Xing-Zhou Guo1,2,3†
Lu-Yun Zhang1,2,3†
Zong-Biao Tan1,2,3†Hao-Dong He1,2,3Ya-Fei Liu1,2,3
Bao-Ping Yu1*
Wei-Guo Dong1*- 1Department of Gastroenterology, Renmin Hospital of Wuhan University, Wuhan, Hubei, China
- 2Key Laboratory of Hubei Province for Digestive System Disease, Renmin Hospital of Wuhan University, Wuhan, Hubei, China
- 3Central Laboratory, Renmin Hospital of Wuhan University, Wuhan, Hubei, China
Background: The pathogenesis of ulcerative colitis (UC) involves genetic susceptibility and immune dysregulation. However, a more comprehensive understanding of its underlying mechanisms is still required.
Methods: Weighted correlation network analysis (WGCNA) and Mendelian randomization (MR) were used to investigate the association and causality between lipid metabolism and UC. Hub genes were identified by applying ensemble machine learning (LASSO, SVM, XGBoost) to the merged transcriptomic data, followed by cell-type-specific localization using public single-cell RNA-seq data from UC tissues. Pathophysiological relevance was confirmed in a DSS-induced colitis mouse model. The mechanistic role of the key hub gene was defined through knockdown LPS challenge in colonic epithelial cells, coupled with transcriptomic profiling.
Results: Integrated WGCNA and MR analyses established a strong association between UC and lipid metabolic pathways, which was experimentally confirmed by marked intracellular lipid accumulation in UC models. Mechanistically, this dysregulation was traced to lysosomal dysfunction, a finding solidified by pharmacological modulation with Rapamycin (enhancer) and Chloroquine (inhibitor), which directly demonstrated that lysosomal activity governs lipid metabolic disturbance. Subsequent machine-learning-based feature selection from lysosome-related genes pinpointed TRIM29 as a central regulator. Functional and transcriptomic analyses together demonstrated that TRIM29 knockdown attenuates UC progression by rescuing the lysosome-lipid metabolism axis, as evidenced by restored lysosomal function, enhanced cytoskeletal transport, improved lipid catabolism, a resolved pro-inflammatory immune profile, and downregulated inflammatory signaling.
Conclusion: This study identifies TRIM29 as a master regulator that drives UC progression by disrupting lysosomal function and reprogramming lipid metabolism. Our work delineates the TRIM29-lysosome-lipid metabolism axis, providing a mechanistic rationale for targeting TRIM29 as a promising therapeutic strategy in UC.
1 Introduction
Ulcerative colitis (UC) is a chronic, non-specific inflammatory bowel disease with multifactorial origins (1). It primarily affects the rectal and colonic mucosa and is characterized by recurrent episodes of mucopurulent bloody stools, abdominal pain, and systemic symptoms. In the 21st century, the global incidence of UC has risen rapidly (2). Although its etiology and pathogenesis remain unclear, they are thought to involve genetic predisposition, epithelial barrier dysfunction, immune dysregulation, and environmental factors (3–6). This lack of clear mechanistic understanding continues to pose challenges for diagnosis and treatment. In recent years, computational approaches, particularly in bioinformatics and machine learning (ML), have played increasingly important roles in UC research (7).
By integrating multi-omics data, researchers can identify novel UC-associated biomarkers and potential therapeutic targets (8). In transcriptomics, bioinformatics approaches have matured, with weighted gene co-expression network analysis (WGCNA) being widely applied to identify disease-related gene modules (9, 10). Such analyses provide new perspectives on disease mechanisms, offering opportunities for improved diagnosis and treatment (11, 12). Moreover, ML offers distinct advantages in multi-omics research and shows strong potential for elucidating disease mechanisms. When combined with WGCNA, it enables the identification of key diagnostic genes from large-scale transcriptomic datasets and facilitates the construction of predictive models. These models can support disease subtype classification and predict progression and treatment responses (12–17). Together, WGCNA and ML provide complementary perspectives on the molecular mechanisms of disease pathogenesis and hold promise for clinical translation.
In this study, we analyzed UC-specific mRNA expression profiles from NCBI, employing an integrated approach of WGCNA and ML to investigate key molecular mechanisms underlying UC pathogenesis. Our computational findings were further validated through molecular biology experiments. This work aims to provide new insights into the mechanisms driving inflammation and disease progression in UC.
2 Method
2.1 Data acquisition
All mRNA datasets were retrieved from the Gene Expression Omnibus (GEO) Datasets in the NCBI (National Center for Biotechnology Information) database. Using the keyword “ulcerative colitis” and restricting the organism to “Homo sapiens”, We manually screened the GEO repository and selected two independent mRNA expression datasets, GSE206285 (comprising only placebo-treated patients with stage IV ulcerative colitis, representing a pure UC cohort) and GSE87466, for model construction. Two additional datasets, GSE75214 and GSE107597, were prospectively reserved for independent external validation.
2.2 Cell lines and model inflammatory model construction
The human intestinal epithelial cell line NCM460 was obtained from the BeNa Culture Collection (BNCC, Beijing, China) and authenticated by short tandem repeat (STR) profiling to exclude cross-contamination. Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Hyclone, Cytiva, MA, USA) supplemented with 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Biosharp, China) at 37°C in a 5% CO2 incubator.
NCM460 cells were seeded into 6-well plates and 24-well plates, incubated overnight, and then treated with lipopolysaccharide (LPS, 5 µg/mL, 24 hours, Sigma, USA) (18–21).
2.3 DSS-induced colitis mouse model
Six-week-old male C57BL/6J mice were purchased from Shubeili Biotechnology Co., Ltd. (Wuhan, China). Mice (n = 6 per group) were randomly divided into two groups: a control group receiving water and an experimental group receiving 2.5% dextran sodium sulfate (DSS, MP Biomedicals) in drinking water for 7 days.
2.4 WGCNA
WGCNA was used to identify co-expressed gene modules and their associations with phenotypic traits. A weighted co-expression network was built from pairwise gene expression correlations. The adjacency matrix was calculated using a soft-thresholding power (β) to meet scale-free topology criteria (R² > 0.8) and was then transformed into a topological overlap matrix (TOM) to improve robustness. Genes were hierarchically clustered based on TOM dissimilarity, and modules were defined using dynamic tree cutting. Module eigengenes (MEs), represented by the first principal component, summarized module expression profiles. Associations between MEs and clinical traits were tested (p < 0.05, Bonferroni-corrected). Significant modules were further analyzed by functional enrichment (GO/KEGG). All analyses were performed using the WGCNA R package (v1.72) (22) (Supplementary Figure S2).
2.5 Machine learning modeling and validation
The bulk transcriptomic datasets GSE206285 and GSE87466 were integrated and batch effects were removed using the removeBatchEffect function from the limma R package (Supplementary Figure S1), with disease status (UC vs. control) preserved as a biological variable. The combined cohort was then randomly partitioned into training (60%), validation (20%), and test (20%) sets. LASSO regression, support vector machine (SVM), and XGBoost algorithms were applied to the training set to identify key feature genes from the lysosomal gene set that were differentially expressed between UC and HC.
Hyperparameters for SVM and XGBoost were optimized using the validation set. The final performance of all three models was evaluated on the test set.
2.6 LASSO regression
After preprocessing the mRNA expression matrix and binary phenotypic data, we performed L1-regularized logistic regression using the R package glmnet (v4.1-8). The normalized gene expression matrix was used as the independent variable, and the binary phenotype as the dependent variable. Genes with non-zero coefficients were selected by 10-fold cross-validation based on the minimum λ (min λ) criterion to generate the candidate gene set. Key parameters: alpha = 1, nfold = 10, family = “binomial”, type.measure = “class”.
2.7 SVM machine learning
The curated mRNA expression matrix was input into a support vector machine (SVM) implemented in the R package e1071 (v1.7-13). A linear kernel was used, and the hyperparameter C was optimized by 5-fold grid search. Model performance was evaluated by cross-validation area under the curve (AUC). The absolute values of the retained weights were used to define the feature gene set.
2.8 XGBoost machine learning
The whole-genome expression matrix was analyzed using the xgboost package (v1.7.8.1) with the following parameters: eta = 0.1, eval_metric = “logloss”, nthread = 2, eval_metric = “auc”, and nrounds = 100. The top-ranked genes were selected based on gain weights, and their contributions to the model were visualized using SHAP values.
