2-Oxoglutarate Dehydrogenase Complex - 3D Mechanism Simulator

Structure — human E1 chain (shown in 3D) Human 2-oxoglutarate dehydrogenase E1 (OGDH, homodimer; the displayed chain resi 123–1023) with its thiamin pyrophosphate (TPP) cofactor and Mg²⁺ — 8I0K (Wang J, Yu X, Zhong Y et al., Mol Cell 2025, PMID:39889707). Correction: 8I0K is an OGDH–TCAIM co-chaperone complex, not a TPP-holoenzyme paper — the holo TPP/Mg²⁺/Ca²⁺ assignment is cross-cited from 7WGR (Zhong et al., BBRC 2022, PMID:35189555). Residue numbers follow the 8I0K (PDB-author) numbering. Three pocket residues touch TPP in the deposited coordinates — His311, Arg312, His513 (TPP-contacting only; no catalytic role is assigned on this page).
Mechanism review — E1/E2/E3 catalysis The OGDH complex shares the PDH-complex architecture and mechanism (TPP decarboxylation → lipoyl transsuccinylation → FAD/NAD⁺ reoxidation) — Patel MS, Nemeria NS, Furey W & Jordan F, "The pyruvate dehydrogenase complexes: structure-based function and regulation", J Biol Chem 2014, PMID:24798336. E1o substrate-specificity pocket (three-His cluster) — Frank et al., J Mol Biol 2007, PMID:17367808 (E. coli E1o, PDB 2JGD — homology reference, not human residue assignments). Human E1 enamine/radical chemistry — Nemeria et al., J Biol Chem 2014, PMID:25210035. E3 (DLD) is the identical enzyme in both complexes.
E2 — DLST succinyltransferase core Human dihydrolipoyl succinyltransferase (DLST, EC 2.3.1.61, UniProt P36957) — the lipoyl arm is tethered at Lys110; the catalytic core (24-mer octahedral E2o) was resolved by cryo-EM — Nagy et al., BBA 2021, PMID:33684457 (PDB 6H05). Cited here for the channeling principle only; the E2 pocket on screen is drawn schematically.
E3 — DLD, the shared dehydrogenase (no E3BP here) Human dihydrolipoamide dehydrogenase (DLD, EC 1.8.1.4, UniProt P09622) — the genuine homodimer + FAD on screen is PDB 1ZY8 (2.59 Å, X-ray; Ciszak et al., J Biol Chem 2006, PMID:16263718). Fidelity note: 1ZY8 is an E3–E3BP (PDHc) structure — OGDHc has NO E3BP; only the DLD fold + FAD are carried over. Stacked nicotinamide–FAD geometry — Bräutigam et al., J Mol Biol 2005, PMID:15946682 (NADH-bound DLD; 1ZY8 itself holds no nucleotide).
Regulation — Ca²⁺/ADP on, NADH/succinyl-CoA/ATP off E1 is activated by matrix Ca²⁺ (low-µM, half-max ≈ 1 µM; site 1 D114ADLD motif) and ADP, and inhibited by NADH, succinyl-CoA and ATP — Armstrong et al., Biochem J 2014, PMID:24495017; McCormack & Denton 1979/1981; Denton reviews. Ca-insensitive splice variants exist (tissue-specific). Direction matters: pages stating the reverse are wrong — Reactome R-HSA-71401 prose is inverted and is NOT cited here.
Sequence Human OGDH (E1) UniProt Q02218 (1023 aa) · DLST (E2) UniProt P36957 (lipoyl-Lys K110) · DLD (E3) UniProt P09622 · NCBI Gene 4967 (OGDH)
OGDHc vs PDHc — what differs Same E3 (DLD), different everything else: mammalian PDHc cores are 60-mers (48× E2 + 12× E3BP) while OGDHc has NO E3BP and a 24-mer octahedral E2o core; PDH E1 is an α₂β₂ heterotetramer while OGDH E1 is a homodimer; PDH E2 carries two lipoyl domains, OGDH E2 carries one (Lys110). E1/E2 are complex-specific; E3 is shared (also with BCKDHc). — Nagy 2021 (PMID:33684457); Patel 2014 (PMID:24798336).
