Crystal structure of heart 6‐phosphofructo‐2‐kinase/fructose‐2,6‐bisphosphatase (PFKFB2) and the inhibitory influence of citrate on substrate binding

The heart‐specific isoform of 6‐phosphofructo‐2‐kinase/fructose‐2,6‐bisphosphatase (PFKFB2) is an important regulator of glycolytic flux in cardiac cells. Here, we present the crystal structures of two PFKFB2 orthologues, human and bovine, at resolutions of 2.0 and 1.8 Å, respectively. Citrate, a TCA cycle intermediate and well‐known inhibitor of PFKFB2, co‐crystallized in the 2‐kinase domains of both orthologues, occupying the fructose‐6‐phosphate binding‐site and extending into the γ‐phosphate binding pocket of ATP. This steric and electrostatic occlusion of the γ‐phosphate site by citrate proved highly consequential to the binding of co‐complexed ATP analogues. The bovine structure, which co‐crystallized with ADP, closely resembled the overall structure of other PFKFB isoforms, with ADP mimicking the catalytic binding mode of ATP. The human structure, on the other hand, co‐complexed with AMPPNP, which, unlike ADP, contains a γ‐phosphate. The presence of this γ‐phosphate made adoption of the catalytic ATP binding mode impossible for AMPPNP, forcing the analogue to bind atypically with concomitant conformational changes to the ATP binding‐pocket. Inhibition kinetics were used to validate the structural observations, confirming citrate's inhibition mechanism as competitive for F6P and noncompetitive for ATP. Together, these structural and kinetic data establish a molecular basis for citrate's negative feed‐back loop of the glycolytic pathway via PFKFB2. Proteins 2016; 85:117–124. © 2016 Wiley Periodicals, Inc.     

Site-directed mutagenesis of a potential catalytic and formyl phosphate binding site and substrate inhibition of N10-formyltetrahydrofolate synthetase

Structural studies of N(10)-formyltetrahydrofolate synthetase (FTHFS) have indicated the involvement of Arg 97 in the binding of the formyl phosphate intermediate. Two site-directed mutants were constructed to test this hypothesis: R97S (Ser substitution) and R97E (Glu substitution). The k(cat) of R97S was approximately 60% that of the wild-type enzyme and had K(m) for ATP and formate twofold higher than those of wild type. R97E was completely inactive and had a K(m) for ATP nearly six times that of wild type. Substrate inhibition by tetrahydrofolate was shown to occur in wild-type and R97S enzymes using both steady-state and transient-state kinetic approaches. These results lend greater insight into the mechanistic function of FTHFS by confirming the interaction of both ATP and formate with Arg 97 and introducing the aspect of substrate inhibition by tetrahydrofolate with regard to substrate binding and dissociation.     

Binding of adenosine diphosphate to human blood platelets and to isolated blood platelet membranes

The equilibrium binding of 14C-labeled ADP to intact washed human blood platelets and to platelet membranes was investigated. With both intact platelets and platelet membranes a similar concentration dependence curve was found. It consisted of a curvilinear part below 20 microM and a rectilinear part above this concentration. At high ADP concentrations, the rectilinear part appeared to be saturable. Because of this, two classes of saturable ADP binding sites were proposed. ADP was partly converted to ATP and AMP with intact platelets while this conversion was virtually absent in isolated platelet membranes. ADP was bound to platelet membranes with the same type of curves found for intact platelets. The ADP binding to the high affinity system, which was stimulated by calcium ions, was nearly independent of temperature and had a pH optimum at 7.8. A number of agents were investigated for inhibiting properties. Of the sulfhydryl reagents only p-chloromercuribenzene sulfonate inhibited both high and low affinity binding systems while iodoacetamide and N-ethylmaleimide were without effect. Compounds acting via cyclic AMP on platelet aggregation, such as adenosine and cyclic AMP itself, had no influence on binding. Some nucleosidediphosphates and nucleotide analogs at a concentration of 100 microM had no, or only a slight, effect on high affinity ADP binding. For some other nucleotides inhibitor constants were determined for both platelet ADP aggregation and ADP binding. The inhibitor constants of ATP, adenyl-5'-yl-(beta,gamma-methylene)diphosphate, IDP, adenosine-5'(2-O-thio)diphosphate, for aggregation and high affinity binding were in good correlation with each other. Exceptions formed fluorosulfonylbenzoyl adenosine and AMP. The ATP formation found with intact platelets could be attributed to a nucleosidediphosphate kinase. It was investigated in some detail. The enzyme was magnesium dependent, had a Q10 value of 1.41, a pH optimum at 8.0, was competitively inhibited by AMP and reacted via a ping pong mechanism. All findings described in this paper indicate that platelets as well as platelet membranes bind ADP with the same characteristics and they suggest that the high affinity binding of ADP is involved in platelet aggregation induced by ADP. The results on nucleosidediphosphate kinase did not permit a firm conclusion about the role of the enzyme in induction of platelet aggregation by ADP.     

