Binding dynamics and energetic insight into the molecular forces driving nucleotide binding by guanylate kinase

Plasmodium deoxyguanylate pathways are an attractive area of investigation for future metabolic and drug discovery studies due to their unique substrate specificities. We investigated the energetic contribution to guanylate kinase substrate binding and the forces underlying ligand recognition. In the range from 20 to 35°C, the thermodynamic profiles displayed marked decrease in binding enthalpy, while the free energy of binding showed little changes. GMP produced a large binding heat capacity change of -356 cal mol(-1) K(-1), indicating considerable conformational changes upon ligand binding. Interestingly, the calculated ΔCp was -32 cal mol(-1) K(-1), indicating that the accessible surface area is not the central change in substrate binding, and that other entropic forces, including conformational changes, are more predominant. The thermodynamic signature for GMP is inconsistent with rigid-body association, while dGMP showed more or less rigid-body association. These binding profiles explain the poor catalytic efficiency and low affinity for dGMP compared with GMP. At low temperature, the ligands bind to the receptor site under the effect of hydrophobic forces. Interestingly, by increasing the temperature, the entropic forces gradually vanish and proceed to a nonfavorable contribution, and the interaction occurs mainly through bonding, electrostatic forces, and van der Waals interactions.     

Functional analysis of the nucleotide binding domain of membrane-associated guanylate kinases

Membrane-associated guanylate kinases (MAGUKs) regulate cellular adhesion and signal transduction at sites of cell-cell contact. MAGUKs are composed of modular protein-protein interaction motifs including L27, PDZ, Src homology (SH) 3, and guanylate kinase domains that aggregate adhesion molecules and receptors. Genetic analyses reveal that lethal mutations of MAGUKs often occur in the guanylate kinase domain, indicating a critical role for this domain. Here, we explored whether GMP binding to the guanylate kinase domain regulates MAGUK function. Surprisingly, and in contrast to previously published studies, we failed to detect GMP binding to the MAGUKs postsynaptic density-95 (PSD-95) and CASK. Two amino acid residues in the GMP binding pocket that differ between MAGUKs and authentic guanylate kinase explain this lack of binding, as swapping these residues largely prevent GMP binding to yeast guanylate kinase. Conversely, these mutations restore GMP binding but not catalytic activity to PSD-95. Protein ligands for the PSD-95 guanylate kinase domain, guanylate kinase-associated protein (GKAP) and MAP1A, appear not to interact with the canonical GMP binding pocket, and GMP binding does not influence the intramolecular SH3/guanylate kinase (GK) interaction within PSD-95. These studies indicate that MAGUK proteins have lost affinity for GMP but may have retained the guanylate kinase structure to accommodate a related regulatory ligand.     

Structural requirements for calmodulin binding to membrane-associated guanylate kinase homologs

Effector molecules such as calmodulin modulate the interactions of membrane-associated guanylate kinase homologs (MAGUKs) and other scaffolding proteins of the membrane cytoskeleton by binding to the Src homology 3 (SH3) domain, the guanylate kinase (GK) domain, or the connecting HOOK region of MAGUKs. Using surface plasmon resonance, we studied the interaction of members of all four MAGUK subfamilies--synapse-associated protein 97 (SAP97), calcium/calmodulin-dependent serine protein kinase (CASK), membrane palmitoylated protein 2 (MPP2), and zona occludens (ZO) 1--and calmodulin to determine interaction affinities and localize the binding site. The SH3-GK domains of the proteins and derivatives thereof were expressed in E. coli and purified. In all four proteins, high-affinity calmodulin binding was identified. CASK was shown to contain a Ca2+-dependent calmodulin binding site within the HOOK region, overlapping with a protein 4.1 binding site. In ZO1, a Ca2+-dependent calmodulin binding site was detected within the GK domain. The equilibrium dissociation constants for MAGUK-calmodulin interaction were found to range from 50 nM to 180 nM. Sequence analyses suggest that binding sites for calmodulin have evolved independently in at least three subfamilies. For ZO1, pulldown of GST-calmodulin was shown to occur in a calcium-dependent manner; moreover, molecular modeling and sequence analyses predict conserved basic residues to be exposed on one side of a helix. Thus, calmodulin binding appears to be a common feature of MAGUKs, and Ca2+-activated calmodulin may serve as a general regulator to affect the interactions of MAGUKs and various components of the cytoskeleton.     

