Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2019 Aug 9;10(1):3600.
doi: 10.1038/s41467-019-11435-y.

A switch element in the autophagy E2 Atg3 mediates allosteric regulation across the lipidation cascade

Yumei Zheng  1   2 , Yu Qiu  1   3 , Christy R R Grace  1 , Xu Liu  4   5   6 , Daniel J Klionsky  4 , Brenda A Schulman  7   8   9
Affiliations

A switch element in the autophagy E2 Atg3 mediates allosteric regulation across the lipidation cascade

Yumei Zheng et al. Nat Commun. .

Abstract

Autophagy depends on the E2 enzyme, Atg3, functioning in a conserved E1-E2-E3 trienzyme cascade that catalyzes lipidation of Atg8-family ubiquitin-like proteins (UBLs). Molecular mechanisms underlying Atg8 lipidation remain poorly understood despite association of Atg3, the E1 Atg7, and the composite E3 Atg12-Atg5-Atg16 with pathologies including cancers, infections and neurodegeneration. Here, studying yeast enzymes, we report that an Atg3 element we term E123IR (E1, E2, and E3-interacting region) is an allosteric switch. NMR, biochemical, crystallographic and genetic data collectively indicate that in the absence of the enzymatic cascade, the Atg3E123IR makes intramolecular interactions restraining Atg3's catalytic loop, while E1 and E3 enzymes directly remove this brace to conformationally activate Atg3 and elicit Atg8 lipidation in vitro and in vivo. We propose that Atg3's E123IR protects the E2~UBL thioester bond from wayward reactivity toward errant nucleophiles, while Atg8 lipidation cascade enzymes induce E2 active site remodeling through an unprecedented mechanism to drive autophagy.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interests.

