J Biol Chem, Vol. 274, Issue 46, 32733-32737, November 12, 1999

Proton Transfer from Histidine 244 May Facilitate the 1,2 Rearrangement Reaction in Coenzyme B₁₂-dependent Methylmalonyl-CoA Mutase^*

Nilesh Maiti, Lusiana Widjaja, and Ruma Banerjee

From the Biochemistry Department, University of Nebraska, Lincoln, Nebraska 68588-0664

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Methylmalonyl-CoA mutase is an adenosylcobalamin-dependent enzyme that catalyzes the 1,2 rearrangement of methylmalonyl-CoA to succinyl-CoA. This reaction results in the interchange of a carbonyl-CoA group and a hydrogen atom on vicinal carbons. The crystal structure of the enzyme reveals the presence of an aromatic cluster of residues in the active site that includes His-244, Tyr-243, and Tyr-89 in the large subunit. Of these, His-244 is within hydrogen bonding distance to the carbonyl oxygen of the carbonyl-CoA moiety of the substrate. The location of these aromatic residues suggests a possible role for them in catalysis either in radical stabilization and/or by direct participation in one or more steps in the reaction. The mechanism by which the initially formed substrate radical isomerizes to the product radical during the rearrangement of methylmalonyl-CoA to succinyl-CoA is unknown. Ab initio molecular orbital theory calculations predict that partial proton transfer can contribute significantly to the lowering of the barrier for the rearrangement reaction. In this study, we report the kinetic characterization of the H244G mutant, which results in an acute sensitivity of the enzyme to oxygen, indicating the important role of this residue in radical stabilization. Mutation of His-244 leads to an ~300-fold lowering in the catalytic efficiency of the enzyme and loss of one of the two titratable pK_a values that govern the activity of the wild type enzyme. These data suggest that protonation of His-244 increases the reaction rate in wild type enzyme and provides experimental support for ab initio molecular orbital theory calculations that predict rate enhancement of the rearrangement reaction by the interaction of the migrating group with a general acid. However, the magnitude of the rate enhancement is significantly lower than that predicted by the theoretical studies.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Methylmalonyl-CoA mutase catalyzes the 1,2 rearrangement of methylmalonyl-CoA to succinyl-CoA and is the only coenzyme B₁₂-dependent enzyme that is present in both microbial and animal kingdoms (1, 2). The role of the B₁₂ cofactor or adenosylcobalamin (AdoCbl)¹ is to function as a free radical reservoir responsible for the controlled generation of a substrate radical at the initiation of the reaction cycle (3). The first common step in AdoCbl-dependent reactions is the homolytic cleavage of the cobalt-carbon bond to generate a radical pair (Scheme I). Stopped-flow kinetic studies indicate that the homolysis reaction is coupled to the subsequent step, in which a hydrogen atom is abstracted from the methyl group of the substrate to generate a substrate-centered radical (4). Evidence for kinetic coupling has also been obtained for the related glutamate mutase-catalyzed reaction (5). In ribonucleotide reductase, homolysis of the cobalt-carbon bond appears to be coupled to the generation of a protein-derived thiyl radical (6). Electron paramagnetic resonance studies (7-10) provide evidence for the presence of both metal and carbon-centered radicals in the methylmalonyl-CoA mutase-catalyzed reaction. In the next step, an interchange of the carbonyl-CoA group and a hydrogen atom occurs between vicinal carbons. The precise mechanism by which the substrate radical isomerizes to the product radical is unresolved, with pathways involving free radical intermediates (3), fragmentation products (11), carbocations (12), carbanions (13), and organocobalt adducts (14, 15), each having been proposed (Scheme II). The final step involves reabstraction of a hydrogen atom from deoxyadenosine to generate product and the intact cofactor, AdoCbl.

View larger version (19K): [in this window] [in a new window]   Scheme I.   Postulated reaction mechanism of methylmalonyl-CoA mutase involving a direct rearrangement of the substrate radical. R is CoA.

View larger version (21K): [in this window] [in a new window]   Scheme II.   Alternative pathways for the rearrangement of a methylmalonyl-CoA radical involving a carbonium ion intermediate (a), a carbanion intermediate (b), direct rearrangement (c), dissociative or fragmentation pathway (d), and an organocobalt intermediate (e).

