ATG4B - 3D Mechanism Simulator

Loading structural data… (2Z0D)

References & Databases

START HERE — how to read pro-LC3 + H₂O → LC3-I A Cys-protease cartoon on the human ATG4B frame that actually exists: 2Z0D (1.90 Å) with crystal LC3(1–120). Occupancy is processed LC3 with free Gly120 carboxylate (OXT), not pro-LC3 and not LC3-PE. SEQADV writes ALA 280 ← UNP HIS 280 (H280A). UniProt ACT_SITE is His280 — missing on this mesh. Cys74 ★ is the nucleophile (SG dual occupancy 0.50/0.50; alt A extracted). Teaching LC3 is human; this fold is rat. Schematic uncleaved analog and PE/pro-peptide leaving group are [schematic]. Dual activity (pro-LC3 cut and LC3-PE delipidation) is one id. Blue is the nucleophile, not hydride. Why no PE: PE (phosphatidylethanolamine) is the membrane-lipid anchor of LC3-II; 2Z0D is ATG4B + LC3(1–120) and holds no PE, and no ATG page draws PE — it is always named, never modelled — so delipidation is stated as the same Cys74 chemistry but PE is never drawn. Measured Å are deposited 2Z0D pairs only. 2CY7 is apo ATG4B, not this mesh. ATG4A/C/D other genes. ATG5 next. Not a pure A-page (E~acyl).
Crystal structures PDB 2Z0D — human ATG4B–LC3(1–120) complex, 1.90 Å (Satoo 2009). Coordinates are public. Residue 280 is ALA (H280A). 2CY7 is apo ATG4B — not this mesh. Human ATG4B UniProt Q9Y4P1. LC3 on this deposit is rat Q62625.
Mechanism — Cys74 nucleophile, thioester, water hydrolysis ATG4B is the autophagy Cys protease: pro-LC3 + H₂O → LC3-I, and the same Cys74 chemistry delipidates LC3-PE. Compiled Cys74 thiol attacks the Gly120 carbonyl. This deposit captures product-like LC3-I with free Gly120 carboxylate, not the uncleaved peptide and not a thioester. Satoo et al., EMBO J. 2009, PMID:19322194. EC 3.4.22.-. Dual activity = one id. ATG4A/C/D other genes. Source: Satoo 2009 · UniProt Q9Y4P1
0 — Setup: pocket shown empty, Cys74 present 2Z0D holds processed LC3(1–120). The pocket is shown empty so the cut can start. Opaque, run:null (no autoplay). Cys74 ★ is the nucleophile. Residue 280 is ALA (H280A). Teaching LC3 is human; this fold is rat.Source: PDB 2Z0D (1.90 Å)
① — pro-LC3 analog binds A short pro-LC3 analog docks as a named stand-in for uncleaved Gly120–Xaa — [schematic], not the crystal LC3-I C-terminus. The same chemistry later delipidates LC3-PE; PE is not a second page. Carbonyl faces Cys74 SG.Source: compiled ATG4 processing · Satoo 2009
② — Cys74 attacks C → covalent E~acyl (hypothetical, lit.-based) Blue nucleophile rides from Cys74 SG to the analog carbonyl. Intact analog fades in place to schematic E~acyl plus a leaving group (pro-peptide, or PE on the delipidation beat). This is not crystal occupancy — 2Z0D holds free Gly120 carboxylate. Slow-motion on purpose. Blue = nucleophile, not hydride.Source: compiled Cys protease · UniProt Q9Y4P1 Cys74
③ — Leaving group departs (hypothetical, lit.-based) Schematic leaving group departs — pro-peptide on the processing beat, or PE on delipidation. E~acyl stays on Cys74. Crystal 2Z0D never held that thioester.Source: compiled ATG4 net accounting
④ — Water hydrolyses E~acyl → LC3-I (hypothetical, lit.-based) A schematic water glides in to the E~acyl face. A gold H⁺ leaves water H2 on the yellow arrow and vanishes right at the arrow tip; the remaining OH⁻ attacks E~acyl. Thioester fades in place to crystal LC3 Gly120 — the deposited free carboxylate (OXT). Deposited waters are not this catalytic water.Source: compiled cysteine-protease hydrolysis · 2Z0D Gly120 OXT
⑤ — LC3-I stays — ATG5 next Crystal LC3 Gly120 stays as the deposited free carboxylate. Cys74 SG ⋯ Gly120 C = 4.77 Å (SG alt A). Tyr54 OH ⋯ Gly120 O = 2.69 Å. This is LC3-I occupancy, not pro-LC3 and not LC3-PE. The same Cys74 chemistry delipidates LC3-II — named, not a second page. ATG5 next. 2CY7 is not this fold.Source: 2Z0D Cys74 SG alt A ⋯ Gly120 C

