Chemistry of AdoMet-Dependent Methyltransferases
Saulius Klimasauskas and Grazvydas Lukinavicius, Laboratory of Biological DNA Modification, Institute of Biotechnology, Vilnius, Lithuania
5-adenosyl-L-methionine is a high energy compound and is the major source of methyl groups for a myriad of biologic transmethylation reactions. These highly specific single-carbon transfers onto diverse nucleophilic centers in biomolecules are catalyzed by methyltransferase enzymes and play important regulatory and structural roles in the cell. Here we discuss the chemical mechanism of the methylation reactions, including structural features of the methylsulfonium center in the cofactor molecule, enzyme-assisted activation of diverse nucleophilic targets by deprotonation or covalent catalysis, and spatial constraints of the reaction.
The first literature source describing biologic methylation refers to the methyl donor S-adenosyl-L-methionine (AdoMet) as “ATP-activated form of methionine” (1). The formation of AdoMet from methionine and ATP is catalyzed by AdoMet synthetase (MAT, methionine adenosyltransferase, EC 184.108.40.206) and occurs in a two-step reaction in which PPi and Pi are released along with the product (AdoMet) (2). Although almost any part of AdoMet molecule can be used by the living organisms (3), the most ubiquitous and important role of AdoMet is the participation in biologic transmethylation reactions (Fig. 1). Other methyl group donors such as tetrahydrofolic acid, betaine, and vitamin B12 are used in certain cases, but AdoMet is by far the most often used source of methyl groups and the second-most ubiquitous cofactor after ATP. The methyl transfer reactions from AdoMet are associated with very favorable enthalpies (~—70 kJ/mol) compared with other methyl donors, which permits efficient and selective methylation of a large variety of biologic substrates in all living organisms from bacteriophage to humans (4).
Figure 1. Action of AdoMet-dependent methyltransferases. (Top) Methyltransferases catalyze the transfer of the methyl group from the cofactor AdoMet onto defined nucleophilic targets (Nu = N, O, C, S) in various biomolecules (shown as gray balls). Activation of nucleophiles is often achieved by abstracting a proton (if present) from the target atom by a general base (B:). (Bottom) The methylation reactions proceed via a directSN2 transfer, in which the attacking nucleophile and the substrate are involved in a single transition state structure. These reactions require a transient hybridization change from sp3 to sp2 and back to sp3 with inversion of configuration at the reacting carbon.
Methyl Group as a Biologic Mark
Methyl group transfer to particular targets is directed by enzymes called methyltransferases (MTases, EC 2.1.1.-). These enzymes catalyze more than 150 different reactions (according to SWISS-PROT database). AdoMet-dependent methyltrans- ferases according to their substrates can be classified into four major groups: enzymes acting on small molecules, proteins, nucleic acids, and glycans. The atomic targets in these molecules can be carbon, oxygen, nitrogen, sulfur, or halides (4). Because of this enormous variety of substrates, methyltransferases play vital roles in many cellular processes, including cellular metabolism, signal transduction, storage, and processing of (epi)genetic information.
In a certain sense, the methyl group plays a comparable role with the phosphoryl group in biologic systems. From the chemical standpoint, the methyl group is a small (volume ~20 A3) and uncharged apolar group, which is usually added to replace a hydrogen atom (volume ~5 A3) in a target molecule. Therefore, depending on the role and the chemical context of its predecessor hydrogen, the structural content of the methylation signal will be different. If minimal chemical alterations in the target molecule are brought about, the methylation can be regarded as a “steric” signal. Such subtle additions of a small chemical group can lead to a highly specific recognition by cellular components and dramatic biologic consequences. Alternatively, when methylation leads to altered chemical-physical properties, such as tautomeric forms, H-bonding patterns, and new chiral centers, its role could be assigned as “chemical.” As an extreme example of a “chemical” role, the methyl group can serve as a chemical “activator” for a subsequent chemical conversion of a target molecule.
Out of more than 150 different reactions catalyzed by the methyltransferases more than 90 are carried on small molecules. The methyl groups serve here as important “chemical” building blocks required for the construction of essential cellular components and metabolites. N-methylation of the smallest amino acid glycine is responsible for regulating AdoMet/AdoHcy ratio in the eukaryotic cells (5). Catechol O-methyltransferase (COMT) is responsible for dopamine and other catecholamine neurotransmitters’ methylation hence, it influences nervous signal propagation (6). Another important enzymatic methylation converts norepinephrine to epinephrine (adrenaline) (7). Creatine, which is used by the cells for energy storage, is synthesized by methylation of guanidinoacetate (8). Methylated hydroxycinnamic acid derivatives are precursors of the plant cell wall esterified phenolic compounds, soluble sinapate esters, dimeric lignans, and the extensively cross-linked polymer lignin (9). Biosynthesis of vitamin B12 employs six different methyltransferases that add methyl groups to specific positions of the tetrapyrolle ring (10). Finally, histamine N-methyltransferase inactivates inflammatory and allergenic mediator histamine (11). An example of a fatty acid methylating enzyme is cyclopropane-fatty-acyl-phospholipid synthase from M. tuberculosis. Its action is required for the long-term survival of nongrowing cells and is often associated with environmental stresses (12).
