UC Berkeley

Currently, the majority of biodiesel produced in the United States is derived from oilseed crops such as soybean and rapeseed. The acreage of these oilseed crops that is required to displace a significant fraction of diesel is beyond current sustainable production capacity. Alternative technologies are required to make biodiesel a sustainable, renewable option. Heterotrophic lipid production has the potential to replace the lipids produced by oilseed crops. Two major challenges limit the economic feasibility of biodiesel production from heterotrophically-produced lipids: (1) simple sugars are required as a feedstock and (2) lipid recovery is challenging.

We proposed to overcome these limitations using a cellulolytic filamentous fungus to produce and secrete lipids from a lignocellulosic feedstock. Neurospora crassa is a well-established model filamentous fungus, which benefits from a large toolbox of molecular and genetic techniques, including a near full-genome deletion strain set. Achieving this objective will require N. crassa to efficiently deconstruct and utilize lignocellulosic biomass to produce and secrete lipids at high yields. Three main areas are addressed in this dissertation: (1) hyper-secretion of plant cell wall degrading enzymes, (2) increasing flux through lipid biosynthetic pathways, and (3) engineering lipid export.

Enhanced lignocellulolytic strains of N. crassa were developed and further engineered for lipid production. As a strategy to improve the lignocellulolytic potential of N. crassa, a strain was rationally engineered for increased enzyme secretion, and bioreactor cultivation conditions were optimized to balance fungal growth and production of lignocellulolytic enzymes, which resulted in a four-fold increase in cellulolytic productivity. Further, we made progress towards improving N. crassa as a heterologous protein expression system. Recombinant expression of supplemental lignocellulolytic enzymes from alternate microbial sources and/or engineered enzymes that exhibit improved properties for plant cell wall degradation may result in a fungal host with superior lignocellulolytic capabilities.

Both forward and reverse genetic approaches were employed to understand and harness the mechanism of hyper-accumulation of lipid in N. crassa. Putative genetic targets for the oleaginous phenotype of the morphological mutant (col-2) of N. crassa were identified in our forward genetic study. Complimentary, a rational multi-gene perturbation approach was used to increase triacylglycerol biosynthesis. Carbon flux was re-directed towards fatty acid biosynthesis from competing pathways, and feedback inhibition of fatty acid biosynthesis mediated by a specific pool of metabolic intermediates was reduced.

Finally, a novel mechanism for triacylglycerol secretion is proposed: ABC-transport of triacylglycerol that has been deposited between the leaflets of the plasma membrane. This proposed pathway requires heterologous expression of a triacylglycerol specific ABC-transporter and trafficking and hemi-fusion of lipid droplets with the plasma membrane. Preliminary proof-of-concept construction of this triacylglycerol secretion pathway is underway in S. cerevisiae; however, this novel secretion approach would be applicable to filamentous fungi and other microbes such as algae.

This work takes both a discovery-based and a directed approach to reveal targets for genetic manipulation with the ultimate objective of rationally designing industrially relevant cellulolytic fungi for production and secretion of lipids. The strategies applied here for increasing lignocellulose-derived lipid production and secretion will be foundational in enabling production of economically viable biodiesel from a cheap and widely-available feedstock.