"Wages of fear": transient threefold decrease in intracellular ATP level imposes apoptosis

In HeLa cells, complete inhibition of oxidative phosphorylation by oligomycin, myxothiazol or FCCP combined with partial inhibition of glycolysis by DOG resulted in a steady threefold decrease in the intracellular ATP level. The ATP level recovers when the DOG-containing medium was replaced by that with high glucose. In 48 h after a transient (3 h) [ATP] lowering followed by recovery of the ATP level, the majority of the cells commits suicide by means of apoptosis. The cell death does not occur if DOG or an oxidative phosphorylation inhibitor was added separately, treatments resulting in 10-35% lowering of [ATP]. Apoptosis is accompanied by Bax translocation to mitochondria, cytochrome c release into cytosol, caspase activation, reactive oxygen species (ROS) generation, and reorganization and decomposition of chromatin. Apoptosis appears to be sensitive to oncoprotein Bcl-2 and a pancaspase inhibitor zVADfmk. In the latter case, necrosis is shown to develop instead of apoptosis. The cell suicide is resistant to cyclosporine A, a phospholipase inhibitor trifluoroperazine, the JNK and p38 kinase inhibitors, oligomycin, N-acetyl cysteine and mitoQ, differing in these respects from the tumor necrosis factor (TNF)- and H(2)O(2)-induced apoptoses. It is suggested that the ATP concentration in the cell is monitored by intracellular "ATP-meter(s)" generating a cell suicide signal when ATP decreases, even temporarily, below some critical level (around 1 mM).     

The amounts of adenosine di- and triphosphates bound to H-meromyosin and the adenosinetriphosphatase activity of the H-meromyosin-F-actin-relaxing protein system in the presence and absence of calcium ions. The physiological functions of the two routes of myosin adenosinetriphosphatase in muscle contraction.

The rates of the ATPase [EC 3.6.1.3] reaction of the H-meromyosin-F-actin-relaxing protein system were measured in 2 mM MgCl2, 50mM KC1, and 10mM Tris-HC1 at pH 7.8 and 20 degrees in the presence and absence of 0.05-0.1 mM Ca2+ ions. The concentrations of H-meromyosin (HMM) and the F-actin-relaxing protein (F-A-PR) complex were 3.4 and 3 mg/ml, respectively, and the ATPase reaction was coupled with 4 mg/ml of pyruvate kinase [EC 2.7.1.40] and 1 or 20 mM phosphoenolpyruvate to regenerate ATP. The amount of ADP bound to HMM during the ATPase reaction was determined by measuring the amount of ADP remaining in the reaction mixture. The amount of ATP bound to HMM was determined by subtracting the amount of bound ADP from the total amount of nucleotides bound to HMM, which was measured by a rapid flow-dialysis method. The following results were obtained. 1. The ATPase activity of the HMM-F-A-RP system increased linearly with increase in the amount of ATP added, and was independent of the presence of 0.05 mM Ca2+, when the amount of ATP added was less than 1 mole/mole of HMM. In the presence of 0.05 mM Ca2+, the ATPase activity reached a maximal level when 1.2-1.5 mole of ATP was added per mole of HMM, and maintained this level even at 3 moles of added ATP/mole of HMM. In the presence of 3mM EGTA, the ATPase activity decreased with increase in the amount of ATP added, from 1.5 to 3 moles of ATP/mole of HMM, and reached the level of the HMM ATPase reaction at 3 moles of added ATP/mole of HMM. Similar results were observed when the concentration of HMM was maintained at 3.4 mg/ml and the concentration of the F-A-RP complex was decreased from 3 to 1 or 0.5 mg/ml.     

Involvement of oxyleghaemoglobin and cytochrome P-450 in an efficient oxidative phosphorylation pathway which supports nitrogen fixation in Rhizobium.

Cellular ATP level, ATP/ADP ratio and nitrogenase activity rise when oxyleghaemoglobin is added to respiring suspensions of Rhizobium japonicum bacteroids from soybean root nodules. Increased gaseous O2 tension is much less efficient than oxyleghaemoglobin in stimulation of bacteroid ATP production. Studies with the inhibitor carbonyl cyanide m-chlorophenylhydrazone show this ATP to be generated as a consequence of oxidative phosphorylation. N-Phenylimidazole, a specific cytochrome P-450 inhibitor, also lowers the efficiency of bacteroid oxidative phosphorylation. An approximately linear relationship is observed between ATP/ADP ratio and nitrogenase activity as N-phenylimidazole concentration is lowered. It is suggested that cytochrome P-450 is a component of the leghaemoglobin-facilitated respiration pathway and that it may act as intracellular O2 carrier rather than terminal oxidase. A less efficient oxidase appears to function when cytochrome P-450 is inhibited.     

