Lactylation-related gene signatures identify glioma molecular subtypes with prognostic, immunological, and therapeutic implications

Abstract

Introduction: Lactic acid is a by-product of energy metabolism and a signaling molecule that influences tumor progression by regulating immune cell function, angiogenesis, and epigenetic modifications.

Methods: This study analyzed data from the TCGA database on gliomas to systematically elucidate the expression patterns, prognostic value, and functional regulatory networks of lactylation-related genes.

Results: In this study, 17 lactylation-related prognostic genes were identified through the analysis of TCGA-GBM data. Using non- negative matrix factorization (NMF), two GBM subtypes based on lactylation- related genes (LRGs), termed GBM1 and GBM2, were identified. Survival analysis revealed that the overall survival (OS) of the GBM1 group was significantly lower than that of GBM2 group. Furthermore, notable differences were observed in the expression of key GBM-associated molecular markers between the two subtypes. Tumor microenvironment (TME) analysis demonstrated distinct immune landscapes and genomic characteristics between GBM1 and GBM2. The GBM1 group exhibited higher immune cell infiltration and immune function scores compared to GBM2. Drug sensitivity analysis further revealed differences in response to chemotherapy and targeted therapies between the two subtypes. In vitro data demonstrated that LCP1 knockdown suppressed cell proliferation and invasion, and promoted apoptosis in glioma cells.

Conclusion: In conclusion, our study systematically uncovers the significant role of LRGs in GBM molecular subtyping, prognosis evaluation, and therapeutic guidance. These findings offer new insights and potential strategies for the personalized treatment of GBM.

Keywords: glioblastoma; infiltration; lactylation; prognosis; subtype.

Figure 1

Differential expression analysis of lactylation-related genes in GBM. (A) Volcano plot of differentially expressed genes (DEGs) in GBM. (B) Venn diagram showing the overlap between glioblastoma DEGs and lactylation-related genes. (C) Heatmap displaying the top 40 lactylation-related DEGs. (D) Forest plot of prognostic lactylation-related DEGs, based on univariate Cox regression. (E) Correlation analysis among intersecting genes. (F) Protein–protein interaction (PPI) network illustrating known and predicted interactions among prognostic lactylation-related DEGs.

Figure 2

Molecular subtype classification of GBM based on lactylation-related DEGs. (A) Consensus clustering identified two distinct subtypes (k = 2). (B) Consensus matrix heatmap and cumulative distribution function (CDF) plot illustrating clustering stability. (C) Principal component analysis (PCA) showing clear separation between GBM1 and GBM2 subtypes. (D) Kaplan–Meier survival curves comparing overall survival between the two subtypes. (E) Expression patterns of 17 lactylation-related DEGs across GBM1 and GBM2 subtypes.

Figure 3

Tumor microenvironment characteristics of lactylation-based GBM subtypes. (A) Heatmap of 28 immune cell infiltration profiles across samples. (B) Comparison of 16 immune cell infiltration scores between GBM1 and GBM2 subtypes. (C) Comparison of 13 immune-related pathway activity scores between subtypes. (D) Expression levels of 30 immune checkpoint genes in the two GBM subtypes. (E) ESTIMATE-derived immune and stromal scores in GBM1 and GBM2. (F) Waterfall plot of somatic mutation frequencies in the GBM1 cluster. (G) Waterfall plot of somatic mutation frequencies in the GBM2 cluster. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 4

Functional enrichment analysis of lactylation-related DEGs. (A) Gene Ontology (GO) enrichment analysis of DEGs. (B) KEGG pathway enrichment analysis of DEGs. (C) Heatmap of enriched hallmark gene sets between GBM1 and GBM2. (D) Heatmap of enriched KEGG pathways distinguishing the two subtypes.

Figure 5

Drug sensitivity analysis of GBM subtypes. Comparison of estimated IC50 values for various drugs is illustrated. (A) BMS-509744, (B) AP-24534, (C) GSK-650394, (D) GW843682X, (E) Etoposide, (F) KIN001-260, (G) 17-AAG, (H) AZ628, (I) Lapatinib, (J) AICAR, (K) XAV939, (L) Vinblastine.

Figure 6

Co-expression network analysis of lactylation-related genes. (A) Scale-free topology model fit and soft-thresholding power selection. (B) Gene dendrogram after dynamic tree cutting and module merging. (C) Heatmap showing the correlation between 13 gene modules and GBM subtypes. (D) Key genes in the yellow module with correlation coefficients > 0.5. (E) Protein–protein interaction network of representative genes in the turquoise module.

Figure 7

Heterogeneity in the expression of key genes in GBM at the single-cell level. (A) UMAP plot showing distinct cell subsets identified in GBM samples. (B) Annotation of cell types within GBM tissues based on single-cell RNA sequencing data. (C) Expression profiles of representative marker genes and lactylation-related key genes across different cellular subpopulations. (D) Heatmap displaying the expression levels of key lactylation-related genes across annotated cell types.

Figure 8

LCP1 knockdown suppresses cell viability and proliferation in glioma cells. (A) Western blot analysis was performed to measure the knockdown efficacy of shLCP1 in U251 and LN229 cells. (B) CCK-8 assays were performed to evaluate cell viability at various time points following shLCP1 transfection in U251 and LN229 cells. (C) EdU assays were carried out to assess cell proliferation in U251 and LN229 cells 72 hours post-transfection with shLCP1. (D) Quantification of EdU-positive cells in U251 and LN229 cells following LCP1 knockdown. ***p < 0.001.

Figure 9

LCP1 knockdown induces apoptosis and inhibits cell invasion. (A) Annexin V-FITC/PI staining was used to evaluate apoptosis in U251 and LN229 cells after shLCP1 transfection for 72 hours. (B) Left: Transwell invasion assays were conducted to assess the invasive capacity of U251 and LN229 cells 24 hours post-transfection with shLCP1. Right: Quantification of invaded cells from the Transwell assay. ***p < 0.001..