Properties of tetrahydropteroylpentaglutamate bound to 10-formyltetrahydrofolate dehydrogenase.

A new rapid procedure for purifying 10-formyltetrahydrofolate dehydrogenase results in 90 mg of pure enzyme from two rabbit livers. This abundant liver enzyme is known to bind its product tetrahydropteroylpentaglutamate (H4PteGlu5) so tightly that it does not dissociate during size exclusion chromatography. 10-Formyltetrahydrofolate dehydrogenase is also known to exhibit strong product inhibition by H4PteGlu5. There is a several-fold excess of 10-formyltetrahydrofolate dehydrogenase subunits in liver relative to the concentration of H4PteGlun, suggesting that in vivo this enzyme may bind significant amounts of this coenzyme in a nearly irreversible enzyme. H4PteGlu5 complex. How this tightly bound H4PteGlun is transferred to the other two enzymes in the cytosol, serine hydroxymethyltransferase and C1-tetrahydrofolate synthase, which use H4PteGlu5 as a substrate, is the subject of this investigation. Analysis of the product inhibition curve for 10-formyltetrahydrofolate dehydrogenase shows that H4-PteGlu5 has a dissociation constant near 15 nM which is 60-fold lower than the Ks for 10-formyl-H4PteGlu5. Fluorescence titration studies also yield a Kd of about 20 nM for H4PteGlu5. Coupling the 10-formyltetrahydrofolate dehydrogenase reaction to an excess of either serine hydroxymethyltransferase or C1-tetrahydrofolate synthase not only abolishes product inhibition but also increases the initial rate of its activity by about 2-fold. Passage of a reaction mixture of 10-formyltetrahydrofolate dehydrogenase down a size exclusion column results in enzyme with 1 equiv of H4PteGlu5 bound per subunit. However, addition of either serine hydroxymethyltransferase or C1-tetrahydrofolate synthase results in a rapid transfer of this bound folate to these enzymes. Evidence is also presented that the tightly bound folate is in equilibrium with solvent H4PteGlu5.

The affinity of pyridoxal 5'-phosphate for folding intermediates of Escherichia coli serine hydroxymethyltransferase.

Escherichia coli serine hydroxymethyltransferase is a 94-kDa homodimer. Each subunit contains a covalently attached pyridoxal-P, which is required for catalytic activity. At which step pyridoxal-P binds in the folding pathway of E. coli serine hydroxymethyltransferase is addressed in this study. E. coli serine hydroxymethyl-transferase is rapidly unfolded to an apparent random coil in 8 M urea. Removal of the urea initiates a complete refolding to the native holoenzyme in less than 10 min at 30 degrees C. Several intermediates on the folding pathway have been identified. The most important information was obtained during folding studies at 4 degrees C. At this temperature, the far-UV circular dichroism spectrum and the fluorescence spectrum of the 3 tryptophan residues become characteristic of the native apoenzyme in less than 10 min. Size exclusion chromatography shows that under these conditions the refolding enzyme is a mixture of monomeric and dimeric species. Continued incubation at 4 degrees C for 60 min results in the formation of only a dimeric species. Neither the monomer nor dimer formed at 4 degrees C bind pyridoxal phosphate. Raising the temperature to 30 degrees C results in the formation of a dimeric enzyme which rapidly binds pyridoxal phosphate forming active enzyme. These studies support the interpretation that pyridoxal phosphate binds only at the end of the folding pathway to dimeric apoenzyme and plays no significant role in the folding mechanism.

Domain structure and function of 10-formyltetrahydrofolate dehydrogenase.

