UC Berkeley

It is widely accepted that using petroleum-based fuels and chemicals is unsustainable, both in terms of their limited supply and hazardous greenhouse gas emissions. One alternative is to replace petroleum-derived liquid fuels such as gasoline with biofuels. Biofuels are energy-dense compounds such as medium-chain alcohols or fatty acids produced by microbes from plant-based feedstocks. Replacing fossil fuels with biofuels will have multiple beneficial impacts on the environment and economy, including carbon sequestration, greenhouse gas reduction, and generation of a renewable industry.

Here, we engineer the model yeast Saccharomyces cerevisiae to enhance its potential for biofuel production. Yeast are promising host organisms for many reasons. A sophisticated genetic toolkit is available, which allows for simple manipulation of metabolic pathways and strain characteristics. Industrial fermentations are well- established: S. cerevisiae are currently used to produce a variety of compounds, including bioethanol, amino acids, and therapeutics. In order for yeasts to become commercially viable production organisms for a range of biofuels, some challenges, need to be overcome, from limited production to product toxicity. Here, we used protein engineering methods to address some of these challenges and improve S. cerevisiae as a production organism.

It is important to consider transport at the commodity scales relevant to fuel production. Biofuel efflux is particularly important because it drives production forward, reduces cellular toxicity, and is necessary for continuous process schemes. We therefore set out to understand how medium chain fatty acids exit the cell. We characterized the transporter Tpo1 for activity on the fatty acids octanoic and decanoic acid, and identified two fatty acid substrate binding residues. Moreover, we showed that the native promoter gives the optimal expression level for fatty acid efflux. We also found that localization to the plasma membrane is sufficient for fatty acid tolerance and vacuolar localization is not necessary. We anticipate that this transporter will make an excellent target for increasing efflux of desired non-native substrates such as butanol.

We also investigated the use of transporters to improve cytoplasmic acetyl- Coenzyme A A(AcCoA) levels. As many heterologous bioproducts are constructed from AcCoA by cytosolic enzymes, increasing cytoplasmic AcCoA concentrations should improve product titers. Our long-term goal is to use the AcCoA transporters AT-1 and YBR220C to transport AcCoA directly from the mitochondria into the cytoplasm. To this end, we found that the AT-1 transporter is active in yeast and draws AcCoA into the endoplasmic reticulum. We have designed a suite of mitochondrial targeting tags to direct the transporters to the mitochondria, and are currently testing their effectiveness. We anticipate that the mitochondrially-located transporters will raise cytoplasmic AcCoA levels, and concomitantly improve product titers.

Finally, to address potential issues from product toxicity, we explored the cellular effects of medium-chain alcohols on S. cerevisiae, and used protein engineering to reduce a toxicity response. Specifically, we found that medium chain alcohols, including n-pentanol, n-hexanol, and n-heptanol, inhibit translation initiation. Two mutations in translation initiation factors, Gcd1 D85E and Sui2 D77Y, both reduce inhibition of translation initiation, producing a stable, dominant medium chain alcohol tolerance phenotype. Interestingly, a range of mutations at positions 85 in Gcd1 and 77 in Sui2 improve medium chain alcohol tolerance. These findings will be important for generating efficient medium-chain alcohol production strains.

Through these projects, we have advanced the potential of S. cerevisiae for commodity-scale biofuel production.