LCAT exhibits subtype specific prognostic and immunometabolic associations in renal cell carcinoma Authors: Yanwen Li, Yubin Li Published: 2026 ABSTRACT AbstractBackgroundThe three major renal cell carcinoma subtypes—chromophobe renal cell carcinoma (KICH), clear cell renal cell carcinoma (KIRC), and papillary renal cell carcinoma (KIRP)—show marked differences in molecular programs and tumor microenvironmental features, supporting the need for subtype-resolved evaluation of candidate metabolic biomarkers.MethodsUsing TCGA-based transcriptomic and clinical data, we evaluated LCAT expression and its associations with overall survival (OS), disease-specific survival (DSS), and progression-free survival (PFS). We further assessed LCAT copy-number variation (CNV), DNA methylation, DNA methyltransferase correlations, m6A-related regulators, and relationships with the tumor immune microenvironment and functional pathways using established public platforms.ResultsLCAT expression was decreased in KICH but increased in KIRC and KIRP relative to normal kidney tissues. Kaplan–Meier analyses showed that low LCAT was associated with shorter OS/DSS/PFS in KICH, whereas high LCAT was associated with shorter OS/DSS/PFS in KIRC; in KIRP, higher LCAT was associated with improved DSS but not OS/PFS. Immune analyses revealed subtype-specific changes in ESTIMATE-derived scores, immune-cell enrichment, and immune-related signatures, with significantly lower StromalScore and ESTIMATEScore in the LCAT-high group in KIRP and altered immune-cell subsets across all three subtypes. LCAT correlated with multiple immunomodulators and MHC molecules. LCAT expression was not correlated with tumor mutational burden (TMB) in any subtype, and MSI showed a significant positive association with LCAT only in KICH. CNV analyses demonstrated a positive CNV–mRNA correlation (most pronounced in KIRP); CNV deletions were consistently associated with the poorest survival in KICH and KIRP, whereas CNV status was not prognostically significant in KIRC. DNA methylation was inversely associated with LCAT expression in KIRC, and higher methylation predicted worse OS only in KICH. LCAT expression was positively correlated with multiple m6A regulators in KIRC and KIRP but not in KICH.ConclusionLCAT displays subtype-dependent prognostic and immunometabolic associations in RCC and may serve as a candidate biomarker for stratification; further external-cohort and experimental validation is warranted.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1007/s12672-026-05349-7.Keywords: LCAT, Renal cell carcinoma, Immune modulation, Prognosis, Epigenetic regulation Introduction Renal cell carcinoma (RCC) is a biologically heterogeneous malignancy that encompasses several molecularly distinct entities, most prominently clear cell RCC (KIRC), papillary RCC (KIRP), and chromophobe RCC (KICH) [1–3]. Multi-omics profiling initiatives have demonstrated that these subtypes differ in driver events, transcriptional programs, and microenvironmental composition, indicating that biomarkers and therapeutic vulnerabilities are subtype-specific [1–3]. Targeted therapies and immune checkpoint inhibitors (ICIs) have improved outcomes for subsets of patients with advanced RCC, yet durable benefit remains inconsistent and resistance is common [4, 5]. This clinical variability underscores the need for robust biomarkers that integrate tumor-intrinsic pathways with tumor microenvironment (TME) states, particularly immune infiltration and immunosuppressive programs. Metabolic reprogramming is a hallmark of RCC and is tightly intertwined with immune regulation within the TME [6, 7]. Lipid and cholesterol handling is especially relevant in RCC, where altered cholesterol flux can influence tumor-cell signaling and immune-cell function [7]. Lecithin–cholesterol acyltransferase (LCAT) catalyzes the esterification of free cholesterol on lipoproteins and is essential for high-density lipoprotein maturation and reverse cholesterol transport [8]. Perturbations in cholesterol esterification pathways have been linked to cancer cell proliferation, stress adaptation, and immune modulation, suggesting that LCAT may connect lipid metabolism with antitumor immunity[7]. Gene expression and function of metabolic enzymes can be shaped by genomic and epigenomic alterations, including copy-number variation, DNA methylation, and RNA modifications such as N6-methyladenosine (m6A) [9]. In parallel, immunogenomic features—tumor mutational burden (TMB), microsatellite instability (MSI), and mismatch repair (MMR) status—are widely used to interpret tumor immunogenicity and response to ICIs [10–12]. However, whether LCAT is coordinated with these regulatory layers and how such relationships vary across RCC subtypes remain unclear. Here, we performed a systematic, subtype-stratified analysis of LCAT across KICH, KIRC, and KIRP using TCGA expression and clinical annotations [1–3]. Protein-level patterns were cross-checked using the Human Protein Atlas [13]. We interrogated associations between LCAT and patient survival, immune-related gene correlations (TIMER2.0), immune-cell infiltration and enrichment (CIBERSORT/GSVA/ssGSEA), and stromal/immune scores (ESTIMATE) [14–16]. We further explored potential regulatory layers and immunogenomic correlates, including DNA