Glyceraldehyde-3-Phosphate Dehydrogenase

Frame — 1U8F, human GAPDH (NAD⁺-holo crystal; the animation opens visually empty and replays the deposited NAD⁺ at ①) 1U8F = human cytoplasmic GAPDH, a 1.75 Å NAD⁺-holo complex (Jenkins JL, Tanner JJ, Acta Crystallogr D 62:290–301, 2006, RCSB 1U8F · PMID:16510976 · DOI:10.1107/S0907444905042289; preferred tetramer, chains O/P/Q/R, no mutations, r=0.191 / r_free=0.217). NAD⁺ is bound in chains P, Q and R at author residue 336 (chain O carries none). Protein frame (Cα) and catalytic patch are this deposit; the NAD⁺ pose is also this deposit and is replayed as its ① arrival — the pocket opens visually empty only as staging for the ordered-ter-bi narrative (the frame's own NAD⁺ is the deposited holo). "Ordered ter bi (NAD⁺ in first)" is the classical mechanism, not a primary claim this review re-verified. Residue numbering = human 1U8F (PDB-author 3–335 = UniProt P04406).Source: PDB 1U8F — human NAD⁺-holo 1.75 Å (Jenkins & Tanner 2006); the NAD⁺ pose is the deposited holo, replayed as its ① arrival — the empty pocket is staging only; ordered-ter-bi = classical, not re-verified (Frame card)
Michaelis D-G3P — 1NQA, superposed into the human frame 1NQA = the ternary Bs (Bacillus stearothermophilus) GAPDH·NAD·D-G3P Michaelis complex (Didierjean C et al., J Biol Chem 278:12968–12976, 2003, RCSB 1NQA · PMID:12569100 · DOI:10.1074/jbc.M211040200; 2.2 Å, r=0.182 / r_free=0.214, starting model 1NPT). On this page the 1NQA D-G3P was superposed into the 1U8F human frame by a whole-Cα Kabsch fit (330 Cα, RMSD ≈ 1.04), so the substrate sits in the human pocket; the deposited substrate is the free (noncovalent) aldehyde D-G3P whose C3-phosphate occupies the P_s site (the authors' own assignment), with the C1 carbonyl facing the Bs His176. Disclosure: 1NQA is a C149A variant in all four chains (NAD×4 + D-G3P×4 per tetramer; cf. human Cys152 ★) — the catalytic thiol is drawn back onto the human numbering for the chemistry.
② D-G3P docked — measured Michaelis distances D-G3P is shown at the superposed ternary pose; all Å numbers are measured from the assembled frame (cat block + ligand in one coordinate system): Cys152·SG ⋯ C1 = 2.15 Å — this is the human 1U8F SG against the superposed G3P C1, an assembled-frame value and NOT a seated-thiol distance of a single crystal: 1NQA itself is C149A (its own Ala···C1 window ≈ 4.15–4.41 Å), so this contact is not presented as attack geometry. His179·NE2 ⋯ O1 = 2.76 Å (human numbering; Bs His176···O1 spans 2.69–2.83 in the four deposited chains), Arg234·NH2 ⋯ O2P = 2.70 Å (Bs Arg231 2.70–2.89). NAD C4N ⋯ C1 = 4.71 Å — hydride donor–acceptor pre-geometry in the assembled Michaelis state, within the Bs window 4.50–4.84 (the hydride label is shown as the arrow launches).Source: measured from the assembled common frame (human 1U8F cat + 1NQA G3P, whole-Cα Kabsch, RMSD ≈ 1.04): Cys152·SG ⋯ C1 = 2.15 Å (assembled value — 1NQA is C149A, not attack geometry), His179·NE2 ⋯ O1 = 2.76 Å, Arg234·NH2 ⋯ O2P = 2.70 Å, NAD C4N ⋯ C1 = 4.71 Å (1NQA superposed from Didierjean et al. 2003; see Michaelis card)
③ Hydride transfer — NAD⁺ → NADH; A-side / pro-R stereospecificity [schematic hydride path] The oxidative first half-reaction transfers the A-side / pro-R hydride of the aldehyde carbon C1 to the C4N of the nicotinamide, giving β-NADH (the pro-R / β-NADH stereospecificity is the textbook assignment; the primary paper was not re-examined in the latest review, so it is carried as literature-based) and the thioester R–C(=O)–S–Cys152. This half-reaction is rate-determining. The blue H⁻ ball is a deliberately drawn cartoon (Rule 23) — real hydride transfer is coupled quantum proton/electron motion, not a discrete ball hop; the arrow is a trajectory marker, not curly-arrow electron pushing. NADH reuses the NAD⁺ skeleton with the nomenclature switched at C4N (no NADH complex is co-crystallized here — seam disclosed).Source: [schematic] — no hydride TS observed (blue H⁻ ball = Rule 23 cartoon; Proposal-vs-fact card) · pre-geometry C4N ⋯ C1 = 4.71 Å = measured assembled-frame value (② card) · A-side/pro-R → β-NADH + rate-determining first half-reaction = classical, literature-based (Kinetic-mechanism card)
