CHEMICAL ROUTES TO COENZYME B12 AND ANALOGUES *

When it became known that adenine was present in the coenzyme form of vitamin BIZ, attempts were made in these and other laboratories to induce adenine to combine with reduced vitamin BIZ. These fruitless experiments were based on two misconceptions; first, that the adenine was bound to be linked directly to cobalt through one of its nitrogen atoms and, secondly, that the coenzyme contained divalent cobalt. A more rational approach, based on the 5’-p-toluene sulphonyl (tosyl) ester of adenosine or a protected derivative, became possible when the correct structure was elucidated by X-ray crystallography. It appears to have been conceived independently in three laboratories, Mention of the first indications of success at Glaxo Laboratories, Ltd., in October 1961 to A. W. Johnson brought the immediate response that he was engaged on the same problem, and we at once joined forces. It was not until June 1962 that we communicated some of our results to K. Bernhauer and learned that he also had been working along similar lines. Although the coenzyme probably does not contain divalent cobalt, reduction is still necessary to induce combination: moreover, as Malcolm Clark suggested to one of us (E. L. S.), the reactive species is not the brown intermediate reduction product known as vitamin BIZr but the fully-reduced grey-green product from vitamin BIZ. We have prepared this with chromous acetate in EDTA buffer at pH 10, as proposed by Beaven and Johnson,l with zinc dust and acetic acid, or most conveniently with soduim borohydride in water, methanol or ethanol. Because it reacts extremely rapidly with air to give vitamin and, more slowly, vitamin B,,, (vitamin BIZ1,; hydroxocobalamin), it was necessary to work in an inert atmosphere, purified nitrogen conveniently. To avoid contamination of the products with cyanocobalamin reformed on working up the reaction mixtures, we have usually employed hydroxocobalamin, which is easier to separate from the coenzyme or its analogues. The grey-green fully reduced vitamin Biz needs a trivial name, and much as we now dislike the proliferation of alphabetical subscripts to denote variants of vitamin B12, we shall adopt for convenience the term B12s used by Hill, Pratt, and Williams.2 In a preliminary note,3 we coined the name ‘hydrido-cobalamin’ but this is not entirely satisfactory; we shall provide evidence convincing US that the compound is indeed a cobalt hydride derivative, but the ‘cobalamin’ part implies trivalent cobalt, and the same arrangement of double bonds in the corrin ring as in the cobalamins-points that are being argued in this monograph. Preliminary notes have been p ~ b l i s h e d , ~ , ~ and