methylation, CNV, m6A regulators, TMB, MSI, and MMR genes [9–12]. Recent multi-omics and machine-learning studies in ccRCC and other tumor settings have shown that integrative analytical frameworks can define prognostic programs, characterize tumor heterogeneity, and connect molecular states with immune and therapeutic phenotypes [17–24]. Additional recent RCC-focused studies have further emphasized immune architecture, therapeutic-vulnerability stratification, metabolic progression, broader metabolic targeting, and the emerging cancer-related biology of LCAT [25–29]. However, subtype-resolved evaluation of a single lipid-metabolism gene across KICH, KIRC, and KIRP remains limited. Materials and methods LCAT expression profiling Immunohistochemistry images of LCAT in normal human tissues were retrieved from the Human Protein Atlas (HPA; https://www.proteinatlas.org) [13]. RNA-seq count matrices and corresponding clinical data for tumors and matched adjacent normal tissues were downloaded from The Cancer Genome Atlas (TCGA) [1–3]. Raw counts were converted to transcripts per million (TPM) and log2-transformed (log2[TPM + 1]) for normalization. LCAT expression levels in tumor and paired normal samples were compared. Survival analysis RNA-seq data linked to complete clinical annotations were extracted from TCGA using the Gene Set Cancer Analysis platform (https://guolab.wchscu.cn/GSCA/). Univariate Cox regression was applied to estimate the association between LCAT expression and three survival endpoints—overall survival (OS), disease-specific survival (DSS), and progression-free survival (PFS). Statistical computations were performed in R (v3.6.3) using the “survival” package (v3.2-10), and Kaplan-Meier curves were generated with “survminer” (v0.4.9). Because complete model-level outputs for multivariable adjustment and confidence intervals were not directly available from the online platform used in this exploratory analysis, only univariate Cox results are presented in the current study. Copy-number variation (CNV) analysis In GSCA, “LCAT” queries were conducted for TCGA tumors of the three subtypes of renal cell carcinoma. Three modules were selected: CNV summary, CNV & Expression, and CNV & Survival. GISTIC2.0-processed CNV data from 11 495 TCGA samples were used in this study. The bar charts depict the proportion of samples with heterozygous amplification (red), heterozygous deletion (brown), homozygous amplification (light green), homozygous deletion (dark green), or no CNV (gray). Spearman correlation between LCAT CNV and mRNA abundance in RCC was calculated, and the false discovery rate (FDR) was adjusted. Patients were stratified into wild-type (WT), amplification, or deletion groups, and survival differences were assessed by the log-rank test using the R “survival” package. TIMER2.0 database In TIMER2.0, the “Gene_Corr” module was queried for LCAT [14]. Expression matrices of 47 genes—including mismatch repair genes (EPCAM, MLH1, MSH2, MSH6, PMS2), DNA methyltransferases (DNMT1, DNMT3A, DNMT3B, DNMT3L), m6A regulators, and immune modulators—were retrieved for correlation analysis. Purity-adjusted Spearman correlation coefficients were calculated and visualized as heat maps using ggplot2 (v3.4.0). Immune-related analyses TPM matrices and clinical annotations for KICH, KIRC, and KIRP were obtained from TCGA via the Xiantao Academic zero-code platform [1–3]. Single-sample gene set enrichment analysis (ssGSEA) implemented in the GSVA package (v1.46.0) was used to quantify the relative enrichment of 24 immune-cell signatures [15]. In addition, the ESTIMATE algorithm was used to calculate ImmuneScore, StromalScore, and ESTIMATEScore, and the CIBERSORT algorithm was applied to estimate the relative proportions of 22 infiltrating immune cell types. Samples were divided into LCAT-high and LCAT-low groups, and immune scores were compared using the Wilcoxon rank-sum test. Tumor mutational burden (TMB) and microsatellite instability (MSI) data were downloaded from the Clinical Bioinformatics Home (https://www.aclbi.com), and correlations with LCAT expression were evaluated using Spearman’s method. Methylation analysis The Illumina methylation β-values and corresponding mRNA expression data were retrieved from TCGA via GSCA. When multiple CpG probes were mapped to LCAT, each probe was tested individually. Spearman’s correlation coefficient was used to assess the relationship between LCAT methylation and expression. Samples were dichotomized by median methylation level; survival differences between the high- and low-methylation groups were analyzed using Cox proportional hazards models and log-rank tests in R. To further characterize the methylation landscape of LCAT, the CpG site map and the correlations between individual CpG methylation sites and LCAT expression were additionally explored using the MEXPRESS platform (https://mexpress.ugent) based on TCGA data. Functional enrichment analysis Using CAMOIP (https://www.camoip.net), the “GSEA” module was invoked for TCGA-KICH, TCGA-KIRC, and TCGA-KIRP cohorts with LCAT as the target gene. Enrichment was performed for KEGG pathways and GO terms (biological process, cellular component, and molecular function) at an FDR < 0.05. Pathways were ranked by normalized enrichment score (NES), and the top 15 significant terms were