Search

Novel inhibitors of lupin diadenosine 5',5'''-P(1),P(4)-tetraphosphate (Ap(4)A) hydrolase have been identified by in silico screening of a large virtual chemical library. Compounds were ranked on the basis of a consensus from six scoring functions. From the top 100 ranked compounds six were selected and initially screened for inhibitory activity using a single concentration isothermal titration calorimetry assay. Two of these compounds that showed excellent solubility properties were further analyzed, but only one [NSC51531; 2-((8-hydroxy-4-(4-methyl-2-sulfoanilino)-9,10-dioxo-9,10-dihydro-1-anthracenyl)amino)-5-methylbenzenesulfonic acid] exhibited competitive inhibition with a K(i) of 1 microM. A structural analogue of this compound also exhibited competitive inhibition with a comparable K(i) of 2.9 microM. (1)H, (15)N NMR spectroscopy was used to map the binding site of NSC51531 on lupin Ap(4)A hydrolase and demonstrated that the compound bound specifically in the substrate-binding site, consistent with the competitive inhibition results. Binding of NSC51531 to the human form of Ap(4)A hydrolase is nonspecific, suggesting that this compound may represent a useful lead in the design of specific inhibitors of the plant-like form of Ap(4)A hydrolases.

The DHH superfamily human protein h-prune, a binding partner of the metastasis suppressor nm23-H1, is frequently overexpressed in metastatic cancers. From an evolutionary perspective, h-prune is very close to eukaryotic exopolyphosphatases. Here, we show for the first time that h-prune efficiently hydrolyzes short-chain polyphosphates (k cat of 3-40 s (-1)), including inorganic tripoly- and tetrapolyphosphates and nucleoside 5'-tetraphosphates. Long-chain inorganic polyphosphates (>or=25 phosphate residues) are converted more slowly, whereas pyrophosphate and nucleoside triphosphates are not hydrolyzed. The reaction requires a divalent metal cofactor, such as Mg (2+), Co (2+), or Mn (2+), which activates both the enzyme and substrate. Notably, the exopolyphosphatase activity of h-prune is suppressed by nm23-H1, long-chain polyphosphates and pyrophosphate, which may be potential physiological regulators. Nucleoside triphosphates, diadenosine hexaphosphate, cAMP, and dipyridamole (inhibitor of phosphodiesterase) do not affect this activity. Mutation of seven single residues corresponding to those found in the active site of yeast exopolyphosphatase led to a severe decrease in h-prune activity, whereas one variant enzyme exhibited enhanced activity. Our results collectively suggest that prune is the missing exopolyphosphatase in animals and support the hypothesis that the metastatic effects of h-prune are modulated by inorganic polyphosphates, which are increasingly recognized as critical regulators in cells.

Engineering proteins to withstand a broad range of conditions continues to be a coveted objective, holding the potential to advance biomedicine, industry, and our understanding of disease. One way of achieving this goal lies in elucidating the underlying interactions that define protein stability. It has been shown that the hydrophobic effect, hydrogen bonding, and packing interactions between residues in the protein interior are dominant factors that define protein stability. The role of surface residues in protein stability has received much less attention. It has been believed that surface residues are not important for protein stability particularly because their interactions with the solvent should be similar in the native and unfolded states. In the case of surface charged residues, it was sometimes argued that solvent exposure meant that the high dielectric of the solvent will further decrease the strength of the charge-charge interactions. In this paper, we challenge the notion that the surface charged residues are not important for protein stability. We computationally redesigned sequences of five different proteins to optimize the surface charge-charge interactions. All redesigned proteins exhibited a significant increase in stability relative to their parent proteins, as experimentally determined by circular dichroism spectroscopy and differential scanning calorimetry. These results suggest that surface charge-charge interactions are important for protein stability and that rational optimization of charge-charge interactions on the protein surface can be a viable strategy for enhancing protein stability.

