' ' +2H20 - Semantic Scholar

1 downloads 0 Views 409KB Size Report
$Department of Pharmaceutical Chemistry, School of Pharmacy, University of California, ... ment” in accordance with 18 U.S.C. Section 1734 solely to indicate.
THEJOURNAL OF BIOLOGICAL CHEMISTRY Vol. 257, No.20, Igsue of October 25. pp. 11942-11945, 1982 Printed in U.S.A.

Steric Inhibitionof Phenylboronate Complex Formationof 2’-(5”-Ph0~ph0rib0~yl)-5’-AMP* (Received for publication, May 27, 1982)

Takeyoshi MinagaS, Jerome McLickS,N. Pattabiramang, and ErnestKun$l From the *Cardiovascular ResearchInstitute and theDepartment of Pharmacology, School of Medicine and the $Department of Pharmaceutical Chemistry, School of Pharmacy, University of California, Sun Francisco Medical Center, San Francisco, California 94143

Unlike all other adenine nucleotides which contain by polyacrylamide-based dihydroxyboryl affinity chromatogcis-diols in their ribose groups, the degradation product raphy has also been achieved (6, 7). of polyadenosine diphosphoribose, 2’-(5“-phosphoribo- Polyadenosine diphosphoribose, the homopolymer of 5’syl)-5’-AMP, does not bind to boronate affinity columns. ADP-R,’ is a covalently bound macromolecular-modifying This anomalous behavior occurs even under conditions constituent of a variety of nuclear proteins and a biological of high ionic strength (1 M salt), indicating that the correlation was observed between the degree of polyadenosine absence of binding is not due to generalized repulsion diphosphoribosylation of selected nonhistone proteinsand the between the negatively charged phosphate groups of 2’-(5”-phosphoribosy1)-5’-AMP and the negatively action of developmental hormones and carcinogens (7,8). An charged boronatesites of the resin. Despite the lack of age and ontogenic development-dependent effect was also found on the magnitude of protein polyadenosine diphosphoboronate binding the presence of a chemically funcribosylation in cardiocyte nuclei (9). Interpretation of these tional cis-diol group in 2’-(5”-phosphoribosyl)-5’-AMP was verified by quantitative periodate oxidation. Sub- phenomena in molecular terms requires the detailed knowlsequent dephosphorylation renderedthe cis-diol in ri- edge of the nature of interactions between polymer-protein adducts andbetween these adducts andnucleic acids. We are bosyladenosine accessible to binding to the boronate sites; therefore, one or both of the phosphate groups approaching the resolution of this problem by the identifcation of chromatin proteins that exhibit large changes in the sterically interfere with complexation of the cis-diol degree of their polyadenosine diphosphoribosylation under with boronate. Computer model-building studies based physiologically meaningful experimental conditions (7-9) and on NMR data implicate the 5”phosphate group of the AMP moiety which,in a favorable conformation of the by the studyof hitherto unresolved structural andmacromomolecule, is located sterically near the cis-diol group, lecular properties of the protein-free polymer itself. The presinhibiting the complex formation with boronate. ent paper deals with a structural problem of the homopolymer

It is well established that in aqueous media arene boronic acids, ArB(OH)*, reactwith water to form negatively charged boronate species, ArB-(OH)3, which are tetrahedral about their boron centers (1).Organic molecules containing 1,2-cisdiol are relatively strong bidentate ligands toward boronate centers, forming a chelate complex with boron by displacing two hydroxyls (2) according to the following equation: OH

I@ Ar-B-OH

I

OH

+

HO-CH-C-

’ I

I

OH

I@ Ar-B-0 I

=

oI

\

‘c’ /

+2H20

\C/ \

/ \

that was recognized by studying the boronate complex formation of its enzymatic degradation products. Enzymatic hydrolysis of the a-glycosidic bond by a specific glycohydrolase (13) yields 5’-ADP-R, whereas cleavage of the pyrophosphate bond results in PR-AMP, which is the specific degradation product of polyadenosine diphosphoribose. We pursued the application of boronate affinity chromatography to the isolation of both enzymatic degradation products of polyadenosine diphosphoribose and observed that 2’-(5”phosphoribosyl)-5‘-AMP did not bind to affinity columns under conditions when all other adenine nucleotides did. The present paper is concerned with the clarification of this apparently anomalous behavior of 2’-(5”-phosphoribosyl)-5’AMP and evidence will be presentedindicating that a terminal phosphate of this nucleotide in a specific conformation can sterically hinder the formation of the boronate-ribose complex.

