Abstract: Objective To explore the quality evaluation system of Bupleuri Radix decoction pieces based on the ‘Shape-Chemistry-Biology' trinity so as to ensure the safety and effectiveness of clinical medications with Bupleuri Radix decoction pieces. Methods Twenty-six batches of Bupleuri Radix decoction pieces from different medicinal materials markets and from licensed decoction piece companies were purchased online, and 6 batches of special samples (water boiled and alcohol extracted) were prepared at the same time, so there were a total of 32 batches of samples. Bupleuri Radix decoction pieces from different sources were identified empirically using the Delphi method. The contents of 5 types of saikosaponins (a, b₁, b₂, c, d) were determined with HPLC, and the content differences of index components of Bupleuri Radix decoction pieces from different sources were analyzed. Orthogonal partial least squares discriminant analysis (OPLS-DA) and variable importance in projection (VIP) were carried out. The lipopolysaccharide-induced RAW264.7 cell inflammation model was used to investigate its anti-inflammatory activity by detecting its effect on expressions of interleukin 6 (IL-6) and tumor factor α (TNF-α) inflammatory factors. At the same time, the cytotoxicity of Bupleuri Radix decoction pieces from different sources was determined, and the safety range was clarified to effectively identify unqualified samples. Results The traditional trait identification method could effectively identify the Bupleurum marginatum var. stenophyllum with high contents of index components and the Bupleurum marginatum with indistinguishable traits. However, the special prepared samples (boiled and alcohol extracted) could not be effectively distinguished during trait identification. The results showed that the contents of saikosaponin a and d in the 32 batches of Bupleuri Radix decoction pieces except 4 batches of special prepared samples (TS2, TS4, TS5, TS6) did not meet the 0.3 % limit stipulated by the ‘Pharmacopoeia of the People's Republic of China' (2020 edition). However, the total contents of five saponins in all the six batches of special prepared samples (TS1-TS6) were lower than the lowest content of Bupleurum chinense samples, which were 0.53% to 1.09% and 3.40% to 4.01% in Bupleurum chinense and Bupleurum marginatum var. stenophyllum, respectively. The total content of Bupleurum marginatum var. stenophyllum was 3 to 4 times that of Bupleurum chinense, while that of the two doped varieties ranged from 2.02% to 3.41%. The total contents of 5 kinds of saikosaponins in boiled and alcohol-extracted samples decreased significantly to 50% and 25% of the contents of the original samples. Further biological evaluation showed that Bupleuri Radix decoction pieces from different sources had obvious anti-inflammatory activity (P<0.05,0.01), which was better in Bupleurum marginatum var. stenophyllum than in Bupleurum chinense (P<0.05,0.01). The cytotoxicity of Bupleurum marginatum var. stenophyllum was significantly higher than that of Bupleurum chinense (P<0.01). The anti-inflammatory activity and cytotoxicity of Bupleurum marginatum Wall. were close to those of Bupleurum chinense. Conclusion The sequential detection method of ‘traditional traits-chemical content determination-biological evaluation' can more accurately reflect the quality of Bupleuri Radix decoction pieces, more effectively detect unqualified samples, ensure the consistency and stability of the quality of Bupleuri Radix decoction pieces, and improve the effectiveness and safety of clinical medications.

Key words: Bupleuri Radix, Character Identification, Anti-Inflammatory Activity, Cytotoxicity, Chemical Content Determination, Biological Evaluation, Quality Evaluation and Control