A redox-independent stress response mediated by phase-separated SQSTM1/p62

ABSTRACT

The KEAP1 (kelch like ECH associated protein 1)− NFE2L2/NRF2 (NFE2 like bZIP transcription factor 2) pathway is a major antioxidative stress pathway that contributes to cellular homeostasis. KEAP1 acts as a sensor and attenuates degradation of the transcription factor NRF2, which induces gene expression for a network of enzymes involved in the antioxidant response. When cells are exposed to various electrophiles and reactive oxidative species, they modify one or more selective cysteine residues in KEAP1, resulting in conformational changes that disable its NRF2-inhibitory function. In addition to this redox-dependent pathway, SQSTM1/p62 (sequestosome 1), which is a selective autophagy receptor for ubiquitinated proteins and a driver of liquid-liquid phase separation (LLPS) upon binding to ubiquitinated proteins, competitively inhibits the binding between KEAP1 and NRF2, thereby disabling the NRF2-repressive function of KEAP1. Our study showed that phase-separated SQSTM1/p62 bodies are phosphorylated by ULK1 (Unc-51 like autophagy activating kinase 1) and that KEAP1 is retained in the SQSTM1/p62 body, resulting in NRF2-activation in a redox-independent manner.

Proteostasis through autophagic degradation of the phase-separated SQSTM1/p62 bodies

Upon accumulation and ubiquitination of proteins degenerated by oxidative stress or of abnormal nascent peptide chains after translation inhibition, these ubiquitinated proteins bind to the SQSTM1/p62 protein and undergo LLPS in combination with p62 oligomerization, forming membrane-less organelles called the SQSTM1/p62 bodies. The binding affinity of SQSTM1/p62 to ubiquitin chains is enhanced when Ser407 in the ubiquitin-associated (UBA) domain of SQSTM1/p62 is phosphorylated by ULK1 and Ser403 is phosphorylated by TBK1 (TANK-binding kinase 1) or CK2 (casein kinase 2). The SQSTM1/p62 bodies then mature while incorporating client proteins such as KEAP1 and the vaults, large ribonucleoprotein particles, as well as selective autophagy receptors such a s NBR1 (NBR1 autophagy cargo receptor) and TAX1BP1 (Tax1 binding protein 1). Binding of RB1CC1/FIP200 (RB1-inducible coiled-coil 1), an upstream factor involved in the autophagosome formation, to SQSTM1/p62 and/or TAX1BP1 in the SQSTM1/p62 bodies, leads to the generation of an isolation membrane/phagophore around the SQSTM1/p62 bodies. The SQSTM1/p62 bodies are selectively surrounded by an isolation membrane/phagophore through a wetting effect driven by the binding of ATG8 (autophagy related 8) family proteins present in the isolation membrane/phagophore, to SQSTM1/p62, NBR1, and TAX1BP1 associated to the SQSTM1/p62 bodies. In other words, SQSTM1/p62 and its LLPS in combination with SQSTM1/p62 oligomerization efficiently degrade ubiquitinated proteins and contribute to proteostasis.

Stress response through phase-separated SQSTM1/p62 bodies

Like other liquid droplets inside cells, the SQSTM1/p62 bodies are not just a passive entity degraded by autophagy; they also autonomously perform some specific functions. NRF2 is a master transcription factor that induces the expression of a set of genes encoding antioxidant proteins and detoxifying enzymes. Under normal conditions, NRF2 binds to KEAP1, the substrate recognition adaptor of Cullin 3-type ubiquitin ligases, and is ubiquitinated and degraded by the 26S proteasome. When cells are exposed to a variety of electrophiles and reactive oxidative species such as arsenite and hydrogen peroxide, and a subset of cysteine residues in KEAP1 undergoes to an oxidative modification, which inhibits binding between KEAP1 and NRF2. Consequently, dissociation of KEAP1 stabilizes NRF2, which translocates into the nucleus, thereby upregulating a set of NRF2 target genes.

In another pathway for NRF2 activation, SQSTM1/p62 competitively inhibits the binding of KEAP1 to NRF2, stabilizing NRF2. In our study, we found that this NRF2 activation is regulated by p62 bodies and their phosphorylation (Figure 1). In the absence of stresses, KEAP1 is in equilibrium between a cytoplasm and a SQSTM1/p62 body-associated form. Phosphorylation of Ser349 of SQSTM1/p62 in the phase-separated SQSTM1/p62 bodies by ULK1 or other kinases, increases the binding affinity of KEAP1 for SQSTM1/p62 retaining KEAP1 in the SQSTM1/p62 bodies [1]. Consequently, the cytosolic NRF2 ubiquitination by KEAP1 is suppressed. Because the SQSTM1/p62 gene is a target of NRF2, an emerging notion is that the SQSTM1/p62 bodies can also be formed by oxidative stress to activate NRF2. NRF2 targets include SQSTM1/p62 as well as genes encoding proteasomal subunits and ATG proteins. Thus, the SQSTM1/p62 body-mediated stress response mechanism, which is a redox-independent oxidative stress pathway, may be regulated by a balance between gene expression, proteolysis, and post-translational modifications of the SQSTM1/p62 bodies.

