Negative Regulation of Autophagy by UBA6-BIRC6–Mediated Ubiquitination of LC3

Context

Autophagy is a major degradation pathway, important for detoxifying cells and recycling material. It is a process that is essential for development and protection against neurodegeneration and many other diseases. Upon cellular signals for autophagy induction, members of the ubiquitin-like ATG8 family (LC3A, LC3B, LC3C, GABARAP, GABARAPL1 and GABARAPL2) are recruited to sites of phagophore formation, and help recruit proteins necessary for membrane expansion and curvature, allowing the formation of an autophagosome. LC3 is a small soluble protein, which is cleaved by ATG4 to produce LC3-I. The cleaved LC3-I is then conjugated to the lipid phosphatidylethanolamine, thereby forming LC3-II. LC3-II interacts with LIR domains of several adaptor proteins that are essential for cargo recruitment in autophagic degradation of for example ubiquitinated and aggregated proteins. The measure of LC3-II levels is a common method for assaying levels of autophagy, and increasing or decreasing levels of LC3-II has direct effects on the levels of autophagic flux ¹.

The authors of this preprint used a dual-tagged GFP-mCherry-LC3B construct to monitor autophagy. This “traffic light” approach is a common tool to monitor autophagy flux, as the green GFP signal is quenched when LC3-positive autophagosomes fuse with acidic lysosomes, leaving only the red signal intact. Thus the signal of red autolysosomes versus green+red (yellow) autophagosomes can provide a quantification of autophagic flux. Using this reporter construct, the authors carried out a genome-wide screen to identify new autophagy-regulating genes, and in doing so they identified UBA6 and BIRCH6 as novel regulators of LC3 levels.

 

Major Findings

In this preprint, the authors used an H4 cell line stably expressing GFP-mCherry-LC3B and transfected this cell line with a genome-wide CRISPR/Cas9 knockout library. Using FACS analysis, the authors could isolate cells with an increase in green/red ratio, indicating an increase in LC3 production and/or a block in autophagosome fusion with the lysosome. The isolated cells were then analyzed by next-generation sequencing to identify the mutations that caused this alteration in autophagy. With this approach, the authors identified UBA6 as a novel regulator of autophagy. Using siRNA-mediated knockdown, the authors confirmed that loss of UBA6 increases LC3-I, but not LC3-II levels, without affecting any other known autophagy regulators, or other proteins from the ATG8 family. Thus, stabilizing LC3-I levels on its own is not enough to induce autophagy, as the amount of lipidated LC3-II remained unchanged.

UBA6 is one of two E1 enzymes, responsible for activating ubiquitin before it can be conjugated to substrates by E2 and E3 enzymes. The authors find that UBA6 seems to mediate monoubiquitination of lysine 51 in LC3. Using a second CRISPR screen targeting all known E2 and E3 enzymes, they identified the E2 enzyme BIRCH6 as being involved in LC3 ubiquitination. Similar to knockout of UBA6, BIRC6 depletion increased the levels of LC3-I without inhibiting the conversion of LC3-I to LC3-II or affecting the degradation of LC3-II. Immunoprecipitations confirmed the physical interactions of BIRCH6, UBA6 and GFP-LC3, and an in vitro ubiquitination assay showed that UBA6 together with BIRCH6 can monoubiquitinate LC3, indicating that BIRCH6 acts as a combined E2/E3 ligase.

Studying the influence of increased LC3-I levels on autophagy, the authors found that while starvation induced an initial increase in LC3-II followed by its degradation, the levels of LC3-II remained high in UBA6 and BIRCH6 knockout cells. This phenomenon is likely due to continued replenishment of LC3-II from the larger pool of LC3-I. Consequently, stabilizing LC3-I levels together with autophagy inducing signals (by starvation in this experiment) can increase LC3-II levels and increase autophagic flux. In agreement with this, knockout of BIRCH6 caused a more rapid degradation of autophagy receptors p62 and NBR1 under conditions of protein synthesis inhibition, and knockout of either UBA6 or BIRCH6 prevented the accumulation of puromycin-induced aggresome formation in H4 cells as well as aggregation of mutant α-synuclein in rat neurons.

