The thiolate-alkylating protein (Ada protein) is a zinc protein that repairs the defective DNA by transferring methyl group on to itself. We have used the thiophenylphosphine ligand to provide sulfur-rich environment for model study. In this work, we have accidentally found that this zinc complex can activate CH2Cl2 to generate a methylenated complex (4). Carbon-halide activation is commonly used in organic synthesis. However activation of C-Cl bond is comparatively scarce compared to C-Br or C-I bonds. Varying the degree of deprotonation on the PS3 ligand, a simple zinc dimer (2) can be obtained instead of (4). The dimer (2) can even react with much milder alkyl-reagent, methylphosphotriester. Therefore (2) serves as a successful Ada protein model in this sense. Other related alkyl-transfer reactions using different ligand or zinc complexes were discussed to give insight of the methyl-transfer action of Ada protein.

In our previous studies on modeling methyl transfer protein, Ada protein, we found that zinc complexes of tri(2-thiophenyl)phosphine (PS3), (A), have similar methyl transfer behavior as its biological counter part. In order to probe the role of zinc in methylation process, we used tin and cadmium in model study to compare their chemistry relative to that of zinc. We found that all three metal complexes have similar chemistry and assume similar dimeric anion structure. For example, [Cd(PS3)]22-(1) and [Cd(SiPS3)]22-(2) have been successfully characterized crystallographically to possess the same structure as zinc dimer (A). However , in attempt to crystallize the tin analogue, the accidental oxidation product [SnIV(OH)2(SiPS3)]2, (5), was obtained. Its crystal structure gave clue to the mechanism of oxidation of the original tin dimer. The reactions with alkylating reagent of (1) have been compared with those of zinc dimer (A), and we found that the metal (Zn or Cd) center causes the dimers to produce different degree of methylation products toward different alkylating reagents. For the reactions with CH3I, the different degree of methylation between Cd and Zn dimers shows that the presence of zinc center has higher methylation selectivity and weaker reactivity.

We have successfully used the cadmium ion with (2-thiophenyl)phenylphosphine(PS2) or Bis(3-trimethylsilyl-2-thiophenyl)phenylphosphine(SiPS2) to synthesize [(PS2)Cd(TMEDA)](2) and [(SiPS2)Cd(TMEDA)](4) that were structurally similar to [(PS2)Zn(TMEDA)](1). We also obtained a series of zinc complxes [NEt4][(PS2)Zn(SC6H5)](6), [NEt4][{[(PS2)Zn](SC6H11)[(PS2)Zn]}](7) and [NEt4][(PS2)Zn(SCH2C6H5)](8) with systematically varied thiolates. From studying of the methylation reactions of these complexeswith methyliodine or trimethylphosphate, we found that the metal center(zinc or cadmium), ligand(PS2 or SiPS2) or the net charge (neutral ornegative) can influence the reaction time, selectivity of reaction site ormechanism. The zinc complexes (6), (7), and (8) are capable of completingthe catalytic cycle of mthylation-demethylation and hence a good model for related enzyme system.

The Quantum chemical cluster approach has been shown to be quite powerful and efficient in the modeling of enzyme active sites and reaction mechanisms. In this thesis, the reaction mechanisms of several enzymes have been investigated using the hybrid density functional B3LYP. The enzymes studied include four dinuclear zinc enzymes, namely dihydroorotase, N-acyl-homoserine lactone hydrolase, RNase Z, and human renal dipeptidase, two trinuclear zinc enzymes, namely phospholipase C and nuclease P1, two tungstoenzymes, namely formaldehyde ferredoxin oxidoreductase and acetylene hydratase, aspartate α-decarboxylase, and mycolic acid cyclopropane synthase. The potential energy profiles for various mechanistic scenarios have been calculated and analyzed. The role of the metal ions as well as important active site residues has been discussed.   In the cluster approach, the effects of the parts of the enzyme that are not explicitly included in the model are taken into account using implicit solvation methods.   For all six zinc-dependent enzymes studied, the di-zinc bridging hydroxide has been shown to be capable of performing nucleophilic attack on the substrate. In addition, one, two, or even all three zinc ions participate in the stabilization of the negative charge in the transition states and intermediates, thereby lowering the barriers.   For the two tungstoenzymes, several different mechanistic scenarios have been considered to identify the energetically most feasible one. For both enzymes, new mechanisms are proposed.   Finally, the mechanism of mycolic acid cyclopropane synthase has been shown to be a direct methyl transfer to the substrate double bond, followed by proton transfer to the bicarbonate.   From the studies of these enzymes, we demonstrate that density functional calculations are able to solve mechanistic problems related to enzymatic reactions, and a wealth of new insight can be obtained.