Primary structure of the thermostable formyltetrahydrofolate synthetase from Clostridium thermoaceticum

The complete nucleotide sequence of the Clostridium thermoaceticum formyltetrahydrofolate synthetase (FTHFS) was determined and the primary structure of the protein predicted. The gene was 1680 nucleotides long, encoding a protein of 559 amino acid residues with a calculated subunit molecular weight of 59,983. The initiation codon was UUG, with a probable ribosome binding site 11 bases upstream. A putative ATP binding domain was identified. Two Cys residues likely to be involved in subunit aggregation were tentatively identified. No characterization of the tetrahydrofolate (THF) binding domain was possible on the basis of the sequence. A high level of amino acid sequence conservation between the C. thermoaceticum FTHFS and the published sequences of C. acidiurici FTHFS and the FTHFS domains of the Saccharomyces cerevisiae C1-THF synthases was found. Of the 556 residues shared between the two clostridial sequences, 66.4% are identical. If conservative substitutions are allowed, this percentage rises to 75%. Over 47% of the residues shared between the C. thermoaceticum FTHFS and the yeast C1-THF synthases are identical, 57.4% if conservative substitutions are allowed. Hydrophobicity profiles of the C. acidiurici and C. thermoaceticum enzymes were very similar and did not support the idea that large hydrophobic domains play an important role in thermostabilizing the C. thermoaceticum FTHFS.     

Formation and utilization of formyl phosphate by N10-formyltetrahydrofolate synthetase: evidence for formyl phosphate as an intermediate in the reaction

N10-Formyltetrahydrofolate synthetase from bacteria and yeast catalyzes a slow formate-dependent ADP formation in the absence of H4folate. The synthesis of formyl phosphate by the enzyme was detected by trapping the intermediate as formyl hydroxamate. That the "formate kinase" activity was part of the catalytic center of N10-formyltetrahydrofolate synthetase was shown by demonstrating coordinate inactivation of the "kinase" and synthetase activities by heat and a sulfhydryl reagent, similar effects of monovalent cations, similar Km values for substrates, and similar Ki values for the inhibitor phosphonoacetaldehyde for both activities. The relative rates of the kinase activities for the bacterial and yeast enzymes are about 10(-4) and 4 x 10(-6) of their respective synthetase activities. These slow rates for the kinase reaction can be explained by the slow dissociation of ADP and formyl phosphate from the enzyme. This conclusion is supported by rapid-quench studies where a "burst" of ADP formation (6.4 s-1) was observed that is considerably faster than the steady-state rate (0.024 s-1). The demonstration of enzyme-bound products by a micropartition assay and the lack of a significant formate-stimulated exchange between ADP and ATP provide further evidence for the slow release of the products from the enzyme. The synthesis of N10-CHO-H4folate when H4folate was added to the E-formyl phosphate-ADP complex is also characterized by a "burst" of product formation. The rate of this burst phase at 5 degrees C occurs with a rate constant of 18 s-1 compared to 14 s-1 for the overall reaction at the same temperature. These results provide further evidence for formyl phosphate as an intermediate in the reaction and are consistent with the sequential mechanism of the normal catalytic pathway. Positional isotope exchange experiments using [beta,gamma-18O]ATP showed no evidence for exchange during turnover experiments in the presence of either H4folate or the competitive inhibitor pteroyltriglutamate. The absence of scrambling of the 18O label as observed by 31P NMR suggests that the central complex may impose restraints to limit free rotation of the P beta oxygens of the product ADP.     

