Kinetic properties of two starch phosphorylases from pea seeds

Both of the starch phosphorylase fractions from Victory Freezer pea seeds, that can be separated by DEAE—cellulose chromatography and purified by Sepharose 4B-starch affinity chromatography, contain pyridoxal 5′-phosphate. The addition of further quantities of pyridoxal 5′-phosphate causes inactivation. Both enzymes showed similar bi-substrate kinetics with d-Glc-1-P and varying amounts of amylopectin and also with P_i and varying amounts of amylopectin. In the direction of glucan sythesis the K_m for amylopectin with phosphorylase II was much higher than with phosphorylase I. However, the two enzymes differed in their behaviour on glucan degradation at varying concentrations of P_i. With phosphorylase II the K_m for amylopectin was dependent on the concentration of P_i but that for phosphorylase I was constant. Phosphorylase II was strongly inhibited by ADPG in the direction of glucan degradation but only slightly in the direction of glucan synthesis by both ADPG and UDPG. Phosphorylase I was only slightly inhibited by ADPG in both directions and by UDPG in synthesis. UDPG inhibited both enzymes moderately in glucan degradation,     

Covalent modification of the solubilized rat liver vitamin K-dependent carboxylase with pyridoxal-5′-phosphate

The carboxylation of the pentapeptide substrate, Phe-Leu-Glu-Glu-Ile, by a rat microsomal vitamin K-dependent carboxylase was stimulated two- to threefold at pyridoxal-5′-P concentrations between 0.5 and 1.0 mm. This stimulation was reduced at concentrations higher than 1.0 mm. The K_m for the pentapeptide was lowered twofold in the presence of 1 mm pyridoxal-5′-P. The activation by pyridoxal-5′-P is specific, as 1 mm pyridoxal, pyridoxine, pyridoxine-5′-P, pyridoxamine, pyridoxamine-5′-P, or 4-pyridoxic acid did not stimulate the pentapeptide carboxylation rate. All six analogs, as well as formaldehyde and acetaldehyde, inhibited the carboxylation reaction in a concentration-dependent manner. The activation of the carboxylase by pyridoxal-5′-P appeared to be mediated by its direct binding to the enzyme via Schiff base formation. Sodium borohydride reduction of solubilized microsomes in the presence of pyridoxal-5′-P, followed by dialysis to remove unbound material, resulted in a carboxylase preparation with a specific activity twice that of the untreated control microsomes. The derivatized enzyme was not further stimulated by added pyridoxal-5′-P. This derivatized carboxylase could be obtained in the absence of pentapeptide and divalent cations. The stimulation of the carboxylase activity by divalent cations and pyridoxal-5′-P was mediated at separate site(s) on the enzyme. Studies of the NH₂-terminal pyridoxalated pentapeptide with both a normal and PLP-modified enzyme, in the presence and absence of PLP, demonstrated competition of the pentapeptide PLP moiety to a PLP site on the enzyme. It was concluded that pyridoxal-5′-P forms a covalent attachment to an ?-NH₂ of a lysine near the active site of the carboxylase.     

Chemical modification of an essential lysine at the active site of enoyl-CoA reductase in fatty acid synthetase

Fatty acid synthetase from goose uropygial gland was inactivated by treatment with pyridoxal 5′-phosphate. Malonyl-CoA and acetyl-CoA did not protect the enzyme whereas NADPH provided about 70% protection against this inactivation. 2′-Monophospho-ADP-ribose was nearly as effective as NADPH while 2′-AMP, 5′-AMP, ADP-ribose, and NADH were ineffective suggesting that pyridoxal 5′-phosphate modified a group that interacts with the 5′-pyrophosphoryl group of NADPH and that the 2′-phosphate is necessary for the binding of the coenzyme to the enzyme. Of the seven component activities catalyzed by fatty acid synthetase only the enoyl-CoA reductase activity was inhibited. Inactivation of both the overall activity and enoyl-CoA reductase of fatty acid synthetase by this compound was reversed by dialysis or dilution but not after reduction with NaBH₄. The modified protein showed a characteristic Schiff base absorption (maximum at 425 nm) that disappeared on reduction with NaBH₄ resulting in a new absorption spectrum with a maximum at 325 nm. After reduction the protein showed a fluorescence spectrum with a maximum at 394 nm. Reduction of pyridoxal phosphate-treated protein with NaB³H₄ resulted in incorporation of ³H into the protein and paper chromatography of the acid hydrolysate of the modified protein showed only one fluorescent spot which was labeled and ninhydrin positive and had an R_f identical to that of authentic N⁶-pyridoxyllysine. When [4-³H]pyridoxal phosphate was used all of the ³H, incorporated into the protein, was found in pyridoxyllysine. All of these results strongly suggest that pyridoxal phosphate inhibited fatty acid synthetase by forming a Schiff base with the ?-amino group of lysine in the enoyl-CoA reductase domain of the enzyme. The number of lysine residues modified was estimated with [4-³H]pyridoxal-5′-phosphate/NaBH₄ and by pyridoxal-5′-phosphate/NaB³H₄. Scatchard analysis showed that modification of two lysine residues per subunit resulted in complete inactivation of the overall activity and enoyl-CoA reductase of fatty acid synthetase. NADPH prevented the inactivation of the enzyme by protecting one of these two lysine residues from modification. The present results are consistent with the hypothesis that each subunit of the enzyme contains an enoyl-CoA reductase domain in which a lysine residue, at or near the active site, interacts with NADPH.     