Figure 1.

Stress response by the phase-separated SQSTM1/p62 bodies.

Upon stresses, the levels of unfolded and then ubiquitinated proteins increase in the cytoplasm. SQSTM1/p62 oligomers or fibrils undergo multivalent interactions with these ubiquitinated proteins, resulting in their liquid-liquid phase separation and formation of SQSTM1/p62 bodies. KEAP1, a degradation factor of the transcription factor NRF2, is normally in an equilibrium between a cytoplasm and a SQSTM1/p62 body-associated form. Ser349 of SQSTM1/p62 is phosphorylated by ULK1 or other kinases while in the SQSTM1/p62 bodies, leading to a retention of KEAP1 in these membrane-less organelles. As a result, the amount of KEAP1 in the cytoplasm is reduced, and NRF2 degradation is inhibited. The resulting activation of NRF2 induces the gene expression of a series of antioxidant proteins and detoxification enzymes. NRF2 gene targets also include SQSTM1/p62 as well as genes encoding proteasomal subunits and ATG proteins.

Issues remaining to be addressed

Although the redox-independent oxidative stress pathway has been outlined here, several fundamental questions that remain to be answered. Mechanistically, it is unclear whether Ser349 phosphorylation of SQSTM1/p62 occurs before or after SQSTM1/p62 body formation. In vitro, mixing the Ser349 phosphorylation mimetic SQSTM1/p62 mutant with KEAP1, followed by the addition of a ubiquitin chain, suppresses SQSTM1/p62 body formation and results in an aggregate-like structure. Therefore, a possible scenario is that Ser349 is phosphorylated after KEAP1 is incorporated into the SQSTM1/p62 bodies, retaining KEAP1 in these organelles. The existence of this scenario, however, needs to be demonstrated experimentally. Moreover, is the phosphorylation of Ser349 of SQSTM1/p62 reversible? If so, what is the responsible phosphatase(s)? Arsenite treatment induces phosphorylation of Ser349 of SQSTM1/p62, but the amount of phosphorylated SQSTM1/p62 decreases in a time-dependent manner when arsenite is removed. This phenomenon is also observed in autophagy-incompetent cells. Thus, Ser349 phosphorylation of SQSTM1/p62 appears to be reversible. Identification of the responsible phosphatase(s) and the significance of dephosphorylation thus require further investigation. Phosphorylation of Ser349 of SQSTM1/p62 also increases the number of KEAP1 molecules in the SQSTM1/p62 bodies and the number of SQSTM1/p62 molecules bound to KEAP1, which in turn may change the physical properties of the SQSTM1/p62 bodies. That is, soluble proteins that are selective autophagy substrates need to be “concentrated in the cytoplasm” and “converted from a liquid to a gel-like state”. Therefore, retention of KEAP1 molecules in the SQSTM1/p62 bodies may result lead to a transition from a LLPS into a gel-like state that is subject to autophagic degradation. Enhanced degradation of SQSTM1/p62 bodies containing KEAP1 may thus be a negative feedback mechanism that enhances the degradation of NRF2 by the newly synthesized cytosolic KEAP1.

The role of the SQSTM1/p62 body-mediated stress response at the individual level should also be investigated. Ser349 phosphorylation-incompetent Sqstm1/p62 knock-in mice do not show any phenotype until 4 weeks of age under normal breeding conditions. In future, a more precise phenotypic analyses should be performed during aging, also under pathological conditions such as chronic inflammation or in model mice for Alzheimer disease, Parkinson disease, and amyotrophic lateral sclerosis (ALS). Mutations in SQSTM1/p62, which encodes SQSTM1/p62 protein have been identified in patients with ALS, frontotemporal dementia and other human pathologies. Whether these disease-associated mutations affect the redox-independent oxidative stress pathway linked to SQSTM1/p62 bodies is an important issue requiring further investigation.

Funding Statement

This work was supported by JSPS KAKENHI Grant Numbers [JP19H05706, JP21H004771, 23K20044, 24H00060] (to M.K.), [23K06415] (to Y.I.), AMED Grant Number [JP22gm1410004h0003] (to M.K.), Takeda Science Foundation (to M.K.), Uehara Memorial Foundation (to M.K.), Kobayashi Foundation (to M.K.), and Mitsubishi Foundation (to M.K.).

Abbreviations

KEAP1: kelch like ECH associated protein 1; LLPS: liquid-liquid phase separation; NFE2L2/NRF2: NFE2 like bZIP transcription factor 2; SQSTM1/p62: sequestosome 1; TAX1BP1: Tax1 binding protein 1; ULK1: unc-51 like autophagy activating kinase 1

Disclosure statement

No potential conflict of interest was reported by the author(s).