Rosa María Pinto

Cyclic ADP-ribose (cADPR) is a messenger for Ca²⁺ mobilization. Its turnover is believed to occur by glycohydrolysis to ADP-ribose. However, ADP-ribose/CDP-alcohol diphosphatase (ADPRibase-Mn) acts as cADPR phosphohydrolase with much lower efficiency than on its major substrates. Recently, we showed that mutagenesis of human ADPRibase-Mn at Phe³⁷, Leu¹⁹⁶ and Cys²⁵³ alters its specificity: the best substrate of the mutant F37A + L196F + C253A is cADPR by a short difference, Cys²⁵³ mutation being essential for cADPR preference. Its proximity to the ‘northern’ ribose of cADPR in docking models indicates Cys²⁵³ is a steric constraint for cADPR positioning. Aiming to obtain a specific cADPR phosphohydrolase, new mutations were tested at Asp²⁵⁰, Val²⁵², Cys²⁵³ and Thr²⁷⁹, all near the ‘northern’ ribose. First, the mutant F37A + L196F + C253G, with a smaller residue 253 (Ala > Gly), showed increased cADPR specificity. Then, the mutant F37A + L196F + V252A + C253G, with another residue made smaller (Val > Ala), displayed the desired specificity, with cADPR k_cat/K_M ≈20–200-fold larger than for any other substrate. When tested in nucleotide mixtures, cADPR was exhausted while others remained unaltered. We suggest that the specific cADPR phosphohydrolase, by cell or organism transgenesis, or the designed mutations, by genome editing, provide opportunities to study the effect of cADPR depletion on the many systems where it intervenes.

Endogenous cyclic diadenylate phosphodiesterase activity was accidentally detected in lysates of Escherichia coli BL21. Since this kind of activity is uncommon in Gram-negative bacteria, its identification was undertaken. After partial purification and analysis by denaturing gel electrophoresis, renatured activity correlated with a protein identified by fingerprinting as CpdB (cpdB gene product), which is annotated as 3´-nucleotidase / 2´,3´-cyclic-mononucleotide phosphodiesterase, and it is synthesized as a precursor protein with a signal sequence removable upon export to the periplasm. It has never been studied as a recombinant protein. The coding sequence of mature CpdB was cloned and expressed as a GST fusion protein. The study of the purified recombinant protein, separated from GST, confirmed CpdB annotation. The assay of catalytic efficiencies (k_cat/K_m) for a large substrate set revealed novel CpdB features, including very high efficiencies for 3´-AMP and 2´,3´-cyclic mononucleotides, and previously unknown activities on cyclic and linear dinucleotides. The catalytic efficiencies of the latter activities, though low in relative terms when compared to the major ones, are far from negligible. Actually, they are perfectly comparable to those of the ‘average’ enzyme and the known, bona fide cyclic dinucleotide phosphodiesterases. On the other hand, CpdB differs from these enzymes in its extracytoplasmic location and in the absence of EAL, HD and DHH domains. Instead, it contains the domains of the 5´-nucleotidase family pertaining to the metallophosphoesterase superfamily, although CpdB lacks 5´-nucleotidase activity. The possibility that the extracytoplasmic activity of CpdB on cyclic dinucleotides could have physiological meaning is discussed.

Among metallo-dependent phosphatases, ADP-ribose/CDP-alcohol diphosphatases form a protein family (ADPRibase-Mn-like) mainly restricted, in eukaryotes, to vertebrates and plants, with preferential expression, at least in rodents, in immune cells. Rat and zebrafish ADPRibase-Mn, the only biochemically studied, are phosphohydrolases of ADP-ribose and, somewhat less efficiently, of CDP-alcohols and 2´,3´-cAMP. Furthermore, the rat but not the zebrafish enzyme displays a unique phosphohydrolytic activity on cyclic ADP-ribose. The molecular basis of such specificity is unknown. Human ADPRibase-Mn showed similar activities, including cyclic ADP-ribose phosphohydrolase, which seems thus common to mammalian ADPRibase-Mn. Substrate docking on a homology model of human ADPRibase-Mn suggested possible interactions of ADP-ribose with seven residues located, with one exception (Cys²⁵³), either within the metallo-dependent phosphatases signature (Gln²⁷, Asn¹¹⁰, His¹¹¹), or in unique structural regions of the ADPRibase-Mn family: s2s3 (Phe³⁷ and Arg⁴³) and h7h8 (Phe²¹⁰), around the active site entrance. Mutants were constructed, and kinetic parameters for ADP-ribose, CDP-choline, 2´,3´-cAMP and cyclic ADP-ribose were determined. Phe³⁷ was needed for ADP-ribose preference without catalytic effect, as indicated by the increased ADP-ribose K_m and unchanged k_cat of F37A-ADPRibase-Mn, while the K_m values for the other substrates were little affected. Arg⁴³ was essential for catalysis as indicated by the drastic efficiency loss shown by R43A-ADPRibase-Mn. Unexpectedly, Cys²⁵³ was hindering for cADPR phosphohydrolase, as indicated by the specific tenfold gain of efficiency of C253A-ADPRibase-Mn with cyclic ADP-ribose. This allowed the design of a triple mutant (F37A+L196F+C253A) for which cyclic ADP-ribose was the best substrate, with a catalytic efficiency of 3.5´10⁴ M^-1s^-1 versus 4´10³ M^-1s^-1 of the wild type.