The human tumor suppressor Fhit is a homodimeric histidine triad (HIT) protein of 147 amino acids which has Ap(3)A hydrolase activity. We have recently discovered that Fhit is phosphorylated in vivo and is phosphorylated in vitro by Src kinase [Pekarsky, Y., Garrison, P. N., Palamarchuk, A., Zanesi, N., Aqeilan, R. I., Huebner, K., Barnes, L. D., and Croce, C. M. (2004) Proc. Natl. Acad. Sci. U.S.A. 101, 3775-3779]. Now we have coexpressed Fhit with the elk tyrosine kinase in Escherichia coli to generate phosphorylated forms of Fhit. Unphosphorylated Fhit, Fhit phosphorylated on one subunit, and Fhit phosphorylated on both subunits were purified to apparent homogeneity by column chromatography on anion-exchange and gel filtration resins. MALDI-TOF and HPLC-ESI tandem mass spectrometry of intact Fhit and proteolytic peptides of Fhit demonstrated that Fhit is phosphorylated on Y(114) on either one or both subunits. Monophosphorylated Fhit exhibited monophasic kinetics with K(m) and k(cat) values approximately 2- and approximately 7-fold lower, respectively, than the corresponding values for unphosphorylated Fhit. Diphosphorylated Fhit exhibited biphasic kinetics. One site had K(m) and k(cat) values approximately 2- and approximately 140-fold lower, respectively, than the corresponding values for unphosphorylated Fhit. The second site had a K(m) approximately 60-fold higher and a k(cat) approximately 6-fold lower than the corresponding values for unphosphorylated Fhit. The unexpected kinetic patterns for the phosphorylated forms suggest the system may be enzymologically novel. The decreases in the values of K(m) and k(cat) for the phosphorylated forms in comparison to those of unphosphorylated Fhit favor the formation and lifetime of the Fhit-Ap(3)A complex, which may enhance the tumor suppressor activity of Fhit.

Acylphosphatases catalyze the hydrolysis of the carboxyl-phosphate bond in acyl phosphates. Although acylphosphatase-like sequences are found in all three domains of life, no structure of acylphosphatase has been reported for bacteria and archaea so far. Here, we report the characterization of enzymatic activities and crystal structure of an archaeal acylphosphatase. A putative acylphosphatase gene (PhAcP) was cloned from the genomic DNA of Pyrococcus horikoshii and was expressed in Escherichia coli. Enzymatic parameters of the recombinant PhAcP were measured using benzoyl phosphate as the substrate. Our data suggest that, while PhAcP is less efficient than other mammalian homologues at 25 degrees C, the thermophilic enzyme is fully active at the optimal growth temperature (98 degrees C) of P. horikoshii. PhAcP is extremely stable; its apparent melting temperature was 111.5 degrees C and free energy of unfolding at 25 degrees C was 54 kJ mol(-)(1). The 1.5 A crystal structure of PhAcP adopts an alpha/beta sandwich fold that is common to other acylphosphatases. PhAcP forms a dimer in the crystal structure via antiparallel association of strand 4. Structural comparison to mesophilic acylphosphatases reveals significant differences in the conformation of the L5 loop connecting strands 4 and 5. The extreme thermostability of PhAcP can be attributed to an extensive ion-pair network consisting of 13 charge residues on the beta sheet of the protein. The reduced catalytic efficiency of PhAcP at 25 degrees C may be due to a less flexible active-site residue, Arg20, which forms a salt bridge to the C-terminal carboxyl group. New insights into catalysis were gained by docking acetyl phosphate to the active site of PhAcP.

The human fragile histidine triad protein Fhit catalyzes the Mg(2+)-dependent hydrolysis of P(1)-5'-O-adenosine-P(3)-5'-O-adenosine triphosphate, Ap(3)A, to AMP and ADP. The reaction is thought to follow a two-step mechanism, in which the complex of Ap(3)A and Mg(2+) reacts in the first step with His96 of the enzyme to form a covalent Fhit-AMP intermediate and release MgADP. In the second step, the intermediate Fhit-AMP undergoes hydrolysis to AMP and Fhit. The mechanism is inspired by the chain-fold similarities of Fhit to galactose-1-phosphate uridylyltransferase, which functions by an analogous mechanism, and the observation of overall retention in configuration at phosphorus in the action of Fhit (Abend, A., Garrison, P. N., Barnes, L. D., and Frey, P. A. (1999) Biochemistry 38, 3668-3676). Direct evidence in support of this mechanism is reported herein. Reaction of Fhit with [8,8'-(3)H]-Ap(3)A and denaturation of the enzyme in the steady state leads to protein-bound tritium corresponding to 11% of the active sites. Similar experiments with the poor substrate MgATP leads to 0.9% labeling. The mutated protein H96G-Fhit is completely inactive against MgAp(3)A. However, it is chemically rescued by free histidine. H96G-Fhit also catalyzes the hydrolysis of adenosine-5'-phosphoimidazolide, AMP-Im, and of adenosine-5'-phospho-N-methylimidazolide, AMP-N-MeIm. The hydrolyses of AMP-Im and of AMP-N-MeIm by H96G-Fhit are thought to represent chemical rescue of the covalent Fhit-AMP intermediate. Wild-type Fhit is also found to catalyze the hydrolyses of AMP-Im and of AMP-N-MeIm nearly as efficiently as the hydrolysis of MgAp(3)A. The results indicate that Mg(2+) in the reaction of Ap(3)A is required for the first step, the formation of the covalent intermediate Fhit-AMP, and not for the hydrolysis of the intermediate in the second step.