2.9 Mendelian randomization
Genome-wide association study (GWAS) datasets were retrieved from the IEU database (https://gwas.mrcieu.ac.uk/) using the keywords “Ulcerative Colitis” and “Hyperlipidemia”. After screening and confirming disease-related entries, six hyperlipidemia-associated GWAS datasets were selected: ebi-a-GCST90104007, ebi-a-GCST90104006, ebi-a-GCST90104005, ebi-a-GCST90104004, ebi-a-GCST90104003, and ebi-a-GCST90090994. In addition, multiple UC-related datasets were included: ieu-a-970, ukb-b-19386, ukb-b-7584, ukb-a-104, ebi-a-GCST90018933, ebi-a-GCST90038684, ebi-a-GCST90018713, ukb-a-553, ukb-d-ULCERNAS, finn-b-ULCERNAS, finn-b-K11_UC_STRICT_PSC, finn-b-K11_UC_STRICT, finn-b-K11_UC_STRICT2, finn-b-K11_UC_NOCD, finn-b-K11_ULCER, finn-b-ULCEROTH, ieu-a-972, ebi-a-GCST90020072, ieu-a-971, ieu-a-973, ebi-a-GCST000964, ieu-a-968, ieu-a-32, ebi-a-GCST003045, ebi-a-GCST004133, and ukb-a-553.
MR analyses were performed using the R packages TwoSampleMR (v0.6.19), MRPRESSO (v1.0), and ieugwasr (v1.0.4) (23, 24).
Results were filtered in three steps: (1) preliminary screening for causal significance; (2) exclusion of effects with β values close to zero (biologically negligible); and (3) removal of results with β directions opposite to meta-analysis findings, retaining only exposure-outcome pairs with consistent directions and clear biological relevance.
For each exposure-outcome pair, the primary causal estimate was selected based on heterogeneity: if Cochran’s Q test indicated significant heterogeneity (p < 0.05), the inverse-variance weighted model with multiplicative random effects (IVW-MRE) was applied; otherwise, the fixed-effects inverse-variance weighted model (IVW-FE) was used. All significant associations reported herein satisfied this pre-specified robustness criterion (Supplementary Figure S3).
2.10 Single-cell RNA-seq analysis
Single-cell RNA-seq data were retrieved from the scIBD database (http://scibd.cn), a curated repository of inflammatory bowel disease-associated single-cell transcriptomes. We queried the database with the following parameters: Tissue = “largeInt”, Location = “colon, rectum”, and Disease state = “Healthy, UC_non_inflamed, UC_inflamed”. Gene expression patterns of candidate genes were visualized and extracted using the built-in analysis tools of the platform.
2.11 Inflammatory microenvironment profiling
Immune cell infiltration was estimated using the immunedeconv R package (v2.1.0), which provides a unified interface for multiple deconvolution algorithms. Specifically, the quanTIseq and MCP-counter methods were applied to the batch-corrected bulk RNA-seq data to quantify the relative proportions of immune cell subsets. Results from both methods were used for downstream correlation and group-comparison analyses.
2.12 Western blot
Cells were lysed in RIPA buffer (Servicebio, Wuhan, China) supplemented with 1% protease, phosphatase (Servicebio), and acetylase inhibitors (Beyotime, #P1112). Lysates were centrifuged at 12,000 rpm for 15 min at 4°C (Centrifuge 5424R, Eppendorf). Supernatants were mixed with loading buffer and denatured by heating.
2.13 Quantitative real-time polymerase chain reaction
Total RNA was extracted using an RNA extraction solution (Servicebio, Wuhan, China, G3013–100 mL). mRNA was reverse transcribed into complementary DNA (cDNA) using the Evo M-MLV Reverse Transcription Kit II (Accurate Biology, Changsha, China, AG11711). qPCR was performed with a CFX Connect Real-Time PCR System (Bio-Rad, Hercules, CA, USA) and 2X Universal SYBR Green Fast qPCR Mix (Abclonal, China, #RK21203). Relative expression levels were calculated using the 2^-ΔΔCt method, with β-actin as the reference gene.
2.14 Oil red O staining
Oil Red O (Sudan Red 5B) is a lipophilic azo dye that selectively stains neutral lipids, mainly triglycerides, producing red to orange-red droplets with minimal affinity for phospholipids and steroids. In this study, adherent cells treated with a lipid mixture (Sigma-Aldrich, L0288-100ML) were stained with Oil Red O (Servicebio, Wuhan, China, G1015–100 mL) to visualize intracellular lipid deposition.
2.15 BODIPY fluorescence
Cells were seeded on sterile glass coverslips in 24-well plates and treated with LPS for 24 h. The medium was then replaced with a 200 µM lipid mixture, and cells were incubated for an additional 24, 48, or 72 h. After three PBS washes, cells were stained with 40 µM BODIPY™ FL C12 (Thermo Fisher Scientific, #D3822) for 30 min at 37°C, fixed with 4% paraformaldehyde, and counterstained with DAPI.
2.16 LysoSensor fluorescent staining
Cells were seeded on sterile glass coverslips in 24-well plates and treated with LPS for varying durations. They were then incubated with LysoSensor for 30 min and Hoechst 33342 for 15 min to label lysosomes and nuclei, respectively.
2.17 Statistical analysis
All quantitative imaging data were measured using ImageJ, averaged with cell counts obtained from Cellpose segmentation (25), and subsequently visualized along with statistical analyses using GraphPad Prism. The exact sample sizes for each group are detailed in the figure legends.
3 Result
3.1 Integrated WGCNA and Mendelian randomization implicate dysregulated lipid metabolism in ulcerative colitis
In the bioinformatics analysis of mRNA gene sets, heatmaps, principal component analysis (PCA), and volcano plots can effectively illustrate expression differences across groups. Therefore, we conducted these three basic analyses for the GSE206285 and GSE87466 datasets (Figures 1A–C). The merged dataset revealed marked differences in mRNA expression profiles between healthy controls (HC) and UC patients, underscoring its research feasibility and value. Accordingly, we applied WGCNA using the WGCNA R package.
Figure 1. WGCNA revealed that UC was highly correlated with lipid metabolism. (A) The heatmap of GSE206285 and GSE87466. (B) The volcano plot of GSE206285 and GSE87466. (C) The PCA plot of GSE206285 and GSE87466. (D) Sample dendrogram and trait heatmap. (E) Merged dynamic tree. (F) Eigengene dendrogram and Eigengene adjacency heatmap. (G) Module trait relationships. (H) GO enrichment analysis of the darkgreen gene set. (I) KEGG enrichment analysis of the darkgreen gene set. (****p < 0.0001.).
We performed WGCNA to construct a gene co-expression network, identify modules, and assess their associations with the UC phenotype (Figures 1D–F). Among all modules, the darkgreen module demonstrated the strongest positive correlation with the UC phenotype (correlation coefficient = 0.73, p = 2 × 10−50), indicating a highly significant association (Figure 1G).
KEGG and GO enrichment analyses of key genes from the darkgreen module revealed significant enrichment in pathways related to fatty acid metabolism (Figures 1H, I). These results suggest that dysregulated fatty acid metabolism is strongly associated with UC pathogenesis and may play a crucial role in disease development.
Although the WGCNA analysis revealed strong associations between lipid metabolism modules and UC, observational studies cannot establish causal direction or account for confounding factors. To explore whether lipid metabolism abnormalities serve as genetic causal risk factors for UC, we conducted a two-sample MR analysis (Supplementary Figure S4A). Since hyperlipidemia represents a hallmark phenotype of disordered lipid metabolism (26), we focused on the genetic relationship between hyperlipidemia and UC.
Our MR analyses identified 11 significant exposure-outcome pairs that passed stringent quality control: all exhibited strong instrumental variables (minimum F-statistic = 29.9-37.9) and no evidence of horizontal pleiotropy (MR-Egger intercept p = 0.32-0.83) (Table 1). Of these, eight pairs indicated that genetic predisposition to UC was causally associated with a reduced risk of hyperlipidemia (β = -0.043 to -0.095, p < 0.05) (Supplementary Figures S4D, E), while three pairs suggested a reciprocal protective effect of hyperlipidemia on UC risk (β = -0.0003 to -0.246, p < 0.05). The strongest effect was observed for the influence of familial combined hyperlipidemia (Mexico criteria) on UK Biobank-defined UC (β = -0.246, p = 5.2 × 10−4).
