and Myc-LC3B (wildtype) or Myc-LC3B mutant plasmids (n = 3 independent experiments). j Immunoblotting of HEK293T cells transfected with Myc-LC3B and HA-Ub-wildtype or HA-Ub-K48R/K63R mutant plasmids (n = 3 independent experiments). k Recombinantly expressed human GST-TRIM25 was incubated with recombinant Ub, LC3B, UbcH5a/UbcH5c, and His-UBE1 in a reaction system containing MgATP at 37 °C for 4 h. Samples were analyzed by SDS-PAGE and immunoblotting with antibody to Ub epitope (n = 3 independent experiments). As shown in the result of GST-TRIM25, the polyubiquitination of LC3B (LC3B-Ub) is shown only when all the components are present in the mix. e, f, h–j The samples derived from the same experiment were run on parallel gels, with each gel probed for a different antibody. Statistics are provided in the source data. Source data are provided as a Source Data file.

Fig 3: DSS1 enhances ccRCC cell migration and invasion via autophagy inhibition.a, b Transwell migration and Matrigel invasion assays in ACHN and Caki-1 cells treated with shNC, shDSS1, pcDNA3.1, DSS1 plasmids, with or without CQ (Chloroquine, representative images; scale bar: 100 μm). Quantification of migrated/invaded cells (error bar: mean ± SD; two-tailed Welch’s t-test; n = 3 independent experiments). c Morphology images of ccRCC cells between DSS1 knockdown (pebble-shaped) and negative control (spindle-shaped; scale bar: 100 μm; error bar: mean ± SD; two-tailed Welch’s t-test). The average cell aspect ratio was determined from 100 randomly selected cells per group per experiment (n = 3 independent experiments). d, e Immunoblotting (IB) of autophagy markers (LC3-II, p62) and EMT proteins (E-cadherin, N-cadherin, Vimentin) in siNC- and siDSS1-treated (with or without CQ, 0.02% DMSO as control) cells. Endogenous control: β-actin. Densitometry quantification (e, n = 3 independent experiments, error bar: mean ± SD, two-tailed Welch’s t-test). The samples derived from the same experiment were run on parallel gels, with each gel probed for a different antibody. f RT-qPCR analysis of MAP1LC3B and SQSTM1 mRNA levels in siDSS1 vs. siNC cells (normalized to GAPDH; n = 3 independent experiments; error bar: mean ± SD; two-tailed Welch’s t-test). Statistics are provided in the source data. Source data are provided as a Source Data file.

Fig 4: EMT and SPP1 in DSS1-driven cells-related niche.a Multiplex immunohistochemistry showing spatial distribution of DSS1-driven cells with a high epithelial-mesenchymal transition (EMT) feature (E-cadlow, Vimentinhigh) around vascular cells (CD31+, arrows, n = 4 distinct patients). Scale bar: 40 μm. b SPP1 RNA and SPP1 protein levels in tissues of diverse cancer types (kidney cancer: RNA, n = 35 distinct cell lines, protein, n = 110 distinct samples) from the Human Protein Atlas (HPA, www.proteinatlas.org). c SPP1 Protein levels in normal tissues from the HPA database (kidney, high expression in proximal tubules, n = 3 slides). d Pseudo-bulk RNA levels of SPP1 in epithelial and immune cell types. AvgExp, Average expression. e Protein abundance of SPP1 between tumor and normal tissues in Clinical Proteomic Tumor Analysis Consortium (CPTAC)-ccRCC dataset (n = 194 distinct samples, two-tailed Mann-Whitney U test). Boxplot: Center line = median; box = 25th to 75th percentiles; whiskers = minima to maxima. f Spearman correlation between SPP1 and DSS1 mRNA expression (log2 Count Per Million, two-tailed spearman’s rank correlation test). g Immunoblotting showing SPP1 protein levels in ccRCC cells (shDSS1 vs. shNC, n = 3 independent experiments, error bar: mean ± standard deviation [SD], two-tailed Welch’s t-test). The samples derived from the same experiment were run on parallel gels, with each gel probed for a different antibody. h Spearman correlation between CD68 expression and coexpression of SPP1-ITGB1 in spots of sample sections from either tumor core or tumor-stromal interface (Li2022 spatial dataset, n = 13 sections from 10 distinct patients; error bar: mean ± SD). i Schematic diagram (by Figdraw.com) illustrates the proposed mechanistic model of DSS1-driven cells in ccRCC metastasis: DSS1 (oligomer) interacts with pro-LC3B or LC3B-I, promoting LC3B degradation via E3 ubiquitin ligase TRIM25-mediated Lys-63 (K63)-linked polyubiquitination at LC3B-K51, leading to impaired macroautophagic flux and p62 accumulation, TWIST1 stabilization and increased TWIST1 nuclear transport, promoting EMT activation. DSS1 highly expressed (DSS1hi) tumor cells were increased in late-stage tumors and linked to microvascular invasion within a vascularized invasive niche at the tumor-stromal interface, mediated by SPP1-ITGB1 interactions. Statistics are provided in the source data. Source data are provided as a Source Data file.