Robert T. Taylor

Folate-dependent enzymes in cultured Chinese hamster cells: Folylpolyglutamate synthetase and its absence in mutants auxotrophic for glycine + adenosine + thymidine☆

Abstract Dialyzed sonicates from Chinese hamster ovary (CHO) and V-79 lung cells catalyze the addition of l -[U- 14 C]glutamate to tetrahydrofolate (H 4 PteGlu). Catalysis is optimal between pH 8.5 and 10.2 and is dependent on Mg 2+ and a purine nucleotide triphosphate. Cobalamins do not stimulate the system even when the cells are grown in the absence of cyanocobalamin (CN-Cbl). Incubations with dl -H 4 -[G- 3 H]PteGlu + l -[U- 14 C]glutamate show that the product routinely assayed by DEAE-cellulose chromatography is tetrahydropteroyldiglutamate (H 4 PteGluGlu). Higher reduced folylpolyglutamates are formed when the standard assay level of dl -H 4 PteGlu is decreased from 100 μ m to 1–5 μ m . Using either dialyzed extracts or a 25-fold purified enzyme fraction, dATP is 1.6 times more effective than ATP. The folyl specificity for diglutamate synthesis is H 4 PteGlu > H 4 -homofolate > 5-formyl-H 4 PteGlu > 5-MeH 4 PteGlu. dl -5-MeH 4 PteGlu is only about 15% as active as dl -H 4 PteGlu. Extracts from a CHO mutant AUXB1 (requiring glycine + adenosine + thymidine) and a V-79 mutant ght-1 (requiring glycine + hypoxanthine + thymidine) have 4 PteGluGlu synthetase activity. CHO AUXB1 and V-79 ght-1 extracts are also inactive with the other three reduced folyl compounds cited above and PteGlu. Twelve out of 16 revertant clones that were isolated from CHO AUXB1 in media lacking glycine + adenosine + thymidine contained 44–66% of the wild-type level of H 4 PteGluGlu synthetase activity. Both parent CHO and V-79 extracts catalyzed the conversion of H 4 PteGluGlu and tetrahydropteroyl triglutamate to higher glutamyl conjugates. The AUXB1 and ght-1 mutant extracts again lacked these catalytic properties. In contrast, revertants of AUXB1 with about 50% of the wild-type H 4 PteGluGlu synthetase activity displayed a proportionate ability to synthesize higher polyglutamyl conjugates. From our findings and published genetic data, we conclude that in cultured hamster cells a single synthetase can successively add at least three glutamates to H 4 PteGlu. Loss of its function in certain mutants is responsible for their triple auxotrophy.

Aerobic photoiysis of alkylcobalamins: Quantum yields and light-action spectra

Quantum yields (φ) for the aerobic photolysis of 5′-deoxyadenosylcobalamin (dAB12), methylcobalamin (MeB12), propylcobalamin (PrB12), and ethylcobalamin (EtB12) were determined as a function of the irradiation wavelength. φ Determinations were made for both the base-on and base-off forms of each compound (except base-off dAB12) at incident wavelengths from 250 nm to 570 nm. As a rule, the φs were high (0.1–0.5) and they varied significantly with respect to the irradiation wavelength. In general, each alkylcobalamin at pH 7.0 displayed a quantum yield spectrum distinct from its base-off form at pH 1.0. Across most of the spectrum, the φs of the base-off form were appreciably smaller than the base-on φs of the same compound. An exception to this generality was MeB12 for which the φs at pH 1.0 were about the same as, or slightly greater above 450 nm than those at pH 7.0. At pH 7.0 and in the visible region the trend of the φs was dAB12 < MeB12 < PrB12 < EtB12. Under neutral conditions each compound showed a broad quantum yield peak in the 450–470 nm region. From the quantum yield and absorption spectra, photolysis spectra were calculated for 5.0 × 10−5m solutions of each compound. The light-action spectra accurately give the relative rates/μ Einstein that these solutions photolyze at each wavelength. Thus, for example, MeB12 photolyzed faster at pH 7.0 versus pH 1.0 in 510 nm light, but it photolyzed slower at pH 7.0 versus pH 1.0 in 450 nm light. Solutions of each compound photolyzed faster in the ultraviolet region as opposed to the visible (e.g., 310 nm versus 510 nm). Our findings show that the previously reported photolysis rates estimated by others with tungsten lamps provide no valid information about the intrinsic photolability of various alkyl-cobalt bonds. This also applies to the relative white-light photolysis rates reported for the base-on versus the base-off form of MeB12. All such relative rates are artifacts which represent only the extent of overlap between the true action spectrum and the light emission spectrum of an incandescent lamp.

