Glycogen Debranching Enzyme - 3D Mechanism Simulator

Loading structural data… (8ZEQ)

References & Databases

START HERE — how to read two AGL activities on one chain One polypeptide, two catalytic centres, one apo EM mesh. Real frame: human GDE 8ZEQ (3.36 Å cryo-EM, chain A, no ligand). Tab 1 is the N-terminal 4-α-glucanotransferase (EC 2.4.1.25) at Asp526 ★. Tab 2 is the C-terminal amylo-1,6-glucosidase (EC 3.2.1.33), an inverting GH133 enzyme in which water is the nucleophile; the human GC catalytic pair is Asp1261 + Glu1502 and Asp1333 ★ is the base seat drawn here. The two tabs therefore differ in stereochemical class, not just in position. Limit dextrin, acceptor, branch glucose, and water are all [schematic] — 8ZEQ deposits none of them. A covalent glycosyl-enzyme is disclosed for the transferase (not a pure A-page) and is explicitly absent from the glucosidase. Gold H⁺ is schematic. Blue is nucleophile, not hydride. Measured Å are two apo residue pairs only. GSD III (Cori) is named, not simulated. PGM1 is not this enzyme.
Crystal structure PDB 8ZEQ — human glycogen debranching enzyme, 3.36 Å cryo-EM, apo (Guan et al. 2025, PMID 40593796). Coordinates are public. Mesh = chain A, residues 1–1532 (monomer). UniProt P35573. NCBI Gene 178 (AGL). No sugar or water is in this deposit.
Mechanism — transferase then glucosidase After glycogen phosphorylase stops short of a branch, AGL transfers that limit-dextrin maltotriosyl group onto a nearby 4-OH (transferase), then hydrolyses the leftover α-1,6 glucose (glucosidase). The transfer is retaining, so this page shows it as two beats: ② turns the donor's anomeric ring over and parks C1 on Asp526, and ④ slides the acceptor onto that already-placed C1. Compiled GDE writes glycosyl-Asp + transfer as one cycle and water attack with proton transfer as one cycle; this page splits those beats. Source: UniProt P35573 · PDB 8ZEQ · PMID 40593796
Tab 1 · 0 — Setup: empty transferase pocket 8ZEQ apo. Asp526 ★ / His529 / Asp627 are on the mesh. Opaque, run:null (no autoplay). No sugar is deposited.Source: PDB 8ZEQ (3.36 Å)
Tab 1 · ① — Limit dextrin presented [schematic] A limit-dextrin stub docks with C1 toward Asp526 ★. Not a deposited sugar. Binding order is compiled staging, not a kinetic fit.Source: schematic pose off 8ZEQ Asp526
Tab 1 · ② — Ring inversion, then Asp526 ★ attacks C1 (hypothetical, lit.-based) A retaining GH13 transferase has to turn its anomeric ring over first, so the donor C1 is presented to the nucleophile along the line Asp526 will attack. The donor turns 180° about its own C1–C5 bond, then the whole stub slides in until C1 is 1.45 Å from Asp526 ★ (blue nucleophile, not hydride), forming a covalent glycosyl-Asp C–O bond. C1 does not travel again — ④ slides the acceptor onto it instead. Slow-motion on purpose. Schematic C1⋯OD1 is not a geomClaim.Source: compiled EC 2.4.1.25 · retention mechanism · Braun 1996 (doi 10.1021/bi9526488) · Nakayama 2001 (doi 10.1074/jbc.M102192200) · Guan 2025
Tab 1 · ③ — Acceptor chain binds [schematic] The remaining chain docks as the 4-OH acceptor, drawn as an O4 handle. It lines its O4 up with the glycosyl C1, still 3.05 Å away along the same axis Asp526 came in on, and the glycosyl-Asp bond holds. No ternary crystal on 8ZEQ.Source: schematic acceptor on 8ZEQ transferase seat
Tab 1 · ④ — 4-α transfer (hypothetical, lit.-based) Read as a second displacement: the acceptor O4 is the nucleophile (blue) and attacks the anomeric C1, while a gold H⁺ leaves O4. Because the mechanism is retaining, C1 is already where it has to end up, so the acceptor travels onto it (1.45 Å slide) and the donor does not move at all. The glycosyl-Asp bond to Asp526 is released as a new α-1,4 C1–O4 bond is drawn, so the acceptor is elongated by one residue. The stub is kept, not faded — it stays attached and ⑤ carries it off with the acceptor. Asp526 regenerates (inferred). Slow-motion on purpose.
Not a measurement: 8ZEQ is apo with no sugar deposit, so the O4···C1 approach shown (closing 3.05 Å → 1.45 Å on the [schematic] stand-ins) is teaching geometry, not a geomClaim. The drawn 1.45 Å is a nominal C–O glycosidic length, not an observed contact. The acceptor is drawn as an O4 handle rather than a full pyranose because two complete pyranoses cannot be placed clash-free in this apo pocket; the O4···C1 axis is 36° off the ideal 180° in the drawn pose.Source: compiled EC 2.4.1.25 (retained intermediate, Mechanism A). No ternary 4-α-transferase structure on this frame; the transfer geometry is lit.-based staging, not a trapped state.
Tab 1 · ⑤ — Acceptor leaves — glucosidase is next The elongated chain — acceptor plus the transferred stub — departs together as one unit, keeping the α-1,4 bond drawn in ④. The leftover α-1,6 glucose is the substrate of the glucosidase tab on this same polypeptide. Asp526 regenerates (inferred).Source: glycogen-breakdown bookkeeping
Tab 2 · 0 — Setup: empty glucosidase pocket Same 8ZEQ apo mesh. This domain is an inverting GH133 enzyme, not GH31, so water is the nucleophile and there is no covalent glycosyl-Asp. The human GC catalytic pair is Asp1261 + Glu1502; Asp1333 ★ is a real deposited residue used only as the nearby base seat for this staging. Water and glucose are [schematic]. Opaque, run:null.Source: PDB 8ZEQ · CAZy GH133 · Guan et al. Nat Commun 2025;16:5962 (PMID 40593796)
Tab 2 · ① — Branch glucose presented [schematic] Leftover branch glucose docks with C1 toward the Asp1333 ★ base seat. This is hydrolysis of the α-1,6 linkage, not another 4-α transfer, and it inverts at C1 (GH133) — unlike the retaining transferase in tab 1.Source: schematic pose off 8ZEQ · CAZy GH133 (inverting) · Guan et al. Nat Commun 2025
Tab 2 · ② — Catalytic water binds [schematic] Water slides in toward glucose C1. 8ZEQ is apo — this water is not deposited.Source: schematic water on 8ZEQ glucosidase seat
Tab 2 · ③ — Water H⁺ parks — H₂O → OH⁻ (hypothetical, lit.-based) Gold H⁺ rides from water H2 toward the glucosidase base seat and parks. Removing H2 leaves OH⁻. Compiled GH133 references write deprotonation and attack as one event; this page splits them.Source: compiled EC 3.2.1.33 · CAZy GH133
Tab 2 · ④ — OH⁻ attacks C1 → free glucose (hypothetical, lit.-based) Blue nucleophile (not hydride) from water O toward branch C1. The α-1,6 linkage is hydrolysed to free glucose [schematic]. Parked gold H⁺ stays. Slow-motion on purpose.Source: compiled EC 3.2.1.33 · Guan 2025
Tab 2 · ⑤ — Glucose + H⁺ leave Free glucose and the net gold H⁺ depart. Water was consumed. Net debranching: transferase then glucosidase on one chain. PGM1 is not this enzyme.Source: glycogen-breakdown bookkeeping