Immobilization of nucleoside diphosphatase at its allosteric site using immobilized derivatives of pyridoxal 5'-phosphate

3-O-Immobilized and 6-immobilized pyridoxal 5′-phosphate analogs of Sepharose were bound to the allosteric site of nucleoside diphosphatase with very high affinity. Active immobilized nucleoside diphosphatase was prepared by reduction of the Schiff base linkage between the enzyme and pyridoxal 5′-phosphate bound to Sepharose with NaBH₄. 3-O-Immobilized pyridoxal 5′-phosphate analog gave more active immobilized enzyme than the 6-analog; the immobilized enzyme on the 3-O-immobilized pyridoxal 5′-phosphate analog showed about 90% of activity of free enzyme. The immobilized enzyme thus prepared was less sensitive to ATP, an allosteric effector, and showed a higher heat stability than the free enzyme. When an assay mixture containing inosine diphosphate and MgCl₂ was passed through a column of the immobilized enzyme at 37 °C, inosine diphosphate liberated inorganic phosphate almost quantitatively. Properties of the immobilized enzyme on the pyridoxal 5′-phosphate analog were compared with those of the immobilized enzyme on CNBr-activated Sepharose.     

Starch phosphorylase from tapioca leaves: Absence of pyridoxal phosphate

Starch phosphorylase from tapioca leaves has been purified to homogeneity, using the technique of ammonium sulfate fractionation, heat treatment, DEAE-cellulose chromatography, filtration through Sephadex G-100 and Sephadex G-200, and DEAE-Sephadex chromatography. The enzyme has a molecular weight of 450,000, as determined by gel filtration through Sephadex G-200 and contains 22 sulfhydryl groups per mole of the enzyme protein. Several types of evidence indicate the absence of pyridoxal 5′-phosphate as a prosthetic group of the enzyme. The kinetic data show a sequential type of the reaction mechanism. The enzyme activity is inhibited by tyrosine (K_i = 2.15 mm).     

Direct phosphate-phosphate interaction between coenzyme and substrate in the catalytic reaction of glycogen phosphorylase

Glycogen phosphorylase contains firmly bound pyridoxal 5′-phosphate (PLP), and catalyzes the reversible transfer of a glucosyl moiety between glucose-1-phosphate (G-1-P) and α-1,4-glucan. X-ray crystallographic studies revealed that PLP is located in a pocket where the phosphate group of PLP is pointed toward the G-1-P binding site. We have synthesized pyridoxal(5′)diphospho(1)-α-d-glucose, as a model compound for the phosphate-phosphate interaction between PLP and G-1-P, and reconstituted the enzyme with this compound. The resulting enzyme is catalytically inactive in itself, but, in the presence of glucan, the glycosyl moiety of this compound is transferred to the glucan forming a new α-1,4-glucosidic linkage along with the production of pyridoxal 5′-diphosphate. This glucosyltransfer is similar to the normal catalytic reaction in various aspects, although the rate is smaller in the order of three. AMP accelerates the transfer about 24 times compared with the reaction in its absence. We have more recently used pyridoxal(5′)triphospho(1)-α-D-glucose to reconstitute the enzyme. In the presence of glucan, the compound bound to enzyme is gradually degraded to pyridoxal 5′-triphosphate. This reaction is essentially dependent on AMP, and proceeds several times more slowly than the glucosyltransfer from the diphospho compound. These results provide evidence for the direct phosphate-phosphate interaction between the coenzyme and the substrate in the normal enzyme reaction, and seem to reflect a rather wide allowance in regard to this interaction.     

