p62 condensates are a hub for proteasome-mediated protein turnover in the nucleus

The ability to selectively degrade cellular components ranging from proteins to large complexes and organelles is essential for cellular quality control. Failure leads to accumulation of unwanted material that gives rise to neurodegeneration, cancer, and infectious diseases (1). Two main degradative pathways have evolved: the ubiquitin–proteasome system (UPS) and autophagy. While the selective degradation of soluble proteins is normally conducted by the UPS, which needs unfolding of its substrates to traverse through the narrow openings of the proteasome, macroautophagy (hereafter termed autophagy) is able to sequester cytosolic cargo into a newly synthesized double-membrane compartment, called the autophagosome, that later fuses with the vacuole/lysosome for degradation. This empowers autophagy to degrade bulky cargo such as invading pathogens, protein aggregates, or damaged/disused organelles. Both pathways are essential components of the proteostasis network. In PNAS, Fu et al. (2) uncover a role of the protein p62 in orchestrating proteasomal protein turnover in the nucleus (Fig. 1).

Fig. 1.

Depiction of the functional roles of p62 within the cell. The cytoplasmic pool of p62 serves as a cargo receptor for the recognition of ubiquitylated protein aggregates by selective autophagy. Here, p62 targets ubiquitylated misfolded proteins and phase separates into condensates through multivalent interactions established with multiple Ubs linked in chains. The nuclear pool of p62 also forms condensates in a similar fashion as its cytosolic counterpart but recruits the UPS for efficient removal of nuclear proteins. The local p62-dependent concentration of the UPS members facilitates efficient removal of its substrates.

The multifunctional protein p62/SQSTM1 plays an important role in targeting ubiquitin (Ub)-modified proteins to either the proteasome or the autophagy machinery (3). Ubiquitylated proteins are captured by binding to the C-terminal Ub-associated (UBA) domain of p62 (4). To couple cargo recognition with autophagosome biogenesis, p62 interacts with lipidated LC3 via its LC3 interacting region (LIR). This interaction ties the cargo–receptor complex to the growing phagophore membrane. In addition, the Phox and Bem1p (PB1) domain of p62 mediates self-oligomerization, which leads to the formation of helical p62 filaments in vitro (5) and results within cells in clustering of p62 into so-called p62 bodies. Recent studies have shown that cytosolic p62 bodies form by liquid–liquid phase separation (LLPS) in the presence of ubiquitylated cargo proteins through multivalent interactions of p62 with multiple Ubs linked in chains (5–7). LLPS refers to a reversible process of concentrating molecules in a confined liquid-like compartment that stably coexists within the surrounding liquid environment. This mechanism is important to sufficiently concentrate the cargo–receptor complex for autophagic uptake. Moreover, it provides a dynamic surface to establish high-avidity interactions between the receptor and the autophagy machinery protein Atg8/LC3 (8, 9). Besides the well-characterized role of p62 in selective autophagy, it has been shown that the PB1 domain of p62 is able to bind to the 19S proteasomal subunits Rpn10 and Rpn1, which allows p62 to function as a shuttling protein, delivering ubiquitylated substrates to the proteasome (10, 11). Additional to the cytosolic pool of p62, p62 foci within the nucleus have been observed. A pioneering study showed that p62 contains two nuclear localization signals (NLSs) and shuttles continuously between the nuclear and cytosolic compartment (12). Blocking nuclear export leads to p62-dependent accumulation of polyubiquitinated proteins in different nuclear compartments such as promyelocytic leukemia bodies, and a function of p62 in the recruitment of proteasomes to these sites has been suggested (12, 13). Later studies showed that nuclear p62 accumulates also at sites of DNA damage, which affects the DNA damage response by proteasomal degradation of certain DNA damage repair factors (14), suggesting a relationship between nuclear p62 clusters and proteasomal degradation. However, mechanistic insights into the function of nuclear p62 clusters are still missing.

In PNAS, Fu et al. (2) provide insights into the function of p62 clusters by creating a CRISPR/Cas9-engineered p62 knockout HeLa cell line in which a GFP-tagged p62 lacking its nuclear export signal is constitutively expressed (GFP–p62∆NES). With this cellular model system, they are able to enrich the nuclear p62 pool without affecting nuclear export as a whole. They show in a series of experiments that nuclear p62 clusters also form in a ubiquitin-dependent manner by LLPS, similar to their cytosolic counterparts. Interestingly, fluorescence recovery after photobleaching experiments indicated that nuclear p62 foci recover faster than cytosolic foci, hinting at different physical properties and protein compositions. In line with previous results (12), they showed by immunofluorescence experiments that both a subunit of the 20S and the 19S subcomplex of the proteasome enrich at sites of p62 condensates, providing evidence for the accumulation of intact 26S proteasome complexes.

