A preparation of poly(adenosine diphosphoribose) synthase obtained from pigeon liver nuclei has been used to make poly(adenosine diphosphoribose) with an average chain length of 20. Digestion of the purified poly(adenosine diphosphoribose) with snake venom phosphodiesterase (oligonucleate 5'-nucleotidohydrolase; EC 3.1.4.1) gave the monomer, 2'-5"-phosphoribosyl-5'-AMP. After purification of the monomer on a Dowex-1 column, further digestion with alkaline phosphatase [orthophosphoric-monoester phosphohydrolase (alkaline optimum); EC 3.1.3.1] yielded the dephosphorylated product, 2'-ribosyl adenosine. Nuclear magnetic resonance spectra at 360MHz of the 2'-ribosyl adenosine were obtained in [2H6Jdi-methylsulfoxide, which allows direct observation of the hydroxyl protons. These spectra show the absence of the adenosine 2'-hydroxyl proton, thus confirming the 2' position as the site of attachment of the ribose to the adenosine moiety. Comparison of the coupling constants and the chemical shifts of the ribose hydroxyl protons of 2'-ribosyl adenosine with the model compounds a-and fl-methylribofuranoside establishes an a (1" 2') glycosidic linkage in the monomer. No evidence was found for heterogeneity in either the site of attachment or configuration of the linkage in the 2'(5"-phosphoribosyl)5'-AMP. The nuclei of eukaryotic cells contain an enzyme that converts NAD+ into a homopolymer of repeating adenosine diphosphoribose (ADP-Rib) units with a concomitant release of nicotinamide (1-3). The poly(ADP-Rib) formed has a variable chain length and is covalently attached to histones and other nuclear proteins (4, 5). Although poly(ADP-Rib) has been shown to occur in vivo (5), its specific biological function is as yet undetermined, however there is possible evidence for its implication in DNA replication (5).Doly and Petek (6) exists that poly(ADP-Rib) may be heterogeneous with regard to the site of attachment of the ribosyl moiety. Further heterogeneity is possible since the anomeric configuration at R1" is unknown, although a fi configuration has been drawn (1, 5, 7). The implications arising from possible heterogeneity makes a structural determination imperative.Two classes of enzymes are known that can degrade poly(ADP-Rib): poly(ADP-Rib) glycohydrolases, which attack the ribosyl-adenosine bond (8, 9) liberating ADP-Rib, and pyrophosphatases, such as endogenous pyrophosphatases (10, 11) or snake venom phosphodiesterase (oligonucleate 5'-nucleotidohydrolase; EC 3.1.4.1) (12), which cleave the pyrophosphate linkage. The resulting monomer from the latter reactions, 2'-(5"-phosphoribosyl)-5'-AMP [AMP(P-Rib)], retains the glycosidic linkage and thus allows structural analysis. We report the 360 MHz 1H nuclear magnetic resonance (NMR) spectra of AMP(P-Rib) and 2'-ribosyladenosine (Rib-Ado). The analysis of these spectra allow6 the determination of both the site and configuration of the glycosidic linkage for poly(ADP-Rib) synthesized in uitro and provides evidence for the homogeneity of poly(ADP-Rib).MATERIALS AND...