E1 Setup — E1 + TPP·Mg²⁺ The genuine TPP and structural Mg²⁺ sit in the deposited E1 pocket (MG–TPP O3B 1.97 Å); His311/Arg312/His513 touch TPP at deposited distances — positions, not catalytic assignments.Source: E1 chain + TPP + Mg²⁺ — PDB 8I0K (Wang 2025, PMID:39889707, measured; holo assignment cross-cited from 7WGR); TPP-contact distances re-measured from the data (manifest geomClaims)
E1 ① — α-Ketoglutarate arrives α-KG docks beside the TPP thiazolium C2. The TPP scaffold and contacting residues are the real 8I0K coordinates; the α-KG pose is [schematic] — no E1·α-KG Michaelis complex is deposited.Source: TPP scaffold & contacting residues — PDB 8I0K (measured); α-KG pose [schematic] (Fidelity card, no E1·α-KG co-crystal)
E1 ② — Ylide attack → decarboxylation The TPP C2-ylide attacks the α-KG keto carbon; the carboxylate leaves at once as CO₂ — the second decarboxylation of the cycle, making the direction effectively irreversible (a specific rate-limiting assignment is not re-verified this review) — leaving succinyl-TPP, cross-faded in place with CO₂ leaving beside it. The ylide-formation/proton path is (hypothetical, lit.-based); the arrow and traveling marker trace the drawn trajectory.Source: E1 TPP mechanism (ylide attack → decarboxylation → succinyl-TPP) — Patel et al. 2014 (PMID:24798336); ylide/proton path hypothetical, lit.-based — this card; CO₂ departure = drawn schematic
E1 ③ — Lipoamide arm swings in The oxidized lipoamide disulfide (E2's Lys110-tethered swinging arm) rotates to the E1 site to accept the four-carbon unit. [schematic].Source: lipoamide = Lys110-tethered "swinging arm" (E2 card, Nagy 2021, PMID:33684457); arm rotation to E1 — Patel et al. 2014 (PMID:24798336); pose & rotation [schematic] (Fidelity card, mesh static)
E1 ④ — Reductive succinylation → SLIPO The four-carbon unit transfers onto the disulfide arm — S–S reduced, S-succinyldihydrolipoamide cross-fading in as the marker flies — with TPP regenerated behind it. The succinyl is drawn on the S8 dithiolane sulfur (lipoate numbering — the S8-acyl-lipoamide naming convention; [schematic], not a deposited atom). His311★ marks a TPP-contacting residue (deposited 3.47 Å off the thiazolium sulfur — a position, not a catalytic assignment).Source: reductive succinylation → S-succinyldihydrolipoamide, TPP regenerated — Patel et al. 2014 (PMID:24798336); His311 = TPP-contacting position, measured 3.47 Å (8I0K, manifest geomClaim)
E1 ⑤ — Succinylated arm departs S-succinyldihydrolipoamide swings away toward the E2 site — a tethered swing, not free diffusion.Source: tethered arm is covalently bound to E2 Lys110 ε-amino (E2 card); swing [schematic] — Fidelity card (mesh static, no co-crystal); direction toward E2 — Patel et al. 2014 (PMID:24798336)
E2 Setup — E2 succinyltransferase site The E2 pocket is drawn schematically on the E1 frame — no human E2 pocket is co-crystallized in this frame.Source: schematic pocket (Fidelity card); DLST core reference — Nagy 2021 (PMID:33684457, PDB 6H05, cited for channeling only)
E2 ① — Succinylated arm arrives The succinylated arm reaches the E2 pocket — a [schematic] pocket drawn on the E1 frame.Source: E2 pocket drawn schematically on the E1 frame (Fidelity card — no human E2 pocket co-crystallized); channeling principle only from PDB 6H05 (Nagy 2021, PMID:33684457)
E2 ② — CoA-SH arrives CoA-SH docks with its thiol poised at the arm's thioester. [schematic].Source: transsuccinylation step (CoA-SH ↔ succinylated arm thioester) — Patel et al. 2014 (PMID:24798336); CoA-SH pose [schematic] (Fidelity card, no E2·CoA co-crystal)
E2 ③ — Transsuccinylation → succinyl-CoA CoA sulfur attacks the thioester: succinyl-CoA (the substrate-level-phosphorylation fuel for SCS) forms in place while the arm is left reduced as dihydrolipoamide. Magenta marker and blue arrow trace CoA-S → thioester-C.Source: transsuccinylation → succinyl-CoA + dihydrolipoamide — Patel et al. 2014 (PMID:24798336); succinyl-CoA → SCS — Overall reaction card; chemistry drawn schematically