④ Phosphorolysis — 1,3-BPG + NADH off-loading [schematic] Orthophosphate Pᵢ cleaves the C1–S thioester and the C1 acyl phosphate forms; 1,3-BPG is the G3P skeleton with a schematic acyl-phosphate add-on. Ordered release: 1,3-BPG leaves first, then NADH (ordered ter bi). Net: G3P + Pᵢ + NAD⁺ → 1,3-BPG + NADH + H⁺.Source: [schematic] — 1,3-BPG acyl phosphate & Pᵢ are schematic (no deposit captures 1,3-BPG; "Why is the 1,3-BPG acyl phosphate [schematic]?" card) · classical fresh-Pᵢ route drawn, Didierjean P_s-migration recorded as the alternative (Kinetic-mechanism card) · ordered release 1,3-BPG first / NADH second = classical ordered ter bi, not re-verified this review

How to tell a proposal from a fact on screen Deposited geometry (no marker) = the 1U8F human frame, the 1U8F NAD⁺ pose (shown arriving at ①), the superposed 1NQA G3P pose, and the measured Michaelis distances 2.15 / 2.76 / 2.70 / 4.71 Å (from the assembled data) plus the Cα-fit RMSD ≈ 1.04. [schematic] = the hydride ball, the blue arrow, the NAD→NADH re-annotation, the 1,3-BPG acyl phosphate and the Pᵢ sketch, and every inter-state cross-fade — oxo-transfer transition states are never observed, so this half is a literature-based proposal. The blue arrow marks the path traced by the cartoon ball — a trajectory marker, not a curly-arrow electron-pushing step; the real hydride transfer is quantum-level, not a discrete ball hop.
How to read this page — a first-pass primer First pass (ignore the noise): watch the order — Setup (visually empty pocket — staging) → ① NAD⁺ arrives → ② D-G3P docks → ③ a hydride ball travels from the aldehyde carbon C1 to the cofactor (NAD⁺ → NADH, thioester forms; this half is slow / rate-determining) → ④ a phosphate Pᵢ dives onto the thioester carbon, 1,3-BPG forms and leaves first, then NADH. Second pass (the numbers): assembled-frame SG⋯C1 = 2.15 Å (human 1U8F SG into the superposed G3P — not a seated-thiol ternary distance), His179·NE2 ⋯ O1 = 2.76 Å, Arg234·NH2 ⋯ O2P = 2.70 Å cradle, and the hydride pre-geometry C4N ⋯ C1 = 4.71 Å. Third pass: Open the "How to tell a proposal from a fact" and "Precision vs. accuracy" cards below to separate measured crystal geometry from the literature-based reconstruction.
Chemistry primer — the catalytic thiol and the proton bookkeeping Cys152 ★ is the nucleophile in action: its thiol pK_a is lowered (His179 H-bonding + the pocket's positive field; a specific pK figure was not re-verified this review) so at pH ≈ 7 it is largely a thiolate Cys152·S⁻. The aldehyde carbon C1 reacts with S⁻ through a tetrahedral thiohemiacetal R–CH(OH)–S–Cys152 — a short-lived intermediate this page does not draw (no crystal) — then the C1–H bond breaks and its two electrons travel with the pair as the A-side / pro-R hydride H⁻ to the NAD⁺ nicotinamide C4N, giving β-NADH (pro-R / β = textbook assignment, carried as literature-based) and the thioester R–C(=O)–S–Cys152. NAD⁺ is reduced by a hydride (2 e⁻ + 1 H⁺ equivalent) — this is an oxidation of the aldehyde, not a proton transfer. The net equation spends one H⁺ (G3P + Pᵢ + NAD⁺ → 1,3-BPG + NADH + H⁺), but that net proton does not describe any single elementary step.