Figures

Fig. 1
Fig. 1
The Atg12–Atg5 module within the autophagy E3 activates intrinsic reactivity of the Atg3~Atg8 intermediate. a Structures of yeast Atg3 (left, PDB 2DYT), a ubiquitin E2 (middle, UBE2D1, PDB 4AP4), and their superimposition (right). Atg3 unique elements are indicated by colors shown in schematics of Atg3 constructs used in this study (below). b Intrinsic reactivity of Atg3~Atg8 intermediate monitored by pulse-chase discharge to NH2OH in presence of indicated E3 variant. c Close-up of structural superimposition of yeast Atg12–Atg5-Atg16 (PDB 3W1S) and corresponding human complex with ATG3FR (PDB 4NAW). Residues from yeast Atg12 corresponding to the human ATG3FR-binding site are shown as red sticks. d Effects of mutations in minimal yeast E3 (Atg12–Atg5) on stimulating Atg3~Atg8 discharge, monitored by quantification of Atg3~Atg8 remaining after 2.5 min in pulse-chase assays using the scheme as in (b), as a function of NH2OH concentration. (Error bar: STDEV, N = 3) Representative gel is shown in Supplemental Fig. 1b; all points are shown in Source Data File
Fig. 2
Fig. 2
E123IR plays essential role in E3-dependent Atg8 ligation activity. a Effects of indicated Ala mutants in Atg3′s FR on E3 (Atg12–Atg5-Atg16) activation of Atg3~Atg8 intermediate, quantified as percent Atg3~Atg8 remaining in pulse-chase discharge to NH2OH over 2.5 min. Locations of most defective and moderately defective mutants are indicated on schematics, and those corresponding to Atg3E123IR are shown in red. b Effects of indicated Ala mutants in Atg3′s FR on Atg8 lipidation in vitro, in reactions with Atg7, Atg12–Atg5-Atg16, and liposomes generated from E. coli polar lipids as source of PE, and detected by migration of Atg8 in Coommassie-stained SDS-PAGE gel. c Effects of indicated Ala mutants in Atg3′s FR on Atg8 lipidation in vivo, as detected by western blot for Atg8 after 2 h starvation of XLY161 atg3∆pep4∆ strain of S. cerevisiae expressing either WT or mutant HA-tagged Atg3. Pgk1 is loading control for Atg8, Dpm1 is loading control for Atg3
Fig. 3
Fig. 3
E123IR interacts with autophagy E3. a [15N, 1H] TROSY spectra of 15N-labeled Atg3FR titrated with unlabeled Atg12–Atg5-Atg16. b Estimated binding affinities between E123IR residues and Atg12–Atg5, based on chemical shift perturbations (CSPs) observed upon titrating increasing concentrations of Atg12–Atg5. c Chemical shift perturbations plotted as a function of Atg3FR residue numbers, with resonances showing intermediate exchange line broadening indicated by stripes
Fig. 4
Fig. 4
E123IR binds the E1 Atg7 and E2 core domain Atg3cat. a Close-up showing interactions between Atg3E123IR and N-terminal domain (NTD) of Atg7 (PDB 3T7G). b Effects of Atg3E123IR-binding WT Atg7 NTD and non-binding mutant (P283D) on E3-stimulated intrinsic reactivity of Atg3~Atg8 intermediate, as monitored by pulse-chase discharge to NH2OH. c [15N, 1H] TROSY spectra of 15N-labeled Atg3cat alone (red) or in 1:44 mixture with unlabeled Atg3FR (cyan), with chemical shift perturbations (CSPs) plotted per residue below. d [15N, 1H] TROSY spectra of 15N-labeled Atg3FR titrated with unlabeled Atg3∆FR, with representative (1:5 molar ratio Atg3FR vs. Atg3∆FR) CSPs plotted per residue below. e Estimated binding affinities between E123IR residues and Atg3∆FR, based on CSPs observed upon titrating increasing concentrations of Atg3∆FR. f Residues corresponding to greatest CSPs are shown as spheres on the structure of Atg3 (PDB 2DYT), except E308 and G309 that are not visible in the structure
Fig. 5
Fig. 5
Conformational changes upon E123IR removal from Atg3. a Crystal structure of Atg3∆NFR (PDB 6OJJ, from this study). b Superposition of Atg3∆NFR (PDB 6OJJ, light blue with catalytic Cys shown in green) with prior structure of Atg3FL (cyan, PDB 2DYT) with E123IR and catalytic Cys shown in red, and differences highlighted in cartoons. c Fo–Fc map shown at 3σ after omitting the catalytic cysteine region (residues 231–237) of Atg3∆NFR and performing simulated annealing. d Close-ups of catalytic elements from Atg3∆NFR and Atg3FL. e Close-up superposition of Atg3∆NFR (light blue) and Atg3FL (cyan) structures, showing interactions between Atg3′s catalytic domain and E123IR, and conformational rearrangements upon E123IR dislocation
Fig. 6
Fig. 6