Thus, the authors of this preprint conclude that the UBA6-BIRCH6 axis acts to negatively regulate autophagy, via monoubiquitination of LC3-I, targeting it for proteasomal degradation (figure 1). Controlled inhibition of UBA6 or BIRCH6 could therefore be considered a potential target for therapeutic modulation of autophagic flux.

Figure 1. Based on figure 7F from the preprint. The E1 enzyme UBA6 together with the E2/E3 ligase BIRCH6 monoubiquitinate LC3 and target it for proteasomal degradation, thereby decreasing the pool of LC3-I that can be lipidated (PE) and function in autophagosome formation.

 

Why I like this preprint

LC3 is known to be degraded with the autophagic cargo in the lysosome, and previous studies have implicated a role for the ubiquitin-proteasome system (UPS) in the regulation of LC3 levels ². The authors of this preprint provide proof of the proteasomal degradation of LC3 and reveal the mechanism of how this degradation is regulated. This opens up for a possible way of modifying LC3 levels, and thereby autophagy, by targeting the UBA6-BIRC6 enzymes. These data also provide a mechanism to explain the interplay between the two major degradation systems of the cell, where it is known that blocking UPS induces autophagy ³. One way the block in UPS could be translated into an increase in autophagy could be the increased LC3 levels. Interestingly this degradation signal involves monoubiquitination and not polyubiquitination, which is the more common degradation signal.

 

Open questions

• How is the ubiquitination of LC3 regulated? Is it governed by mTORC1 and involved in the negative feedback loop to inhibit autophagy upon prolonged starvation ⁴? Or is the ubiquitination of LC3 regulated by p53 ⁵? What is the deubiqutinase of LC3?
• If UBA6/BIRCH6 works as a break to reduce autophagy and prevent its overactivation – do knockout cells experience more cell death?
• Previous studies have shown that proteolysis of LC3 by the 20S proteasome can be inhibited by p62 ². Is it possible that p62 compete with the E2/E3 ligase BIRCH6 for binding to LC3? Interestingly, the lysine 51, which was found to be monoubiquitinated, is also important for LC3 binding to autophagy cargo receptors such as p62 and NBR1. Is it possible that competing interactions could explain why LC3-II is not monoubiquitinated by BIRCH6? Or could this be explained by the membrane-localization of LC3-II, making it less accessible?
• In this study, it was shown that stabilization of LC3-I by knockout of UBA6 or BIRCH6 decreased aggregate/aggresome formation. Is this due to increased autophagic degradation or due to altered localization/transport of the substrates? What would be the effect in autophagy deficient cells, can UBA6/BIRCH6 knockout still protect cells from the aggregate load?

References

1          Kabeya, Y. et al. LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing. The EMBO journal 19, 5720-5728, doi:10.1093/emboj/19.21.5720 (2000).

2          Gao, Z. et al. Processing of autophagic protein LC3 by the 20S proteasome. Autophagy 6, 126-137, doi:10.4161/auto.6.1.10928 (2010).

3          Dikic, I. Proteasomal and Autophagic Degradation Systems. Annu Rev Biochem 86, 193-224, doi:10.1146/annurev-biochem-061516-044908 (2017).

4          Ye, J. et al. GCN2 sustains mTORC1 suppression upon amino acid deprivation by inducing Sestrin2. Genes & development 29, 2331-2336, doi:10.1101/gad.269324.115 (2015).

5          Scherz-Shouval, R. et al. p53-dependent regulation of autophagy protein LC3 supports cancer cell survival under prolonged starvation. Proc Natl Acad Sci U S A 107, 18511-18516, doi:10.1073/pnas.1006124107 (2010).

Tags: autophagy, lc3, ubiquitin, ups

Posted on: 18 July 2019 , updated on: 20 July 2019

doi: https://doi.org/10.1242/prelights.12178

Read preprint (No Ratings Yet)