Table 1. Mendelian randomization results for causal relationships between ulcerative colitis and hyperlipidemia-related traits.
3.2 Reduced lipid metabolism capacity in an experimental UC model
To further explore the relationship between UC and lipid metabolism, we established a mouse model of UC using 2.5% DSS. After 7 days of DSS treatment, the mice’s average body weight decreased to 79.8% of the initial weight, the average colon length was reduced by 3.1 cm, and both the DAI and spleen indices (spleen weight/initial body weight) significantly increased, confirming the successful induction of UC (Figures 2A–E). HE and AB-PAS staining of colonic sections revealed crypt destruction (Figures 2F, G), inflammatory infiltration, loss of goblet cells, and significantly reduced mucin abundance in the 2.5% DSS group. Transmission electron microscopy (TEM) showed disrupted tight junctions in the intestinal epithelium, lipid droplet accumulation within epithelial cells, and markedly impaired lipid degradation in the 2.5% DSS group (Figures 2I, J).
Figure 2. Impaired lipid metabolism in an experimental model of ulcerative colitis. (A) Body weight ratio of C57BL/6J mice (n = 6 per group) after 7-day administration of 2.5% DSS versus controls. (B) Representative images of colon morphology. (C) Colon length per group. (D) Spleen index per group. (E) Disease activity index (DAI) score. (F) Representative hematoxylin and eosin (H&E) staining of distal colon sections (scale bar: upper panel, 500 μm; lower panel, 100 μm). (G) Representative Alcian Blue/Periodic Acid–Schiff (AB–PAS) staining (scale bar: upper panel, 500 μm; lower panel, 100 μm). (H) Quantitative analysis of BODIPY fluorescence intensity (total lipid content) in the LPS treated group versus the control group after 24 h, 48 h, and 72 h of high-fat (100 μm) stimulation (n = 20). (I) Electron microscopy (EM) images showing ultrastructural changes in tight junctions (scale bar: upper panel, 2 μm; lower panel, 500 nm). (J) Electron microscopy images showing ultrastructural changes in lipid droplets (scale bar = 2 μm). (K) Representative fluorescence staining of BODIPY (green) and nuclei (DAPI, blue) in NCM460 cells from the inflammation group versus the control group after 24 h, 48 h, and 72 h of high-fat (100 μM) stimulation. (L) Western blot analysis of LPS-treated NCM460 cells. (M) Representative Oil Red O staining images of NCM460 cells in the inflammation group versus the control group after 24 h, 48 h, and 72 h of high-fat (100 μM) stimulation. (N) Quantitative analysis of the average lipid droplet area via Oil Red O staining. Statistical analysis was performed using unpaired t-tests (for two-group comparisons) or one-way ANOVA (for multiple-group comparisons) (n =6). (****p < 0.0001).
Our mouse experiments confirmed that DSS-induced UC significantly suppresses lipid metabolism. To investigate the underlying mechanism, we further established an in vitro colitis model using NCM460 cells. The successful establishment of the inflammatory model was validated by Western Blotting to assess the relative levels of specific proteins (Figure 2L). Following LPS induction, we observed an increase in the pro-apoptotic protein BAX, while levels of the anti-apoptotic protein BCL-2 and the barrier function proteins Occludin and ZO-1 decreased, confirming the successful establishment of the in vitro inflammatory model.
Additionally, NCM460 cells were exposed to a high-fat (HF) environment for varying durations after the same duration of inflammatory treatment. Total lipid content and average lipid droplet area were quantified using BODIPY and Oil Red O staining, respectively (Figures 2K, M). Under inflammatory conditions, both total lipid content and average lipid droplet area at 24 hours were significantly lower than in the control group. At 48 hours, both the average lipid droplet area and total lipid content in inflamed cells were higher than those in the control group. However, at 72 hours, total lipid content in the inflamed cells decreased significantly below the control level. Although the average lipid droplet area also decreased, it remained significantly higher than in the control group (Figures 2H, N).
3.3 Lysosomal dysfunction mediates impaired lipid metabolism in UC
Our data demonstrate a distinct lipid metabolism disorder phenotype in UC. As lysosomes are the primary organelles responsible for intracellular lipid degradation, a process known as lipophagy (27), we hypothesized that lysosomal dysfunction may act as a critical upstream event impairing lipid turnover in our colitis model. We next performed GSEA on the merged dataset, which revealed a significant enrichment of the lysosomal gene signature in UC samples, indicating a heightened transcriptional profile of lysosomal pathways (Figure 3A, Supplementary Datasheet 1).
Figure 3. Lysosomal dysfunction mediates lipid metabolism disorders in ulcerative colitis. (A) Gene Set Enrichment Analysis: In datasets GSE206285 and GSE87466, UC samples showed significant enrichment of lysosome-related gene sets (FDR < 0.05). (B) Representative LysoTracker staining images of NCM460 cells after 12 h and 24 h of LPS stimulation (red = lysosome, blue = nuclei; scale bar = 7 μm). (C) Western blot analysis of lysosome-associated proteins in LPS-treated NCM460 cells. (D) Number of lysosomes and mean lysosomal area stained with LysoTracker (n = 6). (E) Representative LysoSensor staining images of cells after 24 h, 48 h, and 72 h of LPS treatment (green = lysosome, blue = nuclei; scale bar = 100 μm). (F) Quantitative analysis of LysoSensor lysosomal fluorescence intensity after LPS treatment for 24 h, 48 h, and 72 h (n =25). (G) Transmission electron microscopy (TEM) of mouse colon: lysosomal content accumulation in control and DSS groups (scale bar: upper panel, 2 μm; lower panel, 500 nm). (H) Number of lysosomes and average lysosomal area stained with LysoTracker in cells from the control and Rapamycin (7 nM) groups at 12 h and 24 h (n = 9). (I) Representative LysoTracker staining images in cells from the control and Rapamycin (100 nM) groups at 12 h and 24 h (red = lysosomes, blue = nuclei; scale bar = 7 μm). (J, K) Quantitative analysis of total lipid content and average lipid droplet area in the LPS control, RA + LPS, and RA + LPS + CQ groups under 24 h, 48 h, and 72 h of high-fat (100 μM) stimulation (n = 16; n = 6). (L) Representative fluorescence images of BODIPY (green) and nuclei (DAPI, blue) in the LPS control, RA + LPS, and RA + LPS + CQ groups after 24 h, 48 h, and 72 h of high-fat (100 μM) stimulation. (M) Representative Oil Red O staining images of the LPS control, RA + LPS, and RA + LPS + CQ groups after 24 h, 48 h, and 72 h of high-fat (100 μM) stimulation. Statistical analysis was performed using unpaired t-tests (for two-group comparisons) or one-way ANOVA (for multiple-group comparisons). (ns, non-significant, *p < 0.05, **p < 0.01, ****p < 0.0001).
We further assessed lysosomal function in an in vitro inflammation model. Lysotracker confocal imaging revealed that, under inflammatory conditions, the number of NCM460 lysosomes increased at 12 hours but decreased at 24 hours, while their size progressively enlarged over time (Figures 3B, D). Concurrently, Lysosensor staining demonstrated that lysosomal pH significantly increased at 24 hours, plateaued thereafter, and reached its peak at 72 hours under inflammatory conditions (Figures 3E, F).
We further employed Western Blotting to evaluate lysosomal autophagic capacity. Under inflammatory conditions, we observed elevated levels of the lysosomal membrane proteins LAMP1 and LAMP2, together with increased expression of the autophagosome marker LC3 and the autophagy substrate p62 (Figure 3C).
Furthermore, electron microscopy of colonic tissues in the in vivo model revealed pronounced accumulation of lysosomal contents under inflammatory conditions. In contrast, the control group displayed smaller lysosomal contents, indicating preserved degradation capacity (Figure 3G).
To further determine the role of lysosomes in lipid metabolism, we manipulated lysosomal biogenesis and function using Rapamycin, an inducer of lysosomal biogenesis, and Chloroquine (CQ), an inducer of lysosomal dysfunction. Following intervention, the number of lysosomes in Rapamycin-treated cells showed a statistically significant difference compared with the inflammation-only control group at 24 hours. Moreover, the relative lysosomal area was significantly reduced at both 12 and 24 hours (Figures 3H, I).