DNA methyltransferases modify nucleobases by depositing methyl groups onto exocyclic amino groups (N6 in adenine and N4 in cytosine) or the intracyclic C5-position of cytosine. Because these methylation reactions occur in the major groove of the DNA helix and without chemical consequences on the DNA structure, they can be viewed as “steric” signals designed for recognition by specialized proteins, enzymes, or large multicomponent complexes. All three types of DNA methylation found in prokaryotes and archaea occur sequence-specifically. A unique DNA methylation pattern (a combination of several methylated sequences) serves as a discriminatory species “self” code. In higher eukaryotes, the cytosine-5 methylation is solely known, which occurs in both a sequence-specific and a locus-specific manner. DNA methylation generally leads to a strong and heritable repression of gene expression and plays numerous essential regulatory roles in cellular differentiation and development, parental imprinting, X-chromosome inactivation, and silencing of endogenous retroviruses (13).
Methylation of RNA is even more diverse and abundant than that of DNA. RNA methyltransferases target nearly all chemically accessible sites on nucleobases and the ribose: N1, N2, and N7 in guanine, N1 and N6 in adenine, C5 in cytosine or uracil, and 2'O in ribose. Most of the known biologic methyla- tion reactions seem to serve as a means for “chemical” tuning of RNA transcripts into biologically active species. For example, guanine-N1 methylation of tRNA prevents frame shifts during protein translation (14) in bacteria. 2'O-ribose methylations at specific positions are essential for stability of tRNA in both eukaryotes and prokaryotes (15). The same modifications guided by small, nucleolar RNAs to specific loci on rRNA is required for ribosome assembly (16). All mRNAs in eukaryotes are capped at their 5' end, and at least two RNA MTases are required for the maturation of the cap structure; these modifications are required for the stabilization and efficient translation of the transcripts (17). Resistance to clinically important antibiotics (macrolide, lincosamide, and streptogramin B) is conferred by adenine-N6 methylation in 23 SrRNA (18).
In proteins, a large variety of methylation targets has been identified: the carboxylate of aspartate and glutamate, the sulfur of cysteine and methionine, the imidazole of histidine, the amide of glutamine and asparagine, the guanidinium of arginine, the ε-amino group of lysine, and the terminal amino and carboxylate groups (19). Arginine and lysine can accept more than one methyl group producing symmetrical and unsymmetrical dimethylarginine, dimethyllysine, and trimethyllysine.
In the focus of a renewed interest are histone MTases. They produce monomethylarginine, symmetrical and unsymmetrical dimethylarginine, and all possible lysine ε-amino group methylation states. Such post-translational modifications of the histones determines whether chromatin adopts a compacted structure and is associated with silenced DNA-heterochromatin, or if it seems to be an extended structure and is associated with transcriptionally active DNA-euchromatin (20). Arginine MTases seem to methylate even more substrates, but the modification effects are not well understood (21). As in DNA, the above-described examples of methylated residues can be viewed as “steric” marks designed for recognition by highly specific proteins.
In contrast, isoaspartate MTases methylate the carboxyl group of isoaspartate residues, which are spontaneously accumulated in proteins via inadvertent isomerisation of aspartates and asparagines. The resulting methyloxycarbonyl group is reactive in the reverse intramolecular trans-esterification reaction, which regenerates an aspartate residue. Therefore, the methylation is the activating step on a chemical repair pathway of aged proteins and plays an important role in extending the life span of organisms from all domains of life (22).
The Methyl Donor: S-Adenosyl-L-Methionine
As discussed, AdoMet is a high energy compound. The methyl group is activated by the neighboring sulfonium center, and AdoHcy serves as a leaving group during the MTase-catalyzed reactions. The sulfonium center induces a partial positive charge on all three adjacent carbon atoms (methyl group, ribose 5'-carbon, and methionine γ-carbon). This charge is clearly observed from a nearly 1-ppm high field shift of proton chemical shifts in AdoMet as compared with those in methionine and AdoHcy (23). Under physiologic conditions, both adjacent methylene groups can be attacked leading to slow decomposition to inactive species. AdoMet is particularly labile under alkaline conditions, forming adenine and S-ribosylmethionine readily. This reaction is initiated by the deprotonation at C-5' (24). A competing pathway for AdoMet degradation, which is prominent at even lower pH values, involves an intramolecular attack of the α-carboxylate group onto the γ-carbon of the methionine moiety, resulting in methylthioadenosine and homoserine lactone (25). Curiously, no significant direct nucleophilic attack on the methyl groups by water is detectable under physiologic conditions—the third route of decay is pH-independent racemization of the chiral sulfonium center.
Chiral methyl groups containing 3H, 2H, and 1H hydrogens have been used to study the reaction mechanism of several methyltransferases (26, 27). In all cases, it was found that the transfer proceeds with inversion of configuration at the methyl group suggesting an SN2 mechanism. Isotopic replacements in the methyl group with 2H or 13C result in slight positive and negative isotope kinetic effects, respectively, in the case of COMT enzyme. Based on these observations a “symmetrical” and “tight” transition state has been proposed (28). It is now generally accepted that enzymatic methylation reactions proceed via a direct SN2 transfer, in which the attacking nucleophile and the substrate are involved in a single transition state structure (see Fig. 1). The catalytic power of many methyltransferases thus largely derives from their ability to bring the two substrates together in correct orientation.
Substitutions in the sulfonium center
AdoMet analogs, in which sulfur of the sulfonium center is replaced with selenium or tellurium, have been synthesized and used to study the reaction mechanism (Fig. 2a-c). Electronegativities of S, Se, and Te, (2.58, 2.55, and 2.10 on the Pauling scale, respectively) suggest that carbons adjacent to telluronium groups are poorer electrophiles as compared with carbons adjacent to selenonium or sulfonium groups atom because of the reduced ability of the tellurium atom to induce a positive dipole at the adjacent carbon. On the other hand, the increase in atomic radii results in weaker heteroatomic bonds in the series S-CH3 > Se-CH3 > Te-CH3, thus making the higher analogs better leaving groups. These two opposing effects result in a slight chemical activation of the Se analog and substantial inactivation of the Te analog, as observed in the two enzymatic systems examined (29, 30). The same reasoning seems to be valid for explaining the relative effects on the heteroatoms on the two modes of decay of the analogs in water: deprotonation at C-5' and intramolecular nucleophilic attack at C-γ. The deprotonation reaction is decreased, whereas decomposition via the C-γ attack is enhanced in the selenium analog as compared with AdoMet. This is understood taking into account that the first pathway involves the activation of the carbon atom, whereas the second requires both electrophilic activation of the carbon and a good leaving group.