Enzymatic studies on the interaction of myosin and heavy meromyosin with 1,N 6 -ethenoadenosine triphosphate ( ATP), a fluorescent analog of ATP

The fluorescent analog of adenosine triphosphate (ATP)¹ 1,N⁶-ethenoadenosine triphosphate, (εATP), has been utilized as a substitute for ATP in the myosin and heavy meromyosin ATPase systems. For myosin, the analog εATP replaced ATP with a somewhat larger K_m (2.6 × 10^?4 mole ?^?1 for εATP as opposed to 8.8 × 10^?5 mole ?^?1 for ATP), indicating that the apparent affinity of the enzyme for εATP is less than for ATP. Perhaps of more interest, further comparison yielded a V_max for εATP about two and one half times the value for ATP (20 μmole PO₄ sec^?1 g protein^?1 as opposed to 8.1 μmole sec^?1 g protein^?1). Results for the HMM-εATPase system were similar, yielding a K_m value of 1.47 × 10^?4 mole ?^?1 and a V_max of 54.2 μmole PO₄ sec^?1 g protein^?1, as opposed to corresponding K_m and V_max values of 1.23 × 10^?4 mole ?^?1 and 20.4 μmole PO₄ sec^?1 g protein^?1, respectively for the HMM-ATP interaction. The pH dependence of εATPase for both systems was comparable to ATP, suggesting a similarity in the mechanism of hydrolysis of the two nucleotides. Activation of εATPase by Ca²⁺ in the presence of 0.5 M KCl was comparable to ATPase for both systems, but inhibition by Mg²⁺ seemed to be more effective for εATPase. These results indicate that εATP is an excellent substitute for ATP in the myosin and heavy meromyosin systems and because of its insertion into the active site of these muscle proteins, it promises to be a very useful probe for conformation studies at this level.     

Involvement of oxyleghaemoglobin and cytochrome P-450 in an efficient oxidative phosphorylation pathway which supports nitrogen fixation in Rhizobium

Cellular ATP level, ATP/ADP ratio and nitrogenase activity rise when oxyleghaemoglobin is added to respiring suspensions of Rhizobium japonicum bacteroids from soybean root nodules. Increased gaseous O₂ tension is much less efficient than oxyleghaemoglobin in stimulation of bacteroid ATP production. Studies with the inhibitor carbonyl cyanide m-chlorophenylhydrazone show this ATP to be generated as a consequence of oxidative phosphorylation. N-Phenylimidazole, a specific cytochrome P-450 inhibitor, also lowers the efficiency of bacteroid oxidative phosphorylation. An approximately linear relationship is observed between ATP/ADP ratio and nitrogenase activity as N-phenylimidazole concentration is lowered. It is suggested that cytochrome P-450 is a component of the leghaemoglobin-facilitated respiration pathway and that it may act as intracellular O₂ carrier rather than terminal oxidase. A less efficient oxidase appears to function when cytochrome P-450 is inhibited.     

Simplified method for estimation of microbial activity in compost by ATP analysis

A simplified method using ATP analysis was proposed to estimate the time course of microbial activity during composting. A compost sample was suspended in distilled water and its filtrate was used to estimate the ATP content in the compost by the luciferine-luciferase fluorescence reaction. The method permitted the rapid and simple determination of ATP in the compost and was successfully applied to monitor the time course of the microbial activity in the laboratory-scale composting process. The analytical simplicity in this method greatly improved the field-applicability of the ATP analysis for the composting process monitoring.     

