MTORC1 regulates autophagic membrane growth by targeting WIPI2

ABSTRACT

WIPI2 contributes to autophagy by recruiting the ATG12–ATG5-ATG16L1 complex to PtdIns3P-rich membranes, which enables the growth and elongation of phagophores. So far, PtdIns3P and ATG16L1 are the best known WIPI2 interaction partners. In the screening of novel binding proteins for WIPI2, we recently identified interactions between WIPI2 and MTORC1 and the E3 ubiquitin ligase HUWE1. With this clue, we uncovered that WIPI2 is a phosphorylation substrate of MTORC1 and a ubiquitination target of HUWE1. Further, by determining the phosphorylation site on WIPI2 targeted by MTORC1, we show that MTORC1-dependent phosphorylation directs WIPI2 to interact with HUWE1, leading to WIPI2 ubiquitination and subsequent degradation by proteasomes. Finally, we provide evidence that this quantity control mechanism of WIPI2 is utilized by cells for not only the regulation of basal autophagy but also nutrient deprivation-stimulated autophagy. Together, these results indicate that WIPI2 protein level regulated by MTORC1 and HUWE1 is a pivotal determinant of cellular autophagy intensity. The findings firm up the role of MTORC1 as a master regulator of autophagy, suggesting that the quantity control of WIPI2 is a potential intervention point in autophagy-related physiopathological processes.

Abbreviations: HUWE1, HECT, UBA and WWE domain containing 1; MTORC1, mechanistic target of rapamycin complex 1; WIPI2, WD repeat domain phosphoinositide interacting protein 2.

KEYWORDS:

• Autophagy
• HUWE1
• MTORC1
• ubiquitination
• WIPI2

The process of macroautophagy/autophagy is regulated by many proteins at different stages. MTORC1, which serves as a major suppressor of autophagy, integrates intracellular and extracellular cues to coordinately control autophagy induction, execution and termination, by targeting a number of proteins required for autophagy initiation and autophagosome-lysosome fusion. In addition, MTORC1 influences the biogenesis of autophagosomes and lysosomes at the transcriptional level through manipulating the subcellular localization and thereby the transcriptional activity of TFEB, a transcription factor responsible for the expression of many autophagy and lysosome genes. However, so far, there is no evidence that MTORC1 directly participates in the stage of autophagic membrane growth and expansion.

One of the key events after membrane nucleation on the endoplasmic reticulum and phagophore formation from the omegasome is the association of MAP1LC3/LC3 with the phagophore. This is required for the growth of phagophores, which eventually become sealed autophagosomes, possibly by LC3-mediated fusion of vesicles from different intracellular organelles with phagophores. In the process, recruitment of WIPI2 to the membranes is of importance. With its affinity to both ATG16L1 and PtdIns3P, WIPI2 recruits the ATG12–ATG5-ATG16L1 complex, and subsequently LC3, to the PtdIns3P-rich phagophores. Nevertheless, up to date, compared to the other proteins functioning in the autophagy process, we know little about WIPI2, including its quality and quantity control, or its interaction partners besides ATG16L1 and PtdIns3P.

In seeking for novel WIPI2 interacting proteins to better understand the role of WIPI2 in autophagy, we screened them by carrying out mass spectrometry analysis of WIPI2 immunoprecipitated from cells stably expressing WIPI2-MYC [1]. Intriguingly, the E3 ubiquitin ligase HUWE1, which was previously shown to have affinity with WIPI2 by affinity proteomics using HUWE1 as bait, was on the list. We then used protein-protein interaction and protein ubiquitination assays, demonstrating that WIPI2 is a target of HUWE1, and ubiquitination by HUWE1 mediates WIPI2 proteasomal degradation. To investigate whether HUWE1-mediated WIPI2 degradation responds to autophagy signaling, we examined WIPI2 expression in cells with starvation or Torin1 (an MTORC1 inhibitor) treatment, two classic methods for autophagy induction. Surprisingly, WIPI2 protein but not WIPI2 mRNA, increases in the cells. In addition, the upregulation of WIPI2 protein in the cells is abolished by MG132 (a proteasome inhibitor) treatment and HUWE1 depletion. Knockdown of HUWE1 in cells cultured at basal conditions also causes accumulation of WIPI2, but not other ATG proteins such as ATG16L1 or ATG7. These results therefore suggest a constitutive regulation of intracellular WIPI2 protein by HUWE1, which is inhibited during autophagy induction.

Unexpectedly, RPTOR and MLST8, 2 MTORC1 subunits, are among the potential interaction candidates for WIPI2. This prompted us to explore the possible association between MTORC1 and WIPI2. By protein-protein interaction assay, in vitro kinase assay, and mass spectrometry analysis, we proved that WIPI2 is a direct phosphorylation substrate of MTORC1, and that Ser395 on WIPI2 is the major (maybe the only) phosphorylation site for MTORC1. In addition, we found that phosphorylation by MTORC1 significantly enhances the affinity of WIPI2 for HUWE1, thereby promoting WIPI2 ubiquitination and proteasomal degradation.

We then assessed the function of MTORC1-mediated WIPI2 phosphorylation in autophagy. In GFP-LC3-expressing cells, under nutrient-rich conditions formation of intracellular GFP-LC3 puncta and degradation of autophagy receptors SQSTM1/p62 and NBR1, are improved by overexpression of WIPI2 or knockdown of HUWE1. Compared to wild-type (WT) WIPI2, WIPI2^S395A, the phosphorylation-disabled WIPI2 mutant, displays a stronger effect, whereas WIPI2^S395D, the phosphorylation-mimicking WIPI2 mutant, demonstrates a much weaker effect, on autophagosome formation and autophagic degradation. In addition, in amino acid-starved cells, knockdown of WIPI2 also causes a block in autophagic degradation, which can be suppressed by introduction of WT WIPI2 or WIPI2^S395A, but not WIPI2^S395D. These results suggest an inhibitory effect of MTORC1-dependent phosphorylation of WIPI2 on autophagy.

The significance of MTORC1-dependent WIPI2 phosphorylation and HUWE1-mediated WIPI2 degradation were also evaluated in mouse. Strikingly, we detected a significant elevation of WIPI2 protein level in various mouse tissues under fasting conditions. Introduction of shHUWE1 to mouse using recombinant adeno-associated virus evidently causes an increase in WIPI2 protein level and LC3 puncta formation, together with a decrease in SQSTM1/p62 and NBR1 protein level in the liver of mice. The decrease of SQSTM1/p62 and NBR1 is also observed by introduction of WT WIPI2 and WIPI2^S395A, but not WIPI2^S395D. Meanwhile, reduction of neutral lipids is also observed in the liver of mice with introduction of WT WIPI2 or WIPI2^S395A.

Therefore, by identifying WIPI2 as a phosphorylation substrate for MTORC1 and a ubiquitination target for HUWE1, and by elucidating the causal relationship between the 2 post-translational modifications, our study uncovered a quantity control mechanism of WIPI2. Involvement of the mechanism in the control of both constitutive and adaptive autophagy suggests its potential utility in the intervention of autophagy-related diseases. Given that WIPI2 shows discrepant affinity for different phospholipids, phosphorylation of WIPI2 by MTORC1 may be involved in intracellular processes other than autophagy, according to its localization on different membranes.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This study was supported by the National Natural Science Foundation of China (31790402, 31530040 and 31671434) and the National Basic Research Program of China (2017YFA0503402).