Functions of the 5'-phosphoryl group of pyridoxal 5'-phosphate in phosphorylase: a study using pyridoxal-reconstituted enzyme as a model system

Pyridoxal-reconstituted phosphorylase was used as a model system to study the possible functions of the 5'-phosphoryl group of pyridoxal 5'-phosphate (PLP) in rabbit muscle glycogen phosphorylase. Kinetic study was conducted by using competitive inhibitors of phosphite, an activator, and alpha-D-glucopyranose 1-phosphate (glucose-1-P) to study the relationship between the PLP phosphate and the binding of glucose-1-P to phosphorylase. Fluorine-19 nuclear magnetic resonance (19F NMR) spectroscopy of fluorophosphate bound to pyridoxal phosphorylase showed that its ionization state did not change during enzymatic catalysis. Evaluation of the apparent kinetic parameters for the activation of pyridoxal phosphorylase with different analogues having varied pKa2 values demonstrated a dependency of KM on pKa2. Molybdate, capable of binding as chelates in a trigonal-bipyramidal configuration, was tested for its inhibitory property with pyridoxal phosphorylase. On the basis of the results in this study, several conclusions may be drawn: (1) The bound phosphite in pyridoxal phosphorylase and, possibly, the 5'-phosphoryl group of PLP in native phosphorylase do not effect the glucose-1-P binding. (2) One likely function of the 5'-phosphoryl group of PLP in native phosphorylase is acting as an anchoring point to hold the PLP molecule and/or various amino acid side chains in a proper orientation for effective catalysis. (3) The force between the PLP phosphate and its binding site in phosphorylase is mainly electrostatic; a change of ionization state during catalysis is unlikely. (4) Properties of the central atoms of different anions are important for their effects as either activators or inhibitors of pyridoxal phosphorylase.(ABSTRACT TRUNCATED AT 250 WORDS)     

Pyridoxylation of essential lysine residues of mitochondrial adenosine triphosphatase

1. Soluble beef-heart mitochondrial ATPase (F₁) was incubated with [³H]pyridoxal 5′-phosphate and the Schiffbase complex formed was reduced with sodium borohydride. Spectral measurements indicate that lysine residues are modified and gel electrophoresis in the presence of detergent shows the tritium label to be associated with the two largest subunits, α and β. 2. In the absence of protecting ligands, the loss of ATP hydrolysis activity is linearly dependent on the level of pyridoxylation with complete inactivation correlating to 10 mol pyridoxamine phosphate incorporated per mol enzyme. Partial inactivation of F₁ with pyridoxal phosphate has no effect on either the K_m for ATP or the ability of bicarbonate to stimulate residual hydrolysis activity, suggesting a mixed population of fully active and fully inactive enzyme. 3. In the presence of excess magnesium, the addition of ADP or ATP, but not AMP, decreases the rate and extent of modification of F₁ by pyridoxal phosphate. The non-hydrolyzable ATP analog, 5′-adenylyl-β, γ-imidodiphosphate, is particularly effective in protecting F₁ against both modification and inactivation. Efrapeptin and P_i have no effect on the modification reaction. 4. Prior modification of F₁ with pyridoxal phosphate decreases the number of exchangeable nucleotide binding sites by one. However, pyridoxylation of F₁ is ineffective in displacing endogenous nucleotides bound at non-catalytic sites and does not affect the stoichiometry of P_i binding. 5. The ability of nucleotides to protect against modification and inactivation by pyridoxal phosphate and the loss of one exchangeable nucleotide site with the pyridoxylation of F₁ suggest the presence of a positively charged lysine residue at the catalytic site of an enzyme that binds two negatively charged substrates.     