E2 ④ — Products depart Succinyl-CoA leaves for substrate-level phosphorylation at SCS; the dihydrolipoamide arm swings on to the E3 site.Source: succinyl-CoA → SCS, dihydrolipoamide → E3 — Overall reaction card / Patel et al. 2014 (PMID:24798336); departures are display poses, no measured exit geometry
E3 Setup — E3 DLD homodimer + FAD (real) The genuine DLD homodimer + FAD (1ZY8) — the same E3 as the PDH complex. 1ZY8 is an E3–E3BP structure; OGDHc has no E3BP.Source: genuine FAD + E3 DLD homodimer — real crystal coordinates, PDB 1ZY8 (2.59 Å, X-ray; E3 card)
E3 ① — Reduced arm arrives Dihydrolipoamide reaches the E3 site beside the genuine FAD (1ZY8). The arm pose is [schematic].Source: genuine FAD + E3 DLD homodimer — real crystal coordinates, PDB 1ZY8 (2.59 Å, X-ray; E3 card, no link in whitelist); reduced-arm pose [schematic] (Fidelity card, no E3·arm co-crystal)
E3 ② — NAD⁺ arrives NAD⁺ docks as the electron acceptor. No nucleotide is co-crystallized in 1ZY8 — [schematic].Source: PDB 1ZY8 (2.59 Å, X-ray, E3 card) — cited in Fidelity card: "No nucleotide is co-crystallized with 1ZY8, so the NAD⁺/NADH poses are schematic"; NAD⁺ acceptor role — Patel et al. 2014 (PMID:24798336)
E3 ③ — Reoxidation → lipoamide + NADH The lipoyl disulfide re-forms; electrons pass from the arm into the FAD isoalloxazine N5 (arrow tip), then on to NAD⁺ — the enzyme disulfide between arm and flavin is text-only (no Cys sidechains in frame). Magenta marker flies both relay legs (arm → N5, FAD → NAD⁺). A stacked nicotinamide–FAD geometry of this kind is documented in the NADH-bound DLD structure (Bräutigam et al.). The released H⁺ is the gold ball — H⁺ is always gold on this page.Source: electron relay arm → FAD N5 → NAD⁺ — Patel et al. 2014 (PMID:24798336); real FAD — PDB 1ZY8 (2.59 Å, E3 card); stacked nicotinamide–FAD geometry — Bräutigam et al. 2005 (PMID:15946682); H⁺ gold-ball convention — repo color rule
E3 ④ — Lipoamide + NADH depart Lipoamide swings back to E1 for the next succinyl group; NADH feeds the electron transport chain.Source: lipoamide recycle to E1, NADH → ETC — Overall reaction card / Patel et al. 2014 (PMID:24798336); departures are display poses, no measured exit geometry
Overall reaction α-KG + CoA + NAD⁺ → Succinyl-CoA + CO₂ + NADH + H⁺ (strongly exergonic — effectively irreversible; no quantitative ΔG°′ is stated on this page). The lipoyl group is covalently bound to the conserved Lys110 ε-amino group on the E2 lipoyl domain (the "swinging arm"). Succinyl-CoA feeds substrate-level phosphorylation at SCS next.

Legend & Fidelity

Mechanism fidelity — measured vs. inferred The measured layer: the E1 chain + TPP + Mg²⁺ + three TPP-contacting residues (8I0K deposited coordinates; residue numbers are 8I0K author numbers) and the E3 DLD homodimer + FAD (1ZY8) are deposited coordinates. The inferred layer: the E2 pocket, every substrate and intermediate pose (α-KG, succinyl-TPP, the three lipoyl-arm states, CoA, succinyl-CoA, NAD⁺/NADH/H⁺), the ylide/proton path, and the arm swing path are literature-based reconstructions, labeled [schematic] where shown.
Why are the pocket residues "contacting" and not "catalytic"? The deposited distances are real (Arg312 NE–TPP 3.01 Å, His513 NE2–TPP 3.15 Å, His311 NE2–TPP 3.47 Å, MG–TPP 1.97 Å — manifest geomClaims), but no primary paper assigns these three residues a catalytic role in human OGDH E1. An earlier page version called them a catalytic pocket — that wording is withdrawn. Substrate-specificity guidance comes from the E. coli E1o structure (Frank 2007, three-His cluster — homology reference only).