visualized using clusterProfiler (v4.6.0). Results LCAT expression in normal kidneys and KICH, KIRC, and KIRP tumors LCAT mRNA expression differed markedly across RCC subtypes. Compared with normal kidney tissues, LCAT expression was significantly reduced in KICH but elevated in KIRC and KIRP (Fig. 1A–C). In paired analyses, tumor tissues exhibited lower LCAT levels in KICH and higher levels in KIRC, whereas paired differences in KIRP were not statistically significant despite an overall increase in unpaired comparisons (Fig. 1D–F). Representative immunohistochemical (IHC) images of LCAT protein expression in the three RCC subtypes (KICH, KIRC, and KIRP) are shown in Fig. 1G. These images were included as descriptive protein-level support for the transcriptomic findings rather than for quantitative comparison. Differential expression of LCAT in renal carcinoma tissues. A–C Differential expression of LCAT between normal tissues and KICH (A), KIRC (B), and KIRP (C) tumor tissues. D–F Differential expression of LCAT between tumor tissues and paired adjacent normal tissues in KICH (D), KIRC (E), and KIRP (F). G Representative Human Protein Atlas immunohistochemical images of LCAT protein expression in KICH, KIRC, and KIRP tissues. Panel G is provided as descriptive protein-level support only, and no statistical analysis was performed because quantitative raw data were not available from the public database. Statistical significance applies to panels A-F only: *p < 0.05, **p < 0.01, ***p < 0.001 Prognostic analysis of LCAT in KICH, KIRC and KIRP Tumors Survival analyses indicated subtype-dependent prognostic associations of LCAT. Kaplan–Meier analysis showed that patients with low LCAT expression in KICH had shorter OS, DSS, and PFS (Fig. 2A-C). In contrast, high LCAT expression in KIRC was associated with shorter OS, DSS, and PFS (Fig. 2D-F). In KIRP, higher LCAT expression was associated with improved DSS, while OS and PFS were not significantly different between expression groups (Fig. 2G-I). To provide quantitative support for these findings, the corresponding hazard ratios and P values from univariate Cox analysis are summarized in Table 1. The numbers of patients included in each survival analysis are provided in Fig. 2 and its legend. Table 1 is therefore presented as a summary of the available univariate Cox outputs returned by the online platform for the endpoints listed. The Cox outputs were directionally consistent with the Kaplan–Meier trends, although not every endpoint yielded a separately significant or exportable Cox estimate from the platform. Association of LCAT expression with overall survival (OS), disease-specific survival (DSS), and progression-free survival (PFS) in renal carcinoma patients. A–C Kaplan-Meier curves of OS (A), DSS (B), and PFS (C) for KICH tumors according to LCAT expression. D–F Kaplan-Meier curves of OS (D), DSS (E), and PFS (F) for KIRC tumors based on LCAT expression. G–I Kaplan-Meier curves of OS (G), DSS (H), and PFS (I) for KIRP tumors based on LCAT expression. The numbers of patients in the high- and low-expression groups are indicated in each panel. Specifically, KICH included 65 cases for OS (high, n = 32; low, n = 33), 63 cases for DSS (high, n = 32; low, n = 31), and 65 cases for PFS (high, n = 32; low, n = 33). KIRC included 535 cases for OS (high, n = 268; low, n = 267), 467 cases for DSS (high, n = 227; low, n = 240), and 531 cases for PFS (high, n = 265; low, n = 266). KIRP included 289 cases for OS (high, n = 144; low, n = 145), 273 cases for DSS (high, n = 135; low, n = 138), and 288 cases for PFS (high, n = 144; low, n = 144) Table 1Available univariate Cox regression outputs for the association between LCAT expression and survival outcomes in renal cancer subtypesCancer typeSurvival typeHazard ratioCox P valueLogrank P valueKICHOS0.10.038.61e-3KICHPFS0.260.050.03KICHDSS0.130.060.03KIRCOS1.582.9e-32.66e-3KIRCPFS1.340.030.03KIRCDSS1.510.030.03KIRPOS0.710.270.26KIRPPFS0.710.140.14Table 1 is restricted to the univariate Cox outputs returned by the online platform used in this exploratory analysis. Accordingly, not every Kaplan–Meier endpoint shown in Fig. 2 was accompanied by a separate exportable Cox estimate in the platform output Available univariate Cox regression outputs for the association between LCAT expression and survival outcomes in renal cancer subtypes Table 1 is restricted to the univariate Cox outputs returned by the online platform used in this exploratory analysis. Accordingly, not every Kaplan–Meier endpoint shown in Fig. 2 was accompanied by a separate exportable Cox estimate in the platform output Correlation between LCAT Expression and Immunity in KICH, KIRC and KIRP tumors Because LCAT showedopposite prognostic associations across RCC subtypes, we next examined whether these differences were accompanied by subtype-specific immune features. To reduce method-dependent bias, we evaluated the tumor immune microenvironment using three complementary approaches: ESTIMATE to assess overall stromal and immune content, CIBERSORT to estimate the relative proportions of 22 infiltrating immune cell subsets, and ssGSEA to quantify the enrichment of major immune-cell signatures. The ESTIMATE analysis (Fig. 3A–C) first provided an overview of the tumor microenvironment at the score level. In KICH, the