Crystal structure of ADP‐dependent glucokinase from Methanocaldococcus jannaschii in complex with 5‐iodotubercidin reveals phosphoryl transfer mechanism

ADP‐dependent glucokinase (ADPGK) is an alternative novel glucose phosphorylating enzyme in a modified glycolysis pathway of hyperthermophilic Archaea. In contrast to classical ATP‐dependent hexokinases, ADPGK utilizes ADP as a phosphoryl group donor. Here, we present a crystal structure of archaeal ADPGK from Methanocaldococcus jannaschii in complex with an inhibitor, 5‐iodotubercidin, d ‐glucose, inorganic phosphate, and a magnesium ion. Detailed analysis of the architecture of the active site allowed for confirmation of the previously proposed phosphorylation mechanism and the crucial role of the invariant arginine residue (Arg197). The crystal structure shows how the phosphate ion, while mimicking a β‐phosphate group, is positioned in the proximity of the glucose moiety by arginine and the magnesium ion, thus providing novel insights into the mechanism of catalysis. In addition, we demonstrate that 5‐iodotubercidin inhibits human ADPGK‐dependent T cell activation‐induced reactive oxygen species (ROS) release and downstream gene expression, and as such it may serve as a model compound for further screening for hADPGK‐specific inhibitors.     

New insights into the catalytic mechanism of human glycine N‐acyltransferase

Even though the glycine conjugation pathway was one of the first metabolic pathways to be discovered, this pathway remains very poorly characterized. The bi‐substrate kinetic parameters of a recombinant human glycine N‐acyltransferase (GLYAT, E.C. 2.3.1.13) were determined using the traditional colorimetric method and a newly developed HPLC–ESI‐MS/MS method. Previous studies analyzing the kinetic parameters of GLYAT, indicated a random Bi–Bi and/or ping‐pong mechanism. In this study, the hippuric acid concentrations produced by the GLYAT enzyme reaction were analyzed using the allosteric sigmoidal enzyme kinetic module. Analyses of the initial rate (v) against substrate concentration plots, produced a sigmoidal curve (substrate activation) when the benzoyl‐CoA concentrations was kept constant, whereas the plot with glycine concentrations kept constant, passed through a maximum (substrate inhibition). Thus, human GLYAT exhibits mechanistic kinetic cooperativity as described by the Ferdinand enzyme mechanism rather than the previously assumed Michaelis–Menten reaction mechanism.     

Structure of human phosphopantothenoylcysteine synthetase at 2.3 A resolution

The structure of human phosphopantothenoylcysteine (PPC) synthetase was determined at 2.3 A resolution. PPC synthetase is a dimer with identical monomers. Some features of the monomer fold resemble a group of NAD-dependent enzymes, while other features resemble the ribokinase fold. The ATP, phosphopantothenate, and cysteine binding sites were deduced from modeling studies. Highly conserved ATP binding residues include Gly43, Ser61, Gly63, Gly66, Phe230, and Asn258. Highly conserved phosphopantothenate binding residues include Asn59, Ala179, Ala180, and Asp183 from one monomer and Arg55' from the adjacent monomer. The structure predicts a ping pong mechanism with initial formation of an acyladenylate intermediate, followed by release of pyrophosphate and attack by cysteine to form the final products PPC and AMP.     

An enzyme from lupin seeds forming alanine derivatives of cytokinins

A new enzyme, which catalyses the conversion of the cytokinin zeatin to the alanine conjugate lupinic acid, has been partly purified from developing lupin seed (Lupinus luteus). Paired-ion, reverse phase HPLC was adapted to analyse the enzyme reaction quantitatively. The enzyme used O-acetyl-l-serine as the source of the amino acid residue, and it interacted with substrates in a ping pong bi bi mechanism. A number of adenine derivatives served as substrates, but preference was shown for compounds with high cytokinin activity. The possible role of the enzyme, tentatively called β-(9-cytokinin)alanine synthase or lupinic acid synthase, in the regulation of hormone activity is discussed.     

Kinetic mechanism and identification of the active site tyrosine residue in Enterobacter amnigenus arylsulfate sulfotransferase.

Bacterial arylsulfate sulfotransferase (ASST) catalyzes the transfer of a sulfate group from a phenyl sulfate ester to a phenolic acceptor. The kinetic mechanism of Enterobacter amnigenus ASST was determined. Plots of 1/v versus 1/[substrate (A)] at different fixed substrate (B) concentrations gave a series of parallel lines. One of the reaction products, p-nitrophenol, inhibited the enzyme noncompetitively with respect to p-nitrophenyl sulfate, but competitively to alpha-naphthol. These results correspond to a ping pong bi bi mechanism. By site-directed mutagenesis, we substituted each conserved tyrosine residue with phenylalanine. Among the mutants, Y123F showed severely reduced catalytic activity. We conclude that Tyr 123 is an essential active site residue. A mechanistic hypothesis is presented to account for these observations.     