AdoMet analogs with nitrogen replacements of the sulfur atom (N-adenosyl-L-azamethionine or aza-AdoMet; see Fig. 2d) (31) seem even less reactive. These compounds are not susceptible to the decomposition reactions peculiar to AdoMet. They can act as charge-switchable mimics of AdoMet because the tertiary amino group (pKa = 7.1) can be protonated by adjusting the pH slightly below the physiologic values (32). However, because dialkylamines are not nearly as good leaving groups as are dialkylsulfides, these compounds are not expected to be efficient methyl group donors in methyltransferase-catalyzed reactions. Another problem may derive from a low inversion barrier at nitrogen as compared with sulfur, leading to lower abundance of the correct epimeric form and thus poor overall positioning of the reactants in the enzyme pocket. Although initial testing suggested that NAM could serve as a substrate (31), this compound functioned as an inhibitor but not as a methyl group donor in the case of a t-RNA (uracil-5-)-methyltransferase (33) and many other systems. Notably, a significantly improved reactivity is achieved in analogs carrying a sterically tense aziridine cycle (see Fig. 2e-f). Upon protonation of the ring nitrogen, enzyme-assisted nucleophilic attack on the ring carbon leads to the opening of the ring, which couples the whole cofactor to the target (Fig. 3a) (34). Alternatively, the aziridine ring can be generated in situ in the active site of an enzyme from various adenosine-derived N-mustards (35).
A swap of the methyl carbon with nitrogen in aza-AdoMet leads to sinefungin (see Fig. 2g)—a natural nucleoside antibiotic found in Streptomyces griseolus. Such “reverse” chemistry additionally enhances the chemical stability of cofactor. Because of the positive charge of the protonated amine and correct chirality at the carbon center, sinefungin has an extremely high inhibitory potential for AdoMet-dependent methyltransferases.
Figure 2. Structural analogs of S,S-AdoMet with replacements in the sulfonium center and the transferable methyl group. Formulas are aligned to show the methyl group (or its equivalent) above, and the C-y and C-5' atoms below to the left and right side from the onium center, respectively. (a) S-adenosyl-L-methionine (AdoMet); (b) Se-adenosyl-L-selenomethionine (Se-AdoMet); (c) 7e-adenosyl-L-telluromethionine (Te-AdoMet); (d) N-adenosyl-L-azamethionine (aza-AdoMet); (e) N-adenosylaziridine; (f) N-adenosyl-N-mustard; (g) sinefungin; (h) S-adenosyl-L-ethionine (AdoEth); (i) S-adenosyl-L-propenthionine; (j) S-adenosyl-L-butynthionine; (k) S-adenosylvinthionine.
Substitutions in the transferable methyl group
AdoMet analogs with larger moieties replacing the methyl group (see Fig. 2h-k) have been obtained previously enzymatically (36) and later via regiospecific chemical S-alkylation of AdoHcy (37). Enzymatic studies of such AdoMet analogs indicate that relatively short chemical groups, such as ethyl and propyl, can be transferred by MTases, but the transfer rates decline drastically with increasing size of the transferable group (methyl » ethyl > propyl) (36). SN2 reactions proceed via a hybridization change at the reacting carbon from sp3 to sp2 and back to sp3 with inversion of configuration (Fig. 1). A transient p orbital formed at the reacting carbon interacts with both the attacking nucleophile and the leaving group. Therefore, unfavorable steric effects within the penta-coordinated transition state occur when the methyl group is extended to ethyl and additionally to propyl, which leads to dramatic progressive decreases in the reaction rate.
Remarkably, it was shown recently that the transalkylation reaction can be rescued by placing n-orbitals near the reaction center. This finding is observed with synthetic AdoMet analogs carrying a double bond (allylic system) or a triple bond (propargylic system) (37) (and likely aromatic systems) next to the reactive carbon in the extended side chain (see Fig. 2i-j and Fig. 3b). The unsaturated bonds seem to stabilize the sp2 transition state via conjugation of their n orbitals, with the transient p orbital in the penta-coordinated reaction center. These new synthetic cofactors are thus termed double-activated AdoMet analogs because the reactive carbon located between the sulfonium center and the unsaturated bond is activated for transfer by both adjacent groups (38).
A different effect is observed when a C=C double bond is attached directly to the sulfonium center (vinylic system; see Fig. 2k and Fig. 3c). Nucleophilic addition prevails because an ylide intermediate generated after addition at the adjacent carbon atom is stabilized by resonance involving d-orbitals of the sulfur atom. Consequently, the whole cofactor is covalently attached to the target molecule and irreversibly inhibits the enzyme (39).
Figure 3. Methyltransferase-directed coupling of extended groups to biomolecules using analogs of AdoMet. (a) Covalent coupling of N-aziridine and N-mustard cofactor mimics; (b) transfer of an extended aliphatic chain from a double-activated cofactor, S-adenosyl-L-propenthionine; (c) covalent coupling of an S-vinyl-analog of AdoMet.
