Elovl6 inhibits colorectal cancer progression through stearic acid–mediated mitochondrial fusion and metabolic reprogramming

Abstract

Lipid metabolic reprogramming is a hallmark of colorectal cancer (CRC), yet the precise molecular mechanisms underlying lipid-mediated oncogenesis and the specific lipid metabolic enzymes involved remain largely elusive. Here, we identify elongation of very-long-chain fatty acid protein 6 (Elovl6) as a critical regulator in CRC progression. Clinical data reveal significant down-regulation of Elovl6 in colon cancer tissues, with low expression levels correlating with unfavorable patient prognosis. We demonstrate that Elovl6 exerts potent tumor-suppressive effects, significantly inhibiting cellular proliferation in vitro and attenuating tumor growth in vivo. Mechanistically, it maintains intestinal microbial homeostasis by preventing the expansion of opportunistic pathogens while simultaneously orchestrating metabolic reprogramming through modulation of phospholipid biosynthesis pathways. Notably, we find that stearic acid, a key Elovl6-derived metabolite, promotes mitochondrial fusion by stabilizing mitofusin 1 protein. These findings not only position Elovl6 as a promising therapeutic target but also suggest that dietary supplementation with stearic acid could represent a viable strategy for CRC prevention and treatment.

Elovl6 suppresses CRC via stearic acid–mediated mitochondrial fusion and metabolic reprogramming.

INTRODUCTION

Colorectal cancer (CRC) is a common malignant tumor in the digestive system, ranking second and third in terms of incidence and mortality rates worldwide (1, 2). Although genetic predisposition and environmental factors are well-established contributors to CRC pathogenesis (3, 4), the molecular mechanisms underlying tumor progression remain incompletely elucidated. Current clinical management is guided by molecular profiling, using multimodal strategies that combine surgery with systemic chemotherapy, molecularly targeted agents, and immunotherapy for biomarker-selected patients (5, 6). Despite these therapeutic advances, 5-year survival rates remain unsatisfactory, particularly for patients with advanced disease (7). Faced with the persistent challenges of high incidence, mortality, and limited treatment efficacy, the identification of previously unidentified molecular targets and mechanistic insights into CRC pathogenesis is urgently needed to develop more effective therapeutic and preventive strategies.

Lipid metabolism reprogramming is increasingly recognized as a hallmark of cancer (8). Accumulating evidence suggests that dysregulation of lipid metabolism, such as abnormal lipid accumulation, lipogenesis, and lipolysis, is closely related to the progression of CRC disease (9–11). Notably, specific lipid species, including fatty acids (12), cholesterol derivatives (13), and oxylipins (14), have been implicated in CRC growth, survival, and metastatic dissemination. Consequently, key lipid metabolic enzymes have emerged as promising therapeutic targets (15, 16). Despite these advances, the precise molecular mechanisms by which lipid metabolic rewiring drives CRC pathogenesis remain incompletely elucidated, representing a critical barrier to the development of targeted metabolic therapies.

The elongation of very-long-chain fatty acid protein 6 (Elovl6) is a rate-limiting enzyme in fatty acid synthesis that catalyzes the elongation of saturated and monounsaturated fatty acids that contain 14, 16, and 18 carbons (17). Previous studies have shown that Elovl6 plays an important role in regulating obesity and related metabolic diseases, including insulin resistance (18), atherosclerosis (19), and nonalcoholic fatty liver disease (20). Recent research has demonstrated a strong association between Elovl6 and inflammatory diseases, including attenuated high-fat diet–induced hepatic inflammation (20), suppressed pulmonary fibrosis (21), and regulated mechanical skin damage (22). In addition, Elovl6 has been implicated in cell proliferation and apoptosis in various cancers, including CRC (23, 24). However, the precise mechanistic involvement of Elovl6 in CRC progression and its impact on metabolic reprogramming remain poorly understood and warrant further investigation.

In this study, we demonstrated that the expression of Elovl6 is decreased in tumor tissues compared with that in adjacent normal tissues in patients with CRC. To further investigate the function of Elovl6 in CRC, we constructed both Elovl6-knockout (KO) mice and colon-specific Elovl6-knockout (CKO) mice. Our findings confirmed that Elovl6 acts as a tumor suppressor in CRC in vivo. Mechanistically, we found that the knockout of Elovl6 leads to gut dysbiosis and metabolic changes that contribute to tumor development. Furthermore, Elovl6 regulated mitochondrial dynamics in a stearic acid (SA)–dependent manner by inhibiting the ubiquitination of mitofusin 1 (MFN1). These findings suggested that Elovl6 may serve as a previously unidentified prognostic biomarker and a potential therapeutic target in CRC.

RESULTS

Low Elovl6 expression correlated with poor survival of patients with colon cancer

To evaluate the role of Elovl6 in CRC, we first examined its expression in clinical colorectal samples by analyzing RNA sequencing data from The Cancer Genome Atlas (TCGA) database. The results indicated that Elovl6 mRNA levels were significantly down-regulated in CRC tissues compared with those in normal tissues (Fig. 1A). Tumor protein p53 (TP53) is the most frequently mutant gene in CRC. Among patients with TP53 mutations (n = 160), Elovl6 expression was found to be lower than in those without mutations (n = 122) (Fig. 1B). In addition, we assessed Elovl6 expression in CRC tissues at different stages and found that significantly increased Elovl6 expression was only present in the primary CRC samples (N0 stage without lymph node and distant metastasis) (Fig. 1C). Furthermore, the level of Elovl6 expression correlated with tumor progression, from TNM stages I to IV (Fig. 1D). Kaplan-Meier survival analysis indicated a positive association between Elovl6 expression levels and overall survival in patients with colon cancer. This association was particularly significant in patients with stage I and III disease, where high Elovl6 expression predicted improved survival outcomes (Fig. 1E). In addition, we examined Elovl6 protein expression in randomly selected patients with CRC. The results showed that 14 of the 20 tumor samples exhibited decreased Elovl6 protein levels compared with adjacent normal tissues (Fig. 1F). Collectively, these results suggested that Elovl6 might be implicated in CRC progression and could be a potential favorable prognostic indicator for patients with CRC.

Fig. 1. Low Elovl6 expression correlated with poor survival of patients with colon cancer.

(A) Elovl6 mRNA expression was determined in unpaired tumor (n = 286) and adjacent normal tissues (n = 41) from the TCGA cohort. (B and C) Patients (n = 286) were stratified on the basis of TP53 mutation status (B) or lymph node metastasis status (C) (N0, no lymph node metastasis; N1, 1 to 3 lymph node metastasis; and N2, ≥4 lymph node metastasis). (D) The expression levels of Elovl6 in colon cancer of different TNM stages: stage I (n = 103), stage II (n = 217), stage III (n = 173), and stage IV (n = 84). (E) Overall survival of patients with colon cancer stratified by high or low Elovl6 expression (using a 50% cutoff). (F) The expression of Elovl6 protein was assessed in paired tumor (T) and adjacent normal (N) tissues (n = 20) using Western blotting. P values were determined using a two-tailed Student’s t test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001. COAD, colon adenocarcinom.