How to read this page — a first-pass primer First pass: Setup (pocket empty, Cys74 ★, H280A mesh) → ① schematic pro-LC3 analog docks → ② blue Cys74-SG → C, analog becomes E~acyl + leaving group → ③ leaving group leaves → ④ gold H⁺ then OH⁻; fade to crystal Gly120 → ⑤ LC3-I stays; ATG5 next. Second pass: 4.77 / 4.66 / 3.35 / 3.32 / 2.69 / 4.18 Å are crystal facts on 2Z0D; analog, thioester, leaving group, water, and H spheres are [schematic]. Third pass: open “Disagreements” for H280A ≠ His280, product-like ≠ pro-LC3, dual activity = one id, rat vs human LC3, and 2CY7 apo.
Chemistry primer — Cys protease, processing and delipidation ATG4B cuts after LC3 Gly120 via a cysteinyl thioester, then water hydrolyses E~acyl. The same Cys74 chemistry later removes PE from LC3-II. PE is a membrane lipid absent from 2Z0D (and from every ATG page), so delipidation is named, never drawn. Gold = H⁺. Blue = Cys74-SG nucleophile (not hydride). Teaching substrate = pro-LC3 analog. This is not ATG3 (E2) and not ATG7 (E1). ATG4A/C/D are other genes.
Words used on screen — H280A · LC3-I · dual activity · OXT H280A — residue 280 is ALA on this deposit; UniProt ACT_SITE is His280. LC3-I — processed LC3 with free Gly120 carboxylate; this crystal. Dual activity — pro-LC3 cut and LC3-PE delipidation share Cys74; one id, not two pages. OXT — deposited Gly120 C-terminus is a free carboxylate, not a peptide bond and not PE.

Legend & Fidelity

Mechanism fidelity — measured vs. inferred The measured layer: 2Z0D (1.90 Å) human ATG4B with rat LC3B; Cys74 SG ⋯ Gly120 C = 4.77 Å (SG alt A); Cys74 SG ⋯ Gly120 O = 4.66 Å; Cys74 CB ⋯ Gly120 C = 3.35 Å; Tyr54 OH ⋯ Gly120 C = 3.32 Å; Tyr54 OH ⋯ Gly120 O = 2.69 Å; Trp142 NE1 ⋯ Gly120 C = 4.18 Å. The inferred layer ([schematic]): uncleaved analog, E~acyl, leaving group, water, transferring H⁺/nucleophile spheres, product identity as LC3-I. Binding order is compiled staging, not a kinetic fit. The protein mesh is static. Cys74 SG alt B is spoken, not extracted.
Disagreements on record (a) H280A ≠ UniProt His280. Residue 280 is ALA; the catalytic histidine is missing. (b) Product-like occupancy ≠ pro-LC3. Crystal holds LC3(1–120) with free Gly120 carboxylate, not an uncleaved Gly120–Xaa peptide. (c) Not LC3-PE. Delipidation is named as the same Cys74 chemistry, not a second crystal. (d) Cys74 SG dual occupancy. Alt A (0.50) is extracted; alt B (0.50) is spoken, not a geomClaim. (e) Rat vs human LC3. Teaching LC3 is human; this fold is rat Q62625. (f) 2CY7 is apo ATG4B, not this mesh. (g) ATG4A/C/D are other genes. (h) EC 3.4.22.- is a class, not a fully assigned number. (i) no protein-domain swing. (j) gaps 187–217 / 289–293 / 340–344 named, not drawn.
Regulation — dual activity · ATG4 family · starvation This mesh is ATG4B holding product-like LC3. The same Cys74 chemistry processes pro-LC3 and later delipidates LC3-II — dual activity is one id, not two pages. ATG4A/C/D are other genes. Starvation / redox gating of ATG4 is drawer, not motion. Flexible loops 187–217 / 289–293 / 340–344 are missing on 2Z0D.Source: UniProt Q9Y4P1 · Satoo 2009
Kinetics scope (no measured timecourse on this page) One pro-LC3 + H₂O → LC3-I event is drawn, not a steady-state run. No measured k_cat/K_m is re-asserted here — no measured k_cat/K_m on this page [schematic]. Pair with a kinetics chapter.Source: Satoo 2009 — no k_cat/K_m re-asserted
Sequence Human ATG4B UniProt Q9Y4P1 · NCBI Gene 23192. This screen uses PDB numbering (Cys74 on 2Z0D = UniProt Cys74). Residue 280 is ALA (H280A); UniProt His280 is missing. Chain A Cα 5–354, gaps 187–217 / 289–293 / 340–344. LC3 chain B Cα 5–119 offset +1000. Gly120 is the ligand, not on the tube. Rat LC3B UniProt Q62625.
Simulation map — what you see is what is written Setup = 2Z0D pocket empty (opaque, no autoplay). ① schematic pro-LC3 analog moves in. ② slow-mo: blue Cys74-SG → C; analog fades to E~acyl + leaving group. ③ leaving group leaves. ④ gold H⁺ then OH⁻; fade to crystal Gly120. ⑤ Gly120 stays; ATG5 next. Only these events move.
Sources — bibliography of this page S1 — PDB 2Z0D: human ATG4B–LC3(1–120), 1.90 Å — public coordinates; measured SG⋯C = 4.77 Å · OH⋯O = 2.69 Å S2 — Satoo et al., EMBO J. 2009, PMID:19322194 — ATG4B–LC3 complex; LC3 processing and delipidation S3 — Sequence: UniProt Q9Y4P1 (ATG4B, ACT_SITE Cys74 / His280) · Q62625 (rat LC3B) · NCBI Gene 23192 S4 — Kinetics: no k_cat/K_m re-asserted on this page

› Backbones and ligand poses are taken directly from the listed PDB entries, superposed into one common frame. Where no crystalline state exists for a step (e.g., uncleaved analog, E~acyl, leaving group, water, H⁺ flight), the pose is drawn schematically and labeled as such in the narration. PE is a membrane lipid anchored to LC3-II in the membrane and is absent from 2Z0D, so delipidation is named, never drawn — no ATG page carries PE coordinates.

ATG4BLC3ONH⁺ (gold)nucleophile (blue)