Nucleophilic Methyl Groups Acceptors
As discussed, AdoMet is a high energy compound, and therefore, AdoMet-dependent methylation reactions are known to be irreversible. However, enzymatic catalysis is often required to enhance the rate of the reaction. Enzymes employ a variety of ways to enhance the nucleophilicity on the attacking atom in a substrate. Often, the reaction results in a proton exchange for the methyl group. The proton can be removed before, in concert with, or after the methyl transfer; this step usually requires the presence of a general base in the active site.
Nitrogen is a relatively good nucleophile but also a good base. In biologic substrates, nitrogen occurs in the form of alkyl amines (glycine, lysine, or phenylethanolamine), aromatic amines (cytosine-N4, adenine-N6, or guanine-N2), heterocyclic systems (guanine-N1 or -N7, adenine-N7, or histidine), or conjugated amines (glutamine or guanidine). Generally, an enzyme needs only to orient a lone pair of the nitrogen for SN2 in line attack onto the methyl group of the AdoMet. If the nitrogen is protonated (alkylammonium), the proton needs to be removed before the reaction can take place.
A well-studied example of aliphatic primary amino group methylation is provided by histone lysine MTases. The critical step in methylation of this amino acid is its deprotonation, because a protonated lysine is a very bad nucleophile. There are two possible ways to achieve this, as follows: “passive” catching of deprotonated species or “active” deprotonation of the lysine by a base in the active site of the enzyme. It turns out that both mechanisms are in use. In SET domain lysine MTases, the absence of an apparent general base in the active site and a very high and sharp pH optimum (~9-10) (40) suggest a passive mechanism. In contrast, classic fold lysine MTases, Dotlp, are active in a broader pH interval (6-9.5). These enzymes contain a conserved essential Asn residue in the active site, which may facilitate the deprotonation of the target lysine (41, 42).
The methylation of exocyclic (aromatic) nitrogens proceeds in a different fashion, because pKa values of such nitrogens are usually low enough to stay unprotonated under physiologic conditions. The best representatives of this kind of N-methylation are cytosine-N4 and adenine-N6 MTases. These enzymes make hydrogen bonds to the target amino group directing its hydrogens to positions corresponding to sp3 hybridization. Such a “forced” hybridization change is thought to enhance the nucleophilicity of the nitrogen lone pair and to accelerate the reaction (43). A similar mechanism may also be used in nonaromatic conjugated amines, for example, in arginine MTases. Because of a very high pKa value of the guanidinium group (~12), arginine methyltransferases are unlikely to achieve its complete deprotonation. Instead, they could polarize the guanidinium group and redistribute the positive charge away from the target nitrogen such that the nucleophilic attack on the methyl group of AdoMet is accelerated (44).
Oxygen is less nucleophilic as compared with nitrogen. However its nucleophilicity can be enhanced by generating a (partial) negative charge on it. In biologic substrates, modifiable oxygens are often found in three types of groups: phenolic, ribose hydroxyl, and carboxyl groups.
No general base catalysis is needed in the case of a carboxyl group, because it is usually deprotonated (pKa ~ 4.5) under physiologic conditions. Indeed, a well-characterized example of such an enzyme, protein-L-isoaspartate O-MTase, contains no acidic or basic residues in the substrate-binding cleft (22). Phenolic hydroxyl groups have a modestly high pKa (~10.5), and thus, they remain largely protonated under physiologic conditions. Enzymes can enhance the methylation reaction rate by abstracting a proton. For example, in the catechol MTase, the target hydroxyl group is coordinated with essential Mg2+, which apparently replaces the proton to generate a nucleophilic phenolate (6). Interestingly, electron-withdrawing groups in the phenol ring, such as nitro groups, lead to strong inactivation of the substrates converting them to potent inhibitors of the enzyme. Alternatively, caffeate, isolflavone, and chalcone O-methyltransferases use a well-positioned His residue for proton abstraction (45).
As in aromatic alcohols, ribose 2'-hydroxyls are protonated under physiologic conditions (pKa ~ 14.5). However, it seems less likely that enzymes can catalyze the methylation reaction by abstracting directly the proton from the hydroxyl group. Alternatively, rotation freezing and steering an oxygen lone pair toward the methyl group could sufficiently enhance the reaction. Such a mechanism is exemplified in the VP39 cap-2 mRNA (nucleoside-2'-O-)-methyltransferase. The enzyme is thought to form a nondeprotonating hydrogen bond between a Lys side chain and the 2'-OH proton. That would be possible if the lysine was unprotonated beforehand; i.e., its pKa value was much lower than usual. An 15N-labeled steric mimic of the lysine was employed to confirm that the pKa is indeed perturbed (~8.5) by adjacent Arg and Asp residues (46). The ultimate removal of the proton from the ribose oxygen onto the lysine occurs synchronously with or right after the methyl transfer. A similar mechanism may be valid in the case of tRNA guanosine-2'-O-methyltransferase, which uses an Arg residue to scavenge the proton (47).
Because of the low intrinsic nucleophilicity of carbon and a high energetic cost of generating an intermediate carbanion or its equivalent, the formation of a C-C bond in aqueous milieu seems a challenging task. However, single carbon transfers to carbon centers are a common event in biologic systems. Among the best studied examples are the methylation of the C5-position in pyrimidine nucleobases, cytosine and uracil, the methylation of tetrapyrrole system during synthesis of vitamin B12, and the cyclopropan ring formation in unsaturated fatty acids. A common motif of these reactions is the addition of the methyl group at a C=C double bond; however, it is achieved using different mechanisms.
