Periodontitis is a chronic inflammatory, non-infectious disease caused by bacterial biofilms in periodontal tissue that lead to irreversible damage to susceptible hosts (alveolar bone and periodontal ligament) (Dannewitz et al., 2021, Lin et al., 2021). Globally, the prevalence of periodontitis is approximately 11%, affecting 743 million people (Kwon et al., 2021). The pathological features of periodontitis include gingival inflammation, clinical attachment loss, alveolar bone loss, bleeding on probing, and pathological tooth migration (Papapanou et al., 2018). It is generally accepted that the pathogenic mechanisms of periodontitis arise from bacterial-host interactions, including immune and inflammatory responses caused by periodontal pathogens and their products as well as host susceptibility (Bartold, 2018, Sun et al., 2021). As one of the most common dental diseases, the risk of periodontitis should not be underestimated because severe periodontitis can lead to facial collapse, chewing difficulties, impaired digestion, and complications of systemic and distal inflammatory diseases (Yang et al., 2021). Therefore, a humoral-based diagnostic test (saliva and gingival crevicular fluid [GCF]) is particularly important for identifying and treating patients with early periodontitis.

In patients with periodontitis, factors that determine immune adaptation are often related to individual genetic factors. Studies have shown that the expression of inflammatory cytokines in periodontitis is regulated at the epigenetic level by DNA methylation, histone modifications, and microRNAs (Luo et al., 2018, Shen et al., 2023, Wichnieski et al., 2019). Another study reported at least 65 genetic variants associated with periodontitis (Loos et al., 2020). In addition, oral metabolites are inextricably related to the periodontal status and plaque microbiome of patients with periodontitis (Na et al., 2021). Several metabolic profiling analyses of periodontitis have revealed significant differences in the composition of metabolites in serum and GCF between patients and healthy controls (Chen et al., 2018, Wei et al., 2022). A meta-analysis summarized more than 40 metabolites that were significantly discriminatory for the identification of periodontitis and were mainly involved in the amino acid and lipid degradation pathways (Baima et al., 2021). Salivary metabolites can also be used to differentiate oral cancer from periodontitis, with an accuracy of up to 80% (Kouznetsova et al., 2021). Therefore, a combined analysis of genes and metabolites is important to refine the upstream and downstream genetic regulatory axes in periodontitis. Although a candidate gene association study proposed 2649 single nucleotide polymorphisms in five genes associated with metabolic traits in periodontitis patients (Moon, 2019), there is still a lack of integrated analysis at the transcriptome and metabolome levels to reveal the key biomarkers affecting the development of periodontitis and the metabolic pathways involved.

In this study, a liquid chromatography/tandem mass (LC/MS)-based metabolomic analysis of GCF samples was carried out, and metabolomics was initially integrated with transcriptomics in periodontitis using a bioinformatics approach. These identified key genes and metabolites could be used as potential biomarkers for periodontitis and reflect disease progression. Furthermore, the gene-metabolite-pathway network constructed in this study comprehensively interprets the molecular regulatory pathways involved in upstream and downstream metabolites, providing genetic evidence for refining the functions of genes and metabolites as markers of periodontitis.