StromalScore, ImmuneScore, and ESTIMATEScore were all numerically higher in the LCAT-high group than in the LCAT-low group, but none of these differences reached statistical significance. In KIRC, none of the three scores differed significantly between groups. In KIRP, the StromalScore and ESTIMATEScore were significantly lower in the LCAT-high group (P < 0.05), whereas the ImmuneScore did not differ significantly between groups. These findings suggest that the global stromal-immune profile associated with LCAT is modest in KICH and KIRC but more apparent in KIRP. Correlation between LCAT expression and immune infiltration in renal carcinoma. A-C Bar plots showing immune cell enrichment scores in the LCAT-high and LCAT-low groups for KICH (A), KIRC (B), and KIRP (C) tumor samples. D-F Distribution of scores for multiple immune cell types in LCAT-high and-low groups for KICH (D), KIRC (E), and KIRP (F) tumors. *p < 0.05, **p < 0.01, ***p < 0.001 We then used CIBERSORT to determine whether the relationship with LCAT became clearer at the level of individual immune-cell populations. As shown in Fig. 3D–F, the association between LCAT expression and the relative abundance of infiltrating immune cells varied substantially across the three RCC subtypes, supporting a subtype-dependent pattern of immune remodeling. The ssGSEA analysis (Fig. 3G–I) provided a more detailed view of immune-cell enrichment patterns. In KICH (Fig. 3G), the LCAT-high group showed significantly higher enrichment of CD8 + T cells, cytotoxic cells, NK CD56bright cells, NK CD56dim cells, NK cells, M1 macrophages, and Th1 cells (all P < 0.05), whereas T helper cells and Th2 cells were significantly reduced. In KIRC (Fig. 3H), the LCAT-high group showed increased enrichment of CD8 + T cells, cytotoxic cells, plasmacytoid dendritic cells (pDCs), and T helper cells, but reduced enrichment of activated dendritic cells (aDCs), B cells, dendritic cells (DCs), eosinophils, immature dendritic cells (iDCs), macrophages, mast cells, neutrophils, effector memory T cells (Tem), follicular helper T cells (TFH), gamma delta T cells (Tgd), and Th1, Th17, and Th2 cells. In KIRP (Fig. 3I), the LCAT-high group showed increased enrichment of eosinophils, mast cells, and T helper cells, whereas activated dendritic cells (aDCs), B cells, dendritic cells (DCs), immature dendritic cells (iDCs), macrophages, neutrophils, NK cells, gamma delta T cells (Tgd), Th1, Th17, Th2, Treg, and TFH cells were decreased. Taken together, the ESTIMATE, CIBERSORT, and ssGSEA results consistently indicate that the immune context associated with LCAT differs across RCC subtypes, although the magnitude of change ranges from broad score-level shifts to more selective alterations in specific immune-cell subsets. Because the cell-composition analyses suggested subtype-dependent immune remodeling, we next examined whether LCAT was also linked to broader immune regulatory programs. Correlation analysis showed that in KICH and KIRC, LCAT expression was positively associated with multiple curated immunostimulatory and immunosuppressive factors, whereas the dominant pattern in KIRP was negative correlation (Fig. 4A-B). Correlation analysis between LCAT expression and immune-related factors in renal carcinoma. A Immunostimulatory factors, B immunosuppressive factors, C immune checkpoints, and D MHC molecules in KICH, KIRC, and KIRP. *p < 0.05, **p < 0.01, ***p < 0.001 We then focused on immune checkpoint molecules. LCAT expression was significantly associated with multiple checkpoint genes across KICH, KIRC, and KIRP, although the detailed correlation patterns differed among subtypes (Fig. 4C). A similar subtype-dependent pattern was observed for MHC-related genes. LCAT expression was positively correlated with multiple MHC molecules in KICH, whereas predominantly negative correlations were evident in KIRC and KIRP (Fig. 4D). Together, these results indicate that LCAT is linked not only to differences in immune-cell abundance but also to differences in immune regulatory signaling across RCC subtypes. Association between LCAT and TMB/MSI Because correlations with immune-cell abundance do not necessarily reflect tumor immunogenicity, we next evaluated whether LCAT expression was related to tumor mutational burden (TMB) and microsatellite instability (MSI). TMB is commonly defined as the number of somatic coding base-substitution and indel mutations per megabase and is used as a proxy for neoantigen load [11]. Microsatellite instability (MSI) arises from slippage mutations at short tandem repeats caused by defective mismatch repair [12]. Both biomarkers have been linked to response to immune checkpoint blockade in several tumor types, particularly in mismatch-repair-deficient settings [10–12]. Scatter plots with fitted regression lines, 95% confidence intervals, and corresponding Spearman correlation coefficients are presented in Fig. 5. Across RCC subtypes, LCAT expression showed no significant association with TMB (Fig. 5A-C). For MSI, a significant positive correlation with LCAT was observed only in KICH, whereas no significant correlations were detected in KIRC or KIRP (Fig. 5D–F). Correlation of LCAT expression with TMB and MSI in renal carcinoma tissues. A–C Scatter plots showing the Spearman correlation between LCAT expression and TMB in A