Why are the arm and the E2 pocket "[schematic]"? Three gaps force it. (1) No human E2 catalytic pocket is co-crystallized in this frame — the E2 tab draws a schematic pocket on the E1 frame, and 6H05 (human E2o core) is cited for the channeling principle only. (2) The lipoyl arm is a protein domain tethered to the E2 core, but the mesh is static — its large rotation is summarized by the moving S–S group, not a true domain swing. (3) No nucleotide is co-crystallized with 1ZY8, so the NAD⁺/NADH poses are schematic even though the FAD beside them is real.
Physiological regulation Activated by matrix Ca²⁺ (low-µM, half-max ≈ 1 µM — site 1 D114ADLD) and ADP; inhibited by NADH, succinyl-CoA and ATP. Ca-insensitive splice variants exist (tissue-specific). OGDHc is a key TCA flux-control point alongside IDH3 — see Armstrong et al. 2014 (PMID:24495017); McCormack & Denton 1979/1981.
Kinetics scope This tool animates a single catalytic event; it does not model timecourses, steady-state kinetics, or Km / kcat / Vmax behavior. The compiled human Km(α-KG) order (0.1–0.2 mM) shown in earlier versions is withdrawn pending re-verification (2026-09-23); what remains is qualitative and sourced — Ca²⁺ lowers Km (McCormack & Denton 1979/1981) — and no kcat is stated, since purified intact-complex kinetics vary. For enzyme parameters see BRENDA (EC 1.2.4.2 · 2.3.1.61 · 1.8.1.4) — this page's outputs are not kinetic data.
Sources — bibliography of this page S1 — PDB 8I0K: human OGDH E1 chain + TPP / Mg²⁺ (deposited coordinates; an OGDH–TCAIM co-chaperone complex — the holo TPP/Mg²⁺ assignment is cross-cited from PDB 7WGR, Zhong et al. 2022, PMID:35189555); residue numbers are 8I0K author numbers S2 — PDB 1ZY8: E3 DLD homodimer with genuine FAD (2.59 Å, X-ray) — deposited coordinates; an E3–E3BP (PDHc) structure so OGDHc carries NO E3BP; no nucleotide is co-crystallized, so the NAD⁺/NADH poses are [schematic]; stacked nicotinamide–FAD geometry documented in the NADH-bound DLD structure (Bräutigam et al. 2005, PMID:15946682) S3 — E2 channeling principle only: PDB 6H05 (human E2o core), Nagy et al., 2021, PMID:33684457 — the human E2 catalytic pocket is not co-crystallized in this frame, so the E2 tab draws a schematic pocket on the E1 frame; lipoyl-Lys K110 (UniProt P36957) S4 — Mechanism review: Patel MS et al. 2014, PMID:24798336 — TPP decarboxylation → lipoyl transsuccinylation → FAD/NAD⁺ reoxidation; E. coli E1o pocket — Frank et al. 2007, PMID:17367808 (homology reference); human E1 enamine/radical — Nemeria et al. 2014, PMID:25210035 S5 — Regulation: Ca²⁺ site 1 D114ADLD — Armstrong et al. 2014, PMID:24495017; Ca/ADP activation vs NADH/succinyl-CoA/ATP inhibition — McCormack & Denton 1979/1981; Reactome R-HSA-71401 prose is inverted and is not cited S6 — Kinetics scope: the compiled human Km(α-KG) order (0.1–0.2 mM) is withdrawn pending re-verification (2026-09-23); the retained statement is qualitative and sourced — Ca²⁺ lowers Km (McCormack & Denton 1979/1981); kcat not stated. BRENDA EC 1.2.4.2 · 2.3.1.61 · 1.8.1.4 S7 — Repo display conventions: measured layer = E1 chain + TPP + Mg²⁺ + contacting residues (8I0K) and the E3 DLD homodimer + FAD (1ZY8); every substrate/intermediate pose (α-KG, succinyl-TPP, the three lipoyl-arm states, CoA, succinyl-CoA, NAD⁺/NADH/H⁺), the E2 pocket, the ylide/proton path and the arm swing path are [schematic] literature-based reconstructions (tools/manifests/kgdh.json: 4 geomClaims) · departures are display poses, no measured exit geometry · OGDHc vs PDHc differences (no E3BP, E2o 24-mer, E1 homodimer)

› The 3D protein is the real E1 chain (PDB 8I0K) with its genuine TPP / Mg²⁺, and the real E3 DLD homodimer with its genuine FAD (PDB 1ZY8 — an E3–E3BP structure, but OGDHc carries no E3BP). Substrates, products and the E2 pocket are drawn as compact schematic molecules posed at the active-site pockets and are labeled [schematic].