Studies on Poly Adenosine Diphosphate-Ribose

The formation of polgadenosine diphosphate-ribose from NXD has been well documented by Chumbon et al. (l), Nishizuka el al. (2), and us (3, 4). It has been suggested that this polymer occurs naturally (5). 1’01~ ADP-ribosel (1) is hydrolyzed at the pyrophosphate bond by snake venom phosphodicstera% (1, 6-8), yielding 1%mlP, that is, 2’-(5”-phosphoribosyl)-5’-AMP (II). A small amount of Y-AMP (111) was also recovered (6, 9) from the hydrolysate. A phosphodiesterase purified from rat liver (10-12) can also hydrolyze poly .IDl’-ribose. From the molar ratio of PR-AMP to 5’-AlIP, we can ca1culut.e the chain length, the repetition of ADP-ribose units, in an intact molecule of poly ADP-ribose. During studies to determine the chain length of purified poly ADP-ribose and of its terminal structure, it was necessary to separate PR-AMP from its dephosphorylated derivatives, 2’(ribosyl)-5’-AhlP (IV) and 2’-(5”-phosphoribosyl)-adenosine (VI). The nuclear enzyme preparation used for the synthesis of poly ADP-ribose from NAD possessed phosphodiesterase and phosphomonoesterase activities and it is likely that the terminal structure of polymer with AMP in its intact form is altered by these enzymes, as indicated in Fig. 1. 2’-(Ribosyl)-5’-QIP can be produced by Pathway a, b, or c in Fig. 1. During incubation

The formation of polgadenosine diphosphate-ribose from NXD has been well documented by Chumbon et al. (l), Nishizuka el al.
From the molar ratio of PR-AMP to 5'-AlIP, we can ca1culut.e the chain length, the repetition of ADP-ribose units, in an intact molecule of poly ADP-ribose. During studies to determine the chain length of purified poly ADP-ribose and of its terminal structure, it was necessary to separate PR-AMP from its dephosphorylated derivatives, 2'-(ribosyl)-5'-AhlP (IV) and 2'-(5"-phosphoribosyl)-adenosine (VI).
The nuclear enzyme preparation used for the synthesis of poly ADP-ribose from NAD possessed phosphodiesterase and phosphomonoesterase activities and it is likely that the terminal structure of polymer with AMP in its intact form is altered by these enzymes, as indicated in Fig. 1. 2'-(Ribosyl)-5'-QIP can be produced by Pathway a, b, or c in Fig. 1 Id l'hosphodiesterase preparations from snake venom arc often cont.aminuted with 5'-nucleotidase (13). 2'-(5"-l'hosphoribosyl)-adenosine and 2'-(ribosyl)-5'.A;\LP may be produced by removul of phosphate from PR-.UIl'.
For precise determination of the chain length uf poly ADPribose and of the terminal structure of this polymer, methods for separation and identification of these compounds are essential. This paper dacribes the separation of these compounds and related substances, including 2'-(ribosyl)-adenosine, by column and paper chromatography.
NAD labeled with V was purified wit.h an appropriate amount of nonlabeled NAD by Dower; 1 column chromatography, eluting with a gradient of formic acid (0 t.o 0.8 al). This NAD preparation gave only a single radioactive spot with two different solvent systems on paper chromatography. To obtain ;?;hD labeled only at the phosphate of its XMI' moiety, WVlabeled ATP and NhIN were incubated with S.W pyrophosphorylase and the resulting N.4D was purified as dcbcribed above. NAD labeled at the adenine moiety was obtained by react.ion of (B-W-adenine)-ATP and NUN. NAU labeled with =l' only at the NMN moiety was prepared from ATP and NMN labeled with V, obtained from SAD labeled with a2P at both phosphates by hydrolysis with snake venom phosphodicsterase (14). Preparation of t'uri$ed 1'01~ AL)P-ribose-A rat liver nuclear enzyme preparation was obtained as described previously (4), with the slight modification that 0.3 n1~1 magnesium ion was added to the 2.4 11 sucrose solution used for isolation of cell INclei.
The enzyme preparation was incubated with labeled NAD in 0.1 Y Tris-HCl buffer solution (pH 8.0) for 10 min at 37", as reported before (4 PR-AMP was purified by Dowex 1 formate column chromatography, as described before (7,8). More than 90% of the total radioactivity in poly ADP-ribose was recovered as PR-AMP.
PR-AMP labeled at the 5'qhosphate of the AMP moiety with s2P was prepared from poly ADP-ribose synthesized from NAD labeled with 32P at the 5'-phosphate of AMP moiety.
