Why Vitamin B12 Contains Cobalt

Vitamin B12 is a family of compounds, the cobalamins, named for the metal they share: each has a single cobalt ion at the centre of a ring system called a corrin. Cyanocobalamin, the member usually meant by "vitamin B12", has the formula C63H88CoN14O14P and a molecular weight of 1355.4 g/mol — one cobalt atom among 181. Cobalt still makes up around 4% of the mass: batches of cyanocobalamin assessed by EFSA contained 38,300 to 45,300 mg of cobalt per kilogram. The structure was solved by X-ray crystallography in Dorothy Hodgkin’s laboratory in the mid-1950s, and Hodgkin received the 1964 Nobel Prize in Chemistry for structure determinations of this kind.

The corrin is a tetrapyrrole, related to the porphyrin of haem and, like other tetrapyrroles, built from 5-aminolaevulinic acid. It differs in one telling detail: two of its four pyrrole units, rings A and D, are joined directly, with no bridging carbon between them. The ring is contracted during biosynthesis, and the cobalt goes in at one of two points — early, before contraction, in the anaerobic pathway, or late, once the ring is fully contracted, in the aerobic one. Either way the metal is installed while the molecule is being built, not attached to a finished product.

The cobalt holds six ligands. Four are the nitrogen atoms of the corrin. Below the ring sits 5,6-dimethylbenzimidazole, a base tethered back to one of the ring’s side chains through a phosphate-containing loop. Above the ring is the one variable position, and what occupies it is the only difference between the named forms. In methyl- and adenosylcobalamin it is an alkyl group bonded directly to the cobalt, which makes them organometallic compounds; the cobalt–carbon bond in adenosylcobalamin was the first of its type described in a biological molecule.

The corrin binds the cobalt through four nitrogens and leaves the positions above and below the ring free for other ligands. The molecule’s functions depend on the cobalt’s redox state — in cyanocobalamin it is formally cobalt(III) — and the cofactor forms put that chemistry to work through their cobalt–carbon bonds: methylcobalamin serves methyltransferases such as methionine synthase, and adenosylcobalamin serves mutases such as methylmalonyl-CoA mutase. The same bond is the weak point. All four forms are light-sensitive in solution, and the two alkyl forms lose their upper group within seconds under UVA.

Cobalamin is made in nature only by some bacteria and archaea, through a pathway of more than thirty steps. Chemistry does not compete: the total synthesis reported by Woodward and colleagues in 1973 took more than sixty steps and yielded less than 1%. Industrial production is fermentation, chiefly with Propionibacterium freudenreichii and Pseudomonas denitrificans, and reported output rose from around 3 tonnes a year in 1989 to about 10 tonnes by 2005.

• Fermentation: supplying cobalt and 5,6-dimethylbenzimidazole is reported to raise production; without enough of the base, cells accumulate incomplete forms
• Extraction: the cobalamins are released by heating to 80–120 °C at pH 6.5–8.5 for 10–30 minutes
• Cyanidation: every form present is converted to cyanocobalamin with potassium cyanide or thiocyanate, usually with sodium nitrite and heat
• Purification: filtration and adsorption, to a specification that limits named relatives such as 8-epi-cyanocobalamin and carboxylated cyanocobalamins