Structural Insights into the Very Low Activity of the Homocoenzyme B₁₂ Adenosylmethylcobalamin in Coenzyme B₁₂‐Dependent Diol Dehydratase and Ethanolamine Ammonia‐Lyase

Abstract

The X-ray structures of coenzyme B 12 (AdoCbl)dependent eliminating isomerases complexed with adenosylmethylcobalamin (AdoMeCbl) have been determined. As judged from geometries, the CoÀ C bond in diol dehydratase (DD) is not activated even in the presence of substrate. In ethanolamine ammonia-lyase (EAL), the bond is elongated in the absence of substrate; in the presence of substrate, the complex likely exists in both pre-and posthomolysis states. The impacts of incorporating an extra CH 2 group are different in the two enzymes: the DD active site is flexible, and AdoMeCbl binding causes large conformational changes that make DD unable to adopt the catalytic state, whereas the EAL active site is rigid, and AdoMeCbl binding does not induce significant conformational changes. Such flexibility and rigidity of the active sites might reflect the tightness of adenine binding. The structures provide good insights into the basis of the very low activity of AdoMeCbl in these enzymes.

Structural Basis for the Activation of the Cobalt‐Carbon Bond and Control of the Adenosyl Radical in Coenzyme B12 Catalysis

“…Based on these structures, we have proposed a steric strain model of the coenzyme's Co−C bond activation and the importance of the positional stabilization of the adenosyl group for catalysis and for prevention of mechanism‐based inactivation. Intriguing difference between these enzymes in the flexibility of active sites upon AdoMeCbl binding has also been reported [11b] …”

“…Recently, the X‐ray structures of the complexes of GM with AdoMeCbl and AdoEtCbl [14b] and of DD and EAL with AdoMeCbl [11b] have been reported one after another. They provide us with important insights into the very low activity of AdoMeCbl in DD and EAL as well as the total inactivity in GM.…”

“…It is thus concluded that the Co−C6′ bond of AdoMeCbl is neither cleaved nor activated enough upon binding to DD (Figure 8A, B). [11b] This would be a primary reason why this homologue is essentially inactive as a coenzyme for DD. In addition, AdoMeCbl binding to DD induces global conformational changes of overall structure and large positional shifts of active‐site residues which makes the enzyme unable to adopt the catalytic conformation (Figure 8A, B).…”

“…The use of precise coenzyme analogues as probes for exploring the active sites has also provided a powerful method of studying AdoCbl‐dependent enzymes [13] . Especially useful are homocoenzymes B 12 (Figure 1F(1)), because they contain both adenine and ribose moieties [11b,14] . In adenosylmethylcobalamin (AdoMeCbl; n =2) [15] and adenosylethylcobalamin (AdoEtCbl; n =3), [15b,16] one and two CH 2 group(s), respectively, are inserted between the Co and C5′ atoms of AdoCbl.…”

“…We have previously reported the structures of the eliminating isomerases DD, GD, and EAL complexed with inactive coenzyme analogues, such as CN‐Cbl [6b,17b,18c,20] and AdePeCbl (Figure 2A), [17b,18c] and recently with AdoCbl itself (Figures 2B and 3A, B) [11a] or its homologue AdoMeCbl [11b] . Based on these structures, we have proposed a steric strain model of the coenzyme's Co−C bond activation and the importance of the positional stabilization of the adenosyl group for catalysis and for prevention of mechanism‐based inactivation.…”