UC Berkeley

Polyketide synthases (PKSs) are remarkable enzymes. The diverse products from naturally ocurring PKSs have saved lives due to their useful medicinal qualities and inspired chemists to push the envelope on size and complexity in their synthetic pursuits. More recently, these enzymes have been modified to produce "unnatural" natural products--including close analogs of naturally occurring products as well as molecules wholly unknown to nature. These engineering pursuits have been enabled by an increasingly sophisticated understanding of PKS mechanics as well as technologies of the genomics era like inexpensive DNA sequencing and synthesis. In the doctoral work presented in this thesis, I will describe studies that have expanded our understanding of PKS biochemistry and demonstrated the production of an extremely economic and environmentally important small molecule from an engineered polyketide synthase.

Chapter 1 will begin with a discussion of the current state of the art and challenges associated with engineering these enzymes as well as offering opinions about routes forward to narrow the considerable gap between the promise and reality of engineered PKSs. This will be followed by a brief review of the PKS system responsible for borrelidin biosynthesis with a focus on features that make it unusual and attractive for engineering purposes.

In Chapter 2, I will present a study describing the first in vitro characterization of the unusual carboxyacyl-processing borrelidin polyketide synthase. It was found that the loading module has a specific requirement for carboxylated substrates of a specific stereochemistry, whereas the first extension module is comparatively promiscuous in its ability to extend carboxy- and descarboxy substrates with no apparent stereoselectivity.

The following chapter describes engineering efforts which exploit borrelidin PKS' diacid processing abilities towards the production of the commodity chemical, adipic acid. Extensive intermediate analysis using mass spectrometry revealed an unexpected difficulty in dehydrating the 3-hydroxyadipic-ACP intermediate, which was overcome in part by judicious chimeric junction selection and utilization of a dehydratase domain that may better tolerate carboxylated substrates. Provision of a thioesterase led to the first demonstration of adipic acid production from an engineered polyketide synthase.

In the final data chapter, I describe a preliminary characterization of a module in the borrelidin cluster which performs three iterative condensations rather than the single extension performed in canonical type I modular PKS systems. It was proven in vitro for the first time that the isolated module is necessary and sufficient for iteration to occur and that the identity of the starter substrate has a profound effect on the distribution of extension products. Based on observations from time course experiments and co-incubation with the downstream extension module, a model for chain-length control in this unusual module is proposed.

The last chapter provides a brief summary of the principal findings described in this work and suggests experiments that will form the foundation of future efforts to improve our understanding of and ability to engineer these fascinating molecular assembly lines.