A novel enzyme, induced by choline, ethanolamine, glycine betaine or dimethylglycine, was released at low temperature and phosphate from Pseudomonas fluorescens (CECT 7229) suspensions at low cell densities. It is a CDP-ethanolamine pyrophosphatase/(dihexanoyl)glycerophosphoethanolamine phosphodiesterase (CGDEase) less active on choline derivatives, and inactive on long-chain phospholipids, CDP-glycerol and other NDP-X compounds. The reaction pattern was typical of phospholipase C (PLC), as either phosphoethanolamine or phosphocholine was produced. Peptide-mass analyses, gene cloning and expression provided a molecular identity for CGDEase. Bioinformatic studies assigned it to the PLC branch of the phospholipase C/acid phosphatase (PLC/APase) superfamily, revealed an irregular phylogenetic distribution of close CGDEase relatives, and suggested their genes are not in operons or conserved contexts. A theoretical CGDEase structure was supported by mutagenesis of two predicted active-site residues, which yielded essentially inactive mutants. Biological relevance is supported by comparisons with CGDEase relatives, induction by osmoprotectants (not by osmotic stress itself) and repression by micromolar phosphate. The low bacterial density requirement was related to phosphate liberation from lysed bacteria in denser populations, rather than to a classical quorum-sensing effect. The results fit better a CGDEase role in phosphate scavenging than in osmoprotection.

Flavins, comprising flavin mononucleotide (FMN), flavin adenine dinucleotide (FAD), and riboflavin (RF, vitamin B₂), play important roles in numerous redox reactions such as those taking place in the electron-transfer chains of mitochondria in all eukaryotes and of plastids in plants. A selective chemosensor for flavins would be useful not only in the investigation of metabolic processes but also in the diagnosis of diseases related to flavins; such a sensor is presently unavailable. Herein, we report the first bifunctional chemosensor (PTZ-DPA) for flavins. PTZ-DPA consists of bis(Zn²⁺-dipicolylamine) and phenothiazine. Bis(Zn²⁺-dipicolylamine) (referred to here as XyDPA) was found to be an excellent catalyst in the conversion of FAD into cyclic FMN (riboflavin 4′,5′-cyclic phosphate, cFMN) under physiological conditions, even at pH 7.4 and 27 °C, with less than 1 mol % of substrate. Utilizing XyDPA’s superior function as an artificial FMN cyclase and phenothiazine as an electron donor able to quench the fluorescence of an isoalloxazine ring, PTZ-DPA enabled selective fluorescent discrimination of flavins (FMN, FAD, and RF): FAD shows ON(+), FMN shows OFF(−), and RF shows NO(0) fluorescence changes upon the addition of PTZ-DPA. With this selective sensing property, PTZ-DPA is applicable to real-time fluorescent monitoring of riboflavin kinase (RF to FMN), alkaline phosphatase (FMN to RF), and FAD synthetase (FMN to FAD).

An enzyme with FAD-AMP lyase (cyclizing) activity, splitting FAD to AMP and riboflavin 4‘,5‘-phosphate (cFMN), was recently identified [Fraiz, F., et al. (1998) Biochem. J. 330, 881−888]. Now, it has been purified to apparent homogeneity from a rat liver supernatant, by a procedure that includes affinity for ADP-agarose (adsorption required the activating cation Mn²⁺ and desorption required its removal), to a final activity of 2.2 units/mg after a 240-fold purification with a 15% yield. By SDS−PAGE, only one protein band was observed (M_r = 59 000). The correspondence between protein and enzyme activity was demonstrated by renaturation after SDS−PAGE, by gradient ultracentrifugation followed by analytical SDS−PAGE, and by native PAGE with visualization of enzyme activity by fluorescence. A native M_r of 100 000 (ultracentrifugation) or 140 000 (gel filtration) indicated that FAD-AMP lyase could be a dimer. The enzyme required millimolar concentrations of Mn²⁺ or Co²⁺, exhibited different optimum pH values with these cations (pH 8.5 or 7.3, respectively), and was strongly inhibited by ADP or ATP, but not by dADP, dATP, or the reaction products AMP and cFMN. A specificity study was conducted with 35 compounds related to FAD, mostly nucleoside diphosphate-X (NDP-X) derivatives. Besides FAD, the enzyme split 11 of these compounds with the pattern NDP-X → NMP + PX. Structure−activity correlations of substrates, nonsubstrates, and inhibitors, and the comparison of the enzymic reactivities of NDP-X compounds with their susceptibilities to metal-dependent chemical degradation, pinpointed the following specificity pattern. FAD-AMP lyase splits ribonucleoside diphosphate-X compounds in which X is an acyclic or cyclic monosaccharide or derivative bearing an X-OH group that is able to attack internally the proximal phosphorus with the geometry necessary to form a PX product, either a five-atom monocyclic phosphodiester or a cis-bicyclic phosphodiester−pyranose fusion. For instance, NDP-glucose and GDP-α-l-fucose were substrates, but dTDP-glucose, NDP-mannose, and GDP-β-l-fucose were not. Judging from k_cat/K_m ratios, we found the best substrate to be FAD, followed closely by ADP-glucose (k_cat/K_m only 2-fold lower, but not a physiological compound in mammals), whereas other substrates exhibited 50−500-fold lower k_cat/K_m values. However, there was no evidence for specific flavin recognition. Instead, what seems to be recognized is the NDP moiety of NDP-X, with a strong preference for ADP-X. Splitting would then depend on the presence of an adequate X-OH group. The possibility that, besides FAD, there could be in mammals other ADP-X substrates of FAD-AMP lyase is discussed, with emphasis placed on some ADP-ribose derivatives.