Mutations in E123IR-binding residues activate Atg3~Atg8 in the absence of E3 in vitro and in vivo. a [15N, 1H] TROSY spectra of 15N-labeled Atg3FR alone (red) or in 1:2 mixture with unlabeled, disulfide-bonded proxy for Atg3∆FR–Atg8 (purple), with chemical shift perturbations (CSPs) per residue shown below. b Locations of mutations a–e designed to impair interactions between Atg3′s catalytic domain and E123IR shown on crystal structure of Atg3 (PDB 2DYT). c Effects of indicated mutants in Atg3 catalytic domain-E123IR interface on intrinsic E3-independent activity of Atg3~Atg8 intermediate. Quantification is of WT or indicated mutant versions of Atg3~Atg8 remaining after 2.5 min as a function of NH2OH concentration in pulse-chase assays, without E3, repeated 3 times repeats average (Error bar: STDEV, N = 3). Representative gel is shown in Supplemental Fig. 5; all data points are shown in Source Data file. d Effects of indicated mutants in Atg3 catalytic domain-E123IR interface on E3-independent Atg8 lipidation in vitro, in reactions with Atg7, Atg12–Atg5-Atg16, and liposomes generated from E. coli polar lipids as a source PE, and detected by migration of Atg8 in Coommassie-stained SDS-PAGE gel. e Effects of indicated mutants in Atg3 catalytic domain-E123IR interface on E3-independent Atg8 lipidation in vivo, as detected by western blot for Atg8 after 4 h starvation of the YCY131 multi-Atg knockout strain of S. cerevisiae expressing Atg7, Atg10, Atg8∆R (activated in absence of Atg4) and either WT or mutant HA-tagged Atg3
Fig. 7
Fig. 7
Extensive surfaces of Atg8 and the Atg3 catalytic domain are required for activation of the Atg3~Atg8 intermediate. a E3-dependent activation of Atg3~Atg8 intermediate testing roles of indicated surfaces through multiple-Ala scanning mutagenesis over Atg3′s catalytic domain. Quantification is of WT or indicated mutant versions of Atg3~Atg8 remaining after 2.5 min as a function of NH2OH concentration in pulse-chase assays, with Atg12–Atg5-Atg16 as E3. b E3-dependent activation of Atg3~Atg8 intermediate testing roles of indicated surfaces through multiple-Ala scanning mutagenesis over Atg8, performed as in (a). c Effects of covalent Atg3 complex formation with Atg8, as detected by comparing 15N, 1H] TROSY spectra of 15N-labeled Atg8 G116C (red) alone and disulfide-bonded complex with Atg3cat as a proxy for Atg3cat–Atg8 intermediate (cyan), with chemical shift perturbations per Atg8 residue shown below. Stripes indicate resonances with line-broadening due to intermediate exchange
Fig. 8
Fig. 8
Modeling of the active conformation of Atg3~Atg8 intermediate. a Generation of model for a potential closed E2~Ubl conformation for Atg3~Atg8, with structures of Atg3∆NFR (PDB 6OJJ, from this study) and Atg8 (PDB 2ZPN) superimposed on E2 and Ub, respectively, in a RING E3–E2–Ub complex (PDB 4AP4). b Sites of Atg3 mutations impairing E3-dependent activation of Atg3~Atg8, mapped on model for closed conformation. Red—residues corresponding to E2–Ub interface in closed conformation; wheat—residues corresponding to RING E3-binding site; bronze—residues in catalytic segment. c Sites of Atg8 mutations impairing E3-dependent activation of Atg3~Atg8, mapped on model for closed conformation. Red—residues corresponding to E2–Ub interface in closed conformation; orange—residues corresponding to AIM/LIR-binding site. d Atg8 residues showing chemical shift perturbation one standard deviation above the mean upon covalent complex formation with Atg3, mapped on model for closed conformation. Red—residues corresponding to E2–Ub interface in closed conformation; orange—residues corresponding to AIM/LIR-binding site
Fig. 9
Fig. 9
Schematic model for allosteric regulation of Atg3 activity through E123IR interactions across the lipidation cascade. The autophagy E2 Atg3 is autoinhibited by AtgE123IR. When Atg3 encounters E1 Atg7, E123IR is relocated upon binding to Atg7NTD, thereby triggering rearrangement of the Atg3 catalytic core to activate Cys234 for attacking the Atg7–Atg8 intermediate. Atg8 is transferred from Atg7 to Atg8, producing the thioester-bonded Atg3~Atg8 intermediate. Relieved from Atg7, Atg3E123IR protects the Atg3~Atg8 intermediate from wayward discharge to errant nucleophiles. The E3 (Atg12–Atg5-Atg16) binds E123IR and further activates the Atg3~Atg8 intermediate for nucleophilic attack. Numerous interactions with membranes place this complex in proximity to PE for the lipidation reaction
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Yang Z, Klionsky DJ. Eaten alive: a history of macroautophagy. Nat. Cell Biol. 2010;12:814–822. doi: 10.1038/ncb0910-814. - DOI - PMC - PubMed
    1. Nakatogawa H, Suzuki K, Kamada Y, Ohsumi Y. Dynamics and diversity in autophagy mechanisms: lessons from yeast. Nat. Rev. Mol. Cell Biol. 2009;10:458–467. doi: 10.1038/nrm2708. - DOI - PubMed
    1. Galluzzi L, et al. Molecular definitions of autophagy and related processes. EMBO J. 2017;36:1811–1836. doi: 10.15252/embj.201796697. - DOI - PMC - PubMed
    1. Stolz A, Ernst A, Dikic I. Cargo recognition and trafficking in selective autophagy. Nat. Cell Biol. 2014;16:495–501. doi: 10.1038/ncb2979. - DOI - PubMed
    1. Gatica D, Lahiri V, Klionsky DJ. Cargo recognition and degradation by selective autophagy. Nat. Cell Biol. 2018;20:233–242. doi: 10.1038/s41556-018-0037-z. - DOI - PMC - PubMed

Publication types

MeSH terms