How to read this page — a first-pass primer First pass: Tab 1 Setup (empty transferase, Asp526 ★) → ① limit dextrin docks → ② ring inversion, then glycosyl-Asp on Asp526 → ③ acceptor docks as an O4 handle → ④ 4-α transfer, acceptor slides onto the fixed C1 → ⑤ chain leaves. Tab 2 Setup (empty glucosidase, Asp1333 ★) → ① branch Glc docks → ② water docks → ③ gold H⁺ parks → ④ OH⁻ attacks C1 → ⑤ glucose leaves. Second pass: only the two deposited residue-pair distances are printed — sugars and water are schematic. Third pass: open “Disagreements” for two centres / one chain, glycosyl-Asp vs A-page, and GSD III.
Chemistry primer — debranching is two reactions Phosphorylase cannot pass an α-1,6 branch. The transferase moves the leftover α-1,4 stub onto another chain. The glucosidase then clips the single remaining α-1,6 glucose with water. Gold = H⁺. Blue = nucleophile (not hydride). Free glucose from this enzyme is not G1P and not G6P.
Words used on screen — limit dextrin · glycosyl-Asp · O4 handle · schematic Limit dextrin — glycogen after phosphorylase stops short of a branch; the transferred unit is a maltotriosyl group. Glycosyl-Asp — covalent Asp526–sugar intermediate of the retaining transferase; disclosed, not a pure A-page. O4 handle — the acceptor is drawn as a partial sugar carrying its attacking O4, because two complete pyranoses cannot be seated clash-free in this apo pocket. Schematic — sugars and water are teaching poses off Asp526 / Asp1333, not a deposited ternary.