Escherichia coli B 5-methyltetrahydrofolate-homocysteine cobalamin methyltransferase: Resolution and reconstitution of holoenzyme

Abstract Extensively purified E. coli B cobalamin methyltransferase was resolved into an apoenzyme plus an unbound, 1-electron-reduced, cobalamin (B 12r ) upon incubation for 15 min at 37 ° with 6.0 m urea containing dithiothreitol (DTT). Apoenzyme which had been prepared by urea resolution was reconstituted 75–80% into a radioactive holoenzyme complex upon a 5-min incubation at 37 ° with 10 nmoles of either methyl- 14 CB 12 or propyl- 14 CB 12 per mg of protein. Reconstituted methyl- 14 CB 12 enzyme was light-stable (100 W tungsten lamp, 0 °) unless acidified and was almost completely demethylated by homocysteine. Incubation of resolved apoenzyme with deoxyadenosyl-B 12 , hydroxy-B 12 , cyano-B 12 , B 12r , diaquocobinamide, and methylcobinamide resulted in negligible holoenzyme formation. However, under the same conditions sulfito-B 12 yielded 48–52% holoenzyme. Urea-prepared apoenzyme differed from the purified B 12 holoenzyme with respect to its sedimentation coefficient. Based on a sedimentation coefficient of 7.0 S for the purified B 12 holoenzyme the apoenzyme sedimented at a slightly slower rate with a relative coefficient equal to 6.2 S . Binding methyl-B 12 ( 3 H) to the apoenzyme yielded a reconstituted methyl-B 12 ( 3 H) holoenzyme with a sedimentation coefficient of 7.0 S . Unpurified holoenzyme and unpurified apoenzyme in cell-free, sonic extracts also centrifuged differently, like purified holoenzyme and urea-resolved apoenzyme, respectively.

Folate-dependent enzymes in cultured Chinese hamster ovary cells: Induction of 5-methyltetrahydrofolate homocysteine cobalamin methyltransferase by folate and methionine

Abstract The effects of media vitamin B 12 (CNB 12 ), l -methionine, folic acid, dl-5-methyltetrahydrofolate (5-MeH 4 folate), homocysteine, and other nutrients on four one-carbon enzymes in cultured Chinese hamster ovary (CHO) cells were examined. Excess 10 m m methionine elevates the amount of B 12 methyltransferase 1.8 – 2.3-fold at media folate concentrations of 0.2 – 2.0 μ m . Conversely, excess 100 μ m folic acid increases the amount of B 12 holoenzyme by 2.4 – 3.0-fold when the medium contains 0.01 – 0.1 m m methionine. These increases in B 12 methyltransferase promoted by 100 μ m media folate and 10 m m methionine are inhibited by cycloheximide. 5-MeH 4 folate will support growth and induce methyltransferase synthesis more efficiently than folic acid. Upon transfer to methionine-free media, wild-type CHO cells will survive and can be repeatedly subcultured in the absence of exogenous methionine, provided it is supplemented with 1.0 μ m CNB 12 , 0.1 m m homocysteine, and 100 μ m folic acid or 10 μ m dl-5-MeH 4 folate. No growth occurs if homocysteine is omitted, but a requirement for added CNB 12 does not become evident until the cells have undergone at least two or three divisions. Survival upon transfer from 0.1 m m methionine-containing to methionine-free media is dependent upon the B 12 holomethyltransferase content of the cells used as an inoculum. Inoculum cells must have been previously grown in media supplemented with 1.0 μ m CNB 12 to stabilize and convert apo- to holomethyltransferase, and 100 μ m folate (or 10 μ m dl-5-MeH 4 folate) to induce maximal enzyme-protein synthesis. Transfer to methionine-deficient medium does not result in more than a 20–25% increase in the cellular B 12 enzyme content over the level already induced by 100 μ m folate in 0.1 m m methionine-supplemented media. A mutant auxotroph CHO AUXB1 with a triple growth requirement for glycine + adenosine + thymidine (McBurney, M. W., and Whitmore, G. F. (1974) Cell , 2 , 173) cannot survive in media lacking exogenous methionine. High concentrations of media folic acid or dl-5-MeH 4 folate fail to induce elevated amounts of B 12 methyltransferase in this mutant. Excess 10 m m medium methionine does, however, elevate its B 12 enzyme as in the parent CHO cells. An additional mutant AUXB3 that requires glycine + adenosine (McBurney, M. W., and Whitmore, G. F. (1974) Cell , 2 , 173) barely survives in methionine-deficient media. It has a folate-induced B 12 enzyme level intermediate between wild-type CHO cells and AUXB1. The level of B 12 methyltransferase induced by high media folate concentrations is a critical determinant of CHO cell survival in methionine-free media.

