Salivary Metabolomic & Proteomic Biomarkers in Occurrence of Early Childhood Caries

Objectives: To explore salivary biomarkers for occurrence of Early Childhood Caries using metabolomic and proteomic techniques.
Methods: 22 caries-free children aged 3-5 years old were recruited into this longitudinal study. Oral health examination and stimulated salivary samples collection were performed every three months during the 18-month check-up period. Those who developed new caries lesions were selected for untargeted metabolomics and data-independent-acquisition (DIA) proteomics analysis. Functional annotation analysis of differentially proteins and metabolites was performed, while random forest algorithm with cross-validation was employed to construct the prediction model for occurrence of caries.
Results: Compared with the CF (caries-free) group, 25 and 173 differentially expressed metabolites were identified in the PC (pre-caries) and OC (caries-occurrence) groups, respectively. The proteomics profile revealed 85 and 93 significantly different proteins in the PC and OC groups compared with CF group. With the integration of metabolomic and proteomic information, seven key pathways were involved in the full course of occurrence of caries. Five potential biomarkers closely associated with occurrence of caries were screened out using random forest algorithm and 10-fold cross-validation, including three metabolites (estriol, 2'-deoxyuridine 5'-monophosphate, propionic acid) and two proteins (ALOX12, CA1). The prediction model constructed by these biomarkers showed an area under curve (AUC) value of 0.94, which demonstrated fairly good capability for prediction of caries.
Conclusions: This study revealed that specific metabolites and proteins in saliva played an important role in occurrence of Early Childhood Caries, laying the basis for early detection, screening and surveillance of this disease.