Enzyme kinetics and binding studies on inhibitors of MEK protein kinase

Inhibition of the protein kinase, MEK1, is a potential approach for the treatment of cancer. Inhibitors may act by prevention of activation (PoA), which involves interfering with phosphorylation of nonactivated MEK1 by the upstream kinase, B-RAF. Modulation also may occur by inhibition of catalysis (IoC) during phosphorylation of the downstream substrate, ERK2, by activated MEK1. Here, five MEK inhibitors are characterized in terms of binding affinity, PoA, and IoC. The compounds are a butadiene (U-0126), an N-alkoxy amide (CI-1040), two CI-1040 analogues (an anthranilic acid and an N-alkyl amide), and a cyanoquinoline. Some compounds give different mechanisms of inhibition (ATP-competitive, noncompetitive, or uncompetitive) in PoA compared to IoC or show a change in potency between the assays. The inhibitors also exhibit different shifts in potency when either PoA or IoC is compared with binding to nonactivated MEK. The inhibitor potency ranking, therefore, is dependent upon the assay format. When the ATP concentration equals K m, IoC IC 50 increases in the order CI-1040 approximately cyanoquinoline < anthranilic acid approximately U-0126 < alkyl amide. Conversely, the K d from nonactivated MEK1 for four of the compounds varies between more than 6-fold lower and over 18-fold higher than this IC 50, with U-0126 having the lowest K d and CI-1040 having the highest. In PoA when the ATP concentration equals K m, U-0126 has the lowest IC 50, becoming more potent than CI-1040, the cyanoquinoline, and the anthranilic acid. These observations have implications for understanding structure-activity relationships of MEK inhibitors and illustrate how assays can be designed to favor different compounds.     

A pre-equilibrium before nucleotide binding limits fingers subdomain closure by Klentaq1

Numerous studies have been undertaken to establish the mechanism of dNTP binding and template-directed incorporation by DNA polymerases. It has been established by kinetic experiments that a rate-limiting step, crucial for dNTP selection, occurs before chemical bond formation. Crystallographic studies indicated that this step may be due to a large open-to-closed conformational transition affecting the fingers subdomain. In previous studies, we established a fluorescence resonance energy transfer system to monitor the open-to-closed transition in the fingers subdomain of Klentaq1. By comparing the rates of the fingers subdomain closure with that of the rate-limiting step for Klentaq1, we showed that fingers subdomain motion was significantly faster than the rate-limiting step. We have now used this system to characterize DNA binding as well as to complete a more extensive characterization of incorporation of all four dNTPs. The data indicate that DNA binding occurs by a two-step association and that dissociation of the DNA is significantly slower in the case of the closed ternary complex. The data for nucleotide incorporation indicate a step occurring before dNTP binding, which differs for all four nucleotides. As the only difference between the (E x p/t) complexes is the templating base, it would suggest an important role for the templating base in initial ground state selection.     

Physiological correlation between nucleoside-diphosphate kinase and the enzyme-associated guanine nucleotide binding proteins

The physiological correlation between NDP-kinase and the enzyme-associated guanine nucleotide binding proteins (G1 and G2) has been studied in vitro. It was found that incubation of the phosphoenzyme (enzyme-bound high-energy phosphate intermediate) of NDP-kinases with one of the nucleoside 5'-diphosphates (NDPs) in the presence of divalent cations (Mg2+ and Ca2+) results in the formation of nucleoside 5'-triphosphates (NTPs) within 40 sec even at low temperatures (below 4 degrees C) without strict base-specificity; and high-energy phosphates on the phosphoenzyme can transfer preferentially to GDP on the guanine nucleotide binding proteins (G1, G2 and r-p21 protein) in the presence of 0.25 mM Ca2+ or 1 mM Mg2+ even if any other NDPs are present in the reaction mixtures. These observations suggest that NDP-kinase may be responsible for the phosphate-transfer between GDP on the guanine nucleotide binding proteins and its phosphoenzyme.     