Fate of nucleotides bound to reconstituted Fo-F1 during adenosine 5'-triphosphate synthesis activation or hydrolysis: role of protein inhibitor and hysteretic inhibition

The protein ATPase inhibitor entraps about five nucleotides in pig heart mitochondrial F1, one at least being a triphosphate [Di Pietro, A., Penin, F., Julliard, J.H., Godinot, C., & Gautheron, D.C. (1988) Biochem. Biophys. Res. Commun. 152, 1319-1325]. The fate of these nucleotides was studied during ATP synthesis driven by NADH oxidation in reconstituted inverted submitochondrial particles. Iodinated F1, containing 0.7 mol of endogenous nucleotides/mol, was first loaded with tritiated adenine nucleotides in the presence or absence of the protein inhibitor and then reassociated with F1-depleted submitochondrial particles (ASU particles) to reconstitute an efficient NADH-driven ATP synthesis. In the absence of the protein inhibitor, 1.7 mol of labeled nucleotides remained bound per mole of reassociated F1, 0.8-0.9 mol being rapidly exchangeable against medium ADP or ATP, as measured after rapid filtration through nitrocellulose filters. In the presence of the protein inhibitor, as many as 3.25 mol of labeled nucleotides remained bound per mole of reassociated F1. Under hydrolysis conditions where ATPase activity was highly inhibited, no release of tritiated nucleotide occurred. In contrast, under ATP synthesis conditions where the protonmotive force was generated by NADH oxidation, the progressive reversal of inhibition by the protein inhibitor was correlated to a concomitant release of tritiated nucleotide. When ATP synthesis became fully active, about one nucleotide was completely exchanged whereas more than three nucleotides remained tightly bound and did not appear to be directly involved in ATP synthesis.(ABSTRACT TRUNCATED AT 250 WORDS)     

Kinetic properties and inhibition of human T lymphoblast deoxycytidine kinase

The kinetic properties of 50,000-fold purified cultured human T lymphoblast (MOLT-4) deoxycytidine kinase were examined. The reaction velocity had an absolute requirement for magnesium. Maximal activity was observed at pH 6.5-7.0 with Mg:ATP for 1:1. High concentrations of free Mg2+ or free ATP were inhibitory. Double reciprocal plots of initial velocity studies yielded intersecting lines for both deoxycytidine and MgATP2-. dCMP was a competitive inhibitor with respect to deoxycytidine and ATP. ADP was a competitive inhibitor with respect to ATP and a mixed inhibitor with respect to deoxycytidine. dCTP, an important end product, is a very potent inhibitor and was a competitive inhibitor with respect to deoxycytidine and a non-competitive inhibitor with respect to ATP. TTP reversed dCTP inhibition. The data suggest that (a) MgATP2- is the true substrate of deoxycytidine kinase; (b) the kinetic mechanism of deoxycytidine kinase is consistent with rapid equilibrium random Bi Bi; (c) deoxycytidine kinase may be regulated by its product ADP and its end product dCTP as well as the availability of deoxycytidine. While many different nucleotides potently inhibit deoxycytidine kinase, their low intracellular concentrations make their regulatory role less important.     

Nucleotide-binding properties of native and cold-treated mitochondrial ATPase.

1. The bound nucleotides of the beef-heart mitochondrial ATPase (F1) are lost during cold inactivation followed by (NH4)2SO4 precipitation. The release of tightly bound ATP parallels the loss of ATPase activity during this process. 2. During cold inactivation, the sedimentation coefficient (s20, w) of the ATPase first declines from 12.1 S to 9 S, then to 3.5 S. (NH4)2SO4 precipitation of the 9-S component also leads to dissociation into subunits with s20, w of 3.5 S. 3. The 9-S component still contains the bound nucleotides, which are removed when it dissociated into smaller subunits. 4. Reactivation of cold-inactivated ATPase by incubation at 30 degrees C is increased by the presence of 25% glycerol. ATP, however, does not have any clearcut effect on the degree of reactivation in the presence of glycerol. 5. ADP is an inhibitor of the reactivation, probably because it exchanges during reactivation for bound ATP giving rise to an inactive 12-S component. 6. The exchange of tightly bound nucleotides with added adenine nucleotides is more extensive and faster with cold-inactivated ATPase than with the native enzyme. During reactivation up to 1.6 moles of ATP and 1.0 mole ADP can exchange per mole enzyme. 7. Incubation with GTP, CTP or inorganic pyrophosphate induces an increased activity of the ATPase, which, however, soon declines in the presence of ATP. It also disappears on precipitation of GTP-treated enzyme with (NH4)2SO4.     