In lipid metabolism assays, Rapamycin treatment led to increased average lipid droplet area and total lipid content at 24 hours, whereas CQ inhibited these effects. At 48 hours, significant differences emerged, with both parameters showing an overall upward trend. The Rapamycin-treated group exhibited the largest average lipid droplet area, whereas the CQ-supplemented group exhibited the smallest. Concurrently, the inflammation-only group showed the highest total lipid content, with no significant differences between the other two groups, although the Rapamycin-treated group exhibited a slight increase. By 72 hours, the average lipid droplet area had decreased across all three groups. Comparisons revealed that the Rapamycin-treated group had the smallest average lipid droplet area, whereas the inflammation-only group had the largest. However, total lipid content exhibited an opposite trend: it was lowest in the Rapamycin-treated group and highest in the Rapamycin plus CQ-treated group (Figures 3J–M).
Together, these findings indicate that lysosomal dysfunction is closely linked to lipid metabolic disorders during the progression of inflammation.
3.4 Machine learning screening of key genes in the lysosomal gene set: TRIM29
We next analyzed the previously merged mRNA dataset with the lysosomal gene set. Heatmap and volcano plot analyses of differentially expressed genes revealed marked differences between the inflammatory and control groups (Supplementary Figures S5A, B).
To further explore the role of lysosomal pathways in inflammatory progression, we performed protein-protein interaction (PPI) network analysis of differentially expressed lysosomal genes via the STRING database, and identified hub genes (MCC ≥10) using the cytoHubba plugin in Cytoscape.
Complementing this network-based approach, we applied three ML algorithms, LASSO regression, SVM, and XGBoost, to the same mRNA dataset using R software (Figures 4A–C, Supplementary Figure S5C). Integration of ML-derived features with PPI hub genes (MCC ≥10) identified TRIM29 as the top consensus candidate (Table 2, Figure 4D).
Figure 4. Machine learning identifies TRIM29 as a key hub gene in lysosome-related gene sets in UC. (A) The curve plots misclassification error against log(λ). The optimal λ (lambda.1se, right dashed line) was selected, yielding 9 non-zero coefficient genes for downstream analysis. (B) Average rank of TOP 10 features by SVM-RFE. (C) LASSO regression coefficients of selected feature genes. (D) Feature gene intersection analysis: integration of feature genes selected by LASSO, SVM and XGBoost with PPI core genes (MCC ≥ 10) revealed TRIM29 as the sole overlapping key hub gene. (E) Receiver operating characteristic (ROC) curve for the LASSO model in the validation set. (F) ROC curve for the LASSO model in the test set. (G) ROC curve for the SVM model in the test set. (H) ROC curve for the XGBoost model in the test set. (I, J) TRIM29 relative expression of GSE75214 and GSE107597. (K) ROC curve analysis of TRIM29, IL-6, IL-1β, and TNF-α using GSE75214 data. (L) ROC curve analysis of TRIM29, IL-6, IL-1β, and TNF-α using GSE107597 data. (**p < 0.01, ****p < 0.0001.)
Table 2. Step-wise gene filtering workflow.
Feature genes selected by LASSO regression demonstrated excellent diagnostic performance in distinguishing UC from HC, with AUC values of 0.8190 and 0.8259 in the validation and test sets, respectively (Figures 4E, F). Following hyperparameter optimization on the validation set, both the SVM (AUC = 0.869, 95% CI: 0.7991-0.9392) and XGBoost (AUC = 0.703, 95% CI: 0.4559-0.9490) models achieved satisfactory predictive performance in the independent test set (Figures 4G, H).
Subsequent validation confirmed that TRIM29 was upregulated in Ulcerative colitis (UC) across human datasets: its mRNA expression was lowest in healthy colon tissue and highest in active UC in both GSE75214 and GSE107597 (Figures 4I, J). Immunohistochemical data from the Human Protein Atlas showed minimal TRIM29 protein in healthy colon (Supplementary Figure S6A). Additionally, in vitro inflammation modeling via qPCR and Western blotting demonstrated significant upregulation under inflammatory conditions (Supplementary Figures S6B, C).
Notably, TRIM29 exhibited diagnostic accuracy for UC comparable to canonical inflammatory cytokines IL-6, IL-1β, and TNF-α in GSE107597 (Figure 4K), and significantly outperformed IL-1β in GSE75214 (Figure 4L), supporting its role as a lysosome-associated inflammatory biomarker with robust diagnostic and translational potential in UC (Supplementary Figure S6D).
3.5 TRIM29 expression correlates with immune infiltration and is enriched in epithelial cell in UC
To investigate the immunological role of TRIM29 in ulcerative colitis (UC), we first analyzed bulk transcriptomic data. Samples from merged datasets were stratified into TRIM29-high and TRIM29-low groups based on median expression. Immune cell infiltration was quantified using both quanTIseq and MCP-counter algorithms.
Both methods consistently revealed a significant increase in B cell infiltration in the TRIM29-high group (Figures 5A, C). Furthermore, quanTIseq indicated elevated levels of regulatory T cells (Tregs), while MCP-counter detected increased proportions of monocytes/macrophages, endothelial cells, natural killer (NK) cells, Fibroblasts, and neutrophils. Correlation analysis confirmed that TRIM29 expression was significantly associated with five immune cell types by quanTIseq and nine by MCP-counter (Figures 5B, D), underscoring its broad link to the UC immune microenvironment.
Figure 5. Analysis of TRIM29-associated immune infiltration and cellular subtypes. (A) The relative abundances of immune cell types, as estimated by the quantiSeq algorithm, are shown. (B) The lollipop plot shows the correlation coefficient of each immune cell type with TRIM29 expression levels estimated by quantiSeq. Red dots denote statistically significant correlations, while black dots indicate non-significant associations. (C) The relative abundances of immune cell types, as estimated by the MCP-counter algorithm, are shown. (D) The lollipop plot shows the correlation coefficient of each immune cell type with TRIM29 expression levels estimated by MCP-counter. Red dots denote statistically significant correlations, while black dots indicate non-significant associations. (E) UMAP visualization showing TRIM29 expression levels selectively plotted for cells where it is detected. (F) UMAP projection of the same single-cell dataset, with each cell colored by its annotated cell identity, providing the reference framework for interpreting TRIM29 expression. (G) Compares TRIM29 expression in key cell subtypes across disease states (Healthy, UC_non_inflamed, UC_inflamed). (*p < 0.05, **p < 0.01,****p < 0.0001.).
To complement these findings and pinpoint the cellular sources of TRIM29, we leveraged the online platform scIBD to analyze single-cell RNA sequencing data. This analysis confirmed broad TRIM29 expression across epithelial cells in UC, with levels positively correlating with mucosal inflammation (Figures 5E, F). TRIM29 expression showed a progressive increase from HC to UC_non_inflamed (non-inflamed mucosa from UC patients) and UC_inflamed groups, with the highest levels specifically localized to epithelial cells in UC_inflamed lesions (Figure 5G).
3.6 TRIM29 knockdown restores lysosomal biogenesis and function in NCM460 cells and markedly alleviates inflammation
To further validate the functional role of the key gene TRIM29 in inflammatory processes, we generated a stable NCM460 cell line stably expressing GFP-tagged shRNA targeting TRIM29 and confirmed efficient knockdown by qPCR and immunofluorescence analyses (Supplementary Figures S7A–C).
We next assessed lipid metabolic capacity and lysosomal morphology in the stable cell lines under inflammatory conditions. The shTRIM29 inflammation group showed significant differences in lysosomal number and average lysosomal area compared with the shNC group at 12 and 24 hours, characterized by an increased lysosomal number and a reduced average area (Figures 6A–C). Similarly, Western blot analysis of autophagy-related lysosomal proteins revealed increased LC3B levels, decreased p62, and reduced expression of LAMP1 and LAMP2 in the shTRIM29 group (Figure 6H).
