Epsilon Open Archive

Abstract

This thesis describes methods for structural classification of Quillaja saponins based on high performance liquid chromatography hyphenated electrospray ionisation ion trap multiple-stage mass spectrometry, in combination with multivariate analysis. Quillaja saponins from the chromatographic fractions QH-A, QH-B, and QH-C as well as crude bark extract have been investigated. MS1-MS3 spectra in positive ion mode were analysed by multivariate methods such as PCA and PLS-DA. Fragmentation of saponins generally results in loss of fragments from the precursor ion. The essential part of the reported method is the referencing of mass spectral data that allows for comparison of structural features between saponin components based on multiple stage mass spectra. Peaks in the referenced spectra obtain a correlation to loss of common structural elements. The multivariate methods captured the variance corresponding to the common structural elements. Thus, the obtained models have the ability to predict new structures that share the common structural elements. In total 12 new saponin structures have been proposed based on ESI-IT-MSn measurements. In these investigations the reported multivariate method was helpful since the generated structural predictions solved either the full saponin structure or parts of it. The reported multivariate method is in compliance to HPLC on-line coupled MSn data and thus provides a rapid method for screening of saponins in either biologic or chromatographic fractions. Further, a chromatographic method suitable for metabolic fingerprinting of saponins is reported.