Fu et al. (2) then go on by using a fluorescent probe to profile proteasome activity directly in living cells. They found that the probe indeed stained the same areas around the p62 condensates as the antibodies against the proteasome subunits, demonstrating accumulation of an active proteasome pool at p62 condensates. Moreover, they not only observed an outer shell-like proteasome structure around the p62 condensates but additionally found enrichment of the E1/E2/E3 ubiquitin conjugation machinery as well as deubiquitylation enzymes together with the Ub-conjugated core of the condensates. This key finding places the whole UPS machinery necessary for ubiquitylation of substrate proteins as well as proteasome degradation in a confined space, suggesting a hub for efficient nuclear protein turnover. Indeed, one functional role of LLPS is to concentrate reactants to facilitate biological reactions (15).

The authors further elucidated on the degradative capacity of the nuclear p62 condensates by monitoring the degradation of different reporter proteins: 1) an NLS–GFP protein fused to a CL1-degron (a proteasomal degradation signal), 2) the overexpressed β4 subunit of the 20S core-particle, and 3) the short-lived and tightly controlled transcription factor c-Myc. In a set of experiments using the above-mentioned reporters, the authors clearly establish that nuclear p62 condensates are a hub for ubiquitin-mediated proteasomal protein turnover of nuclear proteins. They show in experiments where new protein synthesis is inhibited via addition of cycloheximide that all reporters are rapidly cleared from p62 condensates. Importantly, the authors also demonstrate that the whole SCF^Fbxw7 complex, the main E3 ligase that ubiquitylates c-Myc, is found enriched at p62 condensates.

In all cases, the presence of p62 condensates was important for efficient degradation of the individual reporters. Absence of p62 or presence of mutants lacking the PB1 domain or the UBA domain, unable to establish the p62 condensates, showed a decreased efficiency to remove the reporter proteins when compared to full-length p62. Interestingly, the absence of p62 condensates abolished substrate condensation and triggered accumulation of NLS–GFP–CL1 and c-Myc at the nucleolus. Recently, it was shown that the nucleolus serves as a quality control compartment for transient storage of misfolded nuclear proteins during heat stress (16, 17) and the protein chaperone Hsp70 plays a major role in refolding and extraction of these proteins from the nucleolus during recovery (17). In line, Fu et al. (2) observe that the chaperones Hsp70, Hsp90, and CHIP are enriched at p62 condensates. Various different stresses increased the size and number of p62 condensates. Massive accumulation of the reporter NLS–GFP–CL1, particularly in the nucleolus, was observed in p62KO cells after heat stress. This accumulation was completely reversed in reconstituted p62∆NES cells, suggesting a potential link between the two quality control compartments. The stress-induced formation of p62 condensates also revealed that p62 and the proteasome initially form foci independently of each other, and microscopy analysis further suggested that these two populations subsequently fuse to form proteasome-containing p62 condensates. The ability of proteasomes to undergo LLPS is well established (18, 19). Future studies will be needed to give additional insights into foci composition, formation, and substrate selection.

In situ proximity ligation assays in a previous study revealed that p62 and proteasomes not only colocalize in the nucleus but also in the cytosol (20). Whether the cytosolic interaction between p62 and proteasomes leads to activation or inactivation of proteasomes is still controversial (21, 22). Very recently, it was shown that overexpression of p62 leads to mislocalization of TDP-43 to the cytosol, a hallmark for developing amyotrophic lateral sclerosis (23). Furthermore, p62 overexpression shifted the balance toward disease-relevant TDP-43 aggregates. Interestingly, accumulation of the disease-relevant TDP-43 fraction was strongly abolished in an p62∆NES overexpression (23), further highlighting the quality control function of nuclear p62. It will be key to understand in the future what the distinct roles of these two pools are and how they converge into each other. It has been shown that proteasome inhibition induces p62 phosphorylation through the autophagy kinase complex ULK1/Atg1, making it more potent for autophagic clearance of protein aggregates (3). This activation could potentially be a switch between the proteasomal and autophagic function of p62. It will be important to get insights into the interplay of the different distinct p62 pools to understand its role in health and diseases.