Glycogen Phosphorylase and Synthase Activities of Rumen Ciliates of the Genus Entodinium

Glycogen phosphorylase and synthase activities were detected in the sonic lysate of rumen ciliates of the genus Entodinium. The ciliate phosphorylase had the following properties. The pH optimum was narrow and centered at pH 5.9. The activity was maximum at 30°C; above 40°C a rapid inactivation occurred. The K_m value for glucose-1-phosphate (G-1-P) and for glycogen was 15 mM and 0.069% (w/v), respectively. NaF and ethylenediamine tetraacetic acid had no stimulative effect on the enzyme activity, though adenosine 3′,5′-monophosphate and theophylline activated it. NaHSO₃ inhibited the enzyme activity at a concentration of 1 mM. The inhibition of glucose was noncompetitive for G-1-P. Glycolytic intermediates and nucleotides had a minor effect on phosphorylase activity. Glycogen synthase existed in two forms, glucose-6-phosphate dependent and independent forms: the proportion of the latter form increased with the decrease of reserve polysaccharide levels in the ciliates. Correlations between glycolytic enzyme activities included phosphorylase and synthase activities and reserve polysaccharide contents in the ciliates were determined, and a possible regulatory mechanism of polysaccharide synthesis and degradation was discussed.     

Reaction of pyridoxal 5'-phosphate with Escherichia coli CoA transferase: evidence for an essential lysine residue.

The acetyl-CoA:acetoacetate CoA-transferase of Escherichia coli was reversibly inactivated by pyridoxal 5′-phosphate. The residual activity of the enzyme was dependent on the concentration of the modifying reagent to a concentration of 5 mm. The maximum level of inactivation was 89%. Kinetic and equilibrium analyses of inactivation were consistent with a two-step process (Chen and Engel, 1975, Biochem. J.149, 619) in which the extent of inactivation was limited by the ratio of first-order rate constants for the reversible formation of an inactive Schiff base of pyridoxal 5′-phosphate and the enzyme from a noncovalent, dissociable complex of the enzyme and modifier. The calculated minimum residual activity was in close agreement with the experimentally determined value. The conclusion that the loss of catalytic activity resulted from modification of a lysine residue at the active site was based on the following data, (a) After incubation with 5 mm pyridoxal 5′-phosphate, 3.95 mol of the reagent was incorporated per mole of free enzyme with 89% loss of activity, while 2.75 mol of pyridoxal 5′-phosphate was incorporated into the enzyme-CoA intermediate with a loss of 10% of catalytic activity; the intermediate was formed in the presence of acetoacetyl-CoA; (b) acid hydrolysis of the modified, reduced enzyme-CoA intermediate yielded a single fluorescent compound that was identified as N⁶-pyridoxyllysine by chromatography in two solvent systems; (c) the enzyme was also protected from inactivation by saturating concentrations of free CoA and ADP but not by adenosine. The results suggested that a lysine residue is involved in the electrostatic binding of the pyrophosphate group of CoA. Carboxylic acid substrate did not protect the enzyme from inactivation.     

Evidence that the essential, photosensitive histidyl residue in the beta2 subunit of tryptophan synthetase is in the pyridoxyl peptide

The β₂ subunit of tryptophan synthetase of Escherichia coli is photoinactivated in the presence of pyridoxal 5′-phosphate and L-serine as a result of the destruction of one histidyl residue per chain (1). Two tryptic peptides are found in much lower amounts in the photoinactivated enzyme than in the control enzyme. These peptides have been identified from their amino acid composition as the 9 or 10 residue peptides which terminate with the lysyl residue which forms a Schiff base with pyridoxal 5′-phosphate. These peptides contain two histidyl residues, one of which appears to be photosensitive. Thus pyridoxal 5′-phosphate sensitizes the photooxidation of a nearby, essential histidyl residue.     

The effect of pH and temperature on the kinetics of native and altered glycogen phosphorylase.

