Search

The two exocyclic oxygen atoms at phosphorus of cAMP have been replaced by a sulfur atom or by a dimethylamino group. These substitutions introduce chirality at the phosphorus atom; therefore, two diastereoisomers are known for each derivative: (SP)-cAMPS, (RP)-cAMPS, (SP)-cAMPN(CH3)2, and RP-cAMPN(CH3)2. We have investigated the agonistic and antagonistic activities of these compounds in four cAMP-dependent reactions: activation of the cellular slime mold Dictyostelium discoideum via its cell surface cAMP receptor, and phosphorylation by cAMP-dependent protein kinases type I, type II (both mammalian enzymes), and type D (derived from D. discoideum). The results show that 1) the compounds (SP)-cAMPS and (SP)-cAMPN(CH3)2 are (mostly full) agonists for the four proteins. Half-maximal activation is at micromolar concentrations (0.8-7 microM). 2) (RP)-cAMPS is a full antagonist for the cell surface receptor and protein kinases type I and II, with apparent inhibition constants between 0.8 and 8 microM. This compound is a partial agonist for protein kinase type D, where it induces maximally 50% activation of the enzyme if compared with cAMP. 3) (RP)-cAMPN(CH3)2 is a full antagonist for the cell surface receptor, and for protein kinase type II. This compound is a partial agonist for protein kinase type I (at least 50% activation if compared with cAMP), and inactive for protein kinase type D. This derivative is at least 25-fold less active as an antagonist than (RP)-cAMPS. 4) The activity of mixtures of different concentrations of the antagonist (RP)-cAMPS with different concentrations of cAMP reveals that the compound is a competitive antagonist of cAMP at micromolar concentrations.

Bovine heart cyclic AMP phosphodiesterase, which has a requirement for Mg2+, hydrolyses cyclic AMP with inversion of configuration at the phosphorus atom, but only the (Sp)-diastereoisomer of adenosine cyclic 3′:5′-phosphorothioate is hydrolysed by this enzyme. By contrast, the low-affinity yeast cyclic AMP phosphodiesterase, which contains tightly bound Zn2+, hydrolyses both the (Sp)- and the (Rp)-diastereoisomers of adenosine cyclic 3′:5′-phosphorothioate, the (Rp)-diastereoisomer being the preferred substrate under V max. conditions. Both of the diastereoisomers of adenosine cyclic 3′:5′-phosphorothioate, as well as cyclic AMP, are hydrolysed with inversion of configuration at the phosphorus atom by the yeast enzyme. It is proposed that, with both enzymes, the bivalent metal ion co-ordinates with the phosphate residue of the substrate, and that hydrolysis is catalysed by a direct ‘in-line’ mechanism.

The conversion of L-lysine to L-β-lysine is catalyzed by lysine 2,3-aminomutase. The reaction involves the interchange of the 2-amino group of lysine with a hydrogen at carbon 3. As such the reaction is formally analogous to adenosylcobalamin-dependent rearrangements. However, the enzyme does not contain and is not activated by this coenzyme. Instead it contains iron and pyridoxal phosphate and is activated by S-adenosylmethionine. Earlier experiments implicated adenosyl-C-5′ of S-adenosylmethionine in the hydrogen transfer mechanism, apparently in a role similar or analogous to that of adenosyl moiety of adenosylcobalamin in the B12-dependent rearrangements. The question of whether both hydrogens or only one hydrogen at adenosyl-C-5′ participate in the hydrogen-transfer process has been addressed by carrying out the lysine 2,3-aminomutase reaction with S-[5′-3H] adenosylmethionine in the presence of 10 times its molar concentration of enzyme. Under these conditions allof the tritium appeared in lysine and β-lysine, showing that C-5′-hydrogens participate. To determine whether hydrogen transfer is compulsorily intermolecular and intramolecular, various molar ratios of [3,3-2H2]lysine and unlabeled lysine were submitted to the action of lysine 2,3-aminomutase under conditions in which 10–15% conversion to β-lysine occurred. Mass spectral analysis of the β-lysine for monodeutero and dideutero species showed conclusively that hydrogen transfer is both intramolecular and intermolecular. The results quantitatively support our postulate that activation of the enzyme involves a transformation of S-adenosylmethionine into a form that promotes the generation of an adenosyl-5′ free radical, which abstracts hydrogen from lysine to form 5′-deoxyadenosine as an intermediate.

A single sulfur substitution for either the axial or the equatorial exocyclic oxygen of adenosine cyclic 3', 5'-phosphate (cAMP) results in diastereometric phosphorothioate analogs of cAMP with agonist versus antagonist properties towards activation of cAMP-dependent protein kinase. Sulfur substitutions for both of the exocyclic oxygens of cAMP results in a dithioate analog of cAMP, adenosine cyclic 3', 5'-phosphorodithioate (cAMPS2), which has antagonist properties. cAMPS2 displaced [3H]cAMP from the binding sites on bovine heart Type II cAMP-dependent protein kinase as demonstrated by equilibrium dialysis experiments with an apparent Kd of 6.3 microM. The addition of 10, 30, or 100 microM cAMPS2 when measuring cAMP-induced activation of pure porcine heart Type II cAMP-dependent protein kinase resulted in a concentration-dependent increase in the amount of cAMP required to produce half-maximal activation (EC50). A plot of the EC50 values as a function of the cAMPS2 concentration resulted in a straight line from which a KI value of 4 microM was derived. cAMPS2 had no significant effect on the degree of cooperativity (n) of cAMP activation of the holoenzyme. These data suggest that the most important structural requirement for the dissociation of the holoenzyme is an equatorial exocyclic oxygen.

The diastereomeric forms of adenosine cyclic 3‘,5‘-phosphorothioate, Rp cAMPS and Sp cAMPS, were studied in isolated hepatocytes from fed rats for their ability to interact with the intracellular cAMP-dependent protein kinase and to affect the phosphorylase kinase-phosphorylase glycogenolytic cascade. Incubation of the cells with increasing concentrations of Sp cAMPS produced a concentration-dependent activation of cAMP-dependent protein kinase with a concomitant increase in the glycogenolytic rate. Half-maximal and maximal velocities of glycogenolysis were reached at 8 X 10(-7) and 1 X 10(-5) M Sp cAMPS, respectively. Incubation of the cells with 10(-9) to 10(-4) M Rp cAMPS had no effect on basal glucose production or on cAMP-dependent protein kinase activity. Incubation of the cells simultaneously with 3 X 10(-6) M Sp cAMPS and increasing concentrations of Rp cAMPS produced half-maximal inhibition of glycogenolysis at 1 X 10(-5) M Rp cAMPS and maximal inhibition at 1 X 10(-4) M. The concentrations of Sp cAMPS required for half-maximal and maximal activation of glycogenolysis were increased 10-fold when 1 X 10(-5) M Rp cAMPS was present. These data imply that Sp cAMPS is a cAMP-agonist while Rp cAMPS is a cAMP-antagonist.

Measurements of the 18O‐induced isotope shifts of 31P NMR signals for cyclohexyl‐, tetramethyl‐ and tetrabutylammonium salts of ethylphosphorothioate show that in D2O a P‐O bond order of 1.5 exists consistent with the structural formulation (I).

The protected nucleoside cyclic phosphate (I) reacts with aniline (II) to give the phosphoranilidates (III) and (VI), which after separation are converted in a stereoretentive manner to the phosphorothioates (V) or (VII).