A potentially amyloidogenic protein has to be at least partially unfolded to form amyloid aggregates. However, aggregation of the partially or totally unfolded state of a protein is modulated by at least three other factors: hydrophobicity, propensity to form secondary structure, and net charge of the polypeptide chain. We propose to evaluate the relative importance of net charge, as opposed to the other factors, on protein aggregation and amyloidogenicity. For this aim, we have used two homologous proteins that were previously shown to be able to form amyloid fibrils in vitro, the N-terminal domain of HypF from Escherichia coli (HypF-N) and human muscle acylphosphatase (AcP). The aggregation process from an ensemble of partially unfolded conformations is ca. 1000-fold faster for HypF-N than for AcP. This difference can mainly be attributed to a higher hydrophobicity and a lower net charge for HypF-N than for AcP. By using protein engineering methods, we have decreased the net charge of AcP to a value identical to that of wild-type HypF-N and increased the net charge of HypF-N to a value identical to that of wild-type AcP. Amino acid substitutions were selected to minimize changes in hydrophobicity and secondary structure propensities. We were able to estimate that the difference in net charge between the two wild-type proteins contributes to 20-25% of the difference in their aggregation rates. An understanding of the relative influences of these forces in protein aggregation has implications for elucidating the complexity of the aggregation process, for predicting the effect of natural mutations, and for accurate protein design.

Baculovirus phosphatase (BVP) and mammalian capping enzyme (Mce1) are members of the RNA triphosphatase branch of the cysteine phosphatase superfamily. Although RNA triphosphatases have a core alpha/beta fold similar to other cysteine phosphatases, there is little conservation of primary structure outside of the cysteine-containing P-loop motif, HCxxxxxR(S/T), that comprises the active site. However, there is extensive primary structure conservation between members of the RNA triphosphatase branch, whether from cellular or viral sources and whether they are bifunctional capping enzymes such as Mce1 or monofunctional RNA phosphatases such as BVP. To evaluate the functional significance of such sequence conservation, we performed a mutational analysis of 14 residues of BVP. We identified three side chains (Trp6, Lys25, and Arg153) as essential for triphosphatase activity in vitro, i.e., W6A, K25A, and R153A were <0.1% as active as wild-type BVP, and were unable to complement a yeast RNA triphosphatase null mutant in vivo. Six other BVP residues (Thr62, Tyr67, Tyr68, Lys82, Glu158, and Arg159) were deemed functionally important, i.e., Ala mutations reduced triphosphatase activity to <20% of wild-type. On the basis of the locations of the equivalent amino acids in the Mce1 crystal structure, we surmise that the essential/important BVP residues ensure proper conformation of the catalytic P-loop (e.g., Arg153 and Tyr68) or other elements of the tertiary structure. Our results highlight a conserved Trp6-Lys25 pi-cation pair essential for BVP function.

Ecto-nucleoside triphosphate diphosphohydrolase 3 (eNTPDase-3, also known as HB6 and CD39L3) is a membrane-associated ecto-apyrase. Only a few functionally significant residues have been elucidated for this enzyme, as well as for the whole family of eNTPDase enzymes. Four highly conserved regions (apyrase conserved regions, ACRs) have been identified in all the members of eNTPDase family, suggesting their importance for biological activity. In an effort to identify those amino acids important for the catalytic activity of the eNTPDase family, as well as those residues mediating substrate specificity, 11 point mutations of 7 amino acid residues in ACR1-4 of eNTPDase-3 were constructed by site-directed mutagenesis. Mutagenesis of asparagine 191 to alanine (N191A), glutamine 226 to alanine (Q226A), and arginine 67 to glycine (R67G) resulted in an increase in the rates of hydrolysis of nucleoside diphosphates relative to triphosphates. Mutagenesis of arginine 146 to proline (R146P) essentially converted the eNTPDase-3 ecto-apyrase to an ecto-ATPase (eNTPDase-2), mainly by decreasing the hydrolysis rates for nucleoside diphosphates. The Q226A mutant exhibited a change in the divalent cation requirement for nucleotidase activity relative to the wild-type and the other mutants. Mutation of glutamate 182 to aspartate (E182D) or glutamine (E182Q), and mutation of serine 224 to alanine (S224A) completely abolished enzymatic activity. We conclude that the residues corresponding to eNTPDase-3 glutamate 182 in ACR3 and serine 224 in ACR4 are essential for the enzymatic activity of eNTPDases in general, and that arginine 67, arginine 146, asparagine 191, and glutamine 226 are important for determining substrate specificity for human ecto-nucleoside triphosphate diphosphohydrolase 3.