Elovl6 inhibited cell proliferation and tumorigenesis of CRC

To further elucidate the functional role of Elovl6 in CRC pathogenesis, we modulated its expression in HCT-116 and Caco-2 cell lines (Fig. 2A). Cell Counting Kit-8 (CCK-8) assays showed that Elovl6 silencing significantly promoted cell proliferation, while Elovl6 overexpression reversed this effect and restored proliferative capacity to control levels (Fig. 2B). Consistent with this, colony formation assays demonstrated that Elovl6 depletion markedly enhanced the clonogenic ability of CRC cells (Fig. 2C). Elovl6 knockdown had no significant effect on apoptosis but substantially increased the proportion of cells in S phase while reducing the G₁-phase population (fig. S1, A and B). Correspondingly, Western blot analysis demonstrated that Elovl6 knockdown up-regulated CDK1 and CDK2 expression and down-regulated p27 Kip1 (fig. S1C).

Fig. 2. Elovl6 inhibited cell proliferation and tumorigenesis of CRC.

(A) The knockdown efficiency of Elovl6 was assessed by Western blotting. (B and C) Elovl6 knockdown in HCT-116 and Caco-2 cells promoted cell viability (B) and colony formation (C). (D) Body weight of Elovl6^−/− mice and littermate control wild-type (WT) mice was measured. (E) Three months after administration of AOM/DSS, Elovl6^−/− exhibited severe tumor–related phenotypes, including obvious rectal prolapse, compared with littermate control WT mice. (F) The formation of solid tumors in colon tissues was observed visibly in Elovl6^−/− mice and littermate control WT mice by the naked eye. (G) Tumor number. (H) Tumor volume. P values were determined using a two-tailed unpaired Student’s t test for comparisons between two groups or ordinary one-way analysis of variance (ANOVA) with Tukey’s post hoc test for multiple comparisons. *P < 0.05; **P < 0.01; ***P < 0.001; ****P< 0.0001. NC, negative control; KD, Elovl6 knockdown.

To investigate the physiological role of Elovl6 in CRC pathogenesis in vivo, we generated Elovl6-KO (Elovl6^−/−) mice on a C57BL/6J background (fig. S2A). Comprehensive analysis confirmed successful ablation of Elovl6 expression across all examined tissues (fig. S2B). Phenotypic characterization revealed that Elovl6 deficiency resulted in significantly reduced body weight compared with wild-type (WT) littermates (fig. S2, C and D). In agreement with previous findings (18), Elovl6^−/− mice demonstrated markedly improved insulin sensitivity (fig. S2, E to G). Notably, gross anatomical analysis showed no significant differences in organ mass indices for major organs including liver, spleen, and kidneys (fig. S3, A and B). Comprehensive hematological and biochemical profiling showed that Elovl6 ablation kept blood parameters and hepatorenal function within normal physiological ranges (table S1 and fig. S3C). However, Elovl6^−/− mice exhibited changes in their lipid profile compared with WT mice, particularly affecting triglycerides and total cholesterol levels (fig. S3D). Most notably, gas chromatographic analysis demonstrated substantial remodeling of hepatic fatty acid profiles, with particularly pronounced effects on the C16/C18 fatty acid ratio (fig. S3E).

Genetic ablation of Elovl6 modestly exacerbated acute colitis progression, augmenting inflammatory severity in the small intestinal mucosa (fig. S4). In the azoxymethane/dextran sulfate sodium (AOM/DSS)–induced colorectal carcinogenesis model, Elovl6 deficiency not only accelerated colonic tumorigenesis but also induced hepatic steatosis. Pathological examination revealed that Elovl6^−/− mice developed pronounced colonic hyperemia and significant colon shortening following the third DSS cycle compared with WT controls (fig. S5, A to C). In addition, Elovl6^−/− mice exhibited marked splenic necrosis and hepatic lipid accumulation (fig. S5, D to G). Three months post–AOM/DSS induction, Elovl6^−/− mice exhibited significant body weight loss and increased rectal prolapse incidence (Fig. 2, D and E). As expected, Elovl6^−/− groups showed more tumors and larger tumor volumes compared with their control mice (Fig. 2, F and G).

Colon-specific deletion of Elovl6 in mice accelerated colorectal tumorigenesis

To elucidate the tumor-suppressive role of Elovl6 in colorectal carcinogenesis, we generated colon-specific Elovl6-KO mice (CKO) through Cre-loxP–mediated recombination between Elovl6^loxp/loxp and Villin-Cre mice (Fig. 3A and fig. S6, A to C). Successful ablation of colonic Elovl6 expression was confirmed by reverse transcription quantitative real-time fluorescence polymerase chain reaction (RT-qPCR) and Western blotting (fig. S6, D and E). Physiological assessment revealed no significant alterations in body weight, hepatic/renal function, or serum lipid profiles between CKO and control mice (fig. S7). In DSS-induced colitis, CKO mice exhibited exacerbated inflammation, characterized by increased disease activity index (P < 0.05), colon shortening (P < 0.01), tissue damage, and elevated histopathology scores (P < 0.05) (fig. S8, A to E). Notably, male CKO mice exhibited higher mortality rates in the DSS-induced colitis model, demonstrating enhanced susceptibility to DSS-mediated toxicity (fig. S8F). To investigate tumorigenesis, we subjected CKO mice and littermate controls to the AOM/DSS carcinogenesis protocol (Fig. 3B). Endoscopic evaluation demonstrated notably greater colonic mucosal damage in CKO mice, with universal tumor development by day 90 (Fig. 3C). Quantitative analysis confirmed significant differences in tumor numbers and tumor volumes between the CKO and WT groups (Fig. 3, D to F). By day 120, CKO mice showed widespread colonic tumors, while WT mice only had tumors in the distal colons (Fig. 3G). Tumor progression was markedly accelerated in CKO mice, as evidenced by heightened tumor numbers (Fig. 3H), increased tumor volumes (Fig. 3I), and elevated histopathology scores (Fig. 3J). To extend these findings, we also generated Apc-mutant mice with an Elovl6 deletion. Notably, compared with the AOM/DSS model, the APC^min/+ spontaneous model itself exhibited a consistently milder inflammatory response, with significantly lower expression of tumor necrosis factor–α, interleukin-6 (IL-6), and IL-1β in both serum and colon tissue (fig. S9). Despite this distinct inflammatory landscape, genetic ablation of Elovl6 in the Apc-mutant background also robustly enhanced tumorigenesis (fig. S10). Complementary genetic approaches using conventional and conditional Elovl6-KO mice consistently demonstrated increased tumor burden, establishing Elovl6 as a previously unidentified tumor suppressor in CRC pathogenesis.

Fig. 3. Colon-specific deletion of Elovl6 in mice accelerated colorectal tumorigenesis.

(A) Schematic diagram of the strategy for generating intestinal-specific Elovl6-KO (CKO) mice. (B) Schematic representation of the AOM/DSS-induced colitis-associated colon cancer mouse model. BW, body weight; IP, intraperitoneal. (C) Representative colonoscopy images from mice at days 0, 50, and 90. CKO indicates intestinal-specific Elovl6-KO mice; loxp^+/+ indicates littermate WT mice. (D) Representative images of colon tumors from loxp^+/+ and CKO mice on day 90 post–AOM/DSS treatment. (E and F) Quantification of tumor number (E) and tumor volume (F) in colon tissues from loxp^+/+ and CKO mice on day 90 post–AOM/DSS treatment (n ≥ 5). (G) Representative images of colon tumors from loxp^+/+ and CKO mice on day 120 post–AOM/DSS treatment. (H and I) Quantification of tumor number (H) and tumor volume (I) in colon tissues from loxp^+/+ and CKO mice on day 120 post–AOM/DSS treatment (n = 5). (J) Colonic pathology score of loxp^+/+ and CKO mice on day 120 post–AOM/DSS treatment. P values were determined using a two-tailed unpaired Student’s t test. *P < 0.05; ***P < 0.001.