5-methyltetrahydrofolate homocysteine cobalamin methyltransferase in human bone marrow and its relationship to pernicious anemia.

Abstract Dialyzed extracts from human bone marrow catalyze [5- 14 C]methyltetrahydrofolate homocysteine transmethylation at slow but significant rates which can be detected by using substrate with a very high specific radioactivity. Enzymatic activity is associated with nucleated marrow cells rather than mature, nondividing erythrocytes. Extract transmethylase activities in 15 marrow specimens from patients without B-12 deficiency ranged from 157–1020 pmoles of [ Me - 14 C]methionine formed/hr/10 7 nucleated cells. Catalysis is dependent on S -adenosyl- l -methionine and a flavin-reducing system, typical for the presence of a cobalamin (B-12) methyltransferase. No in vitro requirement for exogenous B-12 was observed except for the marrow extracts from two patients known to be B-12 deficient. One of these extracts was markedly stimulated by methyl-B-12 indicative that mostly apomethyltransferase was present. These tracer assays with cell-free extracts provide the first direct evidence that human bone marrow contains B-12 methyltransferase; they also afford further evidence for a 5-methyltetrahydrofolate trap in B-12 deficiency with its associated megaloblastic anemia. In addition, we have observed that in normal peripheral blood leukocytes the mononuclear fraction contains 10–30 times as much B-12 methyltransferase per nucleated cell as the polymorphonuclear granulocyte fraction.

5-Methyltetrahydrofolate aromatic alkylamine N-methyltransferase: an artefact of 5,10-methylenetetrahydrofolate reductase activity.

Abstract A previously reported stimulation of brain 5-methyltetrahydrofolate (5-MeH 4 -folate) N-methyltransferase by FAD and methylcobalamin (MeB 12 is attributed to their roles as nonspecific electron acceptors. Evidence is presented that the catalyst involved is not an aromatic alkylamine methyltransferase, but the widely distributed enzyme, 5,10-methyleneH 4 -folate reductase. In the presence of an electron acceptor it catalyzes the oxidation of [5- 14 C]MeH 4 -folate to [5,10- 14 C]methyleneH 4 -folate which equilibrates to yield dimedone reactive H 14 CHO. The material being measured when incubation systems containing β-phenylethylamine or tryptamine are extracted with tolueneisoamyl alcohol is a condensation product of the H 14 CHO and the aromatic alkylamine. The aromatic alkylamine is not a co-substrate in the enzymic oxidation mechanism. It is required to react nonenzymically with reductase formed H 14 CHO and render it extractable. Our failure and that recently of others to detect significant N-methylation using [5- 14 C]MeH 4 -folate as a Me group donor make the existence of a folate-biogenic amine methyltransferase seem highly improbable.

Platinum-induced mutations to 8-azaguanine resistance in Chinese hamster ovary cells.