10-Formyltetrahydrofolate dehydrogenase catalyzes the NADP(+)-dependent oxidation of 10-formyltetrahydrofolate to CO2 and tetrahydrofolate. Previous studies have shown that the enzyme binds the physiological pentaglutamate form of tetrahydrofolate product so tightly that it remains bound during size exclusion chromatography (Cook, R. J., and Wagner, C. (1982) Biochemistry 21, 4427-4434). In addition to the dehydrogenase activity, the enzyme from rat liver has been reported to exhibit both 10-formyltetrahydrofolate hydrolase and aldehyde dehydrogenase activities (Cook, R. J., Lloyd, R. S., and Wagner, C. (1991) J. Biol. Chem. 266, 4965-4973). We have purified the enzyme from rabbit liver and found that it catalyzes the same three reactions with similar kinetic constants and that it is a 99-kDa homotetramer, as reported previously for the rat and pig enzymes. Previous studies have suggested that the enzyme is composed of three domains and has separate folate binding sites for the dehydrogenase and hydrolase activities. We have investigated the domain structure of the rabbit enzyme. Differential scanning calorimetry reveals two thermal transitions, indicating the presence of two independently folded domains. The pentaglutamate form of tetrahydrofolate and NADP+ each stabilize one of the thermal transitions, showing that these ligands bind to separate domains. Limited proteolytic digestions by several proteases cleave the enzyme in a linker region between the two domains. After proteolytic cleavage, the domains no longer remain associated and do not catalyze the 10-formyltetrahydrofolate dehydrogenase reaction. Isolation and characterization of the intact domains revealed that the N-terminal domain only catalyzes the NADP(+)-independent 10-formyltetrahydrofolate hydrolase activity and the C-terminal domain only catalyzes the NADP(+)-dependent aldehyde dehydrogenase activity. The kinetic constants of these isolated domains are similar to those of the intact enzyme. Binding studies on the native enzyme using fluorescence and isothermal titration calorimetry indicated that the enzyme binds one molecule of tetrahydrofolate and two molecules of NADP+ per tetramer. Dissociation constants for both ligands were also determined by these methods.

Function of the active-site lysine in Escherichia coli serine hydroxymethyltransferase.

Serine hydroxymethyltransferase has a conserved lysine residue (Lys-229) that forms the internal aldimine with pyridoxal 5'-phosphate. In other pyridoxal 5'-phosphate enzymes investigated so far, this conserved lysine residue also plays a catalytic role as a base that removes the alpha-proton from the amino acid substrate. Three mutant forms of Escherichia coli serine hydroxymethyltransferase (K229Q, K229R, and K229H) were constructed, expressed, and purified. The absorbance spectra, rapid reaction kinetics, and thermal denaturation of the mutant analogs were studied. Only the K229Q mutant serine hydroxymethyltransferase resembled the wild-type enzyme. The results indicate that Lys-229 plays a critical role in expelling the product by converting the external aldimine to an internal aldimine. In the absence of Lys-229, ammonia can also catalyze the same function at a much slower rate. However, Lys-229 apparently is not the base that removes the alpha-proton from the amino acid substrate. The K229Q mutant enzyme could catalyze one turnover of either serine to glycine or glycine to serine at rates approaching those of the wild-type enzyme. After one turnover, the mutant enzyme could not expel the product and bind new substrate. The K229Q mutant enzyme can also transaminate D-alanine, which, like the hydroxymethyltransferase activity, also requires removing the alpha-proton from the substrate. The absorbance spectra of the K229R and K229H serine hydroxymethyltransferases showed that their pyridoxal 5'-phosphate could not readily form an external aldimine with substrates, suggesting that Lys-229 in the wild-type enzyme may never bear a positive charge, further evidence that it is not the base that removes the alpha-proton.

Rat brain hexokinase: amino acid sequence at the substrate hexose binding site is homologous to that of yeast hexokinase.

