Abstract: The research presented in this dissertation deals with the phenomena of bacterial efflux pump inhibition by natural products and plant extracts. Bacterial efflux pumps are active transport proteins, primarily deriving their energy source from the proton motive force, which functions to export toxic compounds outside the cell. This is a natural defense mechanism that bacteria utilize to protect themselves from toxic environments. The over-production of efflux pumps is one mechanism by which bacteria can evolve resistance to clinical antibiotics, as well as other antimicrobials. Thus, the study of efflux pump inhibitors is important because it holds the potential to reverse some forms of antibiotic resistance. In light of this importance, a series of studies were designed to improve the ability to study this phenomenon, to investigate the distribution of efflux pump inhibition in land plants, and to improve our ability to identify an important class of efflux pump inhibitors, the flavonoids. The first aim of this research was to develop an improved method for experimentally quantifying efflux pump inhibitory activity of small molecules. Preexisting methods made this difficult due to several limitations including: the collection of indirect results, time consuming materials handling techniques, and/or matrix interference problems pertaining to the quenching of fluorescent signal. An improved method relying on mass spectrometry measurements was developed that addressed the aforementioned limitations. The importance of this improved method lies in its ability to produce data sets useful in calculating IC 50 values for a wider range of samples than was previously possible. The second aim was to evaluate the presence of efflux pump inhibitor production across the land plant lineage. This is important to botanical science and the understanding of plant-microbe interactions and plant evolutionary biology. The most ancient lineages of land plants have not been previously evaluated for efflux inhibitory activity. Additionally, land plants play an important role in many traditional medicinal systems and in modern complementary and alternative medicine. Thus, understanding the distribution of efflux pump inhibitor production in this group increases our understanding of these common forms of medical treatment. In order to gain these data, a set of 14 plant species spanning the major lineages within the land plant group (bryophytes, pterophytes and lycophytes, gymnosperms, and angiosperms) were extracted and assayed to determine efflux pump inhibitory activity of the extracts. Positive results (indicating the presence of an efflux pump inhibitor) were observed for many (but not all) of the plant species tested. The observation of activity in extracts prepared from the most ancient plants tested (bryophytes--the liverworts and mosses) lends credence to the hypothesis that the production of efflux pump inhibitors is of great antiquity in land plants. The last component of this work was the evaluation of methods for the analysis of flavonoids via mass spectrometry. This is of importance to this study due to the commonality of flavonoids in the literature pertaining to efflux pump inhibitors, and the consistency activity of the flavonoids evaluated in Chapter II. The goal of this work was to compare two methods for the tentative identification of signals in complex data produced via high-resolution mass spectrometry that could be labeled as "possible flavonoids." The methods evaluated were firstly the use of mass spectrometry fragmentation spectra to identify key diagnostic fragments of the flavonoid ring structure and secondly the use of mass defect filtering directly applied to high resolution data to select a short list of signals for further processing. The former method was not fruitful due to a combination of the frequent poor fragmentation and the dependency on standards for all samples. The latter proved more useful, successfully producing a list of potential flavonoids to be carried forward to other methods such as database searching and molecular formula calculation. This method was also successfully applied to a complex extract of Hydrastis canadensis , identifying three flavonoids known from previous work to be present. Further, the mass defect method is an intrinsic property of molecules, and therefore does not change with experimental conditions. For all of these reasons mass defect was selected as the more useful of the two methods evaluated for the identification of "possible flavonoid" signals in crude extracts.
Notes: Ph.D. thesis at University of North Carolina at Greensboro; advisor: Nadja Cech