Elovl6-KO mice displayed gut microbiota dysbiosis

Growing evidence implicates gut microbiota dysbiosis as a major contributor to colorectal tumorigenesis. We hypothesized that Elovl6 deletion might exacerbate tumor progression by altering intestinal microbial homeostasis. To test this, we performed longitudinal 16S ribosomal RNA (rRNA) sequencing on fecal samples collected from KO and WT mice at baseline (day 0), intermediate (day 50), and end point (day 90) of AOM/DSS treatment. Alpha diversity analysis revealed significantly higher microbial diversity in KO mice compared with that in controls (Shannon and Simpson indices, P < 0.05; Fig. 4A). Notably, while WT mice exhibited progressive reduction in microbial richness [abundance-based coverage estimator (ACE), Chao1, and Observed species indices] during CRC development, KO mice demonstrated the opposite trend (Fig. 4A). Beta diversity analysis indicated that the composition of microbial communities in the WT-0 group was clearly distinct from all other groups, among which no statistically significant differences were detected (Fig. 4B). Furthermore, we analyzed the differences in microbiome composition between KO and control mice. At the phylum level, both groups maintained Bacteroidota and Firmicutes as dominant phyla, but KO mice displayed a markedly reduced Firmicutes/Bacteroidota ratio (Fig. 4, C and D). Microbiome profiling revealed a significant enrichment of genera associated with intestinal inflammatory conditions in KO mice, including Alistipes, Bacteroides, Parasutterella, Alloprevotella, and Prevotellaceae_UCG-001 (fig. S11). To test whether the microbiota from tumor-bearing Elov6^−/− mice has an enhanced ability to drive tumorigenesis, we performed fecal microbiota transplantation (FMT). Fecal samples from donor Elovl6^−/− mice that had been subjected to AOM/DSS treatment were transferred into antibiotic-pretreated WT recipient mice (Fig. 4E). The oral gavage of different stools had no impact on body weight (Fig. 4F). Within 2 months, mice receiving Elovl6^−/− microbiota exhibited a notably higher tumor incidence (10 of 11 mice, 91%) compared with those receiving WT microbiota (3 of 11 mice, 27%). Consistently, both tumor number and volume were significantly increased in the Elovl6^−/−-FMT group (Fig. 4, G to I). These results demonstrated that the change of gut microbial composition induced by Elovl6 deficiency could at least, in part, contribute to CRC development.

Fig. 4. Elovl6-KO mice displayed gut dysbiosis.

We used fecal samples from mice collected at days 0, 50, and 90 in the AOM/DSS-induced colitis-associated colon cancer mouse model to analyze gut microbial communities via 16S rRNA sequencing (n = 4). (A) Alpha diversity was measured by the Shannon, Simpson, ACE, Chao1, and Observed species indices. (B) Principal components analysis (PCA) showing beta diversity. (C) Microbial community composition at the phylum level for each mouse. (D) Firmicutes-to-Bacteroidetes (F/B) ratios in the Elovl6^−/− (KO) mice and littermate control WT mice. (E) Flowchart of the FMT experiment. (F) Body weight of recipient mice that received fecal microbiota from KO or WT donors. (G) Representative images of colon tumors from recipient mice after AOM/DSS treatment. (H and I) Quantification of tumor number (H) and tumor volume (I) in colon tissues from these recipient mice (n = 11 mice per group). P values were determined using a two-tailed unpaired Student’s t test for comparisons between two groups or ordinary one-way ANOVA with Tukey’s post hoc test for multiple comparisons. *P < 0.05.

Elovl6-KO mice showed altered gut metabolomics

Given the established metabolic versatility of enriched pathogenic microbiota and the potential involvement of Elovl6 in metabolic regulation, we conducted comprehensive nontargeted metabolomic profiling on fecal specimens from both the AOM/DSS and APC^min/+ models. Principal components analysis (PCA) revealed a consistent trend of metabolic shift upon tumor development in both models. In the AOM/DSS model, the metabolomes of WT and KO mice were distinctly separated at baseline but clustered in closer proximity following tumor induction (Fig. 5, A and B). This pattern was mirrored in the APC^min/+ model, where non–tumor-bearing control and KO mice were fully segregated, but tumor-bearing mice from both genotypes showed substantial overlap in their metabolomes (fig. S12A). Differential metabolite analysis quantified genotype-specific alterations across models: In the AOM/DSS model, KO mice exhibited 1258 (564 up and 694 down) and 726 (327 up and 399 down) changes at baseline and tumor stages, respectively (Fig. 5, C and D); a comparable scale of dysregulation was confirmed in the APC^min/+ model (fig. S12B). Crucially, pathway enrichment analysis demonstrated remarkable concordance between the two genetically distinct models. In the non–tumor-bearing state, differential metabolites in both models were consistently enriched in a core set of pathways, including central carbon metabolism in cancer; adenosine triphosphate (ATP)–binding cassette transporters; alanine, aspartate, and glutamate metabolism; linoleic acid metabolism; β-alanine metabolism; and cysteine and methionine metabolism (Fig. 5E and fig. S12C). During tumorigenesis, the metabolic differences between genotypes in both models further converged on shared pathway alterations, most notably in primary bile acid biosynthesis; glycine, serine, and threonine metabolism; and sphingolipid metabolism (Fig. 5F and fig. S12D).

Fig. 5. Elovl6 KO altered the gut metabolome.

Global metabolomics profiling of stool samples from Elovl6^−/− (KO) and WT littermate controls was performed by liquid chromatography–mass spectrometry (LC-MS; n = 4 per group). (A and B) PCA of metabolites from WT and KO mice at day 0 (A) and day 90 (B) after AOM/DSS treatment. (C and D) Volcano plots illustrating differential metabolites between WT and KO mice at day 0 (C) and day 90 (D). Differential metabolites were screened with variable importance in projection (VIP) ≥ 1 from the OPLS-DA (orthogonal partial least-squares discriminant analysis) model, fold change (FC) > 1.0, and P < 0.05. (E and F) Pathway enrichment analysis of differential metabolites at day 0 (E) and day 90 (F) after AOM/DSS treatment. KEGG, Kyoto Encyclopedia of Genes and Genomes; ABC, adenosine triphosphate (ATP)–binding cassette; PPAR, peroxisome proliferator–activated receptor; CoA, coenzyme A; TCA, tricarboxylic acid. GABAergic, γ-aminobutyric acid–mediated. (G) Mfuzz analysis showing dynamic changes in metabolites during colon cancer progression. The x axis represents three time points: 0, 50, and 90 days after AOM/DSS treatment, for both WT and KO mice. The clustering reveals distinct temporal patterns between genotypes. (H) Pathway enrichment analysis of differentially lipid metabolites between KO and WT mice after AOM/DSS induction. GPI, glycosylphosphatidylinositol. (I) Heatmap of differential lipid metabolites between KO and WT mice after AOM/DSS induction. (J) Relative mRNA expression of phospholipid synthesis–related genes in HCT-116 and Caco-2 cells with or without Elovl6 knockdown, measured by RT-qPCR. P values were determined using a two-tailed Student’s t test for comparisons between two groups or ordinary one-way ANOVA with Tukey’s post hoc test for multiple comparisons. *P < 0.05; **P < 0.01; ***P < 0.001.