Comparative thermodynamic analysis of cyclic nucleotide binding to protein kinase A

We have investigated the thermodynamic parameters and binding of a regulatory subunit of cAMP-dependent protein kinase (PKA) to its natural low-molecular-weight ligand, cAMP, and analogues thereof. For analysis of this model system, we compared side-by-side isothermal titration calorimetry (ITC) with surface plasmon resonance (SPR). Both ITC and SPR analyses revealed that binding of the protein to cAMP or its analogues was enthalpically driven and characterised by similar free energy values (DeltaG=-9.4 to -10.7 kcal mol-1) for all interactions. Despite the similar affinities, binding of the cyclic nucleotides used here was characterised by significant differences in the contribution of entropy (-TDeltaS) and enthalpy (DeltaH) to DeltaG. The comparison of ITC and SPR data for one cAMP analogue further revealed deviations caused by the method. These equilibrium parameters could be complemented by thermodynamic data of the transition state (DeltaHnot equal, DeltaGnot equal, DeltaSnot equal) for both association and dissociation measured by SPR. This direct comparison of ITC and SPR highlights method-specific advantages and drawbacks for thermodynamic analyses of protein/ligand interactions.     

Site-specific cyclic nucleotide binding and dissociation of the holoenzyme of cAMP-dependent protein kinase

The photoaffinity reagent 8-azidoadenosine 3':5'-monophosphate (8-N3cAMP) was previously shown to modify a single tyrosine residue on the type II regulatory subunit of cAMP-dependent protein kinase (Kerlavage, A.R., and Taylor, S.S. (1980) J. Biol. Chem, 255, 8483-8488). In the present studies, the binding stoichiometries of type II holoenzyme for cAMP and 8-N3cAMP were determined using Millipore filtration assays in the absence (Assay A) and presence (Assay B) of 2 M NaCl and histone. The binding stoichiometry of holoenzyme for cAMP was 2 mol/mol with Assay A, and 4 mol/mol with assay B. The binding stoichiometry for 8-N3cAMP was 2 mol/mol with Assay B or with Assay A following photolysis of the holoenzyme:8-N3cAMP mixture. In the absence of photolysis, the binding stoichiometry for 8-N3cAMP was 0.4 mol/mol with Assay A. Both 8-N3cAMP and cAMP fully dissociated the holoenzyme. Holoenzyme, labeled with 8-N3[3H]cAMP on a preparative scale, incorporated 1 mol of 8-N3[3H]cAMP/mol of regulatory subunit (RII) monomer. The labeled RII was separated from catalytic subunit, cleaved with cyanogen bromide, and the resultant peptides were separated by high performance liquid chromatography. A single radioactive peptide was observed which had the same NH2 terminal residue and amino acid composition as the peptide obtained when dissociated RII was labeled with 8-N3cAMP.     

Dissection of the nucleotide and metal-phosphate binding sites in cAMP-dependent protein kinase.

The catalytic (C) subunit of cAMP-dependent protein kinase (cAPK) is more stable by several criteria when it is part of a holoenzyme complex. By measuring the thermal stability of the free C subunit in the presence and absence of nucleotides and/or divalent metal ions, it was found that most of the stabilizing effects associated with the type I holoenzyme could be attributed to the nucleotide. The specific requirements for this enhanced stability were further dissected: Adenosine stabilized the C subunit up to 5 degrees C; however, divalent cations (i.e., Mg2+, Ca2+, and Mn2+) do not increase heat stability in combination with adenosine and adenine (1). Divalent cations as well as ATP and ADP have no effect by themselves (2). The enhanced stability derived from both ATP and ADP requires divalent cations. MnATP (12 degrees C) shows a much stronger effect than CaATP (7 degrees C) and MgATP (5 degrees C) (3). In the holoenzyme complex or the protein kinase inhibitor/C subunit complex, metal/ATP is also required for enhanced stability; neither the RI or RII subunits nor PKI alone stabilize the C subunit significantly (4). For high thermal stability, the occupation of the second, low-affinity metal-binding site is necessary (5). From these results, we concluded that the adenine moiety works independently from the metal-binding sites, stabilizing the free C subunit by itself. When the beta- and gamma-phosphates are present, divalent metals are required for positioning these phosphates, and two metals are required to achieve thermostability comparable to adenosine alone. The complex containing two metals is the most stable. A comparison of several conformations of the C subunit derived from different crystal structures is given attributing open and closed forms of the C subunit to less and more thermostable enzymes, respectively.     

