Integrated GC-MS Fingerprinting and Chemometrics for Identifying Antimicrobial Quality Markers in Bidens pilosa L....

Bidens pilosa L. is an herbal medicine rich in volatile oils (VOs) with anti-inflammatory, antioxidant, and antimicrobial activities. However, the volatile constituents responsible for its antimicrobial effect (quality-indicators, Q-markers) have not been clearly defined, limiting product standardization. In this study, VOs from eleven sites in China were obtained by supercritical-fluid extraction and analyzed by GC-MS, yielding 75 compounds and a fingerprint with a similarity > 0.90 based on 43 common peaks. Chemometric screening using relative abundance, followed by hierarchical cluster analysis (HCA) and principal component analysis (PCA) with variable-importance-in-projection (VIP) scores, divided the oils into three geographical clusters and identified 14 differential compounds (VIP ≥ 1, p < 0.05). Antimicrobial assays against E. coli, S. aureus, S. enterica and P. multocida revealed pronounced regional variation, with Guangdong, Fujian and Jiangxi extracts showing broad activity. Spectrum-effect correlation analysis identified methyl laurate (0.832-0.976), (Z)-3-heptadecen-5-yne (0.757-0.997), stearic acid (0.815-0.919) and stearolic acid (0.888-0.960) as candidate quality markers, based on their statistical association with antibacterial activity. Mechanistic studies and validation of these candidate markers are necessary to confirm their individual and synergistic roles and to establish a definitive standardized quality control method.