To elucidate the adaptive metabolic alterations in Elovl6-KO mice during AOM/DSS-induced colon carcinogenesis, we performed comparative metabolomic profiling at three critical phases: baseline (day 0), inflammation (day 50), and tumorigenesis (day 90). Notably, PCA showed distinct clustering of metabolic profiles between healthy and pathological states (inflammation and tumorigenesis) in WT mice. In contrast, the KO mice exhibited considerable overlap in their metabolic profiles across the three time points (fig. S13A). Kyoto Encyclopedia of Genes and Genomes pathway enrichment analysis revealed distinct metabolic adaptations in WT mice during disease progression. During inflammation, WT mice exhibited up-regulation of pentose and glucuronate interconversions, as well as cysteine and methionine metabolism, alongside down-regulation of nicotinate and nicotinamide metabolism and thermogenesis (fig. S13B). In the tumorigenic phase, they further up-regulated cysteine and methionine metabolism and arginine biosynthesis while suppressing primary bile acid biosynthesis and α-linolenic acid metabolism (fig. S13C). In contrast, Elovl6-KO mice displayed divergent metabolic reprogramming during tumorigenesis. Up-regulated metabolites were predominantly enriched in pyrimidine metabolism and linoleic acid metabolism under inflammatory conditions, whereas down-regulated metabolites were associated with arachidonic acid metabolism and primary bile acid biosynthesis (fig. S13D). Notably, compared with that in control mice, the primary bile acid biosynthesis pathway, which was down-regulated in KO mice by day 50, emerged as an up-regulated metabolic pathway by day 90 (fig. S13E). These results suggested a dynamic metabolic remodeling during the inflammation-to-cancer transition, highlighting the remarkable metabolic plasticity of Elovl6-deficient mice throughout disease progression.

We next performed time-series clustering analysis of all metabolites using Mfuzz (Fig. 5G). Our data supported the hypothesis that Elovl6 plays a crucial role in metabolic remodeling, particularly in lipid metabolism. Notably, ~50% of the metabolites showing opposite trends between the two genotypes were lipid-related species. To further investigate this phenomenon, we conducted a comprehensive lipidomic analysis of WT mice and Elovl6-KO mice. The analysis results showed that Elovl6 deletion predominantly affected glycerophospholipid metabolism (Fig. 5H). Specifically, Elovl6 KO induced significant up-regulation of 45 lipid species, including key phospholipids such as phosphatidylethanolamine (PE)(18:0p/20:4), sphingomyelin (SM)(d16:0/18:1), phosphatidylcholine (PC)(18:1p/16:0), and phosphatidylserine (PS)(18:0/22:6), while down-regulating 22 lipid species including PE(18:1p/16:0), PC(31:0), and SM(d34:1) (Fig. 5I). To investigate the cell-autonomous nature of the metabolic reprogramming observed in vivo, we conducted lipidomic profiling in Elovl6-knockdown human CRC cell lines HCT-116 and Caco-2. Elovl6 deficiency resulted in a significant increase in the total cellular abundance of glycerophospholipids in both lines (fig. S14, A and B). Concordantly, pathway analysis identified “Glycerophospholipid metabolism” as a markedly altered pathway in the knockdown cells (fig. S14C). However, heatmap analysis revealed that the specific types of glycerophospholipids up-regulated in the two cell lines differed from the in vivo data (fig. S14D). Furthermore, RT-qPCR analysis showed that Elovl6 knockdown significantly elevated the mRNA expression of key phospholipid biosynthesis–related genes (Fig. 5J and fig. S14E). To determine the functional capacity of accumulated glycerophospholipids to promote tumorigenesis, we used a supplementation-based approach in vitro and in vivo. Exogenous supplementation of four different glycerophospholipids significantly enhanced the proliferative capacity of murine colon cancer cell lines MC38 and CT26 in vitro (Fig. 6, A and B). In a subcutaneous tumor model, administration of individual glycerophospholipids markedly promoted in vivo tumor growth, as evidenced by significant increases in both tumor weight and volume (Fig. 6, C to F). Together, these comprehensive multiomics analyses and functional validation establish that Elovl6 deficiency drives profound metabolic reprogramming, particularly in glycerophospholipid metabolism, which drives accelerated colon tumorigenesis in mice.

Fig. 6. Exogenous glycerophospholipid supplementation promoted tumor growth in vitro and in vivo.

(A and B) MC38 (A) and CT26 (B) cells were cultured in medium supplemented with the specified glycerophospholipid (10 μg/ml) for 36 hours, and cell proliferation was assessed by CCK-8 assay. (C) Schematic diagram of the in vivo supplementation experiment. (D) Representative photographs of excised subcutaneous tumors from each experimental group. (E and F) Quantification of MC38 tumor weight (E) and tumor volume (F) in C57BL/6J mice treated with glycerophospholipids (n = 7 mice per group). P values for multiple comparisons were determined by ordinary one-way ANOVA with Tukey’s post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001; n.s., not significant.

Elovl6 regulated mitochondrial structure and function

Metabolic reprogramming during cellular activation exhibited strict dependence on mitochondrial architectural remodeling (25). To elucidate the role of Elovl6 in mitochondrial regulation, we first assessed key functional parameters. Elovl6 knockdown showed no significant effects on mitochondrial membrane potential, mitochondrial calcium levels, or cellular reactive oxygen species (ROS) levels (fig. S15). However, Seahorse Extracellular Flux analysis revealed that Elovl6 deficiency markedly enhanced mitochondrial oxidative capacity, as evidenced by elevated basal respiration, spare respiratory capacity, and ATP production (Fig. 7A). Corroborating these findings, cellular ATP levels were significantly increased in Elovl6-deficient cells (Fig. 7B), with consistent effects observed in Caco-2 cells (fig. S16, A and B). Notably, ultrastructural analysis by electron microscopy demonstrated that Elovl6 knockdown induced pronounced mitochondrial fragmentation, with substantially shortened organelles (Fig. 7C). Consistent with the in vitro results, colon-specific Elovl6-KO mice exhibited shortened mitochondrial length compared with the WT controls (fig. S16C). Supporting these observations, MitoTracker Red staining revealed both reduced mitochondrial length and disrupted network integrity in Elovl6-deficient cells (Fig. 7D). Given that mitochondrial architecture is dynamically regulated through balanced fission and fusion processes mediated by specialized guanosine triphosphatases (26, 27), we examined key regulators of fission/fusion homeostasis. Western blotting results showed that Elovl6 knockdown specifically down-regulated the fusion proteins MFN1 and MFN2, without altering the expression of the fission regulator dynamin-related protein 1 (DRP1) (Fig. 7E), indicating a shift toward mitochondrial fragmentation through impaired fusion capacity.

Fig. 7. Elovl6 regulated mitochondrial structure and function.

(A) Knockdown of Elovl6 enhanced mitochondrial respiration in HCT-116 cells, as determined by Seahorse Extracellular Flux analysis (n = 6). The oxygen consumption rate (OCR) was quantified. (B) Knockdown of Elovl6 increased total ATP levels in HCT-116 cells. (C) Representative transmission electron microscopy images of HCT-116-NC and HCT-116-KD^Elovl6 cells. Scale bars, 11. (D) Mitochondrial morphology in HCT-116 cells was visualized using MitoTracker Red staining following Elovl6 knockdown. Quantitative analysis of mitochondrial branch length was performed using ImageJ software. (E) Western blotting analysis of MFN1, MFN2, and DRP1 expression in n HCT-116 cells with or without Elovl6 knockdown. β-Actin was used as a loading control. P values for multiple comparisons were determined by ordinary one-way ANOVA with Tukey’s post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001.

