目的 以生品半夏Pinellia ternata毒性大小为标准,寻找标示半夏炮制品相对毒性大小的方法及相对毒性大小与指标成分间的关系。方法 按照《中国药典》2020年版一部方法对半夏生品进行质量检查;将影响半夏炮制品毒性的4个因素(白矾用量、生姜用量、蒸煮时间、蒸煮温度)进行混合均匀设计,建立正棱锥降维模型;将上述4个因素降维成用以表征半夏炮制品相对毒性大小的单一因素S_面,建立其与4个因素之间的回归方程;将得到的S_面与德尔菲评价法得出的相对毒性大小表征值进行相关性分析;采用薄层扫描仪、MATLAB软件测量物性分析指标色度(H、S、I),pH测量仪测定pH值,HPLC法测量半夏有机酸,将测得的以上指标成分与S_面进行回归分析。结果 S_面与4个影响因素间的回归模型为S_面=1-0.011 9 X_白矾-0.005 8 X_生姜-0.000 6 X_时间-0.004 3 X_温度,r²=0.983 8,P<0.001;模型得到的S_面与德尔菲评价法得到表征值均有显著相关性,r=0.985 2,P<0.001;S_面与pH、H、S、I间的逐步回归模型为S_面=0.165 X_pH+2.393 X_H-0.533 X_S+0.372X_I-0.631,r²=0.998 8,P<0.001。S_面与pH值、有机酸间的逐步回归模型为S_面=0.185 X_pH+0.163 X_L-苹果酸+0.186X_草酸-0.508,r²=0.995 5,P<0.001。结论 毒性标示模型具有效性,能够模拟出毒性影响因素与毒性大小间的量化关系,同时也能模拟出物性参数、化学指标与毒性大小的量化关系,对半夏炮制品的炮制方法的确定提供数据支持,为研究确定半夏炮制品毒性大小提供新思路。

Objective Taking the toxicity of raw Banxia (Pinellia ternata) as the standard, search for methods to indicate the relative toxicity of processed P. ternata and the relationship between relative toxicity and index components. Methods According to the first part of the Chinese Pharmacopoeia of the 2020 edition, the quality inspection of P. ternata raw products was carried out. The four factors (alum, ginger, steaming time, steaming temperature) that affect the toxicity of P. ternata processed products were mixed and uniformly designed. By establishing a regular pyramid dimensionality reduction model, the dimensions of the above four factors were reduced to a single factor S-plane which was used to characterize the relative toxicity of P. ternata processed products, and a regression equation between it and the four factors was established. Correlation analysis was carried out on the characterization value of relative toxicity; TLC scanner and MATLAB software were used to measure the chromaticity (H, S, I) of physical property analysis indicators, pH meter was used to measure pH, HPLC was used to measure organic acid of P. ternata, and the above will be measured. Regression analysis was performed between index components and S-plane. Results The regression model between S_surface and four influencing factors was:S_surface=1-0.011 9 X_alum-0.005 8 X_ginger-0.000 6 X_time-0.004 3 X_temperature, r²=0.983 8, P < 0.001; The S_surface obtained by the model had a significant correlation with the representative value obtained by the Delphi evaluation method, r=0.985 2, P < 0.001; The stepwise regression model between the S_surface and pH, H, S, I was:S_surface=0.165 X_pH + 2.393 X_H-0.533 X_S + 0.372 X_I-0.631, r²=0.998 8, P < 0.001. The stepwise regression model between S_surface, pH and organic acid was:S_surface=0.185 X_pH + 0.163 X_{L-malic acid} + 0.186 X_{oxalic acid}-0.508, r²=0.995 5, P < 0.001. Conclusion The toxicity labeling model is effective and can simulate the quantitative relationship between toxicity influencing factors and toxicity, as well as the quantitative relationship between physical parameters, chemical indicators and toxicity, providing data support for the determination of the processing method of P. ternata processed products and provide new ideas for research to determine the toxicity of processed products of P. ternata

R283.6

国家中医药管理局中药炮制技术传承基地项目（国中医药规财发〔2015〕21号）；河南省中医药管理局国家中医临床研究基地科研专项（2017JDZX007）；河南省中医药管理局国家中医临床研究基地科研专项（2019JDZX078）；河南省中医药科学研究专项课题（2018ZY1006）