KICH, B KIRC, and C KIRP. D–F Scatter plots showing the Spearman correlation between LCAT expression and MSI in D KICH, E KIRC, and F KIRP. Fitted regression lines with 95% confidence intervals are shown in each panel, and the corresponding Spearman correlation coefficients are indicated in the figure. *p < 0.05, **p < 0.01, ***p < 0.001 We next profiled LCAT copy-number variation (CNV) across RCC subtypes and evaluated its relationships with expression and clinical outcomes. LCAT CNV events were detected in all three RCC subtypes (Fig. 6A–C). CNV status was positively correlated with LCAT mRNA expression in KICH, KIRC, and KIRP, with the strongest CNV–mRNA association observed in KIRP (Fig. 6D–F). Survival analysis by CNV state further showed that deletion was generally associated with the poorest outcomes in KICH and KIRP, whereas no clear prognostic stratification by CNV status was observed in KIRC (Fig. 6G–O). CNV distribution in different renal carcinoma tissues and its correlation with LCAT expression and survival. A–C Summary pie charts of LCAT CNV proportions in the KICH (A), KIRC (B), and KIRP (C) groups. D–F Scatter plots showing the correlation between LCAT CNV and mRNA expression in KICH (D), KIRC (E), and KIRP (F). G-I Survival outcomes (OS, DSS, and PFS) according to LCAT CNV in patients with KICH. J–L Survival outcomes (OS, DSS, and PFS) according to LCAT CNV in KIRC. M–O Survival outcomes (OS, DSS, and PFS) according to LCAT CNV in KIRP. P Circular correlation plots showing the association between LCAT expression and mismatch-repair (MMR) genes (PMS2, MSH6, MSH2, MLH1, and EPCAM) in KICH, KIRC, and KIRP; red and blue sectors indicate positive and negative Spearman correlations respectively. *p < 0.05, **p < 0.01, ***p < 0.001 Mismatch-repair (MMR) genes act as key guardians of genomic integrity; their impaired expression prevents timely correction of base mismatches during DNA replication, thereby driving genomic instability and mutational accumulation. We evaluated the correlation between LCAT expression and five MMR genes. In KICH, LCAT was significantly negatively correlated with MLH1 expression. In KIRC, LCAT showed significant negative correlations with MSH6 and EPCAM, and a significant positive correlation with PMS2. In KIRP, LCAT expression was significantly positively correlated with PMS2, MLH1, and EPCAM expression (Fig. 6P). Correlation analysis of LCAT methylation and m6A modification To further explore regulatory layers that might underlie subtype-specific LCAT expression, we next examined DNA methylation. DNA methylation analysis suggested that epigenetic regulation contributes to LCAT dysregulation. In KIRC, overall LCAT methylation levels were inversely correlated with LCAT mRNA expression, whereas in KICH and KIRP the overall methylation-expression correlations were weak and did not reach statistical significance (Fig. 7A–C). Prognostic evaluation showed that higher methylation was associated with poorer OS in KICH, while no significant OS differences were observed in KIRC or KIRP (Fig. 7D–F). Correlation analysis with DNA methyltransferases further revealed subtype-specific associations between LCAT expression and DNMT family members (Fig. 7G), suggesting that the methylation-related regulatory context of LCAT differs among RCC subtypes. To further characterize the methylation landscape of LCAT, the CpG site map and site-specific methylation-expression correlations are presented in Figure S1 and Table S1. Among individual CpG sites, cg04582263 and cg06839111 showed the strongest negative correlations with LCAT expression in KIRC, whereas cg07207367 and cg01817009 showed the strongest positive correlations in KIRP; no significant CpG-expression association was observed in KICH. LCAT methylation analysis in different renal carcinomas. A–C Correlation between LCAT methylation and mRNA expression in KICH (A), KIRC (B), and KIRP (C) groups. D–F Survival analysis comparing the high- and low-methylation groups in KICH (D), KIRC (E), and KIRP (F). G Correlation analysis between LCAT expression and DNA methyltransferases in KICH, KIRC, and KIRP tumors. The CpG methylation landscape of the LCAT gene and site-specific correlations between individual CpG probes and LCAT expression are shown in Figure S1 and Table S1. *p < 0.05, **p < 0.01, ***p < 0.001 We then assessed whether LCAT was also linked to the epitranscriptomic layer. LCAT expression showed predominantly positive correlations with multiple m6A regulators in KIRC and KIRP, whereas such correlations were weak and largely absent in KICH (Fig. 8). These findings suggest that epitranscriptomic regulation may contribute to subtype-specific LCAT-associated biology in RCC. Heatmap showing the correlation between LCAT expression and m6A regulators in KICH, KIRC, and KIRP tumors. *p < 0.05, **p < 0.01, ***p < 0.001 Functional pathways associated with LCAT expression in RCC subtypes To better characterize the biological programs associated with LCAT expression in the three RCC subtypes (KICH, KIRC, and KIRP), we stratified patients into LCAT-high and LCAT-low groups and performed functional enrichment analysis of the differentially expressed genes between the two groups. In KICH, the differentially expressed genes were significantly enriched in humoral immunity-related terms (Fig. 9A–D), including