Specialization: Plant Science Degree: Doctor of Philosophy Abstract: Triacylglycerol (TAG) is the major storage lipid in higher plants and has great nutritional and industrial value. Plant storage lipid biosynthesis involves different subcellular compartments and a complex network of enzymes and proteins. Among them, endoplasmic reticulum (ER)-bound diacylglycerol acyltransferase (DGAT) catalyzes the last and committed step in the acyl-CoA-dependent biosynthesis of TAG. Another key enzyme known as long-chain acyl-CoA synthetase (LACS) catalyzes the ATP-dependent formation of acyl-coenzyme As (CoAs) using free fatty acids. The LACS-catalyzed reaction provides an acyl donor to all the acyl-CoA-dependent acyltransferases including DGAT. Both DGAT and LACS are regarded as important targets for manipulating seed oil production. The overall goals of this thesis were to improve the enzyme performance of DGAT and LACS via protein engineering, and to explore their potential roles in TAG biosynthesis and the enhancement of the process. Numerous Brassica napus DGAT1 (BnaDGAT1) variants generated through directed evolution were shown to increase TAG content in yeast. In the first study, the possible roles of the ninth and tenth predicted transmembrane domain (TMD) in affecting BnaDGAT1 performance were revealed by mapping the beneficial amino acid residue substitutions of BnaDGAT1 variants onto a predicted topology model. To further investigate how the amino acid residues would affect enzyme performance, several BnaDGAT1 variants with amino acid residue substitutions residing in predicted TMD9 were characterized. These BnaDGAT1 variants increased yeast TAG content for different reasons including increased enzyme activity, increased polypeptide accumulation and/or possible reduced substrate inhibition. BnaDGAT1 variant L441P was found to display possible weak substrate inhibition and high catalytic efficiency. The beneficial amino acid residue substitutions of BnaDGAT1 variants were then transferred to Camelina sativa DGAT1 (CsDGAT1) and the resulting variant CsDGAT1 enzymes also possessed improved enzyme performance in yeast. Similarly to the equivalent BnDGAT1 variant L441P, possible reduced substrate inhibition was observed for CsDGAT1 variant L460P. Furthermore, the mutagenized libraries of BnaDGAT1 were screened again using linoleic acid and α-linolenic acid (ALA), respectively, and one variant containing seven amino acid residue substitutions exhibited altered preference towards linoleoyl-CoA and α-linolenoyl-CoA. In the second study, error-prone PCR was used to modify the enzyme performance of LACS9 from Arabidopsis thaliana (AtLACS9). Two AtLACS9 variants containing multiple amino acid residue substitutions were identified with improved enzyme activity. Site-directed mutagenesis suggested that increased enzyme activities of the two variants were mainly attributable to the single amino acid reside substitutions of C207F and D238E, respectively. C207 was found as a moderately conserved site among LACS9 from plant eudicots, whereas the unconserved site D238 was predicted under positive selection. Another two AtLACS9 variants, E520D and E630D, were identified to have increased preference toward linoleic acid. Seed oil from flax (Linum usitatissimum) is enriched in ALA, but the biochemical processes underlying ALA enrichment in flax are not fully elucidated. In the third study, a potential process involving the catalytic action of LACS and DGAT was proposed for channeling ALA into TAG. Flax LACS (LuLACS8A) exhibited enhanced specificity for ALA. Flax DGAT2 (LuDGAT2-3) was also found to display approximately 20-times increased preference towards α-linolenoyl-CoA over oleoyl-CoA. Incorporation of ALA into TAG via substrate channeling between LuLACS8A and LuDGAT2-3-catalyzed reactions was supported by both in vitro and in vivo (in yeast) experiments. Finally, membrane yeast two-hybrid assays revealed several interactions among enzymes involved in acyl-editing and TAG assembly in flax. Among these protein-protein interactions, the identification of a physical interaction between flax lysophosphatidylcholine acyltransferase 2 (LuLPCAT2) and LuDGAT1-1 supports previous evidence for biochemical coupling of the LPCAT-catalyzed reverse reaction to form acyl-CoA from phosphatidylcholine to the DGAT1-catalyzed forward reaction. In summary, the findings in this thesis provide insight into the amino acid residues underlying plant DGAT1 and LACS9 function and will benefit the development of innovative strategies to manipulate oil production in oleaginous plants and microorganisms.