Institute of Pharmaceutical Sciences, University of Tübingen, Auf der Morgenstelle 8, 72076 Tübingen, Germany.
University of Vienna, Department of Analytical Chemistry, Währinger Strasse 38, A-1090 Vienna, Austria.
J Chromatogr A. 2014 Oct 10;1363:79-88. doi: 10.1016/j.chroma.2014.04.077. Epub 2014 May 9.
A set of N-derivatized amino acids were separated into enantiomers on a tert-butylcarbamoylated quinine-based chiral stationary phase (CSP). Quantitative structure-property relationship (QSPR) studies were then employed to investigate the retention behavior and factors responsible for enantioselectivity. Computations were performed using a general linear model and a Free-Wilson matrix with indicator variables as structural descriptors. The approach allowed calculations of retention increments for first and second eluted enantiomers as well as group contributions to enantioselectivity. The results demonstrated that the additivity principle of group contributions was obeyed for the majority of solutes in the data set. Only a few basic amino acids (Arg, His) needed to be removed as they did not fit to such a linear model leading to outliers. The model was carefully validated and then utilized to investigate retention and enantioselectivity contributions of different protection groups and individual amino acid residues. It turned out that primarily protection groups were driving retention and enantioselectivity. In contrast, the contribution of amino acid residues to enantioselectivity was only significant for secondary amino acids, α-methylated amino acids, aspartic acid and a few sterically bulky aliphatic amino acid residues (Tle, Ile, allo-Ile). Amongst them only the latter group contributed positively to enantioselectivity while the other residues mentioned reduced enantioselectivity significantly. This type of QSPR model may be valuable to analyze retention/selectivity data of closely related congeneric compound series, is illustrative and straightforward to implement. It is thus valuable for interpretation of retention mechanisms, while its utility for prediction of retention and enantioselectivity data is limited to compounds made up of groups included in the solute set used for deriving the increments.
一组 N-取代的氨基酸在叔丁基carbamoylated 奎宁衍生的手性固定相（CSP）上被分离为对映异构体。然后采用定量构效关系（QSPR）研究来考察保留行为和对映选择性的影响因素。计算使用通用线性模型和带有指示变量的 Free-Wilson 矩阵作为结构描述符进行。该方法允许计算第一和第二洗脱对映异构体的保留增量以及对映选择性的基团贡献。结果表明，对于数据集中的大多数溶质，基团贡献的加和原理是遵守的。只有少数碱性氨基酸（Arg，His）需要被剔除，因为它们不符合这种线性模型，导致异常值。该模型经过仔细验证后，用于研究不同保护基团和单个氨基酸残基对保留和对映选择性的贡献。结果表明，主要是保护基团驱动保留和对映选择性。相比之下，氨基酸残基对对映选择性的贡献仅对二级氨基酸、α-甲基化氨基酸、天冬氨酸和少数空间位阻较大的脂肪族氨基酸残基（Tle，Ile，allo-Ile）显著。其中只有后者基团对对映选择性有积极贡献，而其他提到的残基则显著降低了对映选择性。这种类型的 QSPR 模型对于分析密切相关的同系物化合物系列的保留/选择性数据可能是有价值的，它具有说明性且易于实现。因此，它对于解释保留机制是有价值的，而其用于预测保留和对映选择性数据的用途仅限于由用于推导增量的溶质集中包含的基团组成的化合物。