Carbamoyl phosphate synthetase: closure of the B-domain as a result of nucleotide binding

Carbamoyl phosphate synthetase (CPS) catalyzes the production of carbamoyl phosphate which is subsequently employed in the metabolic pathways responsible for the synthesis of pyrimidine nucleotides or arginine. The catalytic mechanism of the enzyme occurs through three highly reactive intermediates: carboxyphosphate, ammonia, and carbamate. As isolated from Escherichia coli, CPS is an alpha, beta-heterodimeric protein with its three active sites separated by nearly 100 A. In addition, there are separate binding sites for the allosteric regulators, ornithine, and UMP. Given the sizable distances between the three active sites and the allosteric-binding pockets, it has been postulated that domain movements play key roles for intramolecular communication. Here we describe the structure of CPS from E. coli where, indeed, such a domain movement has occurred in response to nucleotide binding. Specifically, the protein was crystallized in the presence of a nonhydrolyzable analogue, AMPPNP, and its structure determined to 2.1 A resolution by X-ray crystallographic analysis. The B-domain of the carbamoyl phosphate synthetic component of the large subunit closes down over the active-site pocket such that some atoms move by more than 7 A relative to that observed in the original structure. The trigger for this movement resides in the hydrogen-bonding interactions between two backbone amide groups (Gly 721 and Gly 722) and the beta- and gamma-phosphate groups of the nucleotide triphosphate. Gly 721 and Gly 722 are located in a Type III' reverse turn, and this type of secondary structural motif is also observed in D-alanine:D-alanine ligase and glutathione synthetase, both of which belong to the "ATP-grasp" superfamily of proteins. Details concerning the geometries of the two active sites contained within the large subunit of CPS are described.     

Mutations in the nucleotide binding loop of adenylate kinase of Escherichia coli

The adk gene of Escherichia coli has been used to overexpress the adenylate kinase protein in two ways: (1) by cloning the adk gene with its own promoter into pEMBL plasmids, which have an increased copy number, and (2) by deleting the adk promoter and cloning the gene behind the regulatable tac promoter. Adenylate kinase comprises up to 40% of the soluble cellular extracts from E. coli strains containing these plasmids. Mutations have been introduced into the gene by site-directed mutagenesis to exchange amino acids in the nucleotide binding loop, which is highly conserved in many mononucleotide binding proteins. The mutation of Lys13----Gln is nearly inactive, whereas the Pro9----Leu and the Gly10----Val mutant proteins have an increased Km for both substrates and a Vmax that is similar to wild type. Proton NMR measurements of the proteins show that a major structural change seems to have taken place for the Pro9----Leu and Gly10----Val mutants. The results are discussed in the light of the kinetic mechanism for adenylate kinase and the three-dimensional structure of the protein.     

Insights into nucleotide binding in protein kinase A using fluorescent adenosine derivatives