Regulation of deoxyadenosine and nucleoside analog phosphorylation by human placental adenosine kinase

The enzymes responsible for the phosphorylation of deoxyadenosine and nucleoside analogs are important in the pathogenesis of adenosine deaminase deficiency and in the activation of specific anticancer and antiviral drugs. We examined the role of adenosine kinase in catalyzing these reactions using an enzyme purified 4000-fold (2.1 mumol/min/mg) from human placenta. The Km values of deoxyadenosine and ATP are 135 and 4 microM, respectively. Potassium and magnesium are absolute requirements for deoxyadenosine phosphorylation, and 150 mM potassium and 5 mM MgCl2 are critical for linear kinetics. With only 0.4 mM MgCl2 in excess of ATP levels, the Km for deoxyadenosine is increased 10-fold. ADP is a competitive inhibitor with a Ki of 13 microM with variable MgATP2-, while it is a mixed inhibitor with a Ki and Ki' of 600 and 92 microM, respectively, when deoxyadenosine is variable. AMP is a mixed inhibitor with Ki and Ki' of 177 and 15 microM, respectively, with variable deoxyadenosine; it is a non-competitive inhibitor with a Ki of 17 microM and Ki' of 27 microM with variable ATP. Adenosine kinase phosphorylates adenine arabinoside with an apparent Km of 1 mM using deoxyadenosine kinase assay conditions. The Km values for 6-methylmercaptopurine riboside and 5-iodotubercidin, substrates for adenosine kinase, are estimated to be 4.5 microM and 2.6 nM, respectively. Other nucleoside analogs are potent inhibitors of deoxyadenosine phosphorylation, but their status as substrates remains unknown. These data indicate that deoxyadenosine phosphorylation by adenosine kinase is primarily regulated by its Km and the concentrations of Mg2+, ADP, and AMP. The high Km values for phosphorylation of deoxyadenosine and adenine arabinoside suggest that adenosine kinase may be less likely to phosphorylate these nucleosides in vivo than other enzymes with lower Km values. Adenosine kinase appears to be important for adenosine analog phosphorylation where the Michaelis constant is in the low micromolar range.     

Study of the kinetics and mechanism of ATP hydrolysis by soluble ATPase of the chloroplasts (CFl) in the presence of Mg2+ ions]

A kinetic study of ATP hydrolysis by soluble ATPase of chloroplasts (CF1) was made. At low concentrations of MgCl2 a linear increase of the reaction rate was observed during the increase in the ATP concentration up to 1 mM. At high concentrations of MgCl2 the dependence was of a more complicated nature. At MgCl2 concentrations lower than 0.1 mM the reaction approached second-order kinetics with respect to Mg2+; the increase in MgCl2 concentration resulted in a decrease of the reaction order. It is assumed that MgATP is the "true" substrate and MgADP the "true" inhibitor of the reaction. A reaction mechanism of ATP hydrolysis is postulated.     

Binding of adenine nucleotides to the F1-inhibitor protein complex of bovine heart submitochondrial particles.

The binding of ATP radiolabeled in the adenine ring or in the gamma- or alpha-phosphate to F1-ATPase in complex with the endogenous inhibitor protein was measured in bovine heart submitochondrial particles by filtration in Sephadex centrifuge columns or by Millipore filtration techniques. These particles had 0.44 +/- 0.05 nmol of F1 mg-1 as determined by the method of Ferguson et al. [(1976) Biochem. J. 153, 347]. By incubation of the particles with 50 microM ATP, and low magnesium concentrations (less than 0.1 microM MgATP), it was possible to observe that 3.5 mol of [gamma-32P]ATP was tightly bound per mole of F1 before the completion of one catalytic cycle. With [gamma-32P]ITP, only one tight binding site was detected. Half-maximal binding of adenine nucleotides took place with about 10 microM. All the bound radioactive nucleotides were released from the enzyme after a chase with cold ATP or ADP; 1.5 sites exchanged with a rate constant of 2.8 s-1 and 2 with a rate constant of 0.45 s-1. Only one of the tightly bound adenine nucleotides was released by 1 mM ITP; the rate constant was 3.2 s-1. It was also observed that two of the bound [gamma-32P]ATP were slowly hydrolyzed after removal of medium ATP; when the same experiment was repeated with [alpha-32P]ATP, all the label remained bound to F1, suggesting that ADP remained bound after completion of ATP hydrolysis. Particles in which the natural ATPase inhibitor protein had been released bound tightly only one adenine nucleotide per enzyme.(ABSTRACT TRUNCATED AT 250 WORDS)     

