Alpha-lipoic acid is an 8-carbon fatty acid with a disulfide ring at its end; the body makes it itself, and unusually it dissolves in both water and fat.Alpha-lipoic acid () is an 8-carbon fatty acid with a disulfide (-S-S-) 5-membered ring at the end. First isolated by Reed's lab in 1951.
Chemically unique:
· 8-carbon fatty chain → lipid-soluble · Disulfide + terminal carboxyl → water-soluble · Both water- and lipid-soluble — most antioxidants only work in the aqueous phase or the lipid phase (vitamin C aqueous only / vitamin E lipid only / GSH aqueous only). ALA is rare in entering both. · In the body it can enter cytosol + cell membranes + inner mitochondrial membrane + nucleus.
Two stereoisomers:
· R-ALA (right-handed, natural): the form used as cofactor in mitochondrial enzymes · S-ALA (left-handed, synthetic byproduct): low biological activity · Industrial racemic = 50:50 mixture
Endogenous synthesis:
· Synthesized endogenously in mitochondria from octanoic acid + sulfur · Not an essential nutrient — unlike B vitamins where deficiency causes disease.
Next step: what it actually does in mitochondria — that is its real day job.
2 · Lipoyl swing arm
Inside mitochondria, is bonded to a lysine on certain enzymes and becomes a swing arm that passes intermediates between three subunits. inside mitochondria is not a free-floating small molecule — it is covalently attached to a specific lysine (Lys) residue of certain enzymes, forming a special cofactor called lipoyl-Lys:
Reaction:
ALA + Lys-NH₂ → amide bond (-CO-NH-) → lipoyl-Lys
The covalent chain plus the Lys side chain together form a ~ 1.4 nm long 'swing arm' — it can physically swing between the three subunits (E1 / E2 / E3) of the enzyme complex, shuttling intermediates back and forth.
· Without the swing arm → intermediates cannot be passed between the active sites of E1, E2, E3 · The whole reaction chain breaks · This is the same mechanism as the biotin (B7) carboxylase swing arm (the 'BC↔CT swing arm' of biotin's carboxylases).
Key facts:
· ALA's 'day job' is as a permanently bound cofactor, not a free molecule · Once ALA is enzymatically synthesized by the body, it is fixed long-term on that enzyme, not 'recycled' for antioxidant duties · The ALA you take orally and the ALA already on enzymes are two separate pools.
3 · Three complexes
The lipoyl-Lys swing arm is a core cofactor shared by three giant enzyme complexes, PDH, KGDH and BCKD, and one PDH reaction needs 5 cofactors at once.The lipoyl-Lys swing arm is a core cofactor in at least 3 giant enzyme complexes:
① PDH (pyruvate dehydrogenase) — exit of glycolysis · pyruvate → acetyl-CoA + CO₂ + · Entry into the cycle · Complex ~ 9 MDa, 60-90 protein subunits
② KGDH (α-ketoglutarate dehydrogenase) — midpoint of the TCA cycle · α-KG → succinyl-CoA + CO₂ + NADH · One of the rate-limiting steps of the TCA
③ BCKD (branched-chain α-keto acid dehydrogenase) — metabolism · Handles leucine / isoleucine / valine (BCAA) · BCKD mutation → maple syrup urine disease (MSUD) — rare but lethal
The three enzymes share the same architecture (E1 / E2 / E3 subunits) + the same set of cofactors:
· E1 ← TPP (vitamin B1) — the active form of thiamin · E2 ← lipoyl () — the star of this drawing · E3 ← FAD (vitamin B2) + NAD⁺ (vitamin B3) — the active forms of riboflavin and niacin · substrate side CoA (B5) — built from pantothenic acid
So one PDH reaction requires 5 cofactors present simultaneously — any one missing slows the whole line down.
This is the silent convergence of 5 B-vitamin stories on the map: thiamin / riboflavin / niacin / pantothenic acid / ALA all gather on this single enzyme PDH.
4 · Energy supplement debunked
Healthy people make their own and it is already bound to the enzymes, so extra ALA adds no enzyme; its truly effective use is diabetic neuropathy.The most common marketing narrative: 'ALA is a mitochondrial cofactor → supplementing ALA boosts energy metabolism → anti-fatigue + fat loss + anti-aging.'
Why this reasoning does not hold in healthy people:
① Endogenous synthesis
· In healthy people, ALA in mitochondria is synthesized endogenously, already bound to PDH/KGDH/BCKD · Adding more ALA does not increase the amount of these enzymes — enzyme quantity is regulated by protein synthesis, not by cofactor supply · This is a common flaw in cofactor marketing: the mistaken belief that 'supplementing cofactor → more enzyme.'
② Five cofactors in parallel
· The PDH reaction requires TPP + lipoyl + FAD + + CoA all five present · Supplementing only ALA while B1/B2/B3 are deficient → reaction still stalls · In a modern healthy diet, B1/B2/B3/B5 + endogenous ALA are usually all adequate → adding ALA has no rate-limiting effect.
③ Antioxidant is a side job
· DHLA (dihydrolipoic acid) = the reduced form of ALA, which does have antioxidant activity + metal chelation · But DHLA half-life ~ 30 minutes — acute effect, not sustained protection · The 'antioxidant network hub' (Packer 1995) is real at the mechanism level, but fails at clinical endpoint — the same story as vitamin E in cell membranes.
④ Real evidence
· Backed by randomized trials, but the endpoint is a 3-week symptom score: diabetic peripheral neuropathy, 600 mg/day by infusion for 3 weeks. In ALADIN I (Ziegler 1995), the share whose symptom score improved by at least 30% was 82.5% on the drug and 57.6% on placebo, about 1 extra responder for every 4 treated ( ≈ 4); the Ziegler 2004 pooled 4 trials of the same design, 1,258 people, and the foot symptom score (TSS) improved 24.1% more than on placebo, with 52.7% vs 36.9% seeing symptoms at least halved — a prescription drug in Germany for more than 50 years · Meta-analysis of randomized trials, but a tiny effect: weight — across the trials Kucukgoncu 2017 pooled (8 weeks to 1 year), people taking it lost on average only about 1.27 kg more than on placebo — too small to be clinically meaningful · No trials behind it: systemic anti-aging / longevity / primary prevention — zero hard-endpoint RCT evidence.
So ALA's truly effective use case is diabetic neuropathy — not 'wellness.' Supplementing it in healthy people is like adding water to a cofactor pool that is already adequate — neither rate-limiting nor stored.