E-NTPDases are extracellular enzymes that hydrolyze nucleotides. The human E-NTPDase gene family currently consists of five reported members (CD39, CD39L1, CD39L2, CD39L3, and CD39L4). Both membrane-bound and secreted family members have been predicted by encoded transmembrane and leader peptide motifs. In this report, we demonstrate that the human CD39L2 gene is expressed predominantly in the heart. In situ hybridization results from heart indicate that the CD39L2 message is expressed in muscle and capillary endothelial cells. We also show that the CD39L2 gene encodes an extracellular E-NTPDase. Flow cytometric experiments show that transiently expressed CD39L2 is present on the surface of COS-7 cells. Transfected cells also produce recombinant glycosylated protein in the medium, and this process can be blocked by brefeldin A, an inhibitor of the mammalian secretory pathway. The enzymology of CD39L2 shows characteristic features of a typical E-NTPDase, but with a much higher degree of specificity for NDPs over NTPs as enzymatic substrates. The kinetics of the ADPase activity exhibit positive cooperativity. The predominance of CD39L2 expression in the heart supports a functional role in regulating platelet activation and recruitment in this organ.

The structural and catalytic importance of the 15-21 and 42-45 loop residues of the acylphosphatase muscular isoenzyme has been investigated by oligonucleotide-directed mutagenesis. Seven mutants involving conserved residues of the two loops have been prepared and characterized for structural, kinetic, and stability features by using different spectroscopic techniques and compared to the wild-type enzyme. The results are discussed in light of the crystal structure of the highly homologous common type acylphosphatase [Thunnissen et al. (1997) Structure 5, 69-79]. A differential role of the two loops has emerged: the 15-21 and the 42-45 loops appear mainly involved in active site formation and enzyme structural stabilization, respectively. These conclusions are supported by a strong impairment of the catalytic efficiency, in terms of enzymatic activity and substrate binding capability, for most of the 15-21 loop mutants. In particular, the Gly15Ala mutant is completely inactive and displays a native-like overall fold, indicating that the correct geometry of the 15-21 loop is an essential requisite for optimal enzymatic catalysis. Instead, the Gly45Ala mutant, though revealing unchanged catalytic properties, shows a considerably reduced conformational stability, as judged by circular dichroism and 1H NMR spectroscopy. This finding confirms previous results relative to Thr42 and Thr46 residues [Taddei et al. (1996) Biochemistry 35, 7077-7083] underlining the structural importance of the 42-45 loop as a linker for the two beta alpha beta units constituting the overall enzyme structure.

Asn41, Thr42, and Thr46 are invariant residues in both muscle and erythrocyte acylphosphatases isolated so far. Horse muscle acylphosphatase solution structure suggests their close spatial relationship to Arg23, the main substrate binding site. The catalytic and structural role of such residues, as well as their influence on muscle acylphosphatase stability, was investigated by preparing several gene mutants (Thr42Ala, Thr46Ala, Asn41Ala, Asn41Ser, and Asn41Gln) by oligonucleotide-directed mutagenesis. The mutated genes were cloned and expressed in Escherichia coli, and the mutant enzymes were purified by affinity chromatography and investigated as compared to the wild-type enzyme. The specific activity and substrate affinity of Thr42 and Thr46 mutants were not significantly affected. On the contrary, Asn41 mutants showed a residual negligible activity (about 0.05-0.15% as compared to wild-type enzyme), though maintaining an unchanged binding capability of both substrate and inorganic phosphate, an enzyme competitive inhibitor. According to the 1H nuclear magnetic resonance spectroscopy and circular dichroism results, all mutants elicited well-constrained native-like secondary and tertiary structures. Thermodynamic parameters, as calculated from circular dichroism data, demonstrated a significantly decreased stability of the Thr42 mutant under increasing temperatures and urea concentrations. The reported results strongly support a direct participation of Asn41 to the enzyme catalytic mechanism, indicating that Asn41 mutants may well represent a useful tool for the investigation of the enzyme physiological function by the negative dominant approach.