Figure 6. TRIM29 knockdown restores lysosomal function and reduces inflammatory response in NCM460 cells. (A) Representative LysoTracker staining images of control and shTRIM29 cells after 12 h and 24 h of LPS treatment (red = lysosomes, blue = nuclei; scale bar = 7 μm). (B, C) Mean lysosomal area and number in control and shTRIM29 cells after 12 h and 24 h of LPS treatment (n = 9). (D) Representative images of control and shTRIM29 cells stained with Oil Red O after 24 h, 48 h, and 72 h of LPS treatment (red = lipid droplets, blue = nuclei; scale bar = 100 μm). (E) Quantitative analysis of average lipid droplet area by Oil Red O staining (n = 5). (F) Representative TUNEL staining images of control and shTRIM29 cells after LPS treatment (red = apoptotic cells, blue = nuclei; scale bar = 100 μm). (G) Quantitative analysis of TUNEL staining fluorescence (n = 8). (H) Western blot detection of lysosome-associated proteins. (I) Western blot detection of apoptosis-related and barrier proteins. Statistical analysis was performed using unpaired t-tests (for two-group comparisons) or one-way ANOVA (for multiple-group comparisons). (*p < 0.05, ***p < 0.001, ****p < 0.0001).
Compared with the shNC inflammation group, the shTRIM29 group had a smaller mean lipid droplet area at 24 hours. Although this value increased at 48 hours, it remained lower than that of the shNC group. By 72 hours, the lipid droplet area in the shTRIM29 group continued to increase, in contrast to the declining trend in the shNC group (Figures 6D, E).
Finally, TUNEL staining and Western Blotting were conducted to evaluate the effects of TRIM29 knockdown on cellular inflammation. The apoptosis rate was significantly lower in the shTRIM29 group than in the shNC group (Figures 6F, G). Additionally, Western Blotting revealed higher expression of the pro-apoptotic protein BAX and lower expression of the anti-apoptotic protein BCL-2 in the shTRIM29 group, along with increased levels of the tight junction proteins ZO-1 and Occludin (Figure 6I).
3.7 TRIM29 depletion induces transcriptomic signatures of membrane-cytoskeleton remodeling and ether lipid metabolism
After comparing the transcriptome sequencing data between the LPS group and shTRIM29+LPS group (Figures 7A–C), we performed KEGG, GO, and GSEA enrichment analyses.
Figure 7. Transcriptomic consequences of TRIM29 knockdown under inflammatory stress. (A) Transcript levels of TRIM29 in LPS-treated control versus shTRIM29 knockdown cells. (B) PCA of the transcriptomic profiles from both groups. (C) Volcano plot displaying genes differentially expressed upon TRIM29 knockdown. (D) KEGG pathway enrichment analysis of the differentially expressed genes. (E) GO enrichment analysis of the differentially expressed genes. (F) GSEA of the lysosome-related gene signature. (G) Functional categorization of differentially regulated genes. (H) GSEA of enriched KEGG pathways. (I) GSEA of enriched Gene Ontology terms. (J) GSEA of enriched Hallmark gene sets. (****p < 0.0001.).
KEGG, GO and GSEA enrichment results showed that lysosome function (Figure 7F, G) and membrane transport-related terms were significantly enriched, including endocytic vesicle lumen and membrane raft-related structural domains such as membrane raft, membrane microdomain, plasma membrane raft, and basolateral plasma membrane (Figure 7E). Simultaneously, cytoskeleton-related pathways including Motor proteins and Cytoskeleton in muscle cells were significantly enriched in KEGG analysis (Figure 7D).
GSEA analysis further revealed that lipid metabolism-related pathways were highly enriched across multiple databases, including Cholesterol metabolism, Arachidonic acid metabolism, alpha-Linolenic acid metabolism, Linoleic acid metabolism, and Ether lipid metabolism in the KEGG database (Figure 7H), as well as sterol transporter activity and lipoprotein particle receptor binding in GO molecular functions (Figure 7I). Inflammation-related pathways were also significantly enriched, including TNFα/NF-κB signaling pathway, interferon alpha response, and interferon gamma response in the Hallmark database, along with Complement and coagulation cascades and NOD-like receptor signaling pathway in the KEGG database. Additionally, KRAS signaling pathway reached significant enrichment levels in Hallmark analysis (Figure 7J).
These transcriptome data, though derived from a limited sample size, systematically suggest potential molecular mechanisms underlying TRIM29 knockdown, spanning from lysosome-membrane transport system alterations to lipid metabolism reprogramming, and ultimately to inflammatory responses.
4 Discussion
In this study, we first applied WGCNA and MR to identify key evidence supporting a potential association between UC and disorders of lipid metabolism. This relationship was subsequently validated using both in vivo and in vitro experimental models of UC. Given the established link between lipid metabolism and lysosomal function, we employed ML algorithms to screen for lysosome-related hub genes, through which TRIM29 was identified as a key candidate.
We then performed bioinformatic analyses to examine TRIM29 associations with immune infiltration and its subtype-specific expression patterns. Finally, a stable TRIM29-knockdown cell line was established. Using this model, we experimentally validated the proposed TRIM29-lysosome-lipid metabolism-UC axis and performed transcriptomic profiling to explore the potential mechanisms underlying this axis.
Prior evidence from WGCNA has implicated lipid metabolism dysregulation in UC pathogenesis.
Unlike the typical association of metabolic disorders with elevated lipid levels (28), patients with UC frequently exhibit hypolipidemia, characterized by reduced total cholesterol, low-density lipoprotein Cholesterol, high-density lipoprotein Cholesterol, and sometimes triglycerides (29), with these alterations being more marked in severe disease (30).
The apparent discrepancy, whereby UC is linked to lower systemic lipids in clinical and MR studies, yet associates with intracellular lipid accumulation in our cellular models, does not represent a true contradiction. Instead, it reflects complementary biological processes operating at different scales: systemic metabolism versus cellular metabolic mechanisms.
This gap is critical: although systemic lipid alterations in UC are well-documented, the intracellular mechanisms governing lipid handling during inflammation remain unclear. Importantly, common comorbidities such as malabsorption and malnutrition (31, 32) can account for systemic lipid reduction, highlighting a disconnect between whole-body and cellular metabolism. We therefore focused on lysosomes, the central organelles for intracellular lipid degradation, as potential mediators linking mucosal inflammation to lipid dysregulation. To address this, we investigated how inflammation reshapes lipid metabolism and lysosomal function in colonic epithelial cells.
Correspondingly, in our in vivo and in vitro UC models, we observed marked intracellular lipid accumulation under inflammatory conditions. This cellular lipid retention likely underpins the MR finding that UC genetically predisposes to a reduced risk of hyperlipidemia, as lipids are sequestered within the epithelium rather than circulating systemically.
Although MR revealed statistically significant causal estimates in both directions, these signals arose from distinct exposure-outcome pairs and differed markedly in magnitude. The effect of hyperlipidemia on UC risk was negligible (β ≈ -0.0003 in most significant pairs), whereas UC consistently reduced hyperlipidemia risk with a substantially larger effect size, aligning with our experimental observation of inflammation-driven intracellular lipid sequestration.
The average lipid droplet area and total cellular lipid content exhibited dynamic, non-linear changes (33). Under high-fat conditions with inflammation, both metrics initially decreased and later increased, but followed distinct temporal patterns. This suggests an early phase in which cells suppress lipid uptake to prioritize inflammatory responses, followed by a subsequent phase of impaired lysosomal degradation that drives lipid accumulation.
The later apparent divergence, where total lipid levels declined while average droplet area increased, can be explained by a combination of lipid droplet fusion (producing larger droplets) and the selective loss of severely lipid-laden cells through lipotoxic apoptosis (34), which our microscopy data support. This aligns with the lipid droplet retention observed in the colonic epithelium of DSS-induced colitis mice.
High-fat diets are standard for inducing hyperlipidemia in animal models. When given before colitis induction, such a diet promotes systemic alterations, including gut microbiota dysbiosis (35–37), immune dysregulation (38), and metabolic disorders (39), that may exacerbate disease. Clinically, without high-fat intake, UC patients typically show reduced serum lipid levels, especially in severe cases. Correspondingly, in animal models, a high-fat diet worsens the pathology of DSS-induced colitis compared to a normal diet (40). Conversely, after colitis onset, switching to a low-fat, high-fiber diet significantly alleviates symptoms, reduces inflammatory markers, and improves gut microbiota composition in patients (41).