The binding of the methylanthraniloyl derivatives of ATP (mant-ATP), ADP (mant-ADP), 2'deoxyATP (mant-2'deoxyATP), and 3'deoxyATP (mant-3'deoxyATP) to the catalytic subunit of protein kinase A was studied to gain insights into the mechanism of nucleotide binding. The binding of the mant nucleotides leads to a large increase in fluorescence energy transfer at 440 nm, allowing direct measurements of nucleotide affinity. The dissociation constant of mant-ADP is identical to that for ADP, while that for mant-ATP is approximately threefold higher than that for ATP. The dissociation constant for mant-3'deoxyATP is approximately fivefold higher than that for 3'deoxyATP while derivatization of 2'deoxyATP does not affect affinity. The time-dependent binding of mant-ATP, mant-2'deoxyATP, and mant-ADP, measured using stopped-flow fluorescence spectroscopy, is best fit to three exponentials. The fast phase is ligand dependent, while the two slower phases are ligand independent. The slower phases are similar but not identical in rate, and have opposite fluorescence amplitudes. Both isomers of mant-ATP are equivalent substrates, as judged by reversed-phase chromatography, although the rate of phosphorylation is approximately 20-fold lower than the natural nucleotide. The kinetic data are consistent with a three-step binding mechanism in which initial association of the nucleotide derivatives produces a highly fluorescent complex. Either one or two conformational changes can occur after the formation of this binary species, but one of the isomerized forms must have low fluorescence compared to the initial binary complex. These data soundly attest to the structural plasticity within the kinase core that may be essential for catalysis. Overall, the mant nucleotides present a useful reporter system for gauging these conformational changes in light of the prevailing three-dimensional models for the enzyme.     

Structures of 5-methylthioribose kinase reveal substrate specificity and unusual mode of nucleotide binding

The methionine salvage pathway is ubiquitous in all organisms, but metabolic variations exist between bacteria and mammals. 5-Methylthioribose (MTR) kinase is a key enzyme in methionine salvage in bacteria and the absence of a mammalian homolog suggests that it is a good target for the design of novel antibiotics. The structures of the apo-form of Bacillus subtilis MTR kinase, as well as its ADP, ADP-PO(4), AMPPCP, and AMPPCP-MTR complexes have been determined. MTR kinase has a bilobal eukaryotic protein kinase fold but exhibits a number of unique features. The protein lacks the DFG motif typically found at the beginning of the activation loop and instead coordinates magnesium via a DXE motif (Asp(250)-Glu(252)). In addition, the glycine-rich loop of the protein, analogous to the "Gly triad" in protein kinases, does not interact extensively with the nucleotide. The MTR substrate-binding site consists of Asp(233) of the catalytic HGD motif, a novel twin arginine motif (Arg(340)/Arg(341)), and a semi-conserved W-loop, which appears to regulate MTR binding specificity. No lobe closure is observed for MTR kinase upon substrate binding. This is probably because the enzyme lacks the lobe closure/inducing interactions between the C-lobe of the protein and the ribosyl moiety of the nucleotide that are typically responsible for lobe closure in protein kinases. The current structures suggest that MTR kinase has a dissociative mechanism.     

Regulation of guanylate cyclase by a guanine nucleotide binding protein, G alpha 2, in Dictyostelium discoideum.

Binding of cyclic AMP (cAMP) to the cell surface receptor induces a transient activation of guanylate cyclase in Dictyostelium discoideum. A frigid mutant (HC85) which lacks G alpha 2, a guanine nucleotide binding protein, does not respond to cAMP. We found that 2,3-dimercapto-1-propanol (BAL) induced a continuous activation both in the frigid and in its parents. Therefore, the BAL-induced continuous activation of guanylate cyclase is independent of G alpha 2. We also found that cAMP enhanced the BAL-induced continuous activation in the frigid mutant. This result suggests that an unidentified signal transduction mechanism from the cAMP-receptor besides the one involving G alpha 2 plays a role in the enhancement of activation. Lastly, we found that the BAL-induced continuous activation was terminated by cAMP in the parental strain, but not in the frigid mutant. Therefore, the cAMP-induced suppression on the BAL-induced continuous activation is mediated through G alpha 2.     

An essential arginyl residue at the nucleotide binding site of creatine kinase.

Treatment of rabbit muscle creatine kinase (EC 2.4.3.2) with either butanedione in borate buffer or phenylglyoxal in Veronal buffer decreases enzymatic activity correlating with the modification of a single arginyl residue per subunit of the dimeric enzyme. Very little activity is lost when modification is performed in the presence of MgATP or MgADP. Nucleotide binding to the modified enzyme is virtually abolished as determined by ultraviolet difference spectroscopy. The data suggest that an arginyl residue plays an essential role in the enzymatic mechanism of creatine kinase, probably as a recognition site for the negatively charged oligophosphate moiety of the nucleotide.     

Interactions of nucleotide analogues with rod outer segment guanylate cyclase.