SA treatment enhanced mitochondrial fusion capacity

In this study, we confirmed the interaction between Elovl6 and MFN1 on both ectopic and endogenous expression levels by coimmunoprecipitation assays (Fig. 8, A and B). Cycloheximide (CHX) chase experiments demonstrated that Elovl6 regulates MFN1 protein stability. As shown in Fig. 8C, Elovl6 knockdown markedly accelerated MFN1 degradation, whereas Elovl6 overexpression conferred protein stabilization. Further experimental investigations are required to elucidate the molecular details of Elovl6-MFN1 interaction. Notably, we identified SA (C18:0) as a specific modulator of Elovl6-MFN1–mediated mitochondrial fusion. The fatty acid–dependent nature of Elovl6’s regulation of mitochondrial dynamics was demonstrated by the inability of Elovl6 overexpression to rescue mitochondrial fragmentation under delipidated serum conditions. Notably, among various tested fatty acids (C16:0, C16:1, C18:0, C18:1, and C18:2), only C18:0 supplementation effectively restored mitochondrial network integrity (Fig. 8D and fig. S16D). Mechanistic investigations revealed that C18:0 modulates MFN1 stability through posttranslational modifications, as evidenced by reduced ubiquitination (Fig. 8, E and F) and prolonged protein half-life in CHX chase assays (Fig. 8G). Collectively, our findings demonstrated that Elovl6-derived SA serves as a critical metabolic regulator of the Elovl6-MFN1–mediated mitochondrial fusion pathway.

Fig. 8. SA treatment enhanced mitochondrial fusion capacity.

(A) Confirmation of Elovl6-MFN1interaction by coimmunoprecipitation. Human embryonic kidney (HEK) 293T cells were transfected, and immunoprecipitation (IP) was performed using an anti-Flag antibody. IgG, immunoglobulin G. HA, hemagglutinin tag. (B) Interaction between endogenous Elovl6 and MFN1 in HCT-116 cells. (C) HEK293T cells were transfected with small interfering RNA or overexpression plasmids of Elovl6 for 24 hours, after which the stability of MFN1 protein was analyzed using a cycloheximide (CHX) (20 μg/ml) assay. h, hours. (D) Mitochondrial morphology was observed by MitoTracker Red CMXRos staining and confocal microscopy. HCT-116 cells were divided into four groups: control (normal serum); KD + delipidated serum; KD + delipidated serum + Elovl6 rescue; and KD + delipidated serum + C18:0 (SA). Quantitative analysis of mitochondrial branch length was performed using ImageJ software. (E and F) HEK293T cells were grown in medium containing delipidated serum for 24 hours. MFN1 protein level (E) and MFN1 ubiquitination (F) were assayed by immunoblotting. (G) The stability of MFN1 in HEK293T cells was analyzed in the presence or absence of SA using a CHX (20 μg/ml) assay. P values for multiple comparisons were determined by ordinary one-way ANOVA with Tukey’s post hoc test. *P < 0.05; **P < 0.01; ****P < 0.0001.

SA supplementation suppressed colorectal tumorigenesis

To determine whether SA functions as the critical downstream effector of Elovl6-mediated tumor suppression, we conducted complementary rescue experiments. In HCT-116 and Caco-2 CRC cells, Elovl6 knockdown significantly enhanced proliferative capacity; this protumorigenic effect was fully abrogated upon SA supplementation, which restored cell viability to control levels (Fig. 9, A and B). To validate these findings in vivo, we performed dietary interventions in the AOM/DSS-induced CRC model. As shown in table S2, SA-treated mice (SA group) received a 1.5% SA-supplemented diet. In the Elovl6-KO mouse model, dietary SA supplementation significantly reduced tumor number and potently restrained tumor growth, with all tumors in SA-treated KO mice remaining below 20 mm³ compared with the large tumors (up to 54.95 mm³) observed in untreated KO controls (Fig. 9, C to E). Moreover, SA supplementation in WT mice significantly suppressed tumorigenesis, reducing both tumor number and volume (Fig. 9, C to E). Collectively, these results established SA as a pivotal tumor-suppressive metabolite, highlighting its potential therapeutic value for CRC intervention.

Fig. 9. SA supplementation suppressed colorectal tumorigenesis.

(A and B) Cell viability was assessed after Elovl6 knockdown and subsequent SA supplementation in HCT-116 (A) and Caco-2 (B) cells. (C and D) Quantification of colon tumor number (C) and tumor volume (D) in WT and Elovl6^−/− mice treated with SA or vehicle control (n ≥ 4). (E) Representative colon images from WT and Elovl6^−/− mice treated with SA or vehicle control (n ≥ 4). (F) Proposed model depicting the molecular mechanism by which Elovl6 attenuates CRC progression. P values for multiple comparisons were determined by ordinary one-way ANOVA with Tukey’s post hoc test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

DISCUSSION

In this study, we demonstrated Elovl6 as a previously unidentified tumor suppressor and favorable prognostic marker in CRC through integrated investigations using CRC cell lines, systemic Elovl6-KO mice, and intestine-specific Elovl6-deficient murine models. Our data indicated that genetic ablation of Elovl6 markedly enhances tumor progression in both the AOM/DSS-induced colitis-associated carcinogenesis model and a spontaneous adenomatous polyposis coli (APC) mutation–driven model. Mechanistic studies revealed that Elovl6 exerts its tumor-suppressive function by coordinately regulating cell-intrinsic metabolic fitness and mitochondrial fusion dynamics, thereby maintaining intestinal metabolic homeostasis and restraining malignant proliferation. These findings not only elucidate a previously unrecognized role of Elovl6 in CRC pathogenesis but also highlight its therapeutic potential as a molecular target for CRC intervention.

The tumor-suppressive role of Elovl6 is further underscored by its marked clinical relevance. Analyses across multiple patient cohorts consistently demonstrate that low Elovl6 expression serves as a powerful independent prognostic marker, correlating strongly with advanced tumor stage and reduced overall survival. Notably, we observed a clear clinical association between TP53 mutation status and diminished Elovl6 expression levels. This correlation links a key genetic driver of CRC with the down-regulation of a pivotal metabolic enzyme. The combined impact of TP53 mutation and low Elovl6 expression appears to foster a permissive microenvironment for tumor progression, likely through the disruption of lipid metabolic homeostasis (28, 29). Having established this clinical correlation, we next sought to determine its functional significance. Genetic ablation of Elovl6 markedly enhanced the proliferative capacity and colony formation of CRC cells. Consistent with these findings, both systemic and intestine-specific Elovl6-deficient mice developed significantly more severe tumors in the AOM/DSS-induced model, providing direct experimental evidence for its tumor-suppressive role. Elovl6 deficiency led to profound systemic metabolic alterations, evidenced by reduced body weight and improved insulin sensitivity even in non–tumor-bearing KO mice (18). Of particular significance, systemic Elovl6-KO mice exhibited profound hepatic steatosis in the colitis-associated CRC model. This finding carries substantial clinical implications given the well-established epidemiological association between nonalcoholic fatty liver disease and increased CRC risk (30). Collectively, these findings define Elovl6 as a pivotal tumor suppressor in colorectal carcinogenesis, operating at the critical interface of metabolic regulation and malignant progression.