humoral immune response, B cell-mediated immunity, immunoglobulin-mediated immune response, regulation of complement activation, and Fc receptor-mediated stimulatory signaling. These findings suggest that the LCAT-high group is associated with a more humoral-immune-oriented transcriptional profile. GO-CC analysis showed enrichment in the immunoglobulin complex, circulating components, the external side of the plasma membrane, blood microparticles, and transmembrane transporter complexes, consistent with altered extracellular and membrane-associated immune functions. GO-MF terms such as antigen binding, immunoglobulin receptor binding, voltage-gated ion channel activity, and cation transmembrane transporter activity further support an association with antigen handling, ion transport, and signaling processes. KEGG analysis linked these genes to carbon metabolism, drug metabolism-cytochrome P450, and pentose and glucuronate interconversions, indicating that immune-related and metabolic programs coexist in LCAT-stratified KICH tumors rather than proving a directmechanistic effect of LCAT itself. Functional enrichment analysis associated with LCAT-stratified KICH, KIRC, and KIRP tumors. A–D Mountain plots showing GO-BP, GO-CC, GO-MF, and GSEA enrichment for the LCAT-high vs. LCAT-low groups in KICH. E–H Mountain plots showing GO-BP, GO-CC, GO-MF, and GSEA enrichment for the LCAT-high vs. LCAT-low groups in KIRC. I–L Mountain plots showing GO-BP, GO-CC, GO-MF, and GSEA enrichment for the LCAT-high vs. LCAT-low groups in KIRP In KIRC, the differentially expressed genes were significantly enriched in immune activation-related terms (Fig. 9E–H), including positive regulation of immune system processes, leukocyte-mediated immunity, myeloid leukocyte activation, neutrophil-mediated immunity, and platelet degranulation. This pattern suggests that the LCAT-high group is linked to an inflammatory and myeloid-enriched transcriptional state. GO-CC analysis highlighted vesicle membranes, cell surfaces, endosomes, lysosomal membranes, and lytic vacuoles, which is compatible with altered vesicle trafficking and granule-related immune functions. GO-MF enrichment for ion transmembrane transporter activity, oxidoreductase activity, lipid binding, and peptidase inhibitor activity further points to concurrent immune and metabolic adaptation. KEGG analysis associated these genes with complement and coagulation cascades, carbon metabolism, drug metabolism-cytochrome P450, steroid hormone biosynthesis, bile secretion, and retinol metabolism, indicating that LCAT expression in KIRC is linked to an immune-metabolic program rather than demonstrating that LCAT directly drives it. In KIRP, the differentially expressed genes were significantly enriched in immune- and hormone-related terms (Fig. 9I–L), including response to hormone, leukocyte-mediated immunity, response to bacteria, humoral immune response, complement activation (classical pathway), and phagocytosis recognition. These findings suggest that the LCAT-high group is associated with a mixed immune-signaling profile rather than a single defined mechanism. GO-CC analysis showed enrichment in vacuoles, lysosomes, lytic vacuoles, cell surfaces, extracellular matrices, and immunoglobulin complexes, consistent with altered immune-complex processing and extracellular interaction. GO-MF terms such as calcium ion binding, transmembrane transporter activity, oxidoreductase activity, antigen binding, and immunoglobulin receptor binding support possible links to calcium signaling, transport, redox balance, and antigen recognition. KEGG analysis further indicated associations with ion-transport-related processes, suggesting that LCAT-stratified KIRP tumors differ in immune and membrane-associated biology, although these results should not be interpreted as direct evidence that LCAT accelerates malignant progression. Discussion This study provides an integrative multi-omics view of LCAT across the three major RCC subtypes (KICH, KIRC, and KIRP) using TCGA-derived cohorts [1–3]. We observed pronounced subtype dependence in expression patterns, prognostic associations, immune contexture, and regulatory correlates. In particular, LCAT was downregulated in KICH but upregulated in KIRC and KIRP relative to normal kidney tissues (Fig. 1), and Kaplan–Meier analyses indicated opposite directions of association: low LCAT was linked to shorter OS/DSS/PFS in KICH, whereas high LCAT was linked to shorter OS/DSS/PFS in KIRC; in KIRP, higher LCAT was associated with improved DSS without significant OS/PFS differences (Fig. 2). Recent RCC-focused multi-omics studies have likewise shown that integrative pipelines can recover prognostic states, therapeutic-vulnerability signatures, and immune-response architectures [17, 21, 25, 26]. Additional reports have emphasized that intra-tumoral heterogeneity, metabolic progression, lipid-metabolic rewiring, and broader metabolic therapeutic targeting are clinically relevant dimensions of RCC biology [22–24, 27, 28]. Emerging pan-cancer analyses of LCAT also support the idea that this enzyme may intersect with epigenetic regulation and immune interactions in a context-dependent manner [29]. Within this broader context, our data suggest that LCAT may capture subtype-dependent biology rather than acting as a uniform