Energy & redox bookkeeping — why this step matters in glycolysis Glycolysis invests 2 ATP (hexokinase, PFK); GAPDH starts the payoff phase. Each D-G3P gives (a) one NADH (≈ 2.5 ATP under oxidative respiration) and (b) a 1,3-BPG whose C1 acyl phosphate stores oxidation energy as an anhydride-like high-energy phosphate (a specific ΔG°′ figure was not re-verified this review). That acyl phosphate is then the substrate-level ATP: phosphoglycerate kinase (PGK, Step 7) transfers it to ADP → ATP. ×2 per glucose — this is where the substrate-level ATP of the payoff phase is earned. GAPDH is not allosterically controlled; it runs on substrate supply from TPI (mass action), NAD⁺ availability and product inhibition (see below).

Legend & Fidelity

Mechanism fidelity — measured vs. inferred The measured layer: the 1U8F human frame with its deposited NAD⁺ pose, the superposed 1NQA D-G3P, and the four Michaelis distances re-derived from that assembled frame (Rule 34 geomClaims). The inferred layer: the hydride cartoon, the coordinate re-annotations, the acyl-phosphate and Pᵢ constructions, and every cross-fade — itemized in the proposal-vs-fact card above. Two disclosures belong here: the frame is human while the substrate deposit is Bs, and that deposit is a C149A variant.
Why is the 1,3-BPG acyl phosphate "[schematic]"? Because it is too reactive to survive in a crystal — no deposit captures 1,3-BPG. The page builds it from the G3P skeleton by adding the C1 acyl phosphate, and shows the orthophosphate that forms it as a converge-and-consume step on the thioester carbon. The skeleton and its Michaelis pose are assembled crystallographic coordinates; the add-on and that step are reconstruction, tagged [schematic] in the narration.
Kinetic mechanism — ordered ter bi and the P_s debate The classical GAPDH kinetic mechanism is ordered ter bi: NAD⁺ binds first → D-G3P → Pᵢ; the first (oxidative) half-reaction (hydride transfer) is rate-determining, and products leave in the order 1,3-BPG (first) → NADH (second) — quoted as the classical mechanism; the ordered-binding/release-order/rate-determining conclusions were not re-derived from the primary steady-state papers in the latest review. Phosphate identity debate: Didierjean 2003 (1NQA) supports the P_s route — the substrate's own 3-phosphate migrates to become the C1 acyl phosphate while the catalytic pocket reorders on phosphorolysis; other proposals (some isotope/positional-isotope-exchange arguments, e.g. lines associated with Zhang & Li 2021 and the classic Cook discussion) support a fresh Pᵢ binding/substitution route. This page draws the classical fresh-Pᵢ route (a Pᵢ converges on the thioester C1 and is consumed into the acyl phosphate) and records both sides here; the Didierjean P_s-migration proposal is the documented alternative. The tetramer is represented by a single subunit pocket; the 3-phosphate sits in the P_s site (Arg234/Thr182/Ser151 cradle) in the deposit.
Precision vs. accuracy — an assembled common frame Distances are single values from the assembled common frame (1U8F human cat + 1NQA G3P transformed by the whole-Cα Kabsch fit), not measured dynamics. The scene joins two structures in one frame; the natural organism difference (human vs Bs) and the C149A variant are disclosed. Treat every Å as a plausible assembled crystallographic value, not a coordinate of the moving state.
Oligomer & dynamics — the tetramer, and why this page shows one pocket GAPDH works as a homotetramer; NAD⁺ binding is classically described as driving an apo (symmetric) ↔ holo (antisymmetric) R↔T-type quaternary transition (a textbook structural model, not re-verified this review), and the four active sites are coupled — classic GAPDH debates include negative cooperativity ("half-of-the-sites" activity in some analyses). This page composites one subunit pocket as a static mesh: no conformer sampling, thermal motion or domain movement is simulated. Animations interpolate between the crystallographic D-G3P-docked state (②) and the [schematic] states (③ hydride/NADH, ④ phosphorolysis/1,3-BPG); every reconstruction is literature-based. The tetramer / R↔T / cooperativity is disclosed here in text because a moving subunit model would require a full rebuild.
Physiological regulation — first-half-reaction control, not allostery GAPDH is a cytoplasmic housekeeping dehydrogenase (also moonlights at the surface as membrane-associated GAPDH). No allosteric activator is recognized here; control is substrate supply (mass action from TPI) + NAD⁺ availability + product inhibition: 1,3-BPG and NADH inhibit the enzyme (NADH competitive vs NAD⁺). G3P substrate inhibition is reported at high concentrations. Regulation also runs through cysteine chemistry on the catalytic Cys152: oxidative glutathionylation / S-nitrosylation inactivate the thiol, and S-nitrosylation at Cys152 triggers Siah1-mediated proteasomal degradation — a post-translational switch linked to glycolysis vs. signaling in oxidative cells. In glycolysis the coupled TPI↔GAPDH pair funnels G3P toward the first substrate-level ATP-maker (PGK, Step 7).
