https://orcid.org/0000-0002-9571-094X
ABSTRACT
Macroautophagy/autophagy cargo receptor SQSTM1/p62 puncta or clustering formation is critical for its function in cargo recognition and LC3 interaction. Evidence suggests that SQSTM1 puncta formation is a process of liquid-liquid phase separation. It is poorly understood how SQSTM1 liquid-liquid phase separation is regulated. We found that cytoplasmic DAXX enhances SQSTM1 puncta formation, and further demonstrated that DAXX drives SQSTM1 liquid phase condensation through increasing SQSTM1 oligomerization. DAXX promotes SQSTM1 recruitment of KEAP1, subsequently activating an NFE2L2/NRF2-mediated stress response. This study suggests a new mechanism of SQSTM1 phase condensation by a protein-protein interaction, and indicates that cytoplasmic DAXX can play a role to regulate redox homeostasis.
In recent years, a range of intracellular body structures, including SQSTM1 bodies, nucleoli, cajal bodies, P granules and promyelocytic leukemia (PML) bodies, have been demonstrated as non-membrane-bound liquid compartments. These non-membrane structures represent open macromolecular assemblies where bio-materials are organized in a dynamic manner. Compelling evidence suggests that these non-membrane-bound assemblies are formed as liquid droplets though a liquid-liquid phase separation (LLPS) process, during which biomacromolecules de-mix from the surrounding solution and form a separate liquid phase. While these droplets comprise hundreds of proteins, only a subset of proteins execute the liquid droplet formation and condensation. It is critical to elucidate the roles of the key proteins in the process. The autophagy receptor SQSTM1 clusters as body-like structures, and recognizes the cargo via its ubiquitin binding. SQSTM1 bodies as non-membrane coherent structures compartmentalize and concentrate specific sets of proteins for selective autophagic clearance. SQSTM1 undergoes phase separation with ubiquitin chain binding. It was tempting to speculate that the LLPS of SQSTM1 is dynamically regulated to meet its multi-functions for cellular homeostasis. As such, we aimed to examine if any interacting partners of SQSTM1 regulate its phase condensation.
Through yeast two-hybrid screening, we initially found that one of the SQSTM1-binding candidates, DAXX, known as a histone H3F3/H3.3 chaperone, is a top potential binding partner of SQSTM1 [Citation1]. We confirmed the physical interaction between DAXX and SQSTM1 by immunoprecipitation, GST affinity isolation and bimolecular fluorescence complementation approaches. DAXX colocalizes with SQSTM1 puncta. Intriguingly, virtually all SQSTM1 puncta with large size colocalize with DAXX body-like structures, but not all DAXX structures colocalize with SQSTM1 puncta. SQSTM1 puncta are consistently larger in the presence of DAXX than those in the absence of DAXX. This suggests that DAXX can regulate SQSTM1 puncta formation. To avoid the potential influence of transfection efficiency on the size of SQSTM1 puncta, we generated a Tet-on inducible SQSTM1-GFP-expressing cell line. Using this cell line, we confirmed that the SQSTM1 puncta sizes are significantly increased in the presence of DAXX. To examine if loss of DAXX can reduce the formation of SQSTM1 puncta, we generated daxx KO cell models as well as DAXX knockdown cells, and observed that SQSTM1 puncta formation is weakened in the cells with DAXX ablation. To exclude any potential effect of an increase or decrease in DAXX expression on autophagic activity and subsequently on SQSTM1 puncta formation, we used a set of autophagy-defective cells for SQSTM1 puncta assays, and found that DAXX increases the size of SQSTM1 puncta independently of autophagy.
Fluorescence recovery after photobleaching (FRAP) indicates the exchange rate of fluorescence-labeled molecules between a droplet and its surrounding environment. Using FRAP and in vitro phase separation assays, we concluded that DAXX increases SQSTM1 phase condensation. We further determined that DAXX promotes SQSTM1 oligomerization by increasing the SQSTM1-SQSTM1 interaction. Using capillary western immunoassays (WES), we observed that less oligomeric SQSTM1 is formed in daxx KO cells in native conditions. In SDS-PAGE conditions, SQSTM1 oligomerization was detected in wild-type mouse embryonic fibroblasts (MEFs), but not in daxx KO MEFs. In vitro assay showed that DAXX is sufficient to induce SQSTM1 oligomerization. This effect appears independent of the PB1 domain that is required for SQSTM1 oligomerization. Our in vitro native WES assay showed that DAXX is sufficient to induce the oligomerization of SQSTM1 lacking the PB1 domain. Moreover, DAXX is required for SQSTM1 recruitment of polyubiquitinated cargos, given that the colocalization between polyubiquitinated proteins and SQSTM1 positively correlates with DAXX levels in cellular and Drosophila models. In DAXX-ablated cells, less polyubiquitinated proteins are affinity isolated with SQSTM1.
The KEAP1-NFE2L2 pathway, which regulates the reactive oxygen species (ROS)-response gene transcription for cytoprotection, is activated by SQSTM1 puncta formation. This is because SQSTM1 puncta retain KEAP1 in the cytoplasm, freeing NFE2L2 for its nuclear translocation to transcribe a battery of target genes encoding anti-oxidant enzymes. SQSTM1 knockdown reduces the levels of nuclear NFE2L2. Likewise, DAXX knockdown decreases the nuclear NFE2L2 levels, given that less SQSTM1 puncta are formed in the absence of DAXX, leading to more NFE2L2 cytoplasmic localization by KEAP1 binding. Therefore, DAXX promotes ROS responses by increasing NFE2L2 nuclear localization. Finally, we found that DAXX is protective against cell death for the cells exposed to ROS and peroxidation.
Our data reveal a new mechanism for SQSTM1 phase separation and puncta formation, and show that DAXX plays a role in selective autophagy in the cytoplasm, in addition to its known nuclear role as a chaperone for histone H3F3/H3.3 deposition onto the chromosome.
Acknowledgments
We are grateful to the Medical Research Council (MR/M023605/1) and the Academy of Medical Sciences (NAF\R1\191045) for funding.
Disclosure statement
No potential conflict of interest was reported by the authors.
Additional information
Funding
Reference
- Yang Y, Willis TL, Button RW, et al. Cytoplasmic DAXX drives SQSTM1/p62 phase condensation to activate Nrf2-mediated stress response. Nat Commun. 2019;10:3759.