A multi-strategy platform for quality control and Q-markers screen of Chaiqin chengqi decoction

Acute pancreatitis (AP) is an inflammatory disorder of the pancreas that is associated with substantial morbidity and mortality. Chaiqin chengqi decoction (CQCQD) has been proven clinically to be an effective treatment for AP for decades in West China Hospital. Quality control for CQCQD containing many hundreds of characteristic phytochemicals poses a challenge for developing robust quality assessment metrics. To evaluate quality consistency of CQCQD with a multi-strategy based analytical method, identify potential quality-markers (Q-markers) based on drug properties and effect characteristics, and endeavor to establish CQCQD as a globally-accepted medicine. A typical analysis of constitutive medicinal plant materials was performed following the Chinese Pharmacopoeia. The extraction process was optimized through an orthogonal array (L9(34)) to evaluate three levels of liquid to solid ratio, soaking time, duration of extraction, and the number of extractions. An ultra-high-performance liquid chromatography (UHPLC) fingerprinting combined with absolute quantitation of multi chemical marker compounds, coupled with similarity, hierarchical clustering analysis (HCA), and principal component analyses (PCA) were performed to evaluate 10 batches of CQCQD. On the basis of systematic analysis of fundamental features of CQCQD in treating AP, the potential Q-marker screen was proposed through detection of quality transfer and efficacy for chemical markers. UHPLC coupled with quadrupole orbitrap mass spectrometry were used to determine compounds in medicinal materials, decoctions and plasma. Network pharmacology and taurolithocholic acid 3-sulfate induced pancreatic acinar cell death were used to evaluate the correlation between chemical markers and anti-pancreatitis activity. A cerulein induced AP murine model was used to validate quality assessed CQCQD batches at clinically-equivalent dose. The effective content of chemical markers was predicted using linear regression analysis on quantitative information between validated batches and the other batches. The chemical markers and other physical and chemical indices in the original materials met Chinese Pharmacopoeia standards. A total of 22 co-existing fingerprint peaks were selected and the similarity varied between 0.946 and 0.990. Batch D10 possessed the highest similarity index. HCA classified the 10 batches into 2 main groups: 7 batches represented by D10 and 3 batches represented by D1. During the initial Q-marker screen stage, 22 compounds were detected in both plant materials and decoctions, while 13 compounds were identified in plasma. Network pharmacology predicted the potential targets and pathway of AP related to the 22 compounds. All 10 batches showed reduced necrosis below 60% with the best effect achieved by D10 (~40%). The spectrum-efficacy relationship analyzed by Pearson correlation analysis indicated that emodin, rhein, aloe emodin, geniposide, hesperridin, chrysin, syringin, synephrine, geniposidic acid, magnolol, physcion, sinensetin, and baicalein showed positive correlation with pancreatic acinar cell death protection. Similar to the in vitro evaluation, batch D10 significantly reduced total histopathological scores and biochemical severity indices at a clinically-equivalent dose but batch D1 did not. The content of naringin, narirutin and baicalin in batches D1, D5 and D9 consistently exceeds the upper limit of the predicted value. Eight markers whose lower limit is predicted to be close to 0 contributed less to the material basis for AP protection. Despite qualified materials used for CQCQD preparation, the clinical effect depends on appropriate content range of Q-markers. Emodin, rhein, aloe emodin, magnolol, hesperidin, synephrine, baicalein, and geniposide are considered as vital Q-markers in the primary screen. This study proposed a feasible platform for producing highly consistent batches of CQCQD in future study.