Background Patients with head and neck squamous cell carcinoma (HNSCC) continue to face poor prognosis, highlighting an urgent need for new diagnostic markers and therapeutic targets. While metabolic reprogramming and immune microenvironment dysregulation are crucial drivers of HNSCC progression, the key causal molecular mechanisms linking these processes remain elusive. Post-translational modifications, especially protein lactylation, may serve as a vital interface for this metabolic-immune “crosstalk”. Methods We developed an integrative analytical framework merging lactylation proteomics, transcriptomics, and Mendelian randomization (MR). Differential expression analysis was conducted on three public transcriptomic cohorts (53 HNSCC vs. 53 controls), and the resulting genes were overlapped with a systematically compiled set of 2, 124 lactylation-related genes. Causal risk genes were then identified using MR analysis with large-scale genetic instruments (from expression quantitative trait locus data) and HNSCC genome-wide association study summary statistics. The functional roles of candidate genes were explored through enrichment analysis, Gene Set Variation Analysis, and immune deconvolution (CIBERSORT). Experimental validation was performed using quantitative real-time PCR and Western blotting in an independent The Cancer Genome Atlas dataset and in HNSCC cell lines. Results We identified 212 lactylation-associated differentially expressed genes. MR analysis established CD44 and APP as genetic causal risk factors for HNSCC, with both genes significantly overexpressed in patient tissues. Functional profiling indicated that high CD44 expression correlated with activation of mTOR signaling and ECM-receptor interaction pathways, and was positively associated with M0 macrophage infiltration. Conversely, high APP expression was linked to activated protein secretion and ECM pathways, and showed a positive correlation with M2 macrophage abundance. The marked upregulation of CD44 and APP in HNSCC was consistently confirmed in the independent validation cohort and in cellular models. Conclusion By pioneering a multi-omics causal inference approach in HNSCC, this study identifies CD44 and APP as genetic causal risk factors for disease susceptibility and progression. These genes connect distinct metabolic pathways with specific immune cell subsets, functioning as central hubs within the HNSCC metabolic-immune crosstalk network. Our work provides a critical theoretical basis for future development of lactylation pathway-based biomarkers and targeted interventions.
Zhang et al. (Thu,) studied this question.