Untargeted Metabolomics and Metabolite–Gene Network Analysis Predict NF-κB Inhibition in Artemisian B-Treated Triple-Negative Breast Cancer Cells

Background: Triple-negative breast cancer (TNBC) remains a formidable clinical challenge due to the scarcity of targeted therapies and profound metabolic heterogeneity. Although Artemisian B, a dimeric sesquiterpene lactone derived from Artemisia argyi, exhibits potent antiproliferative activity, its comprehensive metabolic footprint and the translation of these perturbations into downstream signaling regulation remain poorly characterized. Methods: To address this gap, we employed an integrative analytical framework combining untargeted metabolomics, topology-guided metabolite–gene network mapping, and parallel experimental validation in MDA-MB-231 cells. This workflow systematically profiled cellular phenotypes, global metabolic reprogramming, and key signaling nodes, enabling the prioritization of high-confidence mechanistic links between metabolic alterations and signal transduction. Results: Artemisian B dose-dependently suppressed TNBC cell viability (IC50 = 12.12 μM) and triggered mitochondrial apoptosis, characterized by Bax upregulation, Bcl-2 downregulation, and caspase-9/3 activation. Untargeted metabolomics identified 129 significantly altered metabolites, reflecting extensive dysregulation across lipid peroxidation, bioenergetics, and nucleotide metabolism. Topological analysis of the metabolite–gene network identified the NF-κB pathway as a highly interconnected hub within this perturbed landscape. Parallel experimental validation corroborated this prediction, demonstrating that Artemisian B consistently suppressed the phosphorylation of IKKα/β, IκBα, and p65, while markedly attenuating p65 nuclear translocation. Conclusions: Artemisian B induces TNBC apoptosis through extensive metabolic reprogramming coupled with concurrent inhibition of NF-κB signaling. By seamlessly integrating untargeted metabolomics with network topology, our framework not only successfully bridges metabolic perturbations with signaling outcomes but also establishes a versatile, dual-perspective strategy applicable to both biochemical reaction networks and signal transduction pathways. This approach provides a robust predictive paradigm for decoding the multi-target pharmacological mechanisms of natural products.