How does the mitochondrial ADP/ATP carrier distinguish transportable ATP and ADP from untransportable AMP and GTP?Dynamic modeling of the recognition/translocation process in the major substrate binding region

To understand the transport mechanism of the bovine heart mitochondrial ADP/ATP carrier at the atomic level, we studied the four-dimensional features of the interaction of various purine nucleotides with the adenine nucleotide binding region (ABR) consisting of Arg(151)-Asp(167) in the second loop facing the matrix side. After three-dimensional modeling of ABR based on the experimental results, its structural changes on interaction with purine nucleotides were examined by molecular dynamics computation at 300 K. ATP/ADP were translocated to a considerable degree from the matrix side to the inner membrane region accompanied by significant backbone conformational changes, whereas neither appreciable translocation nor a significant conformational change was observed with the untransportable nucleotides AMP/GTP. The results suggested that binding of the terminal phosphate group and the adenine ring of ATP/ADP with Arg(151) and Lys(162), respectively, and subsequent interaction of a phosphate group(s) other than the terminal phosphate with Lys(162) triggered the expansion and subsequent contraction of the backbone conformation of ABR, leading to the translocation of ATP/ADP. Based on a simplified molecular dynamic simulation, we propose a dynamic model for the initial recognition process of ATP/ADP with the carrier.     

A Methanocaldococcus jannaschii archaeal signature gene encodes for a 5-formaminoimidazole-4-carboxamide-1-beta-D-ribofuranosyl 5'-monophosphate synthetase. A new enzyme in purine biosynthesis

We have identified and characterized a new member of the ATP-grasp enzyme family that catalyzes the ATP- and formate-dependent formylation of 5-aminoimidazole-4-carboxamide-1-beta-D-ribofuranosyl 5'-monophosphate (AICAR) to 5-formaminoimidazole-4-carboxamide-1-beta-D-ribofuranosyl 5'-monophosphate (FAICAR) in the absence of folates. The enzyme, which we designate as PurP, is the product of the Methanocaldococcus jannaschii purP gene (MJ0136), which is a signature gene for Archaea. As is characteristic of reactions catalyzed by this family of enzymes, the other products of the reaction, ADP and P(i), were produced stoichiometrically with the amount of ATP, formate, and AICAR used. Formyl phosphate was found to substitute for ATP and formate in the reaction, yet the methylene analog, phosphonoacetaldehyde, was not an inhibitor or substrate for the reaction. The enzyme, along with PurO, which catalyzes the cyclization of FAICAR to inosine 5'-monophosphate, catalyzes the same overall transformation in purine biosynthesis as is accomplished by PurH in bacteria and eukaryotes. No homology exists between PurH and either PurO or PurP. 1H NMR and gas chromatography-mass spectrometry analysis of an M. jannaschii cell extract showed the presence of free formate that can be used by the enzyme for purine biosynthesis. This formate arises by the reduction of CO2 with hydrogen; this was demonstrated by incorporating 13C into the formate when M. jannaschii cell extracts were incubated with H13CO3- and hydrogen gas. The presence of this signature gene in all of the Archaea indicates the presence of a purine biosynthetic pathway proceeding in the absence of folate coenzymes.     

Structures of dimethylsulfoniopropionate-dependent demethylase from the marine organism Pelagabacter ubique

Dimethylsulfoniopropionate (DMSP) is a ubiquitous algal metabolite and common carbon and sulfur source for marine bacteria. DMSP is a precursor for the climatically active gas dimethylsulfide that is readily oxidized to sulfate, sulfur dioxide, methanesulfonic acid, and other products that act as cloud condensation nuclei. Although the environmental importance of DMSP metabolism has been known for some time, the enzyme responsible for DMSP demethylation by marine bacterioplankton, dimethylsufoniopropionate‐dependent demethylase A (DmdA, EC 2.1.1.B5), has only recently been identified and biochemically characterized. In this work, we report the structure for the apoenzyme DmdA from Pelagibacter ubique (2.1 Å), as well as for DmdA co‐crystals soaked with substrate DMSP (1.6 Å) or the cofactor tetrahydrofolate (THF) (1.6 Å). Surprisingly, the overall fold of the DmdA is not similar to other enzymes that typically utilize the reduced form of THF and in fact is a triple domain structure similar to what has been observed for the glycine cleavage T protein or sarcosine oxidase. Specifically, while the THF binding fold appears conserved, previous biochemical studies have shown that all enzymes with a similar fold produce 5,10‐methylene‐THF, while DmdA catalyzes a redox‐neutral methyl transfer reaction to produce 5‐methyl‐THF. On the basis of the findings presented herein and the available biochemical data, we outline a mechanism for a redox‐neutral methyl transfer reaction that is novel to this conserved THF binding domain.     

