Metabolic Reprogramming-Related Genes in Lung Adenocarcinoma: Identification and Prognostic Model Construction

Abstract

Background: Lung adenocarcinoma (LUAD), the predominant histological subtype of lung cancer, persists in presenting a dismally low 5-year overall survival (OS) rate, notwithstanding advancements in treatment modalities. There exists a pressing necessity for the identification of innovative biomarkers that can enhance prognostic assessments and facilitate individualized therapeutic strategies. The objective of this investigation was to clarify the involvement of genes associated with metabolic reprogramming in the progression of LUAD and to evaluate their viability as prognostic indicators.

Methods: An analysis of differential gene expression was performed utilizing The Cancer Genome Atlas (TCGA)-LUAD dataset, supplemented by a weighted gene co-expression network analysis (WGCNA). Through intersection analysis focusing on metabolic reprogramming genes (MRGs), pivotal differentially expressed metabolic reprogramming genes (hub DEMRGs) were identified. Consensus clustering categorized patients into subtypes based on these genes. Functional enrichment analysis and immune microenvironment characterization were conducted, followed by Cox and least absolute shrinkage and selection operator (LASSO) regression analyses to construct a prognostic risk model.

Results: A total of 31 hub DEMRGs were identified. Patients were classified into two distinct subtypes (C1 and C2), with the C2 subtype exhibiting a markedly reduced OS rate. Functional enrichment revealed significant activation of nuclear division and cell cycle pathways in C2. Immune profiling demonstrated an immunosuppressive phenotype in C2, characterized by elevated M2 macrophage infiltration and reduced CD8⁺ T cells. The risk model based on five critical hub DEMRGs showed robust predictive performance (area under the curve (AUC): 0.68 - 0.71), and high-risk patients displayed unique immune cell infiltration patterns.

Conclusions: This research highlights the critical role of MRGs in LUAD prognosis and their potential for clinical application. The identified subtypes and risk model provide insights into tumor heterogeneity and immunosuppressive mechanisms, offering potential targets for individualized therapy.

Keywords: Consensus clustering; Lung adenocarcinoma; Metabolic reprogramming genes; Prognostic model; Weighted gene co-expression network analysis.

Figure 1

Study workflow. Schematic overview of the analytical pipeline. The workflow encompasses differential expression analysis, WGCNA, MRG integration, consensus clustering, functional enrichment, immune microenvironment characterization, prognostic model construction, and validation. MRG: metabolic reprogramming gene; WGCNA: weighted gene co-expression network analysis.

Figure 2

Identification of DEGs in LUAD. (a) Volcano plot illustrating 1,952 upregulated (red) and 206 downregulated (blue) genes in LUAD tissues compared to normal controls (|log2FC| > 3, FDR < 0.05). (b) Heatmap displaying the expression patterns of top 50 DEGs across LUAD and normal samples. The color scale represents Z-scores of normalized RNA expression levels (blue = low, red = high). DEGs: differentially expressed genes; FC: fold change; FDR: false discovery rate; LUAD: lung adenocarcinoma.

Figure 3

WGCNA and hub gene identification. (a) Module-trait correlation heatmap highlighting the turquoise, blue, brown, and red modules with the strongest associations to LUAD. (b) Venn diagram intersecting WGCNA-derived modules, DEGs, and MRGs to identify 31 hub DEMRGs. DEGs: differentially expressed genes; DEMRGs: differentially expressed metabolic reprogramming genes; LUAD: lung adenocarcinoma; MRG: metabolic reprogramming gene; WGCNA: weighted gene co-expression network analysis.

Figure 4

Consensus clustering and subtype characterization. (a) PCA showing distinct separation between consensus clusters C1 (n = 242) and C2 (n = 261). (b) GO enrichment of subtype-specific DEGs, emphasizing nuclear division and chromosome organization. (c) KEGG pathway analysis showing enrichment in cell cycle and neuroactive ligand-receptor interaction. (d) GSEA plot revealing activation of DNA replication pathways in C2 (FDR < 0.05). (e) Kaplan-Meier survival curves showing poorer overall survival in C2 (log-rank P < 0.001). (f-i) Tumor microenvironment scores: (f) ESTIMATE, (g) immune, (h) stromal, and (i) tumor purity (P < 0.001). DEGs: differentially expressed genes; FDR: discovery rate; GO: Gene Ontology; GSEA: gene set enrichment analysis; KEGG: Kyoto Encyclopedia of Genes and Genomes; PCA: principal component analysis.

Figure 5

Immune microenvironment and checkpoint analysis. (a) CIBERSORT-derived immune cell infiltration profiles between clusters (P < 0.001). (b) Differential expression of immune checkpoint genes (LAG3, PDCD1LG2, HAVCR2; P < 0.05). (c) Reduced HLA-related gene expression in C2 (P < 0.001). HLA: human leukocyte antigen.

Figure 6

Prognostic gene screening and model validation. (a) Forest plot of univariate Cox regression for 19 hub DEMRGs with prognostic significance. (b-g) Expression validation of key DEMRGs (CAV1, FOXM1, FAM83A, PTGES, RECQL4, and SFTPC) in LUAD vs. normal tissues (P < 0.001). (h-j) Risk score distribution and survival status in training, testing, and entire TCGA cohorts. (k-m) Kaplan-Meier survival curves stratified by risk groups (log-rank P < 0.05). (n-p) ROC curves for 1-year survival prediction (AUC: 0.715 training, 0.679 testing, 0.691 overall). AUC: area under the curve; DEMRGs: differentially expressed metabolic reprogramming genes; LUAD: lung adenocarcinoma; ROC: receiver operating characteristic; TCGA: The Cancer Genome Atlas.

Figure 7

Immune landscape and nomogram performance. (a) ssGSEA scores of immune cell subsets in high- versus low-risk groups (P < 0.05). (b) GSVA enrichment of immune-related pathways (APC co-inhibition, MHC class I, etc.). (c) Nomogram integrating five hub MRGs for survival prediction (C-index = 0.623). (d) Calibration curves for 1-, 3-, and 5-year survival probability. APC: antigen-presenting cell; GSVA: gene set variation analysis; MHC: major histocompatibility complex; MRGs: metabolic reprogramming genes; ssGSEA: single-sample gene set enrichment analysis.