A reactive Glc analog, N-(bromoacetyl)-D-glucosamine (GlcNBrAc), has recently been used (D. M. Schirch and J. E. Wilson (1987) Arch. Biochem. Biophys. 254, 385-396) to label the Glc binding site of rat brain Type I hexokinase. This site has been located in a 40-kDa domain at the C-terminus of the enzyme previously shown to be the location of the substrate ATP binding site (M. Nemat-Gorgani and J. E. Wilson (1986) Arch. Biochem. Biophys. 251, 97-103). In the present study, peptide mapping of hexokinase modified by radiolabeled GlcNBrAc yields three labeled peptides (Peptides I-III). Peptides I and III, as well as catalytic activity, are protected by inclusion of Glc or GlcNAc during reaction with GlcNBrAc. These two peptides show considerable homology to contiguous regions in the sequences of yeast hexokinase isozymes A and B. Peptide III is homologous to a sequence which, based on the X-ray crystallographic work by Steitz and co-workers, is located near the Glc binding site of yeast hexokinase; Peptide I is homologous to an immediately adjacent (toward the C-terminus) region of yeast hexokinase. An essential serine residue implicated in the binding of Glc to the yeast enzyme is also conserved in Peptide III from rat brain hexokinase. These results provide strong support for the view that the "catalytic domain" at the C-terminus of the mammalian Type I hexokinase shares a common ancestry with yeast hexokinase. Peptide II appears to be nonspecifically labeled by GlcNBrAc since labeling is insensitive to the presence of protective ligands such as Glc or GlcNAc; the sequence of Peptide II shows no detectable homology with the yeast isozymes.

Rat brain hexokinase: location of the substrate hexose binding site in a structural domain at the C-terminus of the enzyme.

A glucose analog, N-(bromoacetyl)-D-glucosamine (GlcNBrAc), previously used to label the glucose binding sites of rat muscle Type II and bovine brain Type I hexokinases, also inactivates rat brain hexokinase (ATP:D-hexose 6-phosphotransferase, EC 2.7.1.1) with pseudo-first-order kinetics. Inactivation occurs predominantly via a "specific" pathway involving formation of a complex between hexokinase and GlcNBrAc, but significant nonspecific (i.e., without prior complex formation) inactivation also occurs, and equations to describe this behavior are derived. Inactivation is dependent on deprotonation of a residue with an alkaline pKa, consistent with the modified residue being a sulfhydryl group as reported to be the case with the hexokinase of bovine brain. The affinity label modifies three residues (per molecule of enzyme) at indistinguishable rates, but only one of these residues appears to be critical for activity. Amino acid analysis of the modified enzyme indicates derivatization of three cysteine residues; there was no indication of modification of other residues potentially reactive with haloacetyl derivatives. Kinetic analysis and effects of protective ligands were consistent with location of the critical sulfhydryl at the glucose binding site. Peptide mapping techniques permitted localization of the critical residue, and thus the glucose binding site, in a 40-kDa domain at the C-terminus of the enzyme. This is the same domain recently shown to include the ATP binding site. Thus, catalytic function is assigned to the C-terminal domain of rat brain hexokinase.

Serine hydroxymethyltransferase. Effect of proteases on the activity and structure of the cytosolic enzyme.

Homogeneous preparations of cytosolic serine hydroxymethyltransferase from rabbit liver were incubated with several different proteases. Chymotrypsin rapidly cleaves a tetradecapeptide from the NH2-terminal end of the enzyme with the enzyme retaining full catalytic activity. Trypsin digestion results in the release of several small peptides from the NH2-terminal end of the enzyme. The remaining core protein is reduced in molecular mass by about 3500 Da. With L-serine as substrate the core protein has 1.5 times the activity of the native enzyme. The difference in activity is due to a change in Vmax since the Km values for L-serine and tetrahydrofolate are unchanged. When allothreonine is used as the substrate the activity of the trypsin-treated enzyme is unchanged. Ks values for glycine and several folate compounds are also unchanged for the trypsin-digested enzyme. The relative distribution of three glycine-enzyme complexes shows only small differences between the native and trypsin-digested enzyme. Thermal denaturation studies show that the trypsin-digested enzyme has a thermal transition three degrees lower than the native enzyme but the same enthalpy of denaturation. These results suggest that the 25-30 amino acid residues from the NH2-terminal end of the enzyme are not important in determining the catalytic activity and structural stability of the purified enzyme. Several other proteases had no observable effect on the activity and size of the enzyme. All of the proteases tested inactivated the apoenzyme and digested it into small fragments. The loss of enzyme activity in frozen liver is probably the result of the enzyme slowly being converted to the apoenzyme form, which is susceptible to protease degradation.