Quantitative structure-retention relationships of flavonoids unraveled by immobilized artificial membrane chromatography

详细信息    查看全文

• 作者：Adriana Leandra Santoro^a ; Emanuel Carrilho^a ; ^b ; Fernando Mauro Lanç ; as^a ; Carlos Alberto Montanari^a ; ^{Carlos.Montanari@usp.br" class="auth_mail" title="E-mail the corresponding author}
• 关键词：Flavonoids ; Immobilized artificial membrane chromatography ; Partition coefficient logarithm ; Human colon epithelial cancer cell line ; Partial least squares ; Quantitative structure&ndash ; property relationships
• 刊名：European Journal of Pharmaceutical Sciences
• 出版年：2016
• 出版时间：10 June 2016
• 年：2016
• 卷：88
• 期：Complete
• 页码：147-157
• 全文大小：1614 K

文摘

The pharmacokinetic properties of flavonoids with differing degrees of lipophilicity were investigated using immobilized artificial membranes (IAMs) as the stationary phase in high performance liquid chromatography (HPLC). For each flavonoid compound, we investigated whether the type of column used affected the correlation between the retention factors and the calculated octanol/water partition (log P_oct). Three-dimensional (3D) molecular descriptors were calculated from the molecular structure of each compound using i) VolSurf software, ii) the GRID method (computational procedure for determining energetically favorable binding sites in molecules of known structure using a probe for calculating the 3D molecular interaction fields, between the probe and the molecule), and iii) the relationship between partition and molecular structure, analyzed in terms of physicochemical descriptors. The VolSurf built-in Caco-2 model was used to estimate compound permeability. The extent to which the datasets obtained from different columns differ both from each other and from both the calculated log P_oct and the predicted permeability in Caco-2 cells was examined by principal component analysis (PCA). The immobilized membrane partition coefficients (k_IAM) were analyzed using molecular descriptors in partial least square regression (PLS) and a quantitative structure–retention relationship was generated for the chromatographic retention in the cholesterol column. The cholesterol column provided the best correlation with the permeability predicted by the Caco-2 cell model and a good fit model with great prediction power was obtained for its retention data (R² = 0.96 and Q² = 0.85 with four latent variables).