In the methylmalonyl-CoA mutase-catalyzed reaction, a cyclopropyloxy radical is predicted for a pathway involving direct radical rearrangement (Schemes II, pathway c). Ab initio molecular orbital calculations predict that full or partial protonation of the substrate radical can facilitate the rearrangement reaction (12, 16-18). Since the conjugate acid of the thioester carbonyl oxygen has a pK_a of ~6 (19), this precludes complete protonation of the carbonyl oxygen of methylmalonyl-CoA by any active-site residue. However, recent high level ab initio molecular orbital theory-based calculations predict that even partial proton transfer can contribute substantially to lowering the transition state barrier for the putative cyclopropyloxy radical (18).

The crystal structure of methylmalonyl-CoA mutase has been solved recently (20-22). It reveals a buried active site in the  subunit, which opens to the surface via a long tunnel that is plugged by the CoA tail of the substrate. The carbonyl group of the substrate is in hydrogen-bonding distance to an active-site residue, His-244 (Scheme III), which is involved in a hydrogen-bonding network that includes the carboxylate of the substrate as well as Arg-207 (23). His-244 is thus well positioned to function as a general acid that could promote the rearrangement reaction by hydrogen bonding to the carbonyl-CoA group in the ground state and/or in the transition state. Mutation of the aromatic residue Tyr-89 to phenylalanine results in altered energetics with respect to the wild type enzyme, which is consistent with an increase in the barrier to interconversion between the substrate and product radicals (24).

View larger version (13K): [in this window] [in a new window]   Scheme III.   Interactions between the substrate and residues in the active site of methylmalonyl-CoA mutase.

To test the hypothesis that His-244 facilitates the rearrangement reaction by serving as a general acid, we have employed site-directed mutagenesis. Mutation of this residue to glycine results in an acute sensitivity of the mutase-catalyzed reaction to oxygen, in contrast to the relative stability of the wild type enzyme to oxygen, and results from quenching of the radical intermediates. The catalytic efficiency of the H244G mutant is ~300-fold lower than that of the wild type enzyme under anaerobic conditions. The pH dependence of the wild type and mutant enzymes under maximal velocity conditions are distinct and suggests that one of the two kinetic pK_a values associated with the wild type reaction is due to His-244. Our data provide experimental support for a modest rate enhancement of the rearrangement reaction by His-244, in contrast to the large effects predicted by the ab initio molecular orbital theory calculations, and reveal the importance of the aromatic cluster residue, His-244, in radical shielding.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Methylmalonyl-CoA and DEPC were purchased from Sigma. Thiokinase was purchased from Roche Molecular Biochemicals. Radioactive [¹⁴C]CH₃-malonyl-CoA (56.4 Ci/mole) was purchased from New England Nuclear.

Construction of H244G Mutation-- The following primers were employed for the polymerase chain reaction reactions. Sense mutagenic primer: TCCGGCTACGGCATGCAGGAAGCC; Antisense mutagenic primer: GGCTTCCTGCATGCCGTAGCCGGA. The codon specifying Gly-244 is italicized, and the newly created SphI restriction site is underlined. The sequences of the flanking sense and antisense primers containing MluI and PstI restriction sites, respectively, were as follows. Sense: GGCCCGTATGCAACCATGTACGCG; antisense: GTCGTGCCCGATTCCTGCTGCAGG.

After the polymerase chain reaction, a 0.94-kilobase product was cloned into a TA cloning vector using the cloning kit from Invitrogen. MluI and partial PstI restriction digestion were employed to exchange the 0.94-kilobase fragment in the native expression vector, pMEX2 (25) (provided by P. Leadlay at the University of Cambridge) with the corresponding fragment containing the H244G mutation. The presence of the mutation was confirmed by nucleotide sequence determination of both strands as well as by identification of a newly created SphI site.

Enzyme Purification-- The H244G variant was expressed at levels that were comparable to that of the wild type enzyme and was purified using the same conditions described previously for the native enzyme (10).