Colorectal tumorigenesis involves complex interactions between tumor cells and their microenvironment, with the gut microbiota playing an increasingly recognized causal role in CRC development (31–34). Here, we demonstrate that genetic ablation of Elovl6 induces profound gut microbial dysbiosis in mice, characterized by enrichment of several bacterial genera—including Bacteroides, Parasutterella, Alloprevotella, and Prevotellaceae_UCG-001—that have been associated with gastrointestinal disorders (35–38). FMT from Elovl6-deficient mice into WT recipients was sufficient to promote tumorigenesis. It is important to note that our FMT approach used microbiota from tumor-bearing donors. While this demonstrates the functional tumor-promoting capacity of the established microbial community, the experimental design does not allow definitive distinction between whether the dysbiosis acts as a primary driver initiated by the host’s Elovl6-deficient metabolic state or emerges secondarily to the developed tumor microenvironment. Therefore, within the Elovl6-deficient context, the gut microbiota is best interpreted as a major modulator and amplifier of tumorigenesis, rather than its initiator. A key limitation is our incomplete understanding of the specific bacterial strains and host-microbe metabolic interactions responsible for the observed tumor-promoting phenotype. Future research should use metagenomic sequencing and culturomics to isolate and identify the key pathogenic bacteria, followed by gnotobiotic mouse models to validate their individual and synergistic contributions to CRC pathogenesis.

In parallel, and of central importance, Elovl6 deficiency drives extensive metabolic reprogramming, a hallmark of cancer (39, 40). Our comprehensive metabolomic analysis revealed profound perturbations in multiple pathways—most notably arachidonic acid metabolism, primary bile acid biosynthesis, pyrimidine metabolism, and linoleic acid metabolism—during colorectal tumorigenesis in Elovl6^−/− mice. We identified Elovl6 as a master regulator of glycerophospholipid metabolism, with 45 distinct lipid metabolites showing significant alterations. These differentially regulated lipids, essential for membrane dynamics and lipid-mediated signaling (41), were functionally validated as tumor promoters: Exogenous supplementation of these glycerophospholipids enhanced tumor cell proliferation in vitro and accelerated tumor growth in vivo. Collectively, these data establish host-centric metabolic dysregulation as the direct and cell-autonomous core pathway through which Elovl6 loss drives CRC pathogenesis. Future studies should be dedicated to identifying precise microbe-metabolite interactions within this network, which is crucial for further elucidating the functional cross-talk between these two key aspects of CRC pathogenesis.

Having established Elovl6’s role in reshaping the gut microbiome and the systemic metabolome, we next sought to determine how these perturbations converge at the subcellular level to drive tumorigenesis. Our investigation revealed that Elovl6 knockdown significantly enhances oxidative phosphorylation capacity in tumor cells, meeting their heightened metabolic demands during proliferation. We further found that this metabolic shift is underpinned by a profound alteration in mitochondrial architecture: Elovl6 deficiency induces mitochondrial fragmentation through the coordinated down-regulation of the fusion proteins MFN1 and MFN2. MFN1 and MFN2 have been shown to be functionally and mechanistically distinct, despite their high sequence identity and similarity (42–44). Here, we demonstrated that inhibition of MFN1-mediated mitochondrial fusion appears to represent a key mechanism through which Elovl6 regulates cellular metabolism. The mechanistic basis for MFN2 down-regulation and its functional contribution to the phenotype require further investigation. Future studies should use proteomic approaches to identify potential intermediary proteins and use MFN1/MFN2-specific rescue experiments in Elovl6-deficient cells to dissect their distinct roles. Notably, we identified SA, the direct enzymatic product of Elovl6, as a potential regulator of mitochondrial fusion processes. This finding aligns with previous studies that demonstrate SA’s role as a signaling molecule that modulates mitochondrial morphology and function in response to nutritional status (45, 46). Therapeutically, our in vivo experiments demonstrated that SA supplementation significantly reduces tumor burden in Elovl6-deficient hosts. Its efficacy in WT mice underscores a genotype-independent tumor-suppressive role, highlighting the broad translational potential of dietary SA supplementation.

In summary, our study establishes Elovl6 as a pivotal tumor suppressor in CRC. We delineate a core mechanism wherein Elovl6 maintains metabolic homeostasis and promotes mitochondrial fusion, thereby restraining proliferation, with its enzymatic product SA serving as a key mediator of these processes (Fig. 9F). These findings not only position Elovl6 as a promising therapeutic target but also suggest that dietary supplementation with SA or modulation of lipid metabolism could represent a viable strategy for CRC prevention and adjunct therapy.

MATERIALS AND METHODS

Human CRC samples

Human CRC specimens were collected in the Jiangsu Provincial People’s Hospital from 2018 to 2020. The inclusion criteria for patients are as follows: (i) clear imaging and pathological diagnosis, (ii) initial lumpectomy, (iii) no detectable other cancer or other intestinal diseases, and (iv) no familial adenomatous polyposis or hereditary nonpolyposis CRC syndrome. Samples were collected with informed consent from all participants. This study was approved by the Medical Ethics Committee of Nanjing University (approval no. AP202502220015) and complied with all relevant ethical regulations. Table S3 summarizes the information for a subset of patients.

Cell lines and cell culture

The human colorectal adenocarcinoma cell lines HCT-116 and Caco-2 were obtained from the American Type Culture Collection (USA). HCT-116 cells were cultured in Dulbecco’s modified Eagle’s medium (Gibco, USA) supplemented with 10% (v/v) fetal bovine serum (FBS; Gibco, USA). Caco-2 cells were cultured in minimum essential medium (Gibco, USA) containing 20% (v/v) FBS, penicillin (100 U/ml), streptomycin (100 μg/ml), and sodium pyruvate (1 μg/ml) (all from Gibco, USA). All cell lines were incubated at 37°C in a humidified atmosphere with 5% CO₂ and routinely tested for mycoplasma contamination to ensure cell line authenticity and purity.

Silencing of Elovl6 in HCT-116 and Caco-2 cells

Elovl6 KO in HCT-116 and Caco-2 cells was achieved using the CRISPR-Cas9 system. The single guide RNA (sgRNA) sequences were as follows: Elovl6-sgRNA #1: 5′-CAGAATGATGAATATCGTGTCACCTGGAAG-3′; and Elovl6-sgRNA #2: 5′-CGCTGACTCTTGCCGTCTTCAGGTAGGTCG-3′. Plasmid construction was performed using standard procedures, and the correct insertion of the sgRNA sequence was confirmed by DNA sequencing (Jinsirui Biotechnology Co. Ltd., Nanjing, China). HCT-116 and Caco-2 cells were then transfected and selected for stable Elovl6 knockdown using puromycin for more than 2 weeks using puromycin, with vector-transfected cells serving as controls. The efficacy of Elovl6 knockdown was assessed by RT-qPCR and Western blot analysis.

Cell proliferation assay

Cell proliferation was assessed using CCK-8 (Beyotime Biotechnology, China) according to the manufacturer’s instructions. Briefly, 24 hours after transfection with Elovl6 small interfering RNA, cells were seeded into 96-well plates at a density of 5 × 10³ cells per well. To evaluate the effect of SA on cell growth, cells were seeded in 96-well plates and, after 24 hours, treated with 200 μM bovine serum albumin–conjugated SA or an equal volume of vehicle control for 48 hours. Following the respective treatments, the medium was replaced with 100 μl of fresh medium containing 10 μl of CCK-8 reagent, and the cells were incubated for an additional 1 hour. The absorbance at 450 nm was measured using a Synergy H1 microplate reader (BioTek, USA).

Cell cycle and apoptosis assay

Cells were washed with phosphate-buffered saline (PBS), harvested, and then processed using a Cell Cycle and Apoptosis Analysis Kit (Beyotime Biotechnology, China) according to the manufacturer’s instructions. Subsequently, cell cycle distribution and apoptosis rate were analyzed using a NovoCyte flow cytometer (ACEA Biosciences Inc., USA). Data were acquired and processed with the NovoExpress software (Agilent Technologies, Santa Clara, USA).

