Background: Dental caries is a dynamic, multifactorial disease that destroys teeth and can affect anyone’s quality of life because it can cause tooth loss and make chewing difficult. Dental caries involves various factors, such as Streptococcus mutans and host factors. Currently, adjuvant therapies, such as curcumin, have emerged, but how they work has not been adequately described. Therefore, this work aims to identify the molecular mechanism of curcumin in caries and Streptococcus mutans. Methods: We obtained differentially expressed genes from a GEO dataset, and curcumin targets were obtained from other databases. The common targets were analyzed according to gene ontology enrichment, key genes were obtained, and binding to curcumin was verified by molecular docking. Results: Our analysis showed that curcumin presents 134 therapeutic targets in caries. According to the gene ontology analysis, these targets are mainly involved in apoptosis and inflammation. There are seven key proteins involved in the action of curcumin on caries: MAPK1, BCL2, KRAS, CXCL8, TGFB1, MMP9, and IL1B, all of which spontaneously bind curcumin. In addition, curcumin affects metabolic pathways related to lipid, purine, and pyrimidine metabolism in Streptococcus mutans. Conclusions: Curcumin affects both host carious processes and Streptococcus mutans.

Background/Objectives: Early childhood caries (ECC) remains a major public health concern, with Streptococcus mutans as a primary etiological agent. Current treatments rely on broad-spectrum antimicrobials, which can disrupt the oral microbiome and promote resistance. This study applied a structure-based in silico pipeline to identify molecule modulators of Toll-like receptor 2 (TLR2), a key host receptor implicated in ECC, and to explore their binding potential against major S. mutans proteins. Methods: ECC-related genes were collected from public databases and analyzed by functional enrichment and protein–protein interaction (PPI) network analysis. Hub genes were ranked using centrality algorithms. Virtual screening on TLR2 (DrugCLIP) was followed by molecular docking of selected compounds against the TLR1/TLR2 heterodimer and 50 S. mutans proteins, complemented by in silico absorption, distribution, metabolism, excretion, and toxicity (ADMET) profiling. Results: Fifty-four ECC-related genes and nine hub genes were identified, with TLR2 and cathelicidin antimicrobial peptide (CAMP) as central nodes. Virtual screening yielded five lead compounds fulfilling drug-likeness and toxicity criteria. Docking to TLR1/TLR2 showed favorable binding energies, with Z7684613096 showing the most consistent binding. V026-2549 displayed the highest number of strong interactions with S. mutans targets, including dTDP-glucose 4,6-dehydratase (rmlB), NADP-dependent glyceraldehyde-3-phosphate dehydrogenase (gapN), glucosyltransferase C (gtfC), and 5-methyltetrahydropteroyltriglutamate-homocysteine methyltransferase (metE). Conclusions: Five candidate compounds with promising dual activity against TLR1/TLR2 and S. mutans proteins were prioritized for experimental validation, including TLR2 functional assays and in vitro anti-biofilm studies.