It was partially digested by alkaline phosphomonoesterase and subjected to column chromatography with Dowex 1, chloride form, as described below.
derived from poly ilDP-ribose labeled at the adenine, was extensively digested by alkaline phosphomonoesterase and subjected to the same column chromatography as that described in the previous se&ion.
Column Chromatography-Column chromatography on Dowex l-X2 (chloride form) was carried out with convex gradient elution, achieved by placing 420 ml of 0.0035 N HCl in the mixing chamber and 0.25 M NaCl in the reservoir of the column, 0.5 X 20 cm (15  compounds are illustrated in Fig. 2. 2'-(Ribosyl)-adenosine was eluted soon after the start of convex gradient elution.
Inorganic phosphate and ribose 5-phosphate were eluted between 3'-AMP and 5'-UMP. Paper Chromatography-The running distances of various substances are given in Table I Fig. 3. From the curves given in Fig.   3, it is clear that 2'-(5"-phosphoribosyl)-adenosine appeared more rapidly as a dephosphorylat,ed derivative than 2'-(ribosyl)-5'-AMP, while 2'-(ribosyl)-adenosine was formed last. Apparemly, phosphate at position 5' was more susceptible to alkaline phosphomonoesterase than phosphate at position 5" of PR--4MP. Effect of Commercial Preparation of Snake Venom Phosphodiesterase on PR-AllFP-8-W-,\denine-labeled PR-,4MP was incubated with a commercial preparation of snake venom phosphodiesterase (Worthington, Lot VPH 8EC) at a concentration of 5000 wg per ml in 50 mM Tris-HCl buffer (pH 8.0) and 5 mM XgCIZ for 90 min at 37". These conditions did not yield any dephosphorylated derivative of I'R-AMP. This indicates that the commercial preparation of snake venom phosphodiesterase can be used for hydrolysis of poly ADP-ribose to determine its chain length, at least under the present conditions. But it does not exclude the possibility that other specimens of snake venom phosphodiestcrase may be contaminated to a significant extent by enzymes yielding dephosphorylated derivatives of PR-AMP. Thus it is rccommendcd that 5'-nucleotidase activity should be removed by the method of Keller (13). It was also observed that phosphate at position 5' was more susceptible to attack by the phosphomonoesterase activity in the rat liver phosphodiesterase preparation t,han phosphate at posi-tion 5" of PR-AMP.
Recovery The results are given in Table II.
The ratio of PR-AMP to 2'-(ribosyl)-5'-AMP was about 27, representing the chain length of trimmed poly ADPribose. DISCUSSION The present paper describes the separation of PR-AMP and its dephosphorylated derivatives by paper chromatography and column chromatography.
Snake venom phosphodiesterase is known to hydrolyze poly ADP-ribose to yield PR-ANIP (7,8). The commercial preparation of snake venom phosphodiesterase used was free from activity to dephosphorylate PR-AMP. However, a purified preparation of rat liver phosphodiesterase was also able to dephosphorylate PR-AMP, and snake venom phosphodiesterase may also occasionally contain this activity. When poly ADPribose is hydrolyzed by phosphodiesterase contaminated by this activity, the PR-AMP formed will be further converted to 2'-(5"-phosphoribosyl)-adenosine or 2'-(ribosyl)-5'-AMP. These two dephosphorylated compounds behave like 5'-AMP under the conditions of paper chromatography or column chromatography normally used for separation of 5'-AMP and PR-AMP. Thus, the chain length of poly ADP-ribose may be underestimated when measured as the ratio of PR-AMP to 5'-AMP plus dephosphorylated derivatives of PR-AMP instead of that of PR-AMP to 5'-AMP. The true chain length can be obtained by separating dephosphorylated derivatives of PR-AMP from 5'-AMP.
Rat. liver phosphodiesterase hydrolyzed poly ADP-ribose in an exonucleolytic fashion, while snake venom phosphodiesterase hydrolyzed it in both exonucleolytic and endonucleolytic fashions (16). By examining the occurrence of 2'-(5"-phosphoribosyl)adenosine and 2'-(ribosyl)d'-AMP formed by Pathways b and c in Fig. 1, we can determine the terminal structure of poly ADPribose in its intact form or after partial digestion with phosphodiesterase.
Analysis of the terminal structure of poly ADP-ribose after exonucleolytic digestion by rat liver phosphodiesterase would give information on the direction of hydrolysis of poly ADP-ribose by this enzyme.