UC Berkeley

The work presented within this thesis was motivated by the need for a robust S. cerevisiae system that has the capacity to drive the production of an acetyl-CoA derived industrial product. Systems that generated acetyl-CoA derived products have been engineered in E. coli have demonstrated the utility of a robust host engineered with strong acetyl-CoA production. We focus on the production of acetyl-CoA derived advanced biofuels. Chapter one describes the metabolic engineering endeavors of microbial pathways for advanced biofuels. These include short-chain alcohols from fermentative pathways, fuels from isoprenoid pathways, and fuels from fatty acid pathways. We describe the key advantages to metabolic pathway engineering in S. cerevisiae. We then review unique features of native S. cerevisiae central metabolism and its acetyl-CoA generation. Because we focus on engineering an advanced biofuel derived from the isoprenoid pathway, we describe the advantages of using amorphadiene as a proxy for isoprenoid pathway flux.

This dissertation describes the use of a heterologous enzyme, ATP: citrate lyase (ACL), by which acetyl-CoA is formed in the cytosol, whose use was inspired by native mechanisms of lipid accumulating yeast in Chapter two. We aimed to increase production of the sesquiterpene, amorphadiene, by increasing availability of its primary precursor, cytosolic acetyl-CoA. The importance of this aim is underscored by the stoichiometry that dictates that production of one mole amorphadiene biosynthesis requires nine molar equivalents of acetyl-CoA. In S. cerevisiae, acetyl-CoA metabolism takes place in at least four subcellular compartments: nucleus, mitochondria, cytosol and peroxisomes. A challenge lies in increasing precursor flux of cytosolic acetyl-CoA and decreasing acetyl-CoA flux towards other subcellular compartments. Oleaginous, or lipid accumulating yeast, have evolved mechanisms to export units of mitochondrial acetyl-CoA into the cytosol required for lipid biosynthesis. Key to this mechanism is the activity of the ACL enzyme. In these studies, we implemented an ACL from the oleaginous yeast, Aspergillus nidulans, encoded by genes aclA and aclB. In conclusion, we surmised that the expression of ACL genes did indeed alter isoprenoid metabolism. However, it was unable to increase total amorphadiene production. This result is most likely due to poor catalytic activity of the amorphadiene synthase (ADS), the final enzyme synthase leading to the production of amorphadiene, and native mechanisms that work to control increase of acetyl-CoA levels, thus preventing accumulation.

In Chapter three, we increased substrate supply and simultaneously increased carbon flux through the committed isoprenoid biosynthetic steps. We demonstrate the utility of expression of E. faecalis homologs of the upper mevalonate pathway as it sequesters carbon from the central metabolism. In particular, here we sought to determine if 1) expression of native and heterologous enzymes of first committed steps of the mevalonate pathway, E. faecalis genes mvaE and mvaS, are required to sequester acetyl-CoA units toward the committed path of isoprenoid biosynthesis or 2) flux through the mevalonate pathway remained unchanged due do inactivation of the native acetyl-CoA synthase, Acs1p. We conclude that indeed, the native Acs1p demonstrates feedback inhibition, and thus accumulates more acetate than heterologous homologues insensitive to feedback inhibition. Also, indeed we find that overexpression of heterologous E. faecalis genes mvaE and mvaS increase both mevalonate and amorphadiene production. However, the expression of genes mvaE and mvaS in combination with the acetyl-CoA synthase does not further increase mevalonate or amorphadiene production. This result is most likely due to a decrease in total enzyme levels of the first committed step of the mevalonate pathway.

In Chapter four, we perform targeted metabolic characterization of expression of the ATP: citrate lyase in a citrate generating S. cerevisiae host, BY4742 with its isocitrate dehydrogenase gene, IDH1, deleted. In these studies, we profiled changes in central metabolism arising from the expression of Aspergillus nidulans ACL genes, aclA and aclB, in the host wild type S. cerevisiae strain and in the BY4742 ∆IDH1 strain. We have chosen to utilize IDH1 deleted cells, which have previously been shown to generate high levels of citrate, in order to provide increased substrate for ACL utilization. Targeted metabolic profiling demonstrated increased citrate levels in the ∆IDH1 strain. Furthermore, citrate levels vary with nitrogen availability in the medium. We find that expression of ACL decreased total citrate levels and as previously seen in chapter one, acetyl-CoA levels remain unchanged. However, we find that expression of ACL causes large accumulation of two metabolites. The first metabolite has been positively identified as 2-isopropylmalate and is an acetyl-CoA derived intermediate of leucine biosynthesis. The second metabolite has yet to be positively identified. We also demonstrate that combined expression of ACL with E. faecalis genes mvaE and mvaS improve mevalonate production by 48%, thus demonstrating the use of ACL as an acetyl-CoA generating enzyme for production of acetyl-CoA derived products.

In Chapter five, we offer perspectives on future development of S. cerevisiae as cellular host for production of acetyl-CoA derived biofuels. The work presented in this thesis is a first step in re-purposing nature's metabolic mechanisms which lead to lipid accumulation for metabolic engineering acetyl-CoA derived products within S. cerevisiae. Future work aims to improve engineering efforts to mimic oleaginous metabolism and maximize the utilization of cytosolic carbon for the production of a biofuel. Cellular metabolism will be monitored through metabolomic studies and C¹³ metabolic flux analysis.