It is thought that the cyclopropan ring formation in unsaturated fatty acids and methylation of tetrapyrrole system requires no covalent catalysis. In the active site of the cyclopropane fatty acid synthase, the nucleophilic n electrons of the double bond attack the electrophilic methyl group of AdoMet, which leads to a methylated carbocation intermediate. The resulting carbocation is likely to undergo a rapid equilibrium rearrangement to a protonated cyclopropane. The final cis-cyclopropane product is obtained on proton abstraction (apparently from the transferred methyl group itself!) by an active-site base. Interestingly, a bicarbonate ion was identified near the methyl group in one of the enzyme-substrate mimic crystal structures. It seems well positioned for general base catalysis, although bicarbonate is not commonly found in enzyme active sites (48). Less is known about the mechanism of enzymes acting on the tetrapyrole system. It is thought that the methylation of the tetrapyrrole framework is favorable, whereas a residue on the protein probably acts as a general base to initiate catalysis (49).
The catalytic mechanism of the pyrimidine-5 methylation in nucleic acids is more complex as it involves covalent catalysis. The mechanism is common for numerous DNA/RNA cytosine and uracil MTases as well as for thymidylate synthase (although the latter uses tetrahydrofolic acid as the methyl donor) and has been studied in detail in several systems (50). Here, the cytosine-5 methylation in DNA is presented as an example (see Fig. 4a). The C5-position of cytosine, which is part of an aromatic ring, does not carry sufficient nucleophilicity for a direct methyl group transfer. The continuity of the aromatic system is disrupted by a nucleophilic attack of thiolate (from a conserved cysteine residue in the enzyme) on the carbon-6 (51), which is accompanied by protonation of N3 (by a conserved glutamate). The resulting 4-5 enamine structure provides sufficient electron density at C5 (52) for a direct attack on the methyl group of AdoMet. The methyl transfer step is irreversible. The methylated covalent 5,6-dihydrocytidine intermediate is resolved into 5-methylcytosine and free enzyme via deprotonation at C5 and β-elimination of the cysteine residue. The nature of the base responsible for the C5-deprotonation in DNA cytosine MTases remains elusive, whereas an RNA uracil-5 MTase was shown to employ a second conserved Cys residue for that purpose (53). In the absence of cofactor, a proton from bulk water can reversibly bind to the C5 of the target cytosine resulting in MTase-dependent exchange of the C5-hydrogen into solvent (50).
A series of mechanism-based analogs of cytosine was used to elucidate the mechanistic details of covalent activation. The significance of these analyses extends into the realm of MTase inhibitor design for anticancer therapies (54). Among the best known inhibitors of DNA cytosine-5 MTases are 5-aza-2'-deoxycytidine (5-aza-dC or decitabine), 5-fluoro-2'-deoxycytidine (5-F-dC), and 2-pyrimidinone-1-β-D-(2’-deoxyriboside) (zebularine) (Fig. 4, b-d). In 5-aza-dC, the nucleophilic attack of the thiolate on the ring system is strongly facilitated, because a higher electronegative character of nitrogen at position 5 increases the electrophilicity at C6. In the presence of AdoMet, the covalent complex can still accept the methyl group from the cofactor, but it cannot be resolved additionally because no proton is present at N5, which is required to regenerate the double bond and free the enzyme. A similar mechanism is proposed for zebularine. 5-fluoro-2'-deoxycytidine exerts its methylation-dependent inhibitive effect on the action of DNA MTases entirely at the β-elimination step. The covalent 5,6-dihydrocytosine intermediate formed after the methyl group transfer cannot be resolved to products because the fluorine cannot be abstracted as a cation. The enzyme thus becomes irreversibly trapped in a stable covalent complex (55). Such covalent complexes were purified and crystallized to reveal first the structural details of the reaction intermediates (56).
Figure 4. Catalytic activation of cytosine for C5-methylation by nucleophilic addition of a thiolate at the C6 position. (a) The chemical mechanism of enzymatic DNA cytosine-5 methylation. Mechanism-based inhibition of DNA MTases by cytidine analogs 5-fluoro-2'-deoxycytidine (b), 5-aza-2'-deoxycytidine (c), and 2-pyrimidinone-1-p-D-(2'-deoxyriboside) (d).
Spatial Control in Enzymatic Transmethylations
The catalytic power of AdoMet-dependent MTases to a large extent derives from their ability to bring the two substrates, the cofactor AdoMet and a target molecule, together in the right orientation. As most other bisubstrate enzymes, MTases accommodate their substrates next to each other in a concave catalytic pocket. Five different protein folds are known to date that are used to bind AdoMet; however, one of them, similar to the NAD(P)-binding Rossman fold, is highly prevalent (4). Many of these enzymes contain flexible loops that close during catalysis to cover the bound substrates from bulk solvent and, in certain cases, bring in important catalytic residues. This typical arrangement serves to create a proper milieu for the reaction.
However, binding a target molecule in a concave pocket is not always easily achieved, because access to a specific locus that is deeply buried within a large macromolecule is a challenging task. The most extreme and elegant examples are found in DNA MTases, which methylate nucleobases in double-stranded DNA. The target positions for methylation are exocyclic amino groups or the intracyclic C5 position of cytosine, which are located in the major groove of the DNA helix. In the latter case, a covalent catalysis is required for the methylation, meaning that access to both faces of the ring and an edge of the cytosine base is critical (see Fig. 4). It is hard to imagine how that could be achieved in the framework of the DNA helix with nucleobases tightly stacked on each other. Remarkably, this problem is solved by completely rotating the target nucleoside out of the DNA helix and into the active site of a MTase, with minimal distortions to the rest of the DNA (56). Similar mechanisms are operative for many other enzymes acting on nucleic acids, including many RNA MTases.