Kinetics scope This tool animates a single catalytic event; it does not model timecourses, steady-state kinetics, or K_m / k_cat / V_max behavior. Earlier pages quoted BRENDA-compiled reference values (specific activity ≈ 98 µmol·min⁻¹·mg⁻¹, derived k_cat ≈ 60 s⁻¹, K_m(G3P) ≈ 0.021 mM, K_m(NAD⁺) ≈ 0.018 mM, BRENDA r=695298). Those numbers are withdrawn pending verification: UniProt P04406 carries no K_m/k_cat for this reaction, and the BRENDA entries could not be re-confirmed in the latest review (2026-09-19).
Residue numbering — human PDB-author = UniProt; Bs offset On-screen numbers follow the human 1U8F PDB-author = UniProt P04406 numbering (3–335; active-site catalytic residues 151–152, 179, 182, 234, 313, 316). The 1NQA Bs numbers are offset by +3: Bs Cys149 ↔ human Cys152, Bs His176 ↔ human His179, Bs Arg231 ↔ human Arg234. These 1NQA Bs numbers are PDB-author: UniProt P00362 numbers the same residues Cys152 (ACT_SITE), His179 (SITE) and Arg233 (G3P BINDING) — never read the Bs PDB names as UniProt numbers.
Sources — bibliography of this page S1 — PDB 1U8F: human GAPDH NAD⁺-holo frame (chain P; the deposited NAD⁺ pose is replayed at ①); human PDB-author numbering = UniProt P04406 (3–335) S2 — PDB 1NQA (2.2 Å): B. stearothermophilus GAPDH·NAD·D-G3P Michaelis ternary, a C149A variant in all chains; Didierjean C et al., J Biol Chem 278:12968–12976, 2003, PMID:12569100 · DOI 10.1074/jbc.M211040200 — D-G3P superposed into the human frame by the whole-Cα Kabsch fit (as described in the drawerNote); measured assembled-frame distances Cys152·SG⋯C1 / His179·NE2⋯O1 / Arg234·NH2⋯O2P / NAD C4N⋯C1 (4 geomClaims in tools/manifests/gapdh.json) as shown on the ② card S3 — Kinetic mechanism (classical, not re-verified this review): ordered ter bi (NAD⁺ → D-G3P → Pᵢ; 1,3-BPG → NADH out), first half-reaction rate-determining, A-side / pro-R hydride → β-NADH; phosphate-identity debate — classical fresh-Pᵢ route drawn, Didierjean's P_s-migration recorded as the alternative S4 — Kinetics: the BRENDA compiled values quoted in the EvidenceView body (specific activity, k_cat, K_m(G3P), K_m(NAD⁺)) are withdrawn pending verification — UniProt P04406 carries no K_m/k_cat and the BRENDA entries could not be re-confirmed (review 2026-09-19) S5 — Sequence: UniProt P04406 (human GAPDH; numbering on screen) — Bs 1NQA numbers are offset +3 (Bs Cys149 ↔ human Cys152, His176 ↔ His179, Arg231 ↔ Arg234); Bs P00362 counterpart S6 — Repo display conventions: Rule 30 (hydride ball / acyl-P / Pᵢ = trajectory cartoons, TS unobserved) · Rule 34 geometry (geomClaims, own assembled-frame measurement from human 1U8F + superposed Bs 1NQA) · every Å = a single assembled crystallographic value (Precision vs. accuracy card) · [schematic] = coordinate re-annotation / cross-fade, disclosed seam; 1NQA C149A variant disclosed, catalytic thiol drawn back to human numbering

› Backbones and catalytic patch are taken directly from 1U8F (human, chain P); the NAD⁺ cofactor pose is also from the 1U8F holo deposit and is shown arriving at ① (the animation opens visually empty — staging on a frame that is itself an NAD⁺-holo crystal); the D-G3P pose comes from the 1NQA ternary (Bs) superposed into the 1U8F frame by a whole-Cα Kabsch fit (330 Cα, RMSD ≈ 1.04; human numbering used throughout). The hydride path, NAD→NADH re-annotation, 1,3-BPG acyl phosphate and Pᵢ sketch are schematic cartoons; the classical fresh-Pᵢ route is drawn and the phosphate-identity debate is recorded above (both sides). Distances shown on screen (2.15 / 2.76 / 2.70 / 4.71 Å) are measured from the assembled data.