Enzyme Assays-- Activity of the enzyme and the pH dependence of the reaction were determined in the fixed-time radiolabel assay at 37 °C as described previously (10). For the pH studies, the dependence of the reaction velocity was monitored at a fixed and saturating substrate concentration at different pH values. That the enzymes were operating under V_max conditions was confirmed by Michaelis-Menten analysis of the substrate dependence of the reaction velocity at each pH, with substrate concentrations ranging from 0.5× to 10× K_m. The pH profiles obtained from both data sets were comparable. Due to the acute lability of H244G to oxygen, the standard assay had to be modified to make measurements under anaerobic conditions. For this, all reagents were bubbled with N₂ for 3 h before use. The enzyme solution was deoxygenated by gently blowing a stream of N₂ over its surface for 10 min at 4 °C. One unit of enzyme activity produces 1 µmol of succinyl-CoA/min at 37 °C.

DEPC Inhibition Kinetics-- The sensitivity of H244G to alkylation by DEPC was measured under anaerobic assay conditions in the radiolabeled assay described above. The final concentration of DEPC was 200 µM, and 500 µM hydroxylamine was employed to reverse the reaction as described previously (26).

Aerobic Photolysis of Holomethylmalonyl-CoA Mutase-- The stability of the bound cofactor was determined by irradiation of a 13 µM sample containing either wild type or H244G enzyme in 50 mM potassium phosphate buffer, pH 7.5, placed on ice. A 60-W lamp was placed 8 inches away from the enzyme sample. Photolysis of bound AdoCbl was monitored by recording spectra periodically between 300 and 700 nm. The aerobic photolysis product, OHCbl, is easily identified by its sharp absorption maximum at 350 nm.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Purification and Expression of the Mutant Protein-- The mutation of His-244 to glycine does not adversely affect the expression level of the recombinant protein. Since the chromatographic behavior of the mutant was indistinguishable from that of wild type enzyme, it permitted its purification using the previously described protocol (10). The purified enzyme had a specific activity of 0.44 units/mg of protein at 37 °C under V_max conditions in the anaerobic assay. The thermal stabilities of the H244G mutant and wild type enzymes were compared by preincubation of the enzymes for varying lengths of time at 37 °C before their use in the standard assay and were found to be indistinguishable (data not shown).

Spectrum of H244G under Steady-state Turnover Conditions-- Although the rate constants associated with the individual steps of the reaction shown in Scheme I have not been determined, kinetic characterization of the wild type enzyme indicates that product inhibition may be rate-determining (4, 24, 27). Thus, under steady-state conditions, the spectrum of the enzyme is dominated by AdoCbl (4). In contrast, the addition of methylmalonyl-CoA to the holoenzyme form of H244G results in the rapid formation of OHCbl (Fig. 1A, spectrum 3). To determine whether the quenching of the cob(II)alamin intermediate to OHCbl occurred due to accessibility to oxygen or by an active-site water molecule that may reside in the cavity created by the missing imidazole side chain, the same experiment was repeated under anaerobic conditions. For this experiment, 19 µM enzyme was mixed with 14.4 mM (R,S)-methylmalonyl-CoA under anaerobic conditions, and the first spectrum was recorded 1.5 min after the addition of substrate. Under these conditions, the spectrum of the enzyme was obtained under steady-state turnover conditions and revealed the predominance of AdoCbl (Fig. 1A, spectrum 2). It is marked by a decrease in absorption across much of the spectral range. Similar spectral changes have been reported for the wild type mutase under equilibrium conditions (9). In addition, similar spectral changes were noted in the reverse direction, i.e. when H244G was mixed with succinyl-CoA under anaerobic conditions (data not shown). In contrast, the spectrum of wild type enzyme mixed with substrate is identical under aerobic and anaerobic conditions (Fig. 1B).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Spectra of mutant and wild type enzymes in the presence of substrate. A, H244G mutant. Spectral changes resulting from addition of 14.4 mM (R,S)-methylmalonyl-CoA to a solution of H244G (19 µM) in 50 mM potassium phosphate buffer, pH 7.5. Spectra were recorded before (1) and 1.5 min after the addition of substrate under either aerobic (2) or anaerobic conditions (3). B, spectral changes resulting from the addition of 1.85 mM (R,S)-methylmalonyl-CoA to either an aerobic (2) or anaerobic solution of wild type enzyme (12.4 µM) in 50 mM potassium phosphate buffer, pH 7.5. The top spectrum (1) was of holoenzyme before addition of substrate. Essentially the same spectral changes were observed 1.5 min after the addition of substrate to either aerobic or anaerobic enzyme.