Characterization of nucleotide binding sites on the membrane-bound chloroplast ATP synthase (coupling factor CF1)

Phosphorylation of ADP and nucleotide exchange by membrane-bound coupling factor CF₁ are very fast reactions in the light, so that a direct comparison of both reactions is difficult. By adding substrate ADP and phosphate to illuminated thylakoids together with the uncoupler FCCP, the phosphorylation time is limited and the amount of ATP formed can be reduced to less than 1 ATP per enzyme. Low concentrations of medium nucleotides during illumination increase the amount of ATP formed during uncoupling presumably by binding to the tight nucleotide binding site (further designated as ‘site A’) with an affinity of 1 to 7 μM for ADP and ATP. ATP formation itself shows half-saturation at about 30 μM. Loosely bound nucleotides are exchanged upon addition of nucleotides with uncoupler (Schumann, J. (1984) Biochim. Biophys. Acta 766, 334–342). Release depends binding of nucleotides to a second site. The affinity of this site for ADP (in the presence of phosphate) is about 30 μM. It is assumed that phosphorylation and induction of exchange both occur on the same site (site B). During ATP hydrolysis, an ATP molecule is bound to site A, while on another site, ATP is hydrolyzed rapidly. The affinity of ADP for the catalytic site (70 μM) is in the same range as the observed Michaelis constant of ADP during phosphorylation; it is assumed that site B is involved in ATP hydrolysis. Site A exhibits some catalytic activity; it might be that site A is involved in ATP formation in a dual-site mechanism. For ATP hydrolysis, however, direct determination of exchange rates showed that the exchange rate of ATP bound to site A is about 30-times lower than ATP hydrolysis under the same conditions.     

Cloning, expression, purification, cofactor requirements, and steady state kinetics of phosphoketolase-2 from Lactobacillus plantarum

The genes xpk1 and xpk2(Δ1–21) encoding phosphoketolase-1 and (Δ1–7)-truncated phosphoketolase-2 have been cloned from Lactobacillus plantarum and expressed in Escherichia coli. Both gene-products display phosphoketolase activity on fructose-6-phosphate in extracts. A N-terminal His-tag construct of xpk2(Δ1–21) was also expressed in E. coli and produced active His-tagged (Δ1–7)-truncated phosphoketolase-2 (hereafter phosphoketolase-2). Phosphoketolase-2 is activated by thiamine pyrophosphate (TPP) and the divalent metal ions Mg²⁺, Mn²⁺, or Ca²⁺. Kinetic analysis and data from the literature indicate the activators are MgTPP, MnTPP, or CaTPP, and these species activate by an ordered equilibrium binding pathway, with Me²⁺TPP binding first and then fructose-6-phosphate. Phosphoketolase-2 accepts either fructose-6-phosphate or xylulose-5-phosphate as substrates, together with inorganic phosphate, to produce acetyl phosphate and either erythrose-4-phosphate or glyceraldehyde-3-phosphate, respectively. Steady state kinetic analysis of acetyl phosphate formation with either substrate indicates a ping pong kinetic mechanism. Product inhibition patterns with erythrose-4-phosphate indicate that an intermediate in the ping pong mechanism is formed irreversibly. Background mechanistic information indicates that this intermediate is 2-acetyl-TPP. The irreversibility of 2-acetyl-TPP formation might explain the overall irreversibility of the reaction of phosphoketolase-2.     

Carbamyl phosphate-dependent ATP synthesis catalyzed by formyltetrahydrofolate synthetase.

Formyltetrahydrofolate synthetase (formate:tetrahydrofolate ligase (ADP-forming), EC 6.3.4.3) from Clostridium cylindrosporum catalyzes phosphate transfer from carbamyl phosphate to ADP. This activity is lost when monovalent cations are removed and is recovered when K+ is added back. Carbamyl phosphate is an inhibitor of the formyltetrahydrolfolate synthetase forward reaction, and formate as well as phosphate inhibit the ATP synthesis reaction. Acetyl phosphate and phosphonoacetate are inhibitors of both reactions. The results of kinetic studies support the concept that carbamyl phosphate is an analog of the putative intermediate of the formyltetrahydrofolate synthetase reaction, formyl phosphate.     