A Cl(-)-translocating adenosinetriphosphatase in Acetabularia acetabulum. 2. Reconstitution of the enzyme into liposomes and effect of net charges of liposomes on chloride permeability and reconstitution

The Mono Q-III fraction, a Mg2(+)-ATPase, isolated from Acetabularia acetabulum was reconstituted into liposomes of various net charges prepared by the reversed-phase method and tested for a Cl(-)-translocating activity. The liposomes from a mixture of egg lecithin, dicetyl phosphate, and cholesterol (63:18:9 mole ratio, negative liposomes) and from a mixture of egg lecithin and cholesterol (63:9 mole ratio, neutral liposomes) were less leaky than positive liposomes from asolectin, and from a mixture of egg lecithin, stearylamine, and cholesterol (63:18:9 mole ratio). A significant increase in 36Cl- efflux from the negative and neutral liposomes was observed by addition of ATP in the presence of valinomycin after incorporation of the enzyme by short-term dialysis. The ATP-driven 36Cl- efflux was inhibited by addition of azide, an inhibitor of the ATPase. The preincubation of the enzyme with phenylglyoxal, an arginine-modifying reagent, inactivated ATP-mediated 36Cl- efflux, but the ATPase activity of the preparation was not affected. When chloride was replaced by 35SO4(2)-, no ATP-dependent 35SO4(2)- efflux was detectable from the proteoliposomes. Proton-translocating activity of the enzyme was also tested, and no fluorescent quenching of 9-ACMA was observed.     

Studies on the mechanism of action of histone kinase dependent on adenosine 3':5'-monophosphate. Evidence for involvement of histidine and lysine residues in the phosphotransferase reaction.

The reaction of the phosphate residue transfer catalysed by histone kinase dependent on adenosine 3':5'-monophosphate (cyclic AMP) was studied. The phosphotransferase reaction was shown to obey the mechanism of ping-pong bi-bi type. After incubation of the catalytic subunit of histone kinase with [gamma-32P]ATP the incorporation of one mole of [32P]phosphage per mole of protein was observed. The tryptic [32P]phosphohistidine-containing peptide was isolated and its N-terminus and amino acid composition were determined. The 2',3'-dialdehyde derivative of ATP (oATP) was used as the affinity label for the catalytic subunit of cyclic-AMP-dependent histone kinase. The inhibitor formed an alidmine bond with epsilon-amino group of the lysine residue of the active site and was irreversibly bound to the enzyme after reduction by sodium borohydride with concurrent irreversible inactivation of the enzyme. After inactivation, about one mole of 14C-labelled inhibitor was incorporated per mole of the enzyme. ATP effectively protected the catalytic subunit of histone kinase against inactivation by oATP. Tryptic digestion of the enzyme-inhibitor complex led to the isolation of the 14C-labelled peptide of the active site of histone kinase. Basing on these results, the role of histidine and lysine residues in the active site of the catalytic subunit of histone kinase was suggested.     

Characterization of nucleotide binding sites on chloroplast coupling factor 1.

A study of the equilibrium binding of ADP, 1,N6-ethenoadenosine diphosphate, adenylyl imidodiphosphate, and 1,N6-ethenoadenylyl imidodiphosphate to solubilized spinach chloroplast coupling factor 1 (CF1) has been carried out. All four nucleotides were found to bind to two apparently identical "tight" sites, with characteristic dissociation contants generally less than 10 muM. The binding to these "tight" sites is similar in the presence of Mg2+ and Ca2+, is stronger in 0.1 M NaC1 than in 20 mM Tris-C1, and is only slightly altered by heat activation. The slow rate of association of ADP and 1,N6-ethenoadenosine diphosphate at these sites rules out the possibility that they are catalytic sites for ATPase activity on the solubilized enzyme. A third tight site for adenylyl imidodiphosphate was found on the heat-activated enzyme. The dissociation constant for this interaction (7.6 muM) is similar to the adenylyl imidodiphosphate competitive inhibition constant for ATPase activity (4 muM). ADP, which inhibits ATPase activity but is not a strong competitive inhibitor, binds only weakly at a third site (dissociation constant greater than 70 muM). One mole of 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole reacted per mole of CF1 prevents ADP and adenylyl imidodiphosphate binding at the "catalytic" site and abolishes the ATPase activity. A model is proposed in which the "tight" nucleotide binding sites act as allosteric conformational switches for the ATPase activity of solubilizedCF1.     