Previous work by Teixeira et al. showed that a high-fat diet alone induces hyperlipidemia without causing significant colonic pathology. Building on this, we propose that during DSS-induced colitis, a high-fat background creates a specific vulnerability: substantial lipids enter intestinal epithelial cells, but concurrently, inflammation progressively impairs lysosomal function. As the primary site for lipid degradation, dysfunctional lysosomes fail to process this lipid influx.
Consequently, undegraded lipids, such as cholesterol and triglycerides, accumulate and form intra-lysosomal droplets, which further suppress enzymatic activity, creating a self-perpetuating vicious cycle (42). The resulting severe decline in lysosomal activity then engages a bidirectional cascade with inflammatory pathways, leading to cellular damage, pathway activation, and exacerbated colitis (43).
Based on our findings, we propose that under high-fat conditions, the lysosome acts as a critical mediator linking hyperlipidemia to exacerbated colitis. To test this, we employed Rapamycin to enhance lysosomal biogenesis and CQ to inhibit lysosomal function. Rapamycin treatment effectively enhanced lysosomal lipid metabolism, reduced lysosomal swelling, and consequently led to a marked decline in lipid deposition, evidenced by lower total lipid content and a smaller mean lipid droplet area. Conversely, CQ administration abolished these benefits, confirming that lysosomal dysfunction directly mediates the lipid metabolic disorder in cellular inflammation. Together, these opposing pharmacological effects underscore the central role of lysosomal homeostasis in inflammatory progression.
To elucidate potential mechanisms of lysosomal dysfunction in UC, we analyzed the merged expression matrix using ML approaches applied to a lysosome-associated gene set. This approach identified TRIM29 (Tripartite Motif Containing 29) as a candidate key molecule. TRIM29 is located on human chromosome 11q23.3 and encodes a 588-amino-acid protein with a molecular weight of approximately 66 kDa.
Current research on TRIM29 is predominantly in oncology, where it exhibits distinct and sometimes opposing effects across cancer types. It functions as a tumor suppressor in hepatocellular carcinoma and prostate cancer (44), yet promotes tumor progression in thyroid and colorectal cancers (45). In enterovirus-induced intestinal inflammation, TRIM29 targets and degrades NLRP6/NLRP9b, thereby suppressing the host innate immune response and facilitating viral replication (46, 47). Mechanistically, TRIM29 regulates multiple signaling pathways, including Wnt/β-catenin, PI3K/AKT/mTOR, NF-κB, the KRAS-β-catenin synergy, and the STAT3-SNAI1 axis (48). However, the specific role of TRIM29 in UC remains undefined.
Transcriptomic analysis revealed that elevated TRIM29 expression was broadly associated with altered immune cell infiltration. Specifically, TRIM29 levels showed a positive correlation with the proportions of B cells and monocytes/macrophages, but a negative correlation with natural killer (NK) cells. This profile, characterized by an increase in pro-inflammatory lineages and a decrease in NK cells, reflects a pro-inflammatory shift in the immune milieu linked to high TRIM29 expression.
Single-cell analysis via the scIBD platform further revealed that TRIM29 expression is elevated in active UC and predominantly localized to intestinal epithelial cells, supporting the rationale for selecting the human colonic epithelial cell line NCM460 in our experimental validation.
To specifically interrogate TRIM29’s intrinsic role in colonic epithelium, we established stable TRIM29-knockdown NCM460 cells and assessed their response to inflammatory stimulation. TRIM29 knockdown significantly ameliorated lysosomal dysfunction and attenuated inflammatory responses. Mechanistically, the knockdown initially enhanced lysosome-mediated lipolysis, reducing lipid deposition. This was directly evidenced by Oil Red O staining, which showed that after 48–72 hours of high-fat treatment, shTRIM29 cells exhibited smaller individual lipid droplets compared with shNC controls, even as the total lipid droplet area increased, indicating accelerated lipolysis. Paradoxically, TRIM29 knockdown also promoted the survival of metabolically stressed cells, leading to retention of dysfunctional cells and subsequent compensatory lipid overload. As a result, the average lipid droplet area in shTRIM29 cells exhibited a biphasic change: an initial decrease followed by a later increase.
Thus, these findings establish TRIM29 as a critical node coordinating lysosomal function, lipid metabolism, and cell fate in inflamed epithelium. Our data demonstrate that TRIM29 drives pathological lipid accumulation via lysosomal dysfunction and sustains inflammatory responses in UC. This epithelium-intrinsic defect mechanistically explains the pro-inflammatory immune microenvironment linked to high TRIM29. Future studies using intestinal organoids or cell-type-specific models are needed to validate these functions within native colonic cell contexts.
Building on this, we delineate a pathogenic axis in UC: inflammation upregulates epithelial TRIM29, which disrupts lysosomal function and lipid metabolism. This leads to barrier failure that drives a pro-inflammatory microenvironment marked by increased B cells and macrophages alongside decreased NK cells. Our work thereby establishes the TRIM29-lysosome-lipid metabolism-UC axis, providing a mechanistic foundation for targeting TRIM29.
Despite these findings, our study has several limitations. First, due to technical constraints with fluorescently tagged shRNA constructs, we were unable to assess lysosomal pH or total lipid content in the shTRIM29 inflammation model. Second, our functional insights into TRIM29 are primarily derived from LPS-treated NCM460 colonic epithelial cells, an in vitro model that has been widely used to mimic inflammatory conditions in UC (18–21). However, this monoculture system cannot fully recapitulate the complex multicellular architecture, immune-epithelial crosstalk, and microenvironmental heterogeneity characteristic of human UC. To better capture the pathophysiological relevance of the TRIM29–lysosome–lipid metabolism axis in UC, future studies should incorporate more physiologically representative models, such as patient-derived intestinal organoids or primary human colonic epithelial cells. Third, the upstream regulators and downstream pathways mediating TRIM29’s role in colitis were largely inferred from transcriptomic data and require direct experimental validation to elucidate the precise molecular mechanisms involved.
5 Conclusion
In summary, our findings identify TRIM29 as a novel regulatory hub linking lysosomal dysfunction to pathological lipid metabolism in UC. TRIM29 is upregulated in inflamed mucosa and correlates with disease severity. Knockdown of TRIM29 attenuates lysosomal impairment and inflammation in colonic epithelial cells, with accompanying transcriptomic shifts in ether lipid metabolism and membrane-cytoskeleton organization. These results implicate TRIM29 in UC pathogenesis and support therapeutic targeting of the TRIM29-lysosome-lipid metabolism axis.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.
Ethics statement
Ethical approval was not required for the studies on humans in accordance with the local legislation and institutional requirements because only commercially available established cell lines were used. All animal experiments were conducted in compliance with the Hubei Provincial Regulations for Animal Experiment Personnel and were approved by the Animal Ethics Committee (Ethics Approval Number: WDRM-20231005B). The study was conducted in accordance with the local legislation and institutional requirements.
Author contributions
X-ZG: Data curation, Formal Analysis, Investigation, Methodology, Visualization, Writing – original draft, Writing – review & editing. L-YZ: Validation, Visualization, Writing – original draft, Writing – review & editing. Z-BT: Formal Analysis, Investigation, Methodology, Writing – original draft. H-DH: Data curation, Investigation, Writing – original draft. Y-FL: Investigation, Methodology, Writing – original draft. B-PY: Formal Analysis, Writing – review & editing. W-GD: Funding acquisition, Methodology, Writing – review & editing.
Funding
The author(s) declared that financial support was received for work and/or its publication. This work was supported by the National Natural Science Foundation of China under Grant [No. 82370542].
Acknowledgments
We acknowledge the platform support provided by the Key Laboratory of Digestive System Diseases of Hubei Province and the Central Laboratory of Wuhan University People’s Hospital.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fimmu.2025.1728932/full#supplementary-material
Supplementary Figure 1 | Assessment of Batch Effect Correction. (A, B) PCA plots before and after batch effect correction, demonstrating the effectiveness of harmonization. (C, D) Expression distribution of housekeeping genes before and after correction. (E) Statistical significance of batch effects assessed by PERMANOVA.
Supplementary Figure 2 | WGCNA clustering snapshots. (A) Soft-threshold (left) and mean connectivity (right) plots for power selection. (B) Gene dendrogram with dynamic tree-cut module colors. (C) Module-eigengene clustering dendrogram. (D) Network heatmap of selected genes.