marker across RCC histologies. LCAT is a central enzyme in HDL maturation and cholesterol esterification, and cholesterol flux is increasingly recognized as a determinant of cancer cell signaling and antitumor immunity [7, 8]. The divergent prognostic roles observed here may reflect distinct metabolic wiring across RCC subtypes. Clear cell RCC is characterized by prominent lipid storage and altered cholesterol handling, and integrated metabolomic analyses have highlighted extensive lipid remodeling in KIRC [6, 7]. In this context, higher LCAT expression may support a pro-tumor lipid phenotype, whereas in KICH and (partly) KIRP, LCAT may better reflect a more differentiated metabolic state or alternative microenvironmental constraints. Immune deconvolution and immune-gene correlation analyses suggest that LCAT stratifies immunometabolic states in a subtype-specific manner. The LCAT-high group showed reduced global immune-cell enrichment in KIRC and KIRP, while KICH showed no significant difference in ESTIMATE scores but displayed shifts in selected immune-cell subsets (Fig. 3), and LCAT correlated with multiple immune checkpoint molecules and antigen-presentation markers (Fig. 4). Notably, the positive immune-related correlations observed in both KICH and KIRC should not be interpreted as a biological contradiction to their opposite prognostic associations. A gene may show similar directional correlations with immune signatures across subtypes while still carrying different clinical implications, because the broader metabolic, stromal, and cellular contexts of those subtypes are distinct. This point may be particularly relevant in KICH, where LCAT showed a significant positive correlation with MSI and a favorable survival association. One possible interpretation is that LCAT-high KICH tumors retain a biologic setting in which immune visibility or genomic instability coexists with a less adverse immune-metabolic background than that seen in KIRC. By contrast, in KIRC, similar immune-related correlations may accompany a more unfavorable inflammatory-metabolic state. These findings may be clinically relevant given the central role of immune checkpoint blockade in advanced RCC management [4, 5, 10–12]. At the regulatory level, LCAT expression displayed concordant CNV–mRNA associations (most pronounced in KIRP) and subtype-specific prognostic effects of CNV states, with deletions generally tracking poorer outcomes in KICH and KIRP (Fig. 6). Epigenetically, LCAT methylation was strongly inversely correlated with mRNA expression in KIRC but not in KICH/KIRP, whereas methylation-based survival stratification was evident only in KICH (Fig. 7). LCAT also showed subtype-specific correlations with DNA methyltransferases, and prominent positive correlations were observed between LCAT and multiple m6A regulators in KIRC and KIRP (Figs. 7 and 8) [9]. Collectively, these data support a layered genetic–epigenetic–epitranscriptomic framework for LCAT heterogeneity that warrants mechanistic validation. The pathway-level results in Fig. 9 further support this subtype dependence: KICH was enriched for humoral immunity-, B-cell-, and complement-related programs; KIRC showed myeloid, neutrophil, platelet, and inflammatory-metabolic signatures; and KIRP displayed hormone-, lysosome-, and complement-associated patterns. Taken together, these differences are compatible with distinct immunometabolic states across RCC subtypes rather than a single shared LCAT-associated program. Several limitations should be acknowledged. First, the analyses are based on publicdatasets and web-platform outputs; although internally consistent with the figures presented, they remain susceptible to cohort composition, batch effects, and platform-specific algorithms. Second, the study lacks external clinical validation and does not establish causality between LCAT and immune/metabolic phenotypes. Third, whether LCAT can predict response to specific systemic therapies (e.g., immune checkpoint blockade or anti-angiogenic regimens) requires treatment-annotated cohorts and prospective evaluation [4, 5, 10]. Conclusions LCAT exhibits subtype-dependent associations with prognosis, immune contexture, and epigenetic regulation in renal cell carcinoma and may serve as a candidate multidimensional biomarker for risk stratification and biological characterization across RCC subtypes. Given the bioinformatics nature of this work, external cohort validation and functional experiments are needed to clarify the causal role of LCAT and to assess whether modulating LCAT-related pathways has therapeutic value. The representative HPA-based immunohistochemical images shown in Fig. 1G provide descriptive protein-level support for these observations but should not be interpreted as quantitative evidence. Supplementary Information Supplementary Material 1. Acknowledgements Thanks to all researchers and editors who contributed to this study. Author contributions Conceptualization, Li YW and Li YB; Writing—original draft preparation, Li YW; Visualization, Li YW; Methodology, Li YW; Software, Li YW; Validation, Li YW; Formal analysis, Li YW; Investigation, Li YW; Resources, Li YW; Data curation, Li YW; Supervision, Li YW; Project administration, Li YB; Funding acquisition, Li YB. Writing—review and editing: Li YB. All the authors