Kinetic studies on a novel sulfotransferase from Eubacterium A-44, a human intestinal bacterium

A novel sulfotransferase purified from a human intestinal bacterium stoichiometrically catalyzed the transfer of a sulfate group of phenylsulfate esters to phenolic compounds. Vmax values of the enzyme reaction were measured with various concentrations of a sulfate donor substrate, p-nitrophenylsulfate, and of a sulfate acceptor substrate, tyramine. Double reciprocal plots of the acceptor concentration and Vmax showed a linear correlation. One of the reaction products, tyramine O-sulfate, competitively inhibited the enzyme as to a donor substrate, p-nitrophenylsulfate (PNS), but the other reaction product, p-nitrophenol (PNP), noncompetitively inhibited it as to PNS. These kinetic data suggest that the sulfate transfer reaction proceeds according to a ping pong bi bi mechanism. The enzyme was activated by Mg2+ and inhibited by EDTA, which suggests that it is a metalloenzyme.     

Purification and properties of gamma-glutamyltranspeptidase from Proteus mirabilis.

gamma-Glutamyltranspeptidase was purified ca. 15,200-fold from cell-free extracts of Proteus mirabilis to electrophoretic homogeneity and then crystallized. The enzyme has an estimated molecular weight of 80,000 and consists of two different subunits with molecular weights of ca. 47,000 and 28,000. The purified enzyme catalyzed hydrolysis and transpeptidation of various gamma-glutamyl compounds, including the oxidized and reduced forms of glutathione, gamma-glutamyl compounds of L-phenylalanine, L-tyrosine, L-histidine, L-alpha-aminobutyrate, L-leucine, and p-nitroaniline. Glycylglycine, L-phenylalanine, L-methionine, L-histidine, L-tryptophan, and L-isoleucine were good acceptors of the gamma-glutamyl moiety in the transpeptidation reaction. Km values for gamma-glutamyl compounds were on the order of 10(-4) to 10(-5) M, and those for acceptor peptides and amino acids were on the order of 10(-2) to 10(-3) M. The enzyme was inhibited by L-serine plus borate and 6-diazo-5-oxo-L-norleucine, which are inhibitors of gamma-glutamyltranspeptidases isolated from mammals. Various amino acids alone were found to inhibit the transpeptidation competitively with a gamma-glutamyl donor. Kinetic analysis suggested that the reaction sequence of substrate binding and product release proceeds according to a ping pong bi bi mechanism.     

Light-induced exchange of nucleotides into coupling factor 1 in spinach chloroplast thylakoids.

The method of centrifugation of chloroplast thylakoids through silicone fluid, previously used to estimate the uptake of solutes by thylakoids, is shown to be an excellent method for measuring binding of nucleotides to thylakoids. This binding, which is probably an exchange (Harris, D. A. and Slater, E. C. (1975) Biochim. Biophys. Acta 387, 335-348), is enhanced by light and is sensitive to uncoupling. Half-maximal binding of adenosine 5'-triphosphate (ATP) or adenosine 5'-diphosphate (ADP) at 10 mjM was reached within less than 0.1 s. With illumination times sufficient to elicit maximal binding, saturation of the site(s) is approached at 20 muM nucleotide and dissociation constants of 5 muM and 7 muM were calculated for ADP and ATP, respectively. At saturation, the binding corresponds to 1 mol/mol of coupling factor 1 or less. Although the light-dependent binding of ADP does not require Mg2+, that of ATP is markedly enhanced by Mg2+. A 10-fold molar excess of guanosine di- or triphosphate or adenyl-5'-yl imidodiphosphate had little effect on the binding. Adenosine 5'-phosphosulfate, a competitive inhibitor of phosphorylation with respect to ADP, decreases the binding. Thylakoids, previously illuminated in the absence of added nucleotides, retain the capacity to bind ADP or ATP in the dark long after the H+ electrochemical gradient has decayed. The conformation of coupling factor 1 in darkened thylakoids following illumination in the absence of added nucleotides may thus differ from that in thylakoids either illuminated in the presence of nucleotides or kept in the dark. Approximately 20% of the ADP bound to coupling factor 1 in thylakoids is converted to ATP by a 2-s illumination. Bound inorganic phosphate, derived either from ATP or from inorganic phosphate itself, serves as the phosphoryl donor. Bound ADP may, therefore, be of catalytic significance in the mechanism of phosphorylation.