Supplementary Figure 3 | Further filtering of significant causal pairs.
Supplementary Figure 4 | Unidirectional causality was observed in both directions across datasets, yet no cohort exhibited bidirectional causality. (A) Flowchart of MR screening results. (B) Significant causal effects of UC on hyperlipidemia. (C) Significant causal effects of hyperlipidemia on UC. (D)Trait pairs with significant UC-to-hyperlipidemia causal effects whose direction agreed with the meta-analysis. (E) Trait pairs with significant hyperlipidemia-to-UC causal effects whose direction agreed with the meta-analysis. Symbol size is inversely proportional to the absolute beta coefficient; larger points correspond to weaker effects. Horizontal lines represent 95% confidence intervals; endpoints indicate upper and lower bounds.
Supplementary Figure 5 | Machine learning analysis of the lysosome gene set. (A) Heatmap of differentially expressed lysosome-related genes in the merged dataset. (B) Volcano plot of the lysosome gene set. (C) The distribution of predicted probabilities in the independent test set. The histogram displays the probability scores generated by the LASSO regression model for samples in the held-out test set. (D) ROC curve evaluating the SVM classifier (with 5 feature genes) on the independent test set.
Supplementary Figure 6 | TRIM29 as a Biomarker in UC: Expression and Diagnostic Performance. (A) Tissue expression of TRIM29 shown by IHC (Human Protein Atlas). (B, C) Upregulation of TRIM29 at the mRNA (qPCR) and protein (Western blot) levels in LPS-stimulated colonic epithelial cells. (D) Superior diagnostic performance of TRIM29 compared to key inflammatory cytokines in validation datasets.
Supplementary Figure 7 | Stable CMV-GFP-shTRIM490 NCM460 cell line successfully established. (A) Melt curve and melt peak profiles of ACTB primers (upper panel). (B) Melt curve and melt peak profiles of TRIM29 primers (lower panel). (C) TUNEL apoptotic staining of CMV-GFP-shTRIM29 NCM460 stable cell line. (Red: TUNEL; Green: GFP). All the statistical analysis was performed using unpaired t-tests (for two-group comparisons) or one-way ANOVA (for multiple-group comparisons). (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.).
SUPPLEMENTARY DATASHEET 1 | Lysosome related gene set. This lysosome-related gene set was compiled by collecting and integrating all human lysosome-associated genes from the KEGG, GSEA, and GO databases.
References
1. Baumgart DC and Carding SR. Inflammatory bowel disease: cause and immunobiology. Lancet. (2007) 369:1627–40. doi: 10.1016/S0140-6736(07)60750-8
2. Ng SC, Yun SH, Nima H, Fox EU, Whitney T, Eric IB, et al. Worldwide incidence and prevalence of inflammatory bowel disease in the 21st century: a systematic review of population-based studies. Lancet. (2017) 390:2769–78. doi: 10.1016/S0140-6736(17)32448-0
3. de Souza HSP and Fiocchi C. Immunopathogenesis of IBD: current state of the art. Nat Rev Gastroenterol Hepatol. (2016) 13:13–27. doi: 10.1038/nrgastro.2015.186
4. Inflammatory Bowel Disease Group, C. M. A. Chinese Society of Gastroenterology. Chinese consensus on diagnosis and treatment in inflammatory bowel disease (2018, Beijing). J Digestive Dis. (2021) 22:298–317. doi: 10.1111/1751-2980.12994
5. Kudo T and Shimizu T. Mucosal immune systems of pediatric inflammatory bowel disease: A review. Pediatr Int. (2023) 65:e15511. doi: 10.1111/ped.15511
6. Ungaro R, Mehandru S, Allen PB, Peyrin-Biroulet L, and Colombel J-F. Ulcerative colitis. Lancet. (2017) 389:1756–70. doi: 10.1016/S0140-6736(16)32126-2
7. Bu M, Cao X, and Zhou B. Identification of potential biomarkers and immune infiltration characteristics in ulcerative colitis by combining results from two machine learning algorithms. Comput Math Methods Med. (2022) 2022:5412627. doi: 10.1155/2022/5412627
8. Preto AJ, Chanana S, Ence D, Healy MD, Domingo-Fernández D, West KA, et al. Multi-omics data integration identifies novel biomarkers and patient subgroups in inflammatory bowel disease. J Crohn’s Colitis. (2025) 19:jjae197. doi: 10.1093/ecco-jcc/jjae197
9. Hughes SA, Lin M, Weir A, Huang B, Xiong L, Chua NK, et al. Caspase-8-driven apoptotic and pyroptotic crosstalk causes cell death and IL-1beta release in X-linked inhibitor of apoptosis (XIAP) deficiency. EMBO J. (2023) 42:e110468. doi: 10.15252/embj.2021110468
10. Zhang C, Zhang J, Zhang Y, Song Z, Bian J, Yi H, et al. Identifying neutrophil-associated subtypes in ulcerative colitis and confirming neutrophils promote colitis-associated colorectal cancer. Front Immunol. (2023) 14:1095098. doi: 10.3389/fimmu.2023.1095098
11. Xu M, Zhou H, Hu P, Pan Y, Wang S, Liu L, et al. Identification and validation of immune and oxidative stress-related diagnostic markers for diabetic nephropathy by WGCNA and machine learning. Front Immunol. (2023) 14:1084531. doi: 10.3389/fimmu.2023.1084531
12. Yang Y, Hua Y, Zheng H, Jia R, Ye Z, Su G, et al. Biomarkers prediction and immune landscape in ulcerative colitis: Findings based on bioinformatics and machine learning. Comput Biol Med. (2024) 168:107778. doi: 10.1016/j.compbiomed.2023.107778
13. Tang D, Pu B, Liu S, and Li H. Identification of cuproptosis-associated subtypes and signature genes for diagnosis and risk prediction of Ulcerative colitis based on machine learning. Front Immunol. (2023) 14:1142215. doi: 10.3389/fimmu.2023.1142215
14. Greener JG, Kandathil SM, Moffat L, and Jones DT. A guide to machine learning for biologists. Nat Rev Mol Cell Biol. (2022) 23:40–55. doi: 10.1038/s41580-021-00407-0
15. Ge C, Lu Y, Shen Z, Lu Y, Liu X, Zhang M, et al. Machine learning and metabolomics identify biomarkers associated with the disease extent of ulcerative colitis. J Crohn’s Colitis. (2025) 19:jjaf020. doi: 10.1093/ecco-jcc/jjaf020
16. Jahagirdar V, Bapaye J, Chandan S, Ponnada S, Kochhar GS, Navaneethan U, et al. Diagnostic accuracy of convolutional neural network–based machine learning algorithms in endoscopic severity prediction of ulcerative colitis: a systematic review and meta-analysis. Gastrointest Endosc. (2023) 98:145–154.e148. doi: 10.1016/j.gie.2023.04.2074
17. Wang Y, Huang J, Zhang J, Wang F, Tang X, Wang Y, et al. Identifying biomarkers associated with the diagnosis of ulcerative colitis via bioinformatics and machine learning. Math Biosci Eng. (2023) 20:10741–56. doi: 10.3934/mbe.2023476
18. Tao X, Gao Y, Jin Z, Lu S, Li Z, Guo S, et al. Co-administration of berberine and evodiamine: Mitigating evodiamine-induced hepatotoxicity and potentiating colitis treatment. Phytomedicine. (2025) 147. doi: 10.1016/j.phymed.2025.157185
19. Liu W, Yan X, An J, Wang X, Mi H, Liu F, et al. Modified Jiaoqi Powder enhances epithelial autophagy against TNF-triggered apoptosis in chronic ulcerative colitis. Phytomedicine. (2025) 136. doi: 10.1016/j.phymed.2024.155996
20. Hu W, Cai Y, Cai D, Chen Z, Huang S, Zhang S, et al. HIF-1α alleviates ferroptosis in ulcerative colitis by regulation of GPX4. Cell Death Dis. (2025) 16:542. doi: 10.1038/s41419-025-07883-8