6 platinum (Pt) compounds were compared in suspension cultured Chinese hamster ovary (CHO-S) cells with respect to their inhibition of growth, their reduction of cloning efficiency, and their induction of mutants resistant to 200 microM (30 micrograms/ml) 8-azaguanine (8-AG) and 3 mM ouabain (OUA), respectively. The toxicity of these compounds can be ranked by the medium concentrations which decrease suspension growth/or cloning efficiency by 50%: cis-Pt(NH3)2-Cl2 (0.9/1.5 microM) greater than Pt(SO4)2 + methylcobalamin (MeB-12) methylation product (20/10 microM) greater than K2PtCl4 (32/50 microM) = K2PtCl6 (34/50 microM) = MePtCl2-3 (60/50 microM) greater than Pt(SO4)2 (66/105 microM). Following 20 h exposures to concentrations which resulted in relative survivals of 80-2%, none of the foregoing compounds increased consistently the frequency of OUA(R) mutants above the spontaneous frequency (6.0 x 10(-6)). Parallel treatments with 800 microM (100 micrograms/ml) ethyl methanesulfonate (EMS) increased the OUA(R) mutant frequency 10--12-fold. Using 8-AG for mutant selection, dose-dependent increases of 5--7-fold above the spontaneous frequency (3--8 x 10(-5) were obtained with cis-Pt(NH3)2Cl2, Pt(S04)2, and the product from Pt(SO4)2 + MeB-12. Identical 20 h exposures to varying amounts of K2PtCl4, K2PtCl6, and MePtCl2-3 did not induce 8-AG(R) mutants. Optimal detection of Pt-induced 8-AG(R) mutants required 7 post-treatments, expression doublings in suspension culture. Under our selection conditions 8/8 spontaneous and 24/24 Pt-induced 8-AG(R) variants contained reduced hypoxanthine-guanine phosphoribosyl transferase (HGPRT) specific activities (means ranging from 3 to 11% of the parental CHO-S cells). When compared from linear plots of the 8-Ag(r) frequency against the initial medium concentration, cis-Pt(NH3)2Cl2 is 134 times and Pt(SO4)2 si 3.5 times more mutagenic than EMS. However, on a cell-survival basis EMS is 8--10-fold more mutagenic than these two Pt-compounds. 6-Thioguanine (10 microM) can be substituted for 8-AG to assay mutant induction by cis-Pt(NH3)2Cl2 and Pt(SO4)2 in CHO-S cells. The sensitivity of the CHO-S HGPRT locus for detecting mutagenesis by Pt complexes can be increased several fold by continuous subculture in the presence of these agents for 10--25 population doublings. By this procedure K2PtCl6 is seen to be weakly mutagenic and 20 microM Pt(SO4)2 produces 8-AG(R) mutants at frequencies requiring 7--8-fold higher concentrations when a fixed 20 h exposure is used.

Uptake of cyanocobalamin by Escherichia coli B: Some characteristics and evidence for a binding protein☆

Abstract Tritiated cyanocobalamin ([3H]cyano-B12) uptake was maximal at pH 6.0–8.0, required phosphate buffer, was proportional to the cell concentration, was inhibited reversibly by Tris+, and was markedly temperature dependent (optimal, 37–47 °). Starved cells showed a 3–4-fold dependency on glucose and their glucose-dependent uptake was decreased by certain inhibitors of energy metabolism. However, the primary role of glucose in uptake may be to maintain cell wall structure rather than to supply chemical energy for transport per se. The Km for [3H]cyano-B12 uptake was 1.0 × 10−8 m and the Vmax was 2.3 pmoles/min/mg of dry cell wt. Unchanged [3H]-cyano-B12 is the major cobalamin after 5 min of uptake. Treatment with Tris-EDTA reduced the initial uptake rate by 40–50%, but osmotic shock decreased it an additional 5–8-fold. No detectable 5-methyltetrahydrofolate-homocysteine B12 methyltransferase was released by osmotic shock. However, treatment with Tris-EDTA alone, as well as osmotic shock, did release from the cells a heat-labile, ammonium sulfate-precipitable, binding protein. By gel filtration, its molecular weight and Stokes radius were 22,000 and 15.6 A, respectively. A very large binder (mol wt >200,000) was also found in Tris-EDTA and shock fluids.The 4 ° dissociation constant for the reversible complexes between [3H]cyano-B12 and both binders was 0.6 × 10−8 m . At 31 ° it was 0.5 × 10−8 m .

