Abstract
Periodontitis is a prevalent inflammatory disease characterized by immune dysregulation and tissue destruction, yet the molecular mechanisms linking metabolic epigenetic modifications and immune responses remain insufficiently understood. Recent evidence suggests protein lactylation plays a pivotal role in shaping immune landscapes in chronic inflammation, but its relevance in periodontitis is largely unexplored. We systematically analyzed microarray-based bulk gene expression datasets from gingival tissues and single-cell RNA sequencing (scRNA-seq) data derived from peripheral blood mononuclear cells of periodontitis patients and healthy controls. Differentially expressed lactylation-associated genes (DE-LACRGs) were identified using established bioinformatics workflows. Immune infiltration was assessed by single-sample gene set enrichment analysis, while unsupervised subtyping was conducted using nonnegative matrix factorization. Machine learning algorithms, including least absolute shrinkage and selection operator regression, random forest, and extreme gradient boosting, prioritized key diagnostic LACRGs. Single-cell analyses mapped immune cell heterogeneity and pathway activity, with a focus on natural killer (NK) cell subpopulations. Immunohistochemistry (IHC) validated target gene expression in clinical gingival samples. Twelve DE-LACRGs were identified as differentially expressed in periodontitis. Integrative machine learning approaches robustly prioritized ELL-associated factor 1 (EAF1), neurofilament light chain (NEFL), and vimentin (VIM) as the core diagnostic markers. The composite gene signature demonstrated strong diagnostic performance, with area under the receiver operating characteristic curve (AUC) values of 0.919 in the GSE10334 cohort and 0.891 in the GSE16134 cohort. IHC analysis of gingival specimens confirmed elevated VIM expression and reduced EAF1 and NEFL in periodontitis compared to healthy controls, supporting transcriptome-level findings. Molecular subtyping based on LACRG expression separated patients into clusters with distinct immune activation profiles. Immune infiltration analysis revealed significantly increased proportions of NK cells, CD8 + T cells, and other pro-inflammatory subsets in diseased tissues, while regulatory T cells were notably decreased, underscoring a shift toward heightened immune activation in periodontitis. scRNA-seq further demonstrated increased NK cell heterogeneity and identified cell type-specific expression patterns for the diagnostic gene signature. LACRGs are integrally involved in immune remodeling and inflammation in periodontitis. Multi-level integrative analyses highlight their diagnostic value and illuminate new targets for understanding and potentially modulating the metabolic-epigenetic-immune axis in periodontal disease.