21. Guo C, Guo D, Fang L, Sang T, Wu J, Guo C, et al. Ganoderma lucidum polysaccharide modulates gut microbiota and immune cell function to inhibit inflammation and tumorigenesis in colon. Carbohydr Polym. (2021) 267. doi: 10.1016/j.carbpol.2021.118231
22. Langfelder P and Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinf. (2008) 9:559. doi: 10.1186/1471-2105-9-559
23. Hemani G, Zheng J, Elsworth B, Wade KH, Haberland V, Baird D, et al. The MR-Base platform supports systematic causal inference across the human phenome. eLife. (2018) 7. doi: 10.7554/eLife.34408
24. Li J, Hemani G, Tilling K, and Davey Smith G. Orienting the causal relationship between imprecisely measured traits using GWAS summary data. PLoS Genet. (2017) 13. doi: 10.1371/journal.pgen.1007081
25. Stringer C, Wang T, Michaelos M, and Pachitariu M. Cellpose: a generalist algorithm for cellular segmentation. Nat Methods. (2020) 18:100–6. doi: 10.1038/s41592-020-01018-x
26. Mach F, Koskinas KC, Roeters van Lennep JE, Tokgozoglu L, Badimon L, Baigent C, et al. 2025 Focused Update of the 2019 ESC/EAS Guidelines for the management of dyslipidaemias. Eur Heart J. (2025) 46:4359–78. doi: 10.1093/eurheartj/ehaf190
27. Meng Y, Guo D, Lin L, Zhao H, Xu W, Luo S, et al. Glycolytic enzyme PFKL governs lipolysis by promoting lipid droplet–mitochondria tethering to enhance β-oxidation and tumor cell proliferation. Nat Metab. (2024) 6:1092–107. doi: 10.1038/s42255-024-01047-2
28. Chen L, Chen XW, Huang X, Song BL, Wang Y, Wang Y, et al. Regulation of glucose and lipid metabolism in health and disease. Sci China Life Sci. (2019) 62:1420–58. doi: 10.1007/s11427-019-1563-3
29. Soh H, Im JP, Han K, Park S, Hong SW, Moon JM, et al. Crohn’s disease and ulcerative colitis are associated with different lipid profile disorders: a nationwide population-based study. Aliment Pharmacol Ther. (2020) 51:446–56. doi: 10.1111/apt.15562
30. Zhu L, Tian L, Li S, and Huo L. Blood lipid levels and their value as markers of disease activity in severe ulcerative colitis. Int J Gen Med Volume. (2025) 18:671–80. doi: 10.2147/ijgm.S481323
31. Valvano M, Faenza S, Cortellini F, Vinci A, Ingravalle F, Calabrò M, et al. The relationship between nutritional status, micronutrient deficiency, and disease activity in IBD patients: A multicenter cross-sectional study. Nutrients. (2025) 17. doi: 10.3390/nu17162690
32. Bilgen A and Akpınar H. The relationship between disease activity and fecal calprotectin and fecal occult blood in inflammatory bowel disease: the role of nutritional status. Nutrients. (2025) 17:2690. doi: 10.3390/nu17213379
33. Grabner GF, Xie H, Schweiger M, and Zechner R. Lipolysis: cellular mechanisms for lipid mobilization from fat stores. Nat Metab. (2021) 3:1445–65. doi: 10.1038/s42255-021-00493-6
34. Liu J, Fan X, Song Y, and Zhao J. Triglyceride-based lipotoxicity in the pathophysiology of chronic diseases. Trends Endocrinol Metab. (2025) 36:147–60. doi: 10.1016/j.tem.2025.07.004
35. Zhong Y, Xiao Q, Huang J, Yu S, Chen L, Wan Q, et al. Ginsenoside rg1 alleviates ulcerative colitis in obese mice by regulating the gut microbiota–lipid metabolism–th1/th2/th17 cells axis. J Agric Food Chem. (2023) 71:20073–91. doi: 10.1021/acs.jafc.3c04811
36. Tong Y, Gao H, Qi Q, Liu X, Li J, Gao J, et al. High fat diet, gut microbiome and gastrointestinal cancer. Theranostics. (2021) 11:5889–910. doi: 10.7150/thno.56157
37. Poppe J, Boesmans L, Vieira-Silva S, Deroover L, Tito R, Vandeputte D, et al. Differential contributions of the gut microbiota and metabolome to pathomechanisms in ulcerative colitis: an in vitro analysis. Gut Microbes. (2024) 16. doi: 10.1080/19490976.2024.2424913
38. Li P, Xu R, Chen H, Cao S, Xie X, Xu B, et al. The promoting effect of high-fat diet-induced trained immunity on ulcerative colitis and the mechanism of the therapeutic effect of Dahuang Mudan Decoction. Phytomedicine. (2025) 143. doi: 10.1016/j.phymed.2025.156893
39. Frazier K, Kambal A, Zale EA, Pierre JF, Hubert N, Miyoshi S, et al. High-fat diet disrupts REG3γ and gut microbial rhythms promoting metabolic dysfunction. Cell Host Microbe. (2022) 30:809–823.e806. doi: 10.1016/j.chom.2022.03.030
40. Teixeira LG, Leonel AJ, Aguilar EC, Batista NV, Alves AC, Coimbra CC, et al. The combination of high-fat diet-induced obesity and chronic ulcerative colitis reciprocally exacerbates adipose tissue and colon inflammation. Lipids Health Dis. (2011) 10:204. doi: 10.1186/1476-511x-10-204
41. Fritsch J, Garces L, Quintero MA, Pignac-Kobinger J, Santander AM, Fernández I, et al. Low-fat, high-fiber diet reduces markers of inflammation and dysbiosis and improves quality of life in patients with ulcerative colitis. Clin Gastroenterol Hepatol. (2021) 19:1189–1199.e1130. doi: 10.1016/j.cgh.2020.05.026
42. Harvey LD, Alotaibi M, Tai YY, Tang Y, Kim HJ, Kelly NJ, et al. Lysosomal dysfunction and inflammatory sterol metabolism in pulmonary arterial hypertension. Science. (2025) 387:eadn7277. doi: 10.1126/science.adn7277
43. Hu CA, Hou Y, Yi D, Qiu Y, Wu G, Kong X, et al. Autophagy and tight junction proteins in the intestine and intestinal diseases. Anim Nutr. (2015) 1:123–7. doi: 10.1016/j.aninu.2015.08.014
44. Tang Y, Fan S, Peng R, Liu H, Su B, Tu D, et al. TRIM29 reverses lenvatinib resistance in liver cancer cells by ubiquitinating and degrading YBX1 to inhibit the PI3K/AKT pathway. Trans Oncol. (2025) 53. doi: 10.1016/j.tranon.2025.102294
45. Wu Q, Nandi D, and Sharma D. TRIM-endous functional network of tripartite motif 29 (TRIM29) in cancer progression and beyond. Cancer Metastasis Rev. (2024) 44:16. doi: 10.1007/s10555-024-10226-2
46. Lv K, Li Q, Jiang N, and Chen Q. Role of TRIM29 in disease: What is and is not known. Int Immunopharmacol. (2025) 147. doi: 10.1016/j.intimp.2024.113983
47. Wang J, Wang L, Lu W, Farhataziz N, Gonzalez A, Xing J, et al. TRIM29 controls enteric RNA virus-induced intestinal inflammation by targeting NLRP6 and NLRP9b signaling pathways. Mucosal Immunol. (2025) 18:135–50. doi: 10.1016/j.mucimm.2024.10.004
Keywords: lipid metabolism, lysosome, machine learning, TRIM29, ulcerative colitis
Citation: Guo X-Z, Zhang L-Y, Tan Z-B, He H-D, Liu Y-F, Yu B-P and Dong W-G (2026) TRIM29 drives ulcerative colitis by disrupting lipid metabolism via lysosomal dysfunction: a multi-omics and experimental study. Front. Immunol. 16:1728932. doi: 10.3389/fimmu.2025.1728932
Received: 20 October 2025; Accepted: 15 December 2025; Revised: 11 December 2025;
Published: 12 January 2026.
Edited by:
Ruano Juan, University of Cordoba, SpainReviewed by:
Qin-Wei Wu, Anhui University, ChinaMaciej Piernik, Poznań University of Technology, Poland
Copyright © 2026 Guo, Zhang, Tan, He, Liu, Yu and Dong. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Wei-Guo Dong, ZG9uZ3dlaWd1b0B3aHUuZWR1LmNu; Bao-Ping Yu, eXVicDYyQDE2My5jb20=
†These authors share first authorship