Since the H244G mutant is significantly more labile to oxygen under steady-state turnover conditions, it was of interest to determine whether or not the mutation had altered the stability of AdoCbl in the ground state, i.e. in the absence of substrate. Both H244G and wild type enzyme stabilize bound AdoCbl against photolysis to the same extent. Thus, AdoCbl bound to both enzyme forms was resistant to photolysis during ~3 h of irradiation using conditions described under "Experimental Procedures" (data not shown). Slow conversion to OHCbl was observed at longer times. Hence, the lability of the bound cofactor to oxidative escape after cleavage of the cobalt-carbon bond occurs only in the presence of substrate in H244G.

Kinetic Properties of the H244G Variant-- Due to the oxygen lability of the H244G mutant, the steady-state kinetic parameters were determined under anaerobic conditions (Table I). The mutation increases the K_m for methylmalonyl-CoA by a factor of ~3. In contrast, the K_m,app for AdoCbl is unchanged (data not shown). The k_cat for the mutant is ~100-fold lower than for the wild type enzyme. Thus, the H244G variant has an ~300-fold lower kinetic efficiency as compared with the wild type enzyme. The activity of the wild type enzyme measured under aerobic and anaerobic assay conditions is identical, indicating insensitivity to oxygen at least during the duration of the assay.

pH Dependence of Wild Type and H244G Enzymes-- The pH dependences of the wild type and mutant enzymes under V_max conditions are distinct (Fig. 2). The wild type enzyme exhibits a bell-shaped pH dependence with pK_a,1app = 5.7 ± 0.1 and pK_a,2app = 8.4 ± 0.1, respectively, and a pH maximum of 7-7.5 (Fig. 2A). This is very similar to the pH dependence reported previously for the ovine mutase (28). In contrast, a single kinetic pK_a,app (5.2 ± 0.1) controls the activity of the H244G mutant and is similar to pK_a,1app for the wild type enzyme (Fig. 2B). These data support the assignment of the second kinetic pK_a in the wild type enzyme to ionization of His-244. Protonation of His-244 is therefore associated with higher mutase activity (Scheme III).

View larger version (13K): [in this window] [in a new window]   Fig. 2.   pH dependence of methylmalonyl-CoA mutase catalyzed-reaction under Vmax conditions. The activity of wild type (A) and mutant (B) enzymes was measured in the radiolabeled assay at 37 °C. The experimental data were fit to the following equation: V = Vmax (1/1 + 10pK1  pH × (1/1 + 10pH  pK2).

DEPC Inhibition-- The crystal structure of methylmalonyl-CoA mutase reveals the presence of two histidines in the active site: His-244 and His-610 (20). The latter serves as the lower axial ligand to the cobalt. The activity of wild type enzyme is susceptible to the alkylating agent, DEPC (26), and the inhibition is partially reversed by hydroxylamine, consistent with histidine ligation (29). To determine whether the sensitivity of the wild type enzyme results from alkylation at His-244, the effect of DEPC on the activity of H244G was examined (Table II). In contrast to the wild type enzyme, H244G is more resistant to 200 µM DEPC retaining ~30% of its activity versus only 6% for wild type after preincubation of the enzyme with DEPC. The presence of the cofactor, AdoCbl, completely protects against inhibition in both wild type and mutant enzymes. Since the susceptibility of the second histidine mutant, H610A, to DEPC is similar to that of H244G,² these results indicate DEPC inhibition of wild type enzyme results from alkylation at both active-site histidine residues.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The carbon skeleton rearrangements catalyzed by a subgroup of AdoCbl-dependent enzymes represent chemically challenging reactions, and competing hypotheses have been advanced to explain the mechanism of this step (Scheme II). Of these, mechanisms involving carbocation (pathway a) or carbanion (pathway b) intermediates seem unlikely to be important since they would be accompanied by the reduction or oxidation, respectively, of the homolysis product, cob(II)alamin. The absorption spectra of cobalamins containing cobalt in the 3⁺, 2⁺, and 1⁺ oxidation states are distinct, and spectral changes indicative of oxidation state changes have not been observed during enzyme-monitored turnover in either methylmalonyl-CoA mutase (4) or glutamate mutase (5). Based on the crystal structures of methylmalonyl-CoA mutase-containing substrate analogs, the pathway involving an organocobalt adduct (pathway e) appears to be precluded by the distance between the cobalt and the vicinal substrate carbon atoms at which the rearrangement reaction occurs (20, 21, 23).