Data on the effect of pH and temperature on the kinetics of rabbit muscle phosphorylases a and b and reduced phosphorylase b (α-1,4-glucan:orthophosphate glucosyltransferase, EC 2.4.1.1) with glycogen as the saturating and inorganic phosphate the variable substrate are presented. The kinetic profiles as a function of pH are similar for these enzyme species except that the positions of the pH-maximal velocity profiles for reduced phosphorylase b are relatively invariant in the 15 °–30 ° range, whereas the “native” phosphorylases exhibit a substantial shift of the lower pH limb of the profile toward the acid side when the temperature is lowered from 30 to 15 °C. It is proposed that a group with a pK near 6.0 at 30 °C determines the acid limb of maximal velocity profiles. The phosphoryl moiety of enzyme bound pyridoxal 5′-phosphate is suggested for this group. A conformational transition in the protein, which is somehow modified when the aldimine bond between protein and pyridoxal 5′-phosphate is reduced, is invoked to account for the large decrease of this acid side apparent pK for the ternary complex of native phosphorylases when the temperature is lowered. A group with a pK near 7.1 and a heat of ionization of about 8000 cal/mol determines the alkaline limb of maximal velocity profiles at 30 °C. An imidazoyl ring ionization of an enzyme histidyl group is proposed to account for this behavior. In the enzyme-glycogen binary complex, the apparent heat of ionization of this group has an anomalous value of about ?10,000 cal/ mol. It is suggested that a neighboring amino or arginyl guanidinium group is able to interact with the imidazoyl ring in the absence of bound inorganic phosphate to cause this anomalous behavior. The effect of pH on K_m for inorganic phosphate is simply explained by a group with a pK of 6.56 and low heat of ionization. The data are interpreted to indicate that the dianion of inorganic phosphate is the true substrate for all forms of phosphorylase. The kinetic results of this report are closely compared with other kinetic data in the literature on mammalian, plant, and bacterial α-glucan phosphorylases and general overall similarity is demonstrated. Various methods for analyzing pH-kinetic data for enzymes are briefly discussed, and the crucial difference in conclusions the choice of method can make is demonstrated with our data.     

A convenient lactic dehydrogenase-coupled assay for determining pyridoxal 5′-phosphate in plasma

An assay for determining the concentration of pyridoxal 5′-phosphate in plasma from 0.4 ml whole blood is reported. The assay consists of incubating deproteinized plasma with d-serine apodehydratase from Escherichia coli in 0.5 mN-2-hydroxyethyl-piperazine-N′-3-propanesulfonic acid, pH 7.8, at 37°C for 15 min, and then determining the d-serine dehydratase activity of an aliquot of the incubation mixture. A lactic dehydrogenase-coupled assay (at 25°C) was used to measure the rate of enzymically catalyzed conversion of D-serine to pyruvate, wherein depletion of NADH was followed continuously at 340 nm. The concentration of pyridoxal 5′-phosphate in the plasma sample was estimated from the enzymic activity which is a linear function of the amount of pyridoxal 5′-phosphate present in the assay.     

Preparation of pyridoxal 5'-phosphate-bound sepharose and its use for immobilization of tryptophanase

Pyridoxal 5′-phosphate-bound Sepharose (SP) was prepared by coupling pyridoxal 5′-phosphate (PLP) to diazotized p-aminobenzamidohexyl-Sepharose. A derivative of pyridoxine having an absorption maximum at $\text{ca.}$ 316 nm (possibly, 6-amino-pyridoxine 5′-phosphate) was liberated from SP by treatment with 0.1 M sodium dithionite at pH 9.0. SP catalyzed the cleavage of tryptophan in the presence of Cu²⁺, a typical non-enzymatic model of tryptophanase reaction. From the spectrophotometric data and catalytic activity, it was estimated that SP contained about 1.5 μmoles of bound PLP per gram of Sepharose. Tetrameric apotryptophanase was immobilized by incubation with SP, followed by reduction with NaBH₄. The resulting immobilized tryptophanase retained $\text{ca.}$ 60 % of the catalytic activity of free tryptophanase used. This method was much superior to other methods used commonly for preparation of immobilized enzymes.     

A novel coenzymatic function of pyridoxal 5'-phosphate

Pyridoxal 5′-phosphate, the vitamin B₆ derivative, acts as the coenzyme of many enzymes involved in amino acid metabolism. Exceptionally, this compound was found covalently bound to glycogen phosphorylase, the key enzyme in the regulation of glycogen metabolism. Although it is essential for the function of phosphorylase, its direct role has remained an enigma. We have recently found that the glucose moiety of pyridoxal (5′)diphospho (1)-α-D -glucose, a conjugate of pyridoxal 5′-phosphate and glucose 1-phosphate through a pyrophosphate linkage, is transferred to the nonreducing end of glycogen, forming a new α-1,4-glucosidic linkage. This finding emphasizes the importance of the direct phosphate-phosphate interaction between the coenzyme and the substrate in the phosphorylase catalytic reaction. We have proposed a catalytic mechanism for phosphorylase in which the phosphate group of pyridoxal 5′-phosphate acts as an electrophile to the phosphate group of glucose 1-phosphate. This appears to represent the first instance of the direct involvement of a phosphate group in catalysis by enzymes.     