Binding of aquocobalamin to the histidine residues in bovine serum albumin

Abstract Various parameters which affect the binding of aquocobalamin (aquo-B12) to bovine serum albumin (BSA) were examined. Binding at pH 7.4 was markedly dependent on the time of incubation, the temperature and the concentration of aquo-B12. Aquo-B12 was bound to both the monomer and the dimer fractions in the BSA preparation and was not influenced by p-chloromercuribenzoate titration of the SH group. Approximately 70% of the bound aquo-B12 dissociated at pH 4.0 and all of it was readily removed with KCN. The tight binding of aquo-B12 at pH 7.4 was independent of the phosphate buffer concentration but was increased 2-fold in the presence of 5 m urea. Among seven other corrinoid compounds which were tested only diaquocobinamide formed a stable complex with BSA. Isolated complexes containing 0.6–2.6 moles of B12/mole of BSA were red-violet in color. They closely mimicked the complex formed between aquo-B12 and imidazole (or 1-methyl imidazole) in their absorption and their circular dichroism spectra. In addition, the pH curve for the binding of aquo-B12 to BSA was like that for the formation of an aquo-B12-imidazole complex. Equilibrium dialysis indicated that the maximum number of BSA combining sites was 16. From the above findings it was concluded that the binding reaction of aquo-B12 to BSA occurs between the sixth coordinating position of the cobalt and the imidazole side chains.

Acute toxicity of methanol in the folate-deficient acatalasemic mouse.

Formate acidosis is the chief measurable biochemical characteristic of acute methanol toxicity in man. Its marked elevation in the blood stream of primates has been proposed to account for their much greater susceptibility versus rodents to methanol poisoning. Therefore, a study was undertaken to assess whether folic acid deficient (FAD) mice which accumulate formate are much more sensitive to the lethal effects of this alcohol than folic acid sufficient (FAS) mice. Moreover, because some formate is oxidized by catalase-H2O2 in rodents, but not in primates, we also compared the urinary excretion and blood plasma accumulation of formate and the methanol sensitivity of acatalasemic mice. Methanol-dosed C57BL/6Csb (acatalasemic) mice exhibit slightly lower LD50S than CSa (normal catalase) mice, irrespective of their folate state. CSb-FAD mice excreted much more formate and developed higher plasma formate concentrations (11-17 mM) than identically dosed CSa-FAD animals (6 mM). However, in no instance did a folate deficiency produce a large reciprocal decrease in the oral or i.p. LD50 that would be expected from a huge increase (greater than 10-fold) in the 24-h blood plasma formate level. A low methionine (0.2%) intake did not decrease the oral methanol LD50 of CSb-FAD mice, although excess dietary methionine (1.8%) did lower it from 7.1 to 6.4 g/kg. Methanol treated (4 g/kg) Csb-FAD mice excreted 30.8-48.2% of the oral dose as urinary formate, depending on the level of dietary methionine. Csb-FAS and -FAD mice which were given 2 g/kg sodium formate orally (LD50 = 4.7 and 3.7 g/kg) cleared this dose from the blood within 24 h and excreted 58% and 76% of it, respectively, in the urine. Our results indicate that the plasma formate concentration does not correlate well with methanol lethality in Csb-FAS vs. -FAD mice. In addition, urinary excretion, not oxidation, is the primary means by which mice, and probably rats, eliminate high levels of blood formate. Since the Csb-FAD mouse attains high plasma formate levels and low blood pH-values similar to those which have been reported for methanol poisoned monkeys, it appears to be of value as an inexpensive small animal model for further studies of lethal methanol toxicity and the contribution of formate to this process.