The dissociative mechanism (pathway d) predicts the generation of acrylate and a formyl-CoA radical as intermediates (11). To account for the observed retentive substitution at both carbon centers (30), rotation of the acrylate around the C₁-C₂ bond has to be invoked (11). The driving force for this conformational rotation is not clear. Furthermore, formyl-CoA and acrylate inhibit methylmalonyl-CoA mutase with a very high K_i, which is >100-fold larger than the K_m for succinyl-CoA (31). Recently ab initio calculations indicate that the fragmentation-reassociation pathway requires significantly greater energy (93.2 kJ mol¹ versus 46.9 kJ mol¹ for direct rearrangement (18)). Thus, there are no compelling reasons at present to consider this mechanism for the methylmalonyl-CoA mutase-catalyzed reaction.

Although there is insufficient evidence to rigorously dismiss the alternative pathways, the most likely mechanism for methylmalonyl-CoA mutase based on the available information, is direct rearrangement (Scheme II, pathway c). This intramolecular rearrangement reaction is estimated to have a barrier of 46.9 kJ mol¹ (18). Although complete protonation is predicted to lower the barrier to migration by 36.9 kJ/mol (2  3 versus 5  6 in Scheme I), an incremental lowering of this barrier is predicted by partial protonation of the carbonyl group with acids of intermediate strengths (18). The crystal structure of methylmalonyl-CoA mutase indicates that His-244 in the large subunit is within hydrogen-bonding distance to the carbonyl oxygen of the carbonyl-CoA moiety (20, 21). Its location is, thus, suggestive of a potential role in facilitation of the rearrangement reaction.

The mutation of His-244 to glycine represents a nonconservative change, and the interpretation of the kinetic results presented here must therefore be treated with caution. However, a preliminary report on the mutation of His-244 to alanine indicates that replacement of the imidazole ring results in the occupancy of a water molecule between the bound substrate and the alanine in the crystal structure determined at a resolution of 2.8 Å (32). The kinetic properties of the H244A and H244G mutants are similar, and they are both susceptible to oxidation under turnover conditions (32). This engenders confidence in the inference that the pH-dependent changes associated with the H244G mutation do not result from unwanted conformational changes elsewhere in the protein.

One of the remarkable features of methylmalonyl-CoA mutase is that it catalyzes a reaction involving radical intermediates under aerobic conditions. This raises the question of how the enzyme shields the reactive organic and metallic radicals. His-244 is involved in a hydrogen-bonding interaction with the carbonyl group of the CoA moiety of the substrate. The addition of substrate to an aerobic solution of wild type holomethylmalonyl-CoA mutase rapidly leads to an equilibrium mixture of substrate and product in which succinyl-CoA is favored by a factor of 23 (33). Under these conditions the spectrum of the enzyme resembles that of the starting enzyme but displays lower absorbance across the entire wavelength range (Fig. 1B). In contrast, addition of substrate to H244G leads to the rapid appearance of OHCbl, an oxidation product of the intermediate, cob(II)alamin (Fig. 1A). This can be prevented by the exclusion of oxygen from the solution. The sensitivity of H244G to oxygen is mirrored by the difference in its activity under aerobic (specific activity = 0.02 units/mg) and anaerobic (specific activity = 0.44 units/mg) conditions. Thus in H244G, cob(II)alamin is intercepted by oxygen once every 22 turnovers under aerobic conditions. In contrast, the same specific activity is measured for the wild type enzyme whether or not oxygen is present in the assay mixture (data not shown). Hence, His-244 buried in the active site, as revealed by the static crystal structures (20-22), nevertheless plays an important role in radical stabilization in the aerobic milieu that the enzyme operates in. The mutation does not, however, alter the susceptibility of AdoCbl to photolysis in the ground state. Thus, the photolysis products cob(II)alamin and the deoxyadenosyl radical show a strong propensity for recombination rather than oxidative escape, as is seen with the wild type enzyme.