Micelles of pyridoxal-5'-phosphate Schiff bases - an improved model for the B6 site of glycogen phosphorylase

The spectral properties of Schiff bases obtained by reaction of pyridoxal-5′-P with n-alkylamines (C_nNH₂) at neutral pH depend upon the length of the hydrocarbon chain of the amine. While short-chain amines (e.g. n = 4,8) yield a product with absorption maxima at 405 and 273 nm (similar to those reported for pyridoxal-5′-P Schiff bases in an aqueous medium), higher members in the n-alkylamine series (e.g. n = 12), which form micelles under the conditions of the experiment, yield a product with absorption maxima at 335 and 252 nm, similar to those of Schiff bases in apolar solvents. Mixed micelles composed of hexadecyltrimethylammonium bromide and n-dodecylamine hydrochloride were found to entrap stoichiometric amounts of pyridoxal-5′-P (one mole per mole of the primary amine) and to yield a Schiff base. The resulting micelles simulate several absorption, fluorescence, and chemical properties of phosphorylase at neutral pH. This micellar model (like the functioning enzyme molecule) puts the pyridoxal-5′-P Schiff base in a hydrophobic microenvironment within an aqueous medium.     

Evidence for the regulation of pyridoxal 5′-phosphate formation in liver by pyridoxamine (pyridoxine) 5′-phosphate oxidase

Pyridoxamine (pyridoxine) 5′-phosphate oxidase (EC 1.4.3.5) purified from rabbit liver is competitively inhibited by the reaction product, pyridoxal 5′-phosphate. The K_i, 3 μM, is considerably lower than the K_m for either natural substrate (18 and 24 μM for pyridoxamine 5′-phosphate and 25 and 16 μM for pyridoxine 5′-phosphate in 0.2 M potassium phosphate at pH 8 and 7, respectively). The K_i determined using a 10% rabbit liver homogenate is the same as that for the pure enzyme; hence, product inhibition $\text{in}\text{vivo}$ is probably not diminished significantly by other cellular components. Similar determinations for a 10% rat liver homogenate also show strong inhibition by pyridoxal 5′-phosphate. Since the reported liver content of free or loosely bound pyridoxal 5′-phosphate is greater than K_i, the oxidase in liver is probably associated with pyridoxal 5′-phosphate. These results also suggest that product inhibition of pyridoxamine-P oxidase may regulate the $\text{in}\text{vivo}$ rate of pyridoxal 5′-phosphate formation.     

Adenosine metabolism: Modification by S-adenosylhomocysteine and 5′-methylthioadenosine

To elucidate potential toxic properties of S-adenosylhomocysteine and 5′-methylthioadenosine, we have examined the inhibitory properties of these compounds upon enzymes involved with adenosine metabolism. S-Adenosylhomocysteine, but not S-adenosylmethionine, was a noncompetitive inhibitor of adenosine kinase with K_i values ranging from 100 to 400 μm. Methylthioadenosine competitively inhibited adenosine kinase with variable adenosine below 1 μm with a K_i of 120 μm, increased adenosine kinase activity when the adenosine concentration exceeded 2 μm, and did not appear to be a substrate for adenosine kinase. Methylthioadenosine inactivated S-adenosylhomocysteine hydrolase from erythrocytes, B-lymphoblasts, and T-lymphoblasts with K_i values ranging from 65 to 117 μm and “k₂” from 0.30 to 0.55 min^?1. Adenosine deaminase was not inhibited by 5′-methylthioadenosine up to 1000 μm. To clarify how 5′-methylthioadenosine might accumulate, 5′-methylthioadenosine phosphorylase was evaluated. This enzyme was not blocked by up to 500 μm adenosine, deoxyadenosine, S-adenosylhomocysteine, or S-adenosylmethionine and was not decreased in erythrocytes from patients with adenosine deaminase deficiency, purine nucleoside phosphorylase deficiency, or hypogammaglobulinemia. These observations suggest that the inhibitory properties of 5′-methylthioadenosine upon adenosine kinase and S-adenosylhomocysteine hydrolase may contribute to the toxicity of the exogenously added compound. The toxicity resulting from S-adenosylhomocysteine accumulation intracellularly may be related to adenosine kinase inhibition in addition to disruption of transmethylation reactions.