Another catalytic challenge relates to product control in cases when several methyl groups can in principle be transferred onto the same residue or atom. As mentioned, lysine methylation can lead to monomethylated, dimethylated, or trimethylated products. Because, for example, secondary amines are better nucleophiles than primary amines, it is tricky to stop a chemical alkylation reaction at the monosubstituted product. This problem is solved by building in an appropriate number of bulky residues (tyrosines) in the active site such that only a defined number of methyl groups is accepted (57, 58). But this alone does not ensure processivity. The production of dimethylated or trimethylated lysine requires several cycles of AdoMet binding, methyl transfer, and AdoHcy release. The protein architecture with two substrates bound in a single active site bears the risk of releasing incompletely methylated products upon reloading the cofactor. If that is to be avoided for biologic reasons, different protein architecture should be used. Examples of such enzymes are SET domain lysine MTases, in which AdoMet and the substrate peptide are bound in two separate grooves located on different sides of the protein (57, 58). The successive cycles of methyl transfer are carried out in the cofactor pocket, whereas independent processing of the substrate and release of a properly methylated product occurs in the other.
Because many MTases play important roles in biologic processes, those are potentially good target candidates for drug design. Inhibitors for numerous enzymes have been produced and studied. Among important examples are inhibitors of COMT (see above), which are used as therapeutic agents in Parkinson’s disease (59). 5-aza-2'-deoxycytidine (5-aza-dC) and related analogs of cytidine are metabolically incorporated into DNA leading to mechanism-based irreversible inhibition of DNA MTases (54). Despite their high toxicity, these DNA-demethylating drugs can be applied in combination therapies with conventional anticancer drugs.
Another interesting application for MTases is the transfer of larger chemical entities from engineered cofactors. Such enlarged AdoMet mimics, in combination with a myriad of AdoMet-dependent MTases available in nature, would provide useful molecular tools for targeted functionalization of biomolecules (38). So far, two chemistries have shown a good promise. First, derivatives of N-aziridine-adenosine (60) and mechanistically related nitrogen mustards (61) (Fig. 2e-f) are covalently coupled to their natural targets in the presence of DNA MTases (Fig. 3a). By attaching chemical groups to these cofactors, they were shown to work as delivery systems for various functional or reporter groups. An inherent feature of this system is potent product inhibition (single-turnovers) by the covalent cofactor-substrate conjugate. A second class of AdoMet analogs circumvents the problem of catalytic product release. In these analogs, the methyl group is replaced with an extended carbon chain that contains an activating double or triple bond (allylic and propargylic systems; Fig. 2i-j) (37). These cofactors were shown to confer catalytic MTase-directed transfer of their activated side chains (Fig. 3b), permitting sequence-specific functionalization and labeling of plasmid DNA (62).
1. Cantoni GL. The nature of the active methyl donor formed enzymatically from L-methionine and adenosinetriphosphate. J. Am. Chem. Soc. 1952; 74:2942-2943.
2. Takusagawa F, Kamitori S, Markham GD. Structure and function of S-adenosylmethionine synthetase: crystal structures of S-adenosylmethionine synthetase with ADP, BrADP, and PPi at 2.8 angstroms resolution. Biochemistry 1996; 35:2586-2596.
3. Fontecave M, Atta M, Mulliez E. S-adenosylmethionine: nothing goes to waste. Trends Biochem. Sci. 2004; 29:243-249.
4. Schubert HL, Blumenthal RM, Cheng X. Many paths to methyltransfer: a chronicle of convergence. Trends Biochem. Sci. 2003; 28:329-335.
5. Balaghi M, Horne DW, Wagner C. Hepatic one-carbon metabolism in early folate deficiency in rats. Biochem. J. 1993; 291:145-149.
6. Vidgren J, Svensson LA, Liljas A. Crystal structure of catechol O-methyltransferase. Nature 1994; 368:354-358.
7. Gunne LM. Relative adrenaline content in brain tissue. Acta Physiol Scand. 1962; 56:324-333.
8. Komoto J, Yamada T, Takata Y, Konishi K, Ogawa H, Gomi T, et al. Catalytic mechanism of guanidinoacetate methyltransferase: crystal structures of guanidinoacetate methyltransferase ternary complexes. Biochemistry 2004; 43:14385-14394.
9. Ferrer JL, Zubieta C, Dixon RA, Noel JP. Crystal structures of alfalfa caffeoyl coenzyme A 3-O-methyltransferase. Plant Physiol. 2005; 137:1009-1017.
10. Warren MJ, Raux E, Schubert HL, Escalante-Semerena JC. The biosynthesis of adenosylcobalamin (vitamin B12). Nat. Prod. Rep. 2002; 19:390-412.
11. Horton JR, Sawada K, Nishibori M, Cheng X. Structural basis for inhibition of histamine N-methyltransferase by diverse drugs. J. Mol. Biol. 2005; 353:334-344.
12. Grogan DW, Cronan JE Jr. Cyclopropane ring formation in membrane lipids of bacteria. Microbiol. Mol. Biol. Rev. 1997; 61:429-441.
13. Brenner C, Fuks F. DNA methyltransferases: facts, clues, mysteries. Curr. Top. Microbiol. Immunol. 2006; 301:45-66.
14. Bjork GR, Wikstrom PM, Bystrom AS. Prevention of translational frameshifting by the modified nucleoside 1-methylguanosine. Science 1989; 244:986-989.