The spectrum of H244G under steady-state conditions reveals the predominance of the AdoCbl state of the enzyme (Fig. 1A). Partitioning of tritium from AdoCbl to methylmalonyl-CoA and succinyl-CoA are consistent with an increase in the barrier for interconversion of the substrate and product radicals in the kinetically similar albeit conservative mutant, H244Q (32). Since the rate constants for the individual steps in the mutase-catalyzed reaction are not known, it is difficult to predict a priori whether or not the H244G mutation should lead to enhanced accumulation of cob(II)alamin under steady-state conditions. The relative energies of the other intermediates along the pathway are unknown, and the steady-state composition of the enzyme will be governed by these and the height of the barriers separating the intermediates.

The wild type enzyme shows a bell-shaped pH dependence with two kinetic pK_a values (Fig. 2). Since the two pK_a values are within three units of each other and a pH independent region is not observed, the measured pK_a values represent estimates and are not likely to be precisely determined. The descending limb of the pH titration is lost when a single residue, His-244, is mutated. The estimated pK_a for the mutant is similar to pk_a,1app for the wild type enzyme. These data support the assignment of the second pK_a to His-244 in the wild type enzyme.

Assuming that pk_a,2app represents the ionization of histidine, then protonation of His-244 is associated with higher mutase activity (Scheme III). A pK_a of 8.4 is somewhat high for an imidazole, which is typically between 6 and 7. The proximity of His-244 to the carboxylate of the bound substrate, which is within hydrogen-bonding distance (23), could account for the perturbation. In papain, a mercaptide-imidazolium ion pair interaction in the active site raises the histidine pK_a to 8.4 (34).

Using ab initio molecular orbital calculations, Radom and co-workers (18) recently demonstrated that the barrier for the intramolecular rearrangement of the 3-propanal radical, employed as a model for the methylmalonyl-CoA radical, can be lowered by 5.5 kJ mol¹ and 22.4 kJ mol¹ by partial proton transfer from HF and NH₄⁺, respectively. This corresponds to a rate increase by approximately 1 and 4 orders of magnitude (for HF and NH₄⁺, respectively) due to stabilization of the transition state by partial protonation by these acids. Histidine in the mutase active site may be expected to approximate NH₄⁺ in the gas phase in terms of acid strength (18). Our data indicate that in methylmalonyl-CoA mutase, the active-site residue, His-244, has a modest effect and enhances the catalytic efficiency approximately 2 orders of magnitude. The conservative mutation, H244Q, is similar to H244G, lowering k_cat by a factor of 50 and raising K_m for succinyl-CoA 2-fold, thus leading to a 100-fold-lower catalytic efficiency for the reverse reaction compared with the wild type enzyme (32). The pH dependence of V_max has not been reported for either the H244A or H244Q mutants (32).

In summary, this study demonstrates that a major role of His-244 is in shielding the reactive radical intermediates formed during the rearrangement reaction from interception by oxygen. In addition, they furnish support for the participation of His-244 as a general acid in the mutase reaction. To our knowledge, these data provide the first experimental support for the ab initio molecular orbital theory-based calculations that partial proton transfer may facilitate the rearrangement reaction in this AdoCbl-dependent enzyme.

	FOOTNOTES

^* This work was supported in part by National Institutes of Health Grant DK45776 and a summer research fellowship from the Howard Hughes foundation (to L. W.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Fax: 402-472-7842; E-mail: rbanerjee1@unl.edu.

² S. Chakraborty and R. Banerjee, unpublished results.

	ABBREVIATIONS

The abbreviations used are: AdoCbl, 5'-deoxyadenosylcobalamin; DEPC, diethyl pyrocarbonate; OHCbl, hydroxocobalamin.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES