Paclitaxel (Taxol®) is a widely used chemotherapeutic drug with additional medical applications in drug-eluting stents as an anti-restenosis treatment. Paclitaxel is a structurally complex natural product with an excellent scaffold for designing analogs with pharmacological properties. To date, clinically approved analogs include docetaxel and cabazitaxel for the treatment of additional cancers. Currently, plant cell fermentation methods produce paclitaxel and large quantities of the precursors 10-deacetylbaccatin III (10-DAB) and baccatin III. The complexity of the semi-characterized ~19-step paclitaxel biosynthetic pathway limits bioengineering attempts. However, the availability of 10-DAB and baccatin III suggests a semi-biosynthetic pathway to paclitaxel starting with these precursors is feasible. We have designed a short, simple biosynthetic pathway, capable of making paclitaxel, analogs, and/or valuable precursors for the semi-synthesis of additional analogs of biological interest. The paclitaxel biosynthesis enzyme baccatin III: 3-amino-13-O-phenylpropanoyl CoA transferase (BAPT) and the bacterial (2R,3S)-phenylisoserinyl CoA ligase (PheAT) produce N-debenzoylpaclitaxel, N-debenzoyldocetaxel, or precursor analogs. The addition of the paclitaxel biosynthetic N-debenzoyltaxol-N-benzoyltransferase (NDTNBT) and the bacterial benzoate CoA ligase (BadA) produce paclitaxel or other N-acylated analogs. In this dissertation, BAPT and BadA are kinetically characterized. The substrate specificity of BadA was systematically investigated with a series of 24 substrates. Six crystal structures of BadA in complex with different substrates, including benzoyl AMP, are used to explain BadA reactivity and propose rational mutations (A227A, H333A, and I334A) that expand substrate specificity and provide insight into the BadA mechanism and connect with established acetylation regulatory mechanisms in bacteria. Major hurdles including solubility and substrate availability, were overcome in order to characterize BAPT activity in the proposed semi-biosynthetic pathway. BAPT was purified as a fusion protein with maltose binding protein and its (2R,3S)-phenylisoserinyl CoA substrate was biosynthesized. To our knowledge this is the first time (2R,3S)-phenylisoserinyl CoA has been isolated in quantitative yields high enough to allow for characterization of the Michaelis-Menten kinetic constants (kcat and KM) for BAPT. This dissertation also describes the combination of BAPT and a bacterial ligase (PheAT) to produce N-debenzoylpaclitaxel and N-debenzoyl-10-deacetylpaclitaxel, precursors of paclitaxel and docetaxel, respectively. Biosynthesis of a biologically active paclitaxel analog, N-2-furanyl-N-debenzoylpaclitaxel, using the aforementioned enzymes, is also demonstrated as proof-of-principle that this semi-biosynthetic pathway may shorten the number of steps required to make certain paclitaxel (and docetaxel) analogs of interest.
Read
