Quantitative ternary network-oriented discovery of Q-markers from traditional Chinese medicine prescriptions: Bu-Zhong-Yi-Qi-Tang as a case study.

Background: The proposal of Q-markers for traditional Chinese medicine (TCM) represents a novel avenue of research pertaining to the quality control of TCM prescriptions. However, prior exploratory studies on Q-markers with multiple properties consistently neglected the consideration of weights, hampering our ability to accurately gauge the significance of each property and potentially leading to a flawed comprehension of Q-markers.

Purpose: In this study, a quantitative ternary network strategy was firstly proposed to visually discover the Q-markers from TCM prescriptions, and it has been successfully applied into the quality control study of Bu-Zhong-Yi-Qi-Tang (BZYQT), a classical TCM prescription.

Methods: Firstly, the contents of 34 components in BZYQT, along with the kinetic features of 17 candidate Q-markers in biosamples (plasma and small intestinal contents), were characterized by UPLC-QqQ-MS/MS, and their immunomodulatory activities in macrophages and splenic lymphocytes were also assessed. Next, the obtained data were integrated into three properties: testability, bioavailability, effectiveness, and their weights were calculated using the entropy weight method to further establish a ternary network for quantitatively screening Q-markers. Subsequently, the identified Q-markers of BZYQT were utilized for the holistic quality evaluation of 36 batches of the commercial BZYQT preparation, Bu-Zhong-Yi-Qi-Pill (BZYQP) produced by three manufacturers, through similarity evaluation of the Q-marker-based fingerprint.

Results: Nine compounds (hesperidin, astragaloside IV, ononin, 18β-glycyrrhizic acid, narirutin, calycosin, cimigenoside, astragaloside II, and liquiritin) showing three core properties, including testability, bioavailability, and effectiveness, were screened out as Q-markers of BZYQT based on their rankings in terms of regression area of the ternary network. Employing Q-markers as common peaks, the similarity values of 36 batches BZYQP ranged 0.914-0.998 under HPLC-UVD mode, and 0.631-1.000 under HPLC-ELSD mode, which were less than the similarity values evaluated by the conventional common peaks (HPLC-UVD mode: 0.946-0.990; HPLC-ELSD mode: 0.957-0.997). This observation suggests that the identified Q-markers are more representative as common peaks in chromatographic fingerprints for the holistic quality evaluation of TCM-related products from different manufacturers.

Conclusion: The quantitative discovery of Q-markers from BZYQT laid an important foundation for holistic quality assessment of its related commercially available products, and our work offering a new strategy for ensuring the consistency and efficacy of TCM prescriptions.