Molecular regulation of proteasome abundance and localization

Doctor of PhilosophyDivision of BiologyJeroen RoelofsProteasomes are molecular complexes that degrade more than 80% of proteins in eukaryotes. This makes these proteases essential for processes associated with cell survival and proliferation. Proteasomes are localized to both the cytosol and nucleus, with the majority in the nucleus of dividing cells. In recent years, much progress has been made in understanding how the 33 unique subunits of proteasomes assemble into active protease complexes. The identification of ten dedicated chaperones that facilitate assembly suggest that this is an intricate process. Thus, quality control dependent degradation for complexes that fail to assemble correctly likely exist. However, there is little known about when and how proteasomes are degraded in cells. Considering the large size and complex assembly of proteasomes, we hypothesized that their degradation would be carried out via autophagy. Autophagy involves the de-novo synthesis of double membrane vesicles capable of engulfing large cytosolic cargo. In this work, we show that proteasomes are indeed degraded through autophagy. This degradation requires the canonical autophagy genes required for general autophagy. Importantly, factors required for selective autophagy play a role in the degradation of a sub-set of proteasomes, and proteasome autophagy is dependent on several factors not required for general autophagy. Further, we show that several general autophagy-inducing stimuli do not result in proteasome autophagy, but instead result in proteasomes remaining nuclear or being targeted to cytosolic puncta termed proteasome storage granules (PSGs). Thus, our data support the model that proteasomes, depending on the physiological conditions, are selectively regulated. To understand the signaling events involved, we sought to determine how cells make the choice between PSG formation and proteasome autophagy. We hypothesized that carbon starvation, a PSG-inducing condition, unlike autophagy-inducing stimuli, affected proteasomes through a strong reduction in cellular energy. Proteasomes need to hydrolyze ATP for normal biological function, and in vitro proteasome complexes are unstable in the absence of ATP. When we tested conditions that inhibit mitochondrial function, and thus cellular energy levels, we observed PSG formation. Consistent with our model, we observed a reduction in ATP levels in several PSG-inducing conditions, though not for all. If destabilization due to low ATP levels is a key step in the re-localization of proteasomes, we predicted factors that stabilize proteasome complexes, such as proteasome inhibitors or association with Ecm29, would prevent PSGs. Indeed, we found that both proteasome inhibitors and Ecm29 reduced the amount of PSGs formed, and restricted proteasomes to the nucleus. This suggests that proteasome disassembly into subcomplexes plays a role in PSG formation. Further, we identified several proteins that play a role in proteasome re-localization. In all, our data provide new insight into the proteasome’s cellular dynamics under a variety of physiological conditions. Our current model is that the re-localization of proteasomes facilitates cell survival under stress. Here, proteasome autophagy-inducing conditions benefit from the conversion of excess proteasomes to amino acids required for survival. On the other hand, proteasome granules restrict ATP use and proteasome activity under PSG-inducing conditions