Search

A mutation analysis of the catalytic functions of active-site residues of coenzyme B 12 -dependent diol dehydratase in the conversion of 1,2-propanediol to 1,1-propanediol has been carried out by using QM/MM computations. Mutants Hisl43Ala, Glul70Gln, Glul70Ala, and Glul70Ala/Glu221Ala were considered to estimate the impact of the mutations of His143 and Glu170. In the Hisl43Ala mutant the activation energy for OH migration increased to 16.4 from 11.5 kcal mol -1 in the wild-type enzyme. The highest activation energy, 19.6 kcal mol -1 , was measured for hydrogen back-abstraction in this reaction. The transition state for OH migration is not sufficiently stabilized by the hydrogen-bonding interaction formed between the spectator OH group and Gln170 in the Glu170Gln mutant, which demonstrates that a strong proton acceptor is required to promote OH migration. In the Glul70Ala mutant, a new strong hydrogen bond is formed between the spectator OH group and Glu221. A computed activation energy of 13.6 kcal mol -1 for OH migration in the Glul70Ala mutant is only 2.1 kcal mol -1 higher than the corresponding barrier in the wild-type enzyme. Despite the low activation barrier, the Glul70Ala mutant is inactive because the subsequent hydrogen back-abstraction is energetically demanding in this mutant. OH migration is not feasible in the Glul70Ala/ Glu221Ala mutant because the activation barrier for OH migration is greatly increased by the loss of COO - groups near the spectator OH group. This result indicates that the effect of partial deprotonation of the spectator OH group is the most important factor in reducing the activation barrier for OH migration in the conversion of 1,2-propanediol to 1,1-propanediol catalyzed by diol dehydratase.

Coenzyme B 1 2 initiates radical chemistry in two types of enzymatic reactions, the irreversible eliminases (e.g., diol dehydratases) and the reversible mutases (e.g., methylmalonyl-CoA mutase). Whereas eliminases that use radical generators other than coenzyme B 1 2 are known, no alternative coenzyme B 1 2 independent mutases have been detected for substrates in which a methyl group is reversibly converted to a methylene radical. We predict that such mutases do not exist. However, coenzyme B 1 2 independent pathways have been detected that circumvent the need for glutamate, β-lysine or methylmalonyl-CoA mutases by proceeding via different intermediates. In humans the methylcitrate cycle, which is ostensibly an alternative to the coenzyme B 1 2 dependent methylmalonyl-CoA pathway for propionate oxidation, is not used because it would interfere with the Krebs cycle and thereby compromise the high-energy requirement of the nervous system. In the diol dehydratases the 5'-deoxyadenosyl radical generated by homolysis of the carbon-cobalt bond of coenzyme B 1 2 moves about 10 A away from the cobalt atom in cob(II)alamin. The substrate and product radicals are generated at a similar distance from cob(II)alamin, which acts solely as spectator of the catalysis. In glutamate and methylmalonyl-CoA mutases the 5'-deoxyadenosyl radical remains within 3-4 A of the cobalt atom, with the substrate and product radicals approximately 3 A further away. It is suggested that cob(ii)alamin acts as a conductor by stabilising both the 5'-deoxyadenosyl radical and the product-related methylene radicals.

Efficient electrochemical syntheses of "homocoenzyme B(12)" (2, Co(beta)-(5'-deoxy-5'-adenosyl-methyl)-cob(III)alamin) and "bishomocoenzyme B(12)" (3, Co(beta)-[2-(5'-deoxy-5'-adenosyl)-ethyl]-cob(III)alamin) are reported here. These syntheses have provided crystalline samples of 2 and 3 in 94 and 77 % yield, respectively. In addition, in-depth investigations of the structures of 2 and 3 in solution were carried out and a high-resolution crystal structure of 2 was obtained. The two homologues of coenzyme B(12) (2 and 3) are suggested to function as covalent structural mimics of the hypothetical enzyme-bound "activated" (that is, "stretched" or even homolytically cleaved) states of the B(12) cofactor. From crude molecular models, the crucial distances from the corrin-bound cobalt center to the C5' atom of the (homo)adenosine moieties in 2 and 3 were estimated to be about 3.0 and 4.4 A, respectively. These values are roughly the same as those found in the two "activated" forms of coenzyme B(12) in the crystal structure of glutamate mutase. Indeed, in the crystal structure of 2, the cobalt center was observed to be at a distance of 2.99 A from the C5' atom of the homoadenosine moiety and the latter was found to be present in the unusual syn conformation. In solution, the organometallic moieties of 2 and 3 were shown to be rather flexible and to be considerably more dynamic than the equivalent group in coenzyme B(12). The homoadenosine moiety of 2 was indicated to occur in both the syn and the anti conformations.

