Every living organism is must balance constant energy requirements against variable nutrient availability. One mechanism that has evolved to serve this purpose is the storage of energy as triglycerides (TGs). However, the nature of these molecules poses a problem; how can cells store an extremely hydrophobic molecule, shielding it from the hydrophilic cell environment, while maintaining easy access to the energy substrate? The answer is the lipid droplet (LD), a unique and ubiquitous organelle consisting of a phospholipid monolayer that shields a neutral lipid core, composed mostly of TGs and sterol esters (SE) in variable ratios. LDs also have a dynamic protein coat that varies among cell types and among LDs in the same cell. In the work presented here, we focus on the enzymes that synthesize TG, the diacylglycerol acyltransferase (DGAT) enzymes. We previously showed that cells lacking DGAT1 and DGAT2 are unable to synthesize TG and also lack LDs.

In chapter 2, we attempted to elucidate the cellular signals that drive TG synthesis and found that there is significant regulation of DGAT activity during the transition between fasting and feeding. Interestingly, DGAT1 and DGAT2 appear to play opposite roles in this regulation; DGAT1 is more active in fasting than during refeeding, while DGAT2 is more active during refeeding. We also observed significant decreases in activity with phosphatase treatment and identified several phosphorylation sites on both human and murine DGAT1. We were unable to determine whether the phosphorylation was linked to the change in activity observed in vivo.

In chapter 3, we identified and characterized family with a novel mutation in the DGAT1 gene. We determined that the mutation induced incorrect splicing of the DGAT1 mRNA and led to the omission of exon 8 from the full length mRNA. Removing exon 8 excises 75 base pairs from the mRNA, leading to an in-frame deletion of 25 amino acids. This protein product was not detectable when expressed as cDNA and was insufficient to rescue TG synthesis in mouse embryonic fibroblasts lacking DGAT1 and DGAT2. These results indicate the mutated gene gives rise to an unstable protein that is likely misfolded and rapidly degraded.

Finally, in chapter 4, we sought to develop model systems in which LD formation could be rapidly induced from a null background. We used adipocytes lacking DGAT2 combined with chemical inhibition of DGAT1 to block LD formation throughout the differentiation process. We found LD formation was not affected by nocodazole, brefeldinA or cycloheximide treatment. Inhibitors of fatty acid synthesis were also insufficient to block LD formation. However, 2-bromopalmitate did block LD formation. In permeabilized cells, we found that short chain diacylglycerols induced a dramatic change in the size of LDs as compared to the usual 1,2-dioleoylglycerol. We believe this is due to the surface active properties of the diacylglycerols. Finally, we found that treating cells with Bodipy-C12, a fluorescent fatty acid analog induced spots that appeared to be nascent LDs. However, these spots were not blocked in cells lacking both DGATs, nor were the spots coated by the LD-specific proteins perilipin 2 or perilipin 3. When assessing the biochemical nature of these spots by TLC, we found they were incorporated into in glycerolipids, but we were unable to identify the lipid species present. Still, we can conclude that with endogenous lipid substrates TG synthesis and LD formation were always absolutely correlated. Of all the tested conditions, we were only able to alter LD formation by addition of short-chain diacylglycerols. Thus, we believe LD formation is likely a biophysical process that may be modified by protein action, but is primarily driven by the physical properties of TG accumulation in the membrane bilayer.

These three chapters highlight the importance and complexity of TG synthesis at the cellular and organismal level. Here we have defined a framework for investigating the signals that control TG synthesis in vivo, a physiological context for when TG synthesis is important, and three cellular systems for understanding how triglycerides are packaged into lipid droplets.