15. Kawai G, Yamamoto Y, Kamimura T, Masegi T, Sekine M, Hata T, et al. Conformational rigidity of specific pyrimidine residues in tRNA arises from posttranscriptional modifications that enhance steric interaction between the base and the 2’-hydroxyl group. Biochemistry 1992; 31:1040-1046.
16. Decatur WA, Fournier MJ. RNA-guided nucleotide modification of ribosomal and other RNAs. J. Biol. Chem. 2003; 278:695-698.
17. Shuman S. What messenger RNA capping tells us about eukaryotic evolution. Nat. Rev. Mol. Cell. Biol. 2002; 3:619-625.
18. Weisblum B. Erythromycin resistance by ribosome modification. Antimicrob. Agents Chemother. 1995; 39:577-585.
19. Clarke S. Protein methylation. Curr. Opin. Cell. Biol. 1993; 5:977-983.
20. Kouzarides T. Histone methylation in transcriptional control. Curr. Opin. Genet. Dev. 2002; 12:198-209.
21. Bedford MT, Richard S. Arginine methylation an emerging regulator of protein function. Mol. Cell 2005; 18:263-272.
22. Bennett EJ, Bjerregaard J, Knapp JE, Chavous DA, Friedman AM, Royer WE Jr, et al. Catalytic implications from the Drosophila protein L-isoaspartyl methyltransferase structure and site-directed mutagenesis. Biochemistry 2003; 42:12844-12853.
23. Stolowitz ML, Minch MJ. S-Adenosyl-L-methionine and 5’-Adenosyl-L-homocysteine, an NMR Study. J. Am. Chem. Soc. 1981; 103:6015-6019.
24. Borchardt RT. Mechanism of alkaline hydrolysis of S-Adenosyl-L-methionine and related sulfonium nucleosides. J. Am. Chem. Soc. 1979; 101:458-463.
25. Hoffman JL. Chromatographic analysis of the chiral and covalent instability of S-adenosyl-L-methionine. Biochemistry 1986; 25:4444-4449.
26. Woodard RW, Ming-Daw T, Heinz GF, Peter AC, James KC. Stereochemical course of the transmethylation catalyzed by catechol O-methyltransferase. J. Biol. Chem. 1980; 255:9124-9127.
27. H DK, Wu JC, Santi DV, Floss HG. Stereochemical studies of the C-methylation of deoxycytidine catalyzed by Hhal methylase and the N-methylation of deoxyadenosine catalyzed by EcoRI methylase. Arch. Biochem. Biophys. 1991; 284:264-269.
28. Hegazi MF, Borchardt RT, Schowen RL. A-deuterium and carbon-13 isotope effects for methyl transfer catalyzed by catechol O-methyltransferase. SN2-like transition state. J. Am. Chem. Soc. 1979; 101:4359-4365.
29. Iwig DF, Booker SJ. Insight into the polar reactivity of the onium chalcogen analogues of S-adenosyl-L-methionine. Biochemistry 2004; 43:13496-13509.
30. Iwig DF, Grippe AT, McIntyre TA, Booker SJ. Isotope and elemental effects indicate a rate-limiting methyl transfer as the initial step in the reaction catalyzed by Escherichia coli cyclopropane fatty acid synthase. Biochemistry 2004; 43:13510-13524.
31. Davis M, Dudman NPB, White HF. The synthesis of the N-methyl analogue of S-adenosylmethionine; NMR observation of diastereomers. Aust. J. Chem. 1983; 36:1623.
32. Thompson MJ, Mekhalfia A, Jakeman DL, Phillips SEV, Phillips K, Porter J, et al. Homochiral synthesis of an aza analogue of S-adenosyl-L-methionine (AdoMet) and its binding to the E. coli methionine repressor protein (MetJ). Chem. Commun. 1996: 791-792.
33. Santi DV, Hardy LW. Catalytic mechanism and inhibition of tRNA (uracil-5-)methyltransferase: evidence for covalent catalysis. Biochemistry 1987; 26:8599-8606.
34. Pignot M, Siethoff C, Linscheid M, Weinhold E. Coupling of a nucleoside with DNA by a methyltransferase. Angew. Chem. Int. Ed. 1998; 37:2888-2891.
35. Weller RL, Rajski SR. Design, synthesis, and preliminary biological evaluation of a DNA methyltransferase-directed alkylating agent. ChemBioChem. 2006; 7:243-245.
36. Schlenk F, Dainko JL. The S-n-propyl analogue of S-adenosylmethionine. Biochim. Biophys. Acta. 1975; 385:312-323.
37. Dalhoff C, Lukinavicius G, Klimasauskas S, Weinhold E. Direct transfer of extended groups from synthetic cofactors by DNA methyltransferases. Nat. Chem. Biol. 2006; 2:31-32.
38. Klimasauskas S, Weinhold E. A new tool for biotechnology: AdoMet-dependent methyltransferases. Trends Biotechnol. 2007; 25:99-104.
39. Zhao G, Zhou ZS. Vinyl sulfonium as novel proteolytic enzyme inhibitor. Bioorg. Med. Chem. Lett. 2001; 11:2331-2335.
40. Xiao B, Jing C, Wilson JR, Walker PA, Vasisht N, Kelly G, et al. Structure and catalytic mechanism of the human histone methyltransferase SET7/9. Nature 2003; 421:652-656.
41. Min J, Feng Q, Li Z, Zhang Y, Xu RM. Structure of the catalytic domain of human DOT1L, a non-SET domain nucleosomal histone methyltransferase. Cell 2003; 112:711-723.
42. Sawada K, Yang Z, Horton JR, Collins RE, Zhang X, Cheng X. Structure of the conserved core of the yeast Dot1p, a nucleo- somal histone H3 lysine 79 methyltransferase. J. Biol. Chem. 2004; 279:43296-43306.