Legend & Fidelity

Mechanism fidelity — measured vs. inferred The measured layer: 8ZEQ (3.36 Å) human apo GDE; Asp526 OD1 ⋯ His529 NE2 = 6.82 Å; Asp526 CA ⋯ Asp627 CA = 13.13 Å. The inferred layer (all [schematic]): limit dextrin, acceptor, branch glucose, water, glycosyl-enzyme, transferring H⁺/nucleophile spheres, and product leave. Binding order is compiled staging, not a kinetic fit. The protein mesh is static. Schematic sugar contacts are not geomClaims.
Disagreements on record (a) One enzyme, two activities. Both centres sit on one polypeptide; this page splits them into tabs, not two URLs. (b) Glycosyl-enzyme vs A-page. The transferase is retaining GH with a covalent intermediate — disclosed, not sold as a pure A-page. (c) No sugar co-deposit. 8ZEQ is apo EM. Overlay sugar Å are not geomClaims. (d) The glucosidase is GH133, not GH31, and it inverts. Water is the nucleophile; there is no covalent glycosyl-Asp in this domain. The human GC catalytic pair is Asp1261 + Glu1502 (Guan et al. Nat Commun 2025); Asp1333 ★ is only the base seat used for this staging, not a catalytic nucleophile. (e) Catalog vs split. Compiled GDE writes glycosyl-enzyme+transfer and water-deprotonation+attack as single cycles; this page splits them. (f) GSD III (Cori) is named, not simulated. (g) Triad vs displayed residues. The compiled transferase catalytic triad is Asp526 / Glu555 / Asp627; His529 is not part of it, and Glu555 is not drawn — the screen shows Asp526 ★ / His529 / Asp627 (UniProt ACT_SITEs).
Regulation — cytosolic GDE · GSD III AGL is the cytosolic debrancher after PYGL. Human AGL (P35573) is this gene. GSD III (Cori / Forbes) is deficiency of this enzyme — named on this card, not animated. No allosteric crystal is claimed here.Source: UniProt P35573 · Guan 2025 PMID:40593796
Kinetics scope (no measured timecourse on this page) One transferase event and one glucosidase event are drawn, not a steady-state run. Compiled BRENDA values for EC 2.4.1.25 / 3.2.1.33 are not re-asserted here — no measured k_cat/K_m on this page [schematic]. Pair with a kinetics chapter.Source: BRENDA EC 2.4.1.25 · EC 3.2.1.33 — no k_cat/K_m re-asserted
Sequence Human AGL UniProt P35573 · NCBI Gene 178. Residue numbers on this screen are 8ZEQ (Asp526, His529, Asp627, Asp1333, Asp1344, His1352).
Simulation map — what you see is what is written Tab 1 Setup = empty transferase pocket (opaque, run:null). ① limit dextrin moves in (schematic). ② slow-mo: Asp526 attacks C1. ③ acceptor docks. ④ slow-mo: 4-α transfer — blue nucleophile rides acceptor O4 → stub C1, gold H⁺ leaves O4, the acceptor slides onto the already-placed C1 (donor does not move), and a new α-1,4 bond mesh is drawn at swap. ⑤ elongated chain (acceptor + stub) leaves together. Tab 2 Setup = empty glucosidase pocket. ① branch Glc docks. ② water docks. ③ gold H⁺ parks. ④ slow-mo: OH⁻ attacks C1. ⑤ glucose + H⁺ leave. Only these events move.
Sources — bibliography of this page S1 — PDB 8ZEQ: human GDE apo EM, 3.36 Å — public coordinates (Guan 2025, PMID 40593796) S2 — Guan et al., Nat Commun 2025, PMID:40593796 — molecular architecture of human GDE S3 — Sequence: UniProt P35573 · NCBI Gene 178 S4 — Kinetics: BRENDA EC 2.4.1.25 · EC 3.2.1.33 — no k_cat/K_m re-asserted on this page

› Backbones are taken directly from the listed PDB entry. 8ZEQ is apo — no sugar or catalytic water is deposited. Where no crystalline state exists on this frame (limit dextrin, acceptor, branch glucose, water, glycosyl-enzyme, product release), the pose is drawn schematically and labeled as such in the narration.

N-terminal / transferaseC-terminal / glucosidaseONH⁺ (gold)nucleophile (blue)