have read and agreed to the published version of the manuscript. Funding This research was entirely self-funded by the author. Data availability All data analyzed in this study are publicly available.TCGA RNA sequencing gene expression quantification, clinical annotations, DNA methylation β-values, and somatic mutation MAF files for kidney chromophobe (TCGA-KICH), kidney renal clear cell carcinoma (TCGA-KIRC), and kidney renal papillary cell carcinoma (TCGA-KIRP) were downloaded from the NCI Genomic Data Commons Data Portal: https://portal.gdc.cancer.gov/projects/TCGA-KICH, https://portal.gdc.cancer.gov/projects/TCGA-KIRC, and https://portal.gdc.cancer.gov/projects/TCGA-KIRP.Immunohistochemistry images for LCAT were obtained from the Human Protein Atlas: https://www.proteinatlas.org/ENSG00000213398-LCAT and kidney tissue staining at https://www.proteinatlas.org/ENSG00000213398-LCAT/tissue/kidney.Gene-level GISTIC2.0 copy-number calls were accessed through GSCA and are available from the UCSC Xena Pan-Cancer Atlas hub (dataset TCGA.PANCAN.sampleMap/Gistic2_CopyNumber_Gistic2_all_thresholded.by_genes; direct download https://tcga.xenahubs.net/download/TCGA.PANCAN.sampleMap/Gistic2_CopyNumber_Gistic2_all_thresholded.by_genes.gz).Immune-related analyses were performed using TIMER2.0 as an analysis portal. The immune infiltration estimation matrix used by TIMER2.0 is publicly available at https://compbio.cn/timer2/infiltration_estimation_for_tcga.csv.gz. All gene expression matrices used in TIMER2.0 analyses are the TCGA RNA-seq data retrieved from the GDC projects listed above; the specific correlation outputs used in this study are provided in the Supplementary Data submitted with this manuscript.The ssGSEA input matrices used in Xiantao Academic were derived from the TCGA RNA-seq expression matrices downloaded from the GDC projects listed above. The exact ssGSEA input matrices and resulting ssGSEA scores used in this study are provided in the Supplementary Data submitted with this manuscript.Tumor mutational burden and microsatellite instability values were derived from TCGA somatic mutation MAF files downloaded from the GDC projects listed above. The per-sample TMB and MSI tables used for downstream analyses are provided in the Supplementary Data submitted with this manuscript.Functional enrichment results were generated using CAMOIP based on the same TCGA cohorts. The full enrichment output tables used in this study are provided in the Supplementary Data submitted with this manuscript. No new sequencing data were generated in this study. Declarations Ethics approval and consent to participate All analyses were performed using publicly available datasets and online resources. No new human participants or animal experiments were involved; therefore, ethical approval was notrequired. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Citations Data Availability Statement All data analyzed in this study are publicly available.TCGA RNA sequencing gene expression quantification, clinical annotations, DNA methylation β-values, and somatic mutation MAF files for kidney chromophobe (TCGA-KICH), kidney renal clear cell carcinoma (TCGA-KIRC), and kidney renal papillary cell carcinoma (TCGA-KIRP) were downloaded from the NCI Genomic Data Commons Data Portal: https://portal.gdc.cancer.gov/projects/TCGA-KICH, https://portal.gdc.cancer.gov/projects/TCGA-KIRC, and https://portal.gdc.cancer.gov/projects/TCGA-KIRP.Immunohistochemistry images for LCAT were obtained from the Human Protein Atlas: https://www.proteinatlas.org/ENSG00000213398-LCAT and kidney tissue staining at https://www.proteinatlas.org/ENSG00000213398-LCAT/tissue/kidney.Gene-level GISTIC2.0 copy-number calls were accessed through GSCA and are available from the UCSC Xena Pan-Cancer Atlas hub (dataset TCGA.PANCAN.sampleMap/Gistic2_CopyNumber_Gistic2_all_thresholded.by_genes; direct download https://tcga.xenahubs.net/download/TCGA.PANCAN.sampleMap/Gistic2_CopyNumber_Gistic2_all_thresholded.by_genes.gz).Immune-related analyses were performed using TIMER2.0 as an analysis portal. The immune infiltration estimation matrix used by TIMER2.0 is publicly available at https://compbio.cn/timer2/infiltration_estimation_for_tcga.csv.gz. All gene expression matrices used in TIMER2.0 analyses are the TCGA RNA-seq data retrieved from the GDC projects listed above; the specific correlation outputs used in this study are provided in the Supplementary Data submitted with this manuscript.The ssGSEA input matrices used in Xiantao Academic were derived from the TCGA RNA-seq expression matrices downloaded from the GDC projects listed above. The exact ssGSEA input matrices and resulting ssGSEA scores used in this study are provided in the Supplementary Data submitted with this manuscript.Tumor mutational burden and microsatellite instability values were derived from TCGA somatic mutation MAF files downloaded from the GDC projects listed above. The per-sample TMB and MSI tables used for downstream analyses are provided in the Supplementary Data submitted with this manuscript.Functional enrichment results were generated using CAMOIP based on the same TCGA cohorts. The full enrichment output tables used in this study are provided in the Supplementary Data submitted with this manuscript. No new sequencing data were generated in this study.