Hydrogen abstraction from 2-aminoethanol by the 5'-deoxyadenosyl radical, which is formed upon Co-C bond homolysis in coenzyme B 1 2 , was investigated by theoretical means with employment of the DFT (B3LYP) and ab initio (MP2) approaches. As a model system for the 5'-deoxyadenosyl moiety the computationally less demanding 1,5-dideoxyribose was employed; two conformers, which differ in ring conformation (C2- and C3-endo), were considered. If hydrogen is abstracted from "free" substrate by the C2-endo conformer of the 1,5-dideoxyribose-5-yl radical, the activation enthalpy is 16.7 kcalmol - 1 ; with the C3-endo counterpart, the value is 17.3 kcal - mol - 1 . These energetic requirements are slightly above the activation enthalpy limit (15 kcalmol - 1 ) determined experimentally for the rate-determining step of the sequence, that is, hydrogen delivery from 5'-deoxyadenosine to the product radical. The activation enthalpy is lower when the substrate interacts with at least one amino acid from the active site. According to the computations, when a His model system partially protonates the substrate the activation enthalpy is 4.5 kcal mol - 1 for the C3-endo conformer and 5.8 kcal mol - 1 for the C2-endo counterpart. As hydrogen abstraction from the fully as well as the partially protonated substrate is preceded by the formation of quite stable encounter complexes, the actual activation barriers are around 13-15 kcal mol - 1 . A synergistic interaction of 2-aminoethanol with two amino acids where His partially protonates the NH 2 group and Asp partially deprotonates the OH group of the substrate results in an activation enthalpy of 12.4 kcal mol - 1 for the C3-endo conformer and 13.2 kcal mol - 1 for the C2-endo counterpart. However, if encounter complexes exist in the active site, the actual activation barriers are much higher (>25 kcal mol - 1 ) than that reported for the rate-determining step. These findings together with previous computations suggest that the energetics of the initial hydrogen abstraction decrease with an interaction of the substrate with only a protonating auxiliary, but for the rearrangement of the radical the synergistic effects of two auxiliaries are essential to pull the barrier below the limit of 15 kcal mol - 1 .

We describe a novel enantioselective synthesis of aristeromycin, the carbocyclic analogue of adenosine. The seven-step synthesis is also suitable for the preparation of specifically-labelled [6'-(13)C]aristeromycin. Both the unlabelled and (13)C-labelled product was coupled to vitamin B(12) to form aristeromycylcobalamin. This carbocyclic analogue of coenzyme B(12) was examined for its coenzymic activity with several adenosylcobalamin-dependent enzymes. For glyceroldehydratase and dioldehydratase, the reaction rate (k(cat)) was 38 and 44 % of that measured with adenosylcobalamin as coenzyme. In contrast, aristeromycylcobalamin showed no detectable activity with methylmalonyl-CoA mutase and ethanolamine ammonia-lyase. Instead, it was a weak inhibitor of the former and a strong inhibitor of the latter enzyme. The slower turnover rate with glyceroldehydratase raised the hope of detecting the 6'-deoxyaristeromycyl radical intermediate. Comparison of the EPR spectra of the intermediates in the glyceroldehydratase reaction, which used adenosyl- and aristeromycylcobalamines, respectively, as coenzyme, revealed a significant shift and this suggests a different geometric position of these cofactors at the binding site during the cleavage of the carbon-cobalt bond. However, we found no evidence for the existence of a 6'-deoxyaristeromycyl radical during the reaction with [6'-(13)C]aristeromycylcobalamin. We conclude that the lifetime of this radical is still too short to be observed.