43. Goedecke K, Pignot M, Goody RS, Scheidig AJ, Weinhold E. Structure of the N6-adenine DNA methyltransferase M.TaqI in complex with DNA and a cofactor analog. Nat. Struct. Biol. 2001; 8:121-125.
44. Zhang X, Zhou L, Cheng X. Crystal structure of the conserved core of protein arginine methyltransferase PRMT3. EMBO J. 2000; 19:3509-3519.
45. Zubieta C, Kota P, Ferrer JL, Dixon RA, Noel JP. Structural basis for the modulation of lignin monomer methylation by caffeic acid/5-hydroxyferulic acid 3/5-O-methyltransferase. Plant Cell 2002; 14:1265-1277.
46. Li C, Gershon PD. pKa of the mRNA cap-specific 2-O-methyltransferase catalytic lysine by HSQC NMR detection of a two-carbon probe. Biochemistry 2006; 45:907-917.
47. Nureki O, Watanabe K, Fukai S, Ishii R, Endo Y, Hori H, et al. Deep knot structure for construction of active site and cofactor binding site of tRNA modification enzyme. Structure 2004; 12:593-602.
48. Courtois F, Ploux O. Escherichia coli cyclopropane fatty acid synthase: is a bound bicarbonate ion the active-site base? Biochemistry 2005; 44:13583-13590.
49. Vevodova J, Graham RM, Raux E, Schubert HL, Roper DI, Brindley AA, et al. Structure/function studies on a S-adenosyl-L-methionine-dependent uroporphyrinogen III C methyltransferase (SUMT), a key regulatory enzyme of tetrapyrrole biosynthesis. J. Mol. Biol. 2004; 344:419-433.
50. Wu JC, Santi DV. Kinetic and catalytic mechanism of HhaI methyltransferase. J. Biol. Chem. 1987; 262:4778-4786.
51. Chen L, MacMillan AM, Chang W, Ezaz-Nikpay K, Lane WS, Verdine GL. Direct identification of the active-site nucleophile in a DNA (cytosine-5)-methyltransferase. Biochemistry 1991; 30:11018-11025.
52. Perakyla M. A model study of the enzyme-catalyzed cytosine methylation using ab initio quantum mechanical and density functional theory calculations: pKa of the cytosine N3 in the intermediates and transition states of the reaction. J. Amer. Chem. Soc. 1998; 120:12895-12902.
53. King MY, Redman KL. RNA methyltransferases utilize two cysteine residues in the formation of 5-methylcytosine. Biochemistry 2002; 41:11218-21125.
54. Gowher, H, Jeltsch, A. Mechanism of inhibition of DNA methyltransferases by cytidine analogs in cancer therapy. Cancer Biol. Ther. 2004; 3:1062-1068.
55. Osterman DG, DePillis GD, Wu JC, Matsuda A, Santi DV. 5-Fluorocytosine in DNA is a mechanism-based inhibitor of HhaI methylase. Biochemistry 1988; 27:5204-5210.
56. Klimasauskas S, Kumar S, Roberts RJ, Cheng X. HhaI methyl- transferase flips its target base out of the DNA helix. Cell 1994; 76:357-369.
57. Zhang X, Yang Z, Khan SI, Horton JR, Tamaru H, Selker EU, et al. Structural basis for the product specificity of histone lysine methyltransferases. Mol. Cell 2003; 12:177-185.
58. Trievel RC, Flynn EM, Houtz RL, Hurley JH. Mechanism of multiple lysine methylation by the SET domain enzyme Rubisco LSMT. Nat. Struct. Biol. 2003; 10:545-552.
59. Bonifati V, Meco G. New, selective catechol-O-methyltransferase inhibitors as therapeutic agents in Parkinson’s disease. Pharmacol. Ther. 1999; 8:1-36.
60. Pljevaljcic G, Schmidt F, Weinhold E. Sequence-specific methyltransferase-induced labeling of DNA (SMILing DNA). Chembiochem 2004; 5:265-269.
61. Weller RL, Rajski SR. DNA methyltransferase-moderated click chemistry. Org. Lett. 2005; 7:2141-2144.
62. Lukinavicius G, Lapiene V, Stasevskij Z, Dalhoff C, Weinhold E, Klimasauskas S. Targeted labeling of DNA by methyltransferase-directed Transfer of Activated Groups (mTAG). J. Amer. Chem. Soc. 2007; 129:2758-2759.
Cheng X, Blumenthal RM, eds. S-Adenosylmethionine-dependent methyltransferases: structures and functions. 1999. World Scientific, Singapore.
Cheng X, Roberts RJ. AdoMet-dependent methylation, DNA methyltransferases and base flipping. Nucleic Acids Res. 2001; 29:3784- 3795.
Grosjean H, ed. Fine-tuning of RNA Functions by Modification and Editing. 2005. Springer, Heidelberg.
Kozbial PZ, Mushegian AR. Natural history of S-adenosylmethionine-binding proteins. BMC Struct Biol. 2005; 5:1-26.
Martin JL, McMillan FM. SAM (dependent) I AM: the S-adenosyl-methionine-dependent methyltransferase fold. Curr. Opin. Struct. Biol. 2002; 12:783-793.
DNA, Covalent Modifications of
Enzyme Cofactors, Chemistry of
Enzyme Catalysis, Chemical Strategies for
Post-Translational Modification, Regulating Protein Function by
Proteins, in Vivo Chemical Modifications of
Protein-Nucleic Acid Interactions
Tags and Probes in Chemical Biology