Transgene constructs and generation of transgenic mice

Whole-body Elovl6-KO heterozygous mice were generated using CRISPR-Cas9 technology by Shanghai Model Organisms Center Inc. (Shanghai, China). Homozygous Elovl6^−/− mice were obtained by intercrossing these heterozygotes (Elovl6^+/−). Colon-specific Elovl6–conditional knockout (CKO) mice were generated using the Cre-loxP system. Specifically, Villin-Cre mice and C57BL/6J-Elovl6 em1(flox) mice were first generated by Shanghai Model Organisms Center Inc. The colon-specific Elovl6-CKO mice (Villin-Cre; Elovl6^loxp/loxp) were then produced by crossing the homozygous floxed (Elovl6^loxp/loxp) mice with Villin-Cre mice. The Apc ^Min/+ mice were obtained from the Model Animal Research Center of Nanjing University (Gem Pharmatech Co. Ltd.)

Acute colitis model and colitis-associated colon cancer model

To induce acute colitis, 7-week-old transgenic mice and littermate control mice were treated with 3% DSS in their drinking water for 1 week. A daily disease activity index was calculated for each mouse on the basis of percentage weight loss, stool consistency, and rectal bleeding. The scoring criteria were adapted from a previous study (47).

The colitis-associated colon cancer model was induced using the AOM/DSS protocol, as previously described (48) with modifications. Briefly, mice received a single intraperitoneal injection of AOM (10 mg/kg; Sigma-Aldrich, Hamburg, Germany). After 7 days, they were subjected to 1 cycle of 2.0% DSS in drinking water for 7 days, followed by 14 days of regular water. This cycle was repeated twice (Fig. 3B). Colonoscopy was performed using a high-resolution mouse endoscopic system (MiniScope 2 V, Yuyan Instruments, Shanghai, China) to monitor disease progression. All animal experiments in this study were reviewed and approved by the Nanjing University Animal Care and Use Committee (IACUC approval no. 2107005).

Subcutaneous tumor model and treatment

Male C57BL/6J mice (4 to 6 weeks old) were used to establish the subcutaneous tumor model. Under anesthesia induced by an intraperitoneal injection of sodium pentobarbital, each mouse was injected with 5 × 10⁵ MC38 colon carcinoma cells. When the tumor volume reached ~50 to 100 mm³, the tumor-bearing mice were randomly assigned into different groups. Mice then received subcutaneous injections of 100 μl of either sterile PBS (control) or 100 μl of solutions (100 μg/ml) of the following phospholipids: PC(18:0/22:6), PC(16:0/18:1), PE(18:0/20:4), and PS(18:0/22:6). The injections were administered twice a week for 3 weeks. Tumor volume and body weight were recorded regularly. At the end of the experiment, tumors were dissected and collected for further analysis. Tumor volume (in cubic millimeters) was calculated using the formula: Volume = ¹/₂ × length × width². In accordance with ethical guidelines and previous reports, the maximum permitted tumor volume (humane end point) was set at 4000 mm³.

Fecal microbiota analysis by 16S rRNA sequencing

Fecal samples were collected from transgenic mice and littermate control mice at multiple time points. Total genomic DNA was extracted from the samples using the QIAamp DNA Stool Mini Kit (QIAGEN, catalog no. 51504) according to the manufacturer’s protocol. Library construction and sequencing were conducted by Shanghai Ouyi Biotech Company (Shanghai, China). The V3-V4 hypervariable regions of the bacterial 16S rRNA gene were amplified with the primers 343F (5′-TACGGRAGGCAGCAG-3′) and 798R (5′-AGGGTATCTAATCCT-3′). Subsequent bioinformatics analysis was performed using the Omicsmart online platform (www.omicsmart.com/).

Fecal microbiota transplantation

For the FMT experiment, recipient mice were orally administered a broad-spectrum antibiotic cocktail for 2 weeks to deplete the gut microbiota and then divided into two groups (n = 11 per group). One group received fecal material from Elovl6 ^−/ − mice that had been subjected to the AOM/DSS model, while the other group received fecal material from WT littermate controls under the same conditions. To prepare the inoculum, fresh stool samples were collected and homogenized in sterile PBS (1 g in 5 ml). After centrifugation to remove large particulates, the supernatant was collected. Each recipient mouse then received 200 μl of the fecal suspension via oral gavage three times per week. Throughout the FMT period, all mice continued to undergo the AOM/DSS regimen to induce colorectal tumorigenesis.

Fecal metabolomics analysis by nontargeted metabolomic profiling

Sample preparation: Fecal samples were collected from transgenic mice and littermate control mice at different time points and immediately frozen at −80°C until analysis. The sample preparation protocol was slightly modified from a previously described method (49). Briefly, 60 mg of each stool sample was homogenized in 600 μl of methanol/water (4:1, v/v), incubated at −20°C for 2 min, and then ground at 60 Hz for 2 min. Following 10 min of sonication in an ice-water bath, samples were stored at −20°C for 30 min and then centrifuged (13,000 rpm, 4°C, 10 min). The resulting supernatant (200 μl) was transferred into liquid chromatography–mass spectrometry (LC-MS) vials, mixed with 300 μl of methanol/water (1:4, v/v), and stored at −40°C for 2 hours. Last, samples were centrifuged again (13,000 rpm, 4°C, 10 min) and filtered through 0.22-μm membranes before analysis.

LC-MS analysis: Chromatographic separation was performed on an ACQUITY UPLC system equipped with an ACQUITY UPLC HSS T3 column (100 mm by 2.1 mm, 1.8 μm; Waters, USA). The mobile phases consisted of (A) water with 0.1% formic acid and (B) acetonitrile with 0.1% formic acid. The elution gradient was programmed as follows: 0 to 5% B over 2 min; 5 to 30% B from 2 to 4 min; 30 to 50% B from 4 to 8 min, 50 to 80% B from 8 to 10 min; and 80 to 100% B from 10 to 14 min. The flow rate was 0.35 ml/min. Mass spectra were acquired using a VION ion mobility spectrometry quadrupole orthogonal acceleration–time-of-flight mass spectrometer (Waters, USA) in both electrospray ionization positive and negative ion modes. The mass range was set from mass/charge ratio (m/z) 70 to 1000. The resolution was set at 70,000 for full MS scans and 17,500 for higher-energy collision-induced dissociation (HCD) tandem MS scans. The mass spectrometer parameters were as follows: spray voltage, 3800 V (+) and 3000 V (−); sheath gas flow rate, 40 arb (+) and 35 arb (−); aux gas flow rate, 10 arb (+) and 8 arb (−); and capillary temperature, 320°C (+) and 320°C (−).

Data processing: The raw LC-MS data were processed using the software Progenesis QI (V2.3; Nonlinear Dynamics, Newcastle, UK) for baseline filtering, peak identification, integration, retention time correction, peak alignment, and normalization. Metabolites were identified by searching against the Human Metabolome Database, LipidMaps (V2.3), Metlin, EMDB, and PMDB. Differentially abundant metabolites between comparison groups were identified using either Welch’s t test or the Wilcoxon rank sum test. The resulting P values were adjusted for multiple testing using the Benjamini-Hochberg false discovery rate procedure. Bioinformatics analysis of these differential metabolites was conducted on the Omicsmart platform (www.omicsmart.com/).