Studies on avian erythrocyte metabolism: IX. Relationship of changing organic phosphate composition to whole blood oxygen affinity during development of the ostrich (Struthio camelus camelus)

(1) 2,3-Diphosphoglyceric acid (2,3-DPG) is present in erythrocytes (RBC) of the 37-day ostrich embryo at a concentration of 3.7 μmole/ml RBC, representing 23.5% of the cell phosphate. (2) Inositol tetraphosphate (inositol-P₄) is absent in red cells of the 37-day embryo, appears in the RBC about 63 days after hatch, and thereafter gradually accumulates to a level of 2.8 μmole/ml RBC in the adult bird, representing 30–35% of the cell phosphate. (3) Inositol pentaphosphate (inositol-P₅) is present in the erythrocytes of the 37-day embryo at a concentration of 0.8 μmole/ml RBC, reaches a peak level of 2.9 μmole/ml RBC by Day 63 after hatch, and thereafter declines to a level of 1.1 μmole/ml RBC in the adult bird. (4) The crossover time where the molar concentrations of inositol-P₅ and inositol-P₄ are equal in the erythrocyte appears to be about 150 days after hatch. (5) The p₅₀ of whole blood from the 37-day ostrich embryo, 5-day ostrich chick, and adult ostrich was 15.4, 31.8, and 24.9 Torr. The p₅₀ of whole blood immediately after hatch correlates best with an abrupt rise in ATP concentration but after 54 days posthatch correlates best with the appearance of increasing concentrations of inositol-P₄ and the decrease in concentrations of ATP and inositol-P₅. (6) The appearance of inositol-P₄ in the cells could be an adaptive mechanism for regulating oxygen supply by switching to a modulator which maintains a higher oxygen affinity than would a predominance of inositol-P₅.     

Immunosuppressive properties of electrophoretically "slow" and "fast" form alpha 2-macroglobulin. Effects on cell-mediated cytotoxicity and (allo-) antigen-induced T cell proliferation

Lymphokine activated killer cell lysis of K562 cells was inhibited by alpha 2-macroglobulin (alpha 2M), soybean trypsin inhibitor, and alpha 1-proteinase inhibitor. In serum free medium 2 mg/ml alpha 2M suppressed target cell lysis in a 4-h cytotoxic assay with about 40%. Suppression was dose and time dependent. Cytotoxicity was unaffected by alpha 2M concentrations less than 0.25 mg/ml, and by alpha 2M added later than 1.5 h from start of assay. Pre-treatment of effector (but not of target) cells with alpha 2M was even more suppressive than the presence of alpha 2M during assay. Cell-mediated cytotoxicity was not inhibited by alpha 2M treated with methylamine or by various alpha 2M-proteinase complexes. In contrast, alpha 2M-proteinase complex as well as native alpha 2M suppressed the proliferation of Ag-activated T cells. However, methylamine-treated alpha 2M did not inhibit T cell proliferation, and suppression by alpha 2M-proteinase complex was significantly reduced after inhibition of the alpha 2M-bound proteinase. On incubation at 4 degrees C with lymphokine-activated killer cells, alpha 2M reacted with cell associated proteinases and changed from electrophoretically "slow" to "fast" form. Cell associated proteinases bound by alpha 2M showed chymotrypsin- and trypsin-like specificities and their activity surpassed activity caused by cellular leakage and secretion. The present results strongly indicate that alpha 2M mediates immunosuppression in its capacity as a proteinase inhibitor and suggest inhibition of (T)cell surface-associated proteinases as a possible mode of suppression.     

Purification and properties of l(+)-lactate dehydrogenase from potato tubers

1. A purification of l(+)-lactate dehydrogenase is described. 2. The final preparation is active with NADH and NADPH and with a number of keto acids, but evidence is presented to support the view that a single enzyme is involved. 3. NAD(+) showed product inhibition, but at slightly acid pH values there was evidence of co-operative binding. 4. At acid pH values ATP was a potent inhibitor and appears to be an allosteric effector. At neutral or alkaline pH values ATP behaved as a weak competitive inhibitor. 5. The physiological significance of inhibition by ATP is discussed.