Light activation of cyclic GMP hydrolysis in rod outer segments is mediated by a G-protein which is active in the GTP-bound form. Substitution of GTP with a nonhydrolyzable GTP analogue is thought to leave the G-protein in a persistently activated state, thereby prolonging the hydrolysis of cyclic GMP. Restoration of cyclic GMP concentration in the cell also depends upon GTP since it is the substrate for guanylate cyclase, but little is known about the effects of GTP analogues on this enzyme. We report here the effects of the analogues of GTP and ATP as inhibitors and substrates of rod disk membrane guanylate cyclase. The rate of cyclic GMP synthesis from GTP in rod disk membranes was about 50 pmol min-1 (nmol of rhodopsin)-1. Analogues of GTP and adenine nucleotides competitively inhibited the cyclase activity. The order of inhibition, with magnesium as metal cofactor, was ATP greater than GMP-PNP greater than AMP-PNP approximately GTP-gamma-S; with manganese, AMP-PNP was more inhibitory than GTP-gamma-S. The inhibition constants, with magnesium as cofactor, were 0.65-2.0 mM for GTP-gamma-S, 0.4-0.8 mM for GMP-PNP, 1.5-2.3 mM for AMP-PNP, and 0.07-0.2 mM for ATP. The fraction of cyclase activity inhibited by analogues was similar at 1 and 0.03 microM calcium. Besides inhibition of cyclase, the analogues also served as its substrates. GTP-gamma-S substituted GTP with about 85% efficiency while GMP-PNP and ATP were about 5 and 7% as efficient, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)     

Adenine nucleotide regulation of particulate guanylate cyclase from rat lung.

Adenine nucleotides activate basal particulate guanylate cyclase in rat lung membranes. Activation is specific for adenine and not guanine, cytidine or uridine nucleotides. The concentration of adenine nucleotides yielding half-maximum activation of particulate guanylate cyclase is 0.1 mM and this nucleotide activates the enzyme by increasing maximum velocity 11-fold without altering affinity for substrate. Activation is specific for particulate guanylate cyclase, since soluble enzyme is inhibited by adenine nucleotides. Similarly, activation is specific for magnesium as the enzyme substrate cation cofactor, since adenine nucleotides inhibit particulate guanylate cyclase when manganese is used. Adenine nucleotide regulation of particulate guanylate cyclase may occur by a different molecular mechanism compared to other activators, since the effects of these nucleotides are synergistic with those of detergent, hemin and atrial natriuretic peptides. Cystamine inhibits adenine nucleotide activation of particulate guanylate cyclase at concentrations having minimal effects on basal enzyme activity suggesting a role for critical sulfhydryls in mechanisms underlying nucleotide regulation of particulate guanylate cyclase. Purification and quantitative recovery of particulate guanylate cyclase by substrate affinity chromatography results in the loss of adenine nucleotide regulation. These data suggest that adenine nucleotides may be important in the regulation of basal and activated particulate guanylate cyclase and may be mediated by an adenine nucleotide-binding protein which is separate from that enzyme.     

Characterization of nitrosoalkane binding and activation of soluble guanylate cyclase

Soluble guanylate cyclase (sGC) is the primary receptor for the signaling agent nitric oxide (NO). Electronic absorption and resonance Raman spectroscopy were used to show that nitrosoalkanes bind to the heme of sGC to form six-coordinate, low-spin complexes. In the sGC-nitrosopentane complex, a band assigned to an Fe-N stretching vibration is observed at 543 cm(-)(1) which is similar to values reported for other six-coordinate NO-bound hemoproteins. Nitrosoalkanes activate sGC 2-6-fold and synergize with YC-1, a synthetic benzylindazole derivative, to activate the enzyme 11-47-fold. In addition, the observed off-rates of nitrosoalkanes from sGC were found to be dependent on the alkyl chain length. A linear correlation was found between the observed off-rates and the alkyl chain length which suggests that the sGC heme has a large hydrophobic distal ligand-binding pocket. Together, these data show that nitrosoalkanes are a novel class of heme-based sGC activators and suggest that heme ligation is a general requirement for YC-1 synergism.