Fecal lipidomics

Sample preparation: Fecal samples were collected from transgenic and littermate control mice at designated time points and immediately stored at −80°C until analysis. Sample processing was performed as previously described (50). Briefly, 30 mg of fecal samples was homogenized in 200 μl of methanol/water (1:1, v/v). For quantitative normalization and quality control, 300 μl of Lyso PC17:0 (0.1 mg/ml in methanol) was added as an internal standard. The mixture was then ground using steel beads, vortexed for 30 s, and sonicated for 10 min in an ice-water bath. After incubation at −20°C for 20 min, samples were centrifuged at 13,000 rpm for 10 min (4°C). The chloroform layer (200 μl) was carefully transferred to a new 1.5-ml microcentrifuge tube and subjected to a second extraction with 300 μl of chloroform/methanol (2:1, v/v). The combined organic phases were thoroughly mixed, and 200 μl of the final extract was evaporated under a stream of nitrogen gas. The dried residue was reconstituted in 300 μl of isopropanol/methanol (1:1, v/v), vortexed for 30 s, and sonicated for 3 min in an ice-water bath to ensure complete solubilization. Last, samples were filtered through a 0.22-μm organic-phase membrane filter before LC-MS analysis.

LC-MS analysis: Chromatographic separation was performed on an ACQUITY UPLC system equipped with an ACQUITY UPLC BEH C8 column (100 mm by 2.1 mm, 1.7 μm; Waters, USA). The binary gradient elution system consisted of mobile phase (A) acetonitrile/water (6:4, v:v, containing 10 mM ammonium formate) and (B) acetonitrile/isopropanol (1:9, v:v, containing 10 mM ammonium formate). The following gradient was used at a flow rate of 0.26 ml/min: 32% B at 0 min, maintained until 1.5 min; increased to 85% B by 15.5 min; then to 97% B by 15.6 min and held until 18 min; returned to 32% B by 18.1 min and held until 20 min for re-equilibration. The injection volume was 5 μl, and the autosampler temperature was maintained at 4°C. Mass spectrometric detection was performed using a Dionex U3000 ultrahigh performance liquid chromatograph (Thermo Fisher Scientific, USA) equipped with a heated electrospray ionization source (Thermo Fisher Scientific, USA). The instrument parameters were set as follows: heater temperature, 350°C; sheath gas flow rate, 50 arb; auxiliary gas flow rate, 15 arb; sweep gas flow rate, 1 arb; spray voltage, 3800 V (+) and 3000 V (−); capillary temperature, 320°C; MS1 scan ranges, 135 to 2000.

Data processing: The raw mass spectrometry data were processed using LipidSearch software for multi-stage mass spectrometry (MSⁿ) analysis and precise m/z determination of precursor ions. Lipid molecular structures and their adduct formation patterns were identified on the basis of precursor ion spectra and multistage mass spectrometry fragmentation. For relative quantification, all peak intensities were normalized to the internal standard within each sample. Differential lipid metabolites were identified using a two-tailed Student’s t test, with significance defined as a variable importance in projection score >1.0 and a P value <0.05. Subsequent bioinformatics analysis of these differential metabolites was performed using the Omicsmart platform (www.omicsmart.com/).

Transmission electron microscopy

Transmission electron microscopy sample processing and imaging were conducted by Wuhan Servicebio Technology Co. Ltd. according to standardized protocols. For sample preparation, cells and tissues were fixed with 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) at 4°C for 4 hours, followed by postfixation in 1% osmium tetroxide in the same buffer for 2 hours at room temperature. Ultrathin sections were cut using a diamond knife on an ultramicrotome, collected on 200-mesh copper grids, and sequentially stained with uranyl acetate and lead citrate for contrast enhancement. Images were acquired using a Hitachi HT-7800 transmission electron microscope operated at 80 kV.

Mitochondrial staining and imaging

Mitochondria were labeled using MitoTracker Red CMXRos (Beyotime Biotechnology, China; C1035) according to the manufacturer’s instructions. Briefly, cells were seeded in glass-bottom confocal dishes and cultured overnight. Before staining, the culture medium was aspirated, and cells were washed twice with prewarmed PBS. Cells were then incubated with 500 μl of 50 nM MitoTracker Red CMXRos in serum-free medium at 37°C for 20 min in the dark. Following mitochondrial staining, nuclei were counterstained with Hoechst 33342 (Beyotime Biotechnology, China; C1028) for 10 min at 37°C. After two washes with PBS, live-cell imaging was performed using a Zeiss LSM 880 confocal microscope equipped with Airyscan super-resolution capability (Zeiss, Germany).

Flow cytometry analysis

Mitochondrial ROS levels, mitochondrial membrane potential, and intracellular calcium concentrations were assessed by flow cytometry. Mitochondrial ROS was detected using the Reactive Oxygen Species Assay Kit (Beyotime Biotechnology, China; S0033S). Mitochondrial membrane potential was measured with the Mitochondrial Membrane Potential Assay Kit containing tetramethylrhodamine ethyl ester (TMRE) (Beyotime Biotechnology, China; C2001S). Intracellular calcium levels were quantified using the calcium-specific fluorescent probe Rhod-2 acetoxymethyl ester (Rhod-2 AM) (Invitrogen, CA, USA). Following the respective incubation periods, cells were washed, resuspended in PBS, and analyzed on a NovoCyte flow cytometer (ACEA Biosciences, USA). Data acquisition was performed using NovoExpress software.

Mitochondrial respiration analysis

Mitochondrial oxygen consumption rates were measured using a Seahorse XF96 Flux Analyzer (Agilent Technologies, USA) following established protocols from our previous work (51). Briefly, cells were seeded in XF96 microplates at an optimized density and equilibrated in prewarmed, substrate-supplemented XF assay medium (pH 7.4) for 1 hour at 37°C in a non-CO₂ incubator before measurements. The mitochondrial stress test was performed through sequential injection of the following compounds: 1.5 μM oligomycin to assess ATP-linked respiration, 0.5 μM carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone to measure maximal respiratory capacity, and 0.5 μM rotenone/antimycin A to determine nonmitochondrial oxygen consumption. All measurements were normalized to total cellular protein content as quantified by a bicinchoninic acid (BCA) assay (Thermo Fisher Scientific, USA; 23227). Data were acquired and analyzed using the Seahorse Wave Desktop Software (version 2.6, Agilent Technologies).

Coimmunoprecipitation analysis

Protein-protein interactions were analyzed by coimmunoprecipitation using the Protein A+G Magnetic Beads Immunoprecipitation Kit (Beyotime Biotechnology, China; P2179S). Briefly, cells were lysed in cold immunoprecipitation lysis buffer containing protease inhibitor cocktail and centrifuged at 12,000g for 10 min at 4°C. A small aliquot (10%) of the total cell lysate was saved as the “Input” control. The remaining lysate was incubated with the target primary antibody overnight at 4°C with gentle agitation. The antibody-protein complexes were then captured by incubation with Protein A/G magnetic beads for 2 hours at room temperature. After magnetic separation and removal of the supernatant, the bead-bound immunocomplexes were washed three times with cold lysis buffer. Last, the complexes were eluted by boiling in 1× SDS–polyacrylamide gel electrophoresis loading buffer for subsequent Western blot analysis.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 10.1.2. Quantitative data are presented as means ± SD. For comparisons between two groups, a two-tailed Student’s t test was used. Comparisons across multiple groups were analyzed by one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. A P value of less than 0.05 was considered statistically significant.

Supplementary Materials

This PDF file includes:

Figs. S1 to S16

Tables S1 to S3