MATERIALS AND METHODS This chelation is the basis of affinity chromatography methods suitable for the separation of mono- and polynucleotides Polyadenosine diphosphoribose was isolated and purified by pubwhose ribose constituentscontain cis-diols (3-5). Selective lished methods (10, 11).Snake venom phosphodiesterase (EC 3.1.4.1) isolation of protein-polyadenosine diphosphoribose adducts was purifiedas reportedearlier (12).Alkaline phosphatase (EC 3.1.31) was a commercially obtained preparation from Sigma. * This work was supported in part by the United States Air Force The digest of polyadenosine diphosphoribose by phosphodiesterase

Office of Scientific Research (F-49620-81-C-007).The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Recipient of a research career award of the United States Public Health Service. To whom correspondence should be addressed at the Cardiovascular Research Institute.

was purified on AG 1-X2 formate column and was eluted by a linear formic acid gradient (50 ml of Hz0 uersus 50 ml of 6 M formic acid). Purification of polyadenosine diphosphoribose glycohydrolase and analyses of nucleotides by high performance liquid chromatography The abbreviations used are: 5’-ADP-R, 5”adenosine diphosphoribose; PR-AMP, 2”(5”-phosphoribosyI)-5’-AMP.

11942

.Phosphoribosyl)-5’-AMP Steric Inhibition in2’-(5‘’were carried out as reported earlier (7, 13). Periodate oxidation was monitored a t 310 nm (14).The boronate-acrylamide affinity gel (AffiGel 601), a product of Bio-Rad, was washed extensively with 1 M ammonium carbonate (pH 8.8) before use. [U-I4C]Adenine (235mCi/ mmol), [U-’4C]adenosine (567mCi/mmol), [U-14C]5’-AMP (272mCi/ mmol), and [U-I4C]NAD, in the adenine moiety (286 mCi/mmol), were obtained from Amersham. 8[14C]5’-ADP (40 mCi/mmol) was purchased from New England Nuclear and I4C-labeled adenosine diphosphoribose was prepared from labeled NADbycalf spleen NADase obtained from Sigma (15).Boric acid gel (Aldrich Chemicals) was treated like Affi-Gel 601. Computer modeling was performed on an Evans and Sutherland Picture System PS 200 (Color graphics) 92.1computer. 4.6 driven by a PDP 11/70 RESULTS AND DISCUSSION

The anomalous behavior of 2’-(5”-phosphoribosyl)-5”AMP on Affi-Gel601 is illustrated in Fig. 1A. Purified PR-AMP was freeze-dried and aliquots of two samples with different specific activities (2.1 nmol and 427 m o l ) equivalent to 35,000cpm dissolved in 1 ml of 1 M ammonium carbonate (pH 8.8) were applied to a small AffLGel601 column (2.5ml, 1 X 4 cm, with a capacity of 4.8 pmol). Washing the column with the ammonium carbonate buffer (flow rate, 0.5 ml/min) resulted in an almost immediate elution of PR-AMP, a phenomenon that was completely reproducible regardless of the sample size applied, and was contrary to the known behavior of cis-diol-containing substances (1618)on the boronate affinity column. On theother hand, adenosine diphosphoribose (20,000cpm) was retained by the column at pH 8.8, and eluted as a single component by 100 mM K phosphate buffer at pH 4.0 as shown in Fig. 1B. The adenosine diphosphoribose (13, 15) was prepared from the same batch of polyadenosine diphosphoribose that was used for the preparation of PR-AMP.

11943

TABLEI Behavior of adenine-containingnucleotides on Affi-Gel601 Specificity of boronic acid gel toward adenine derivatives. Adenine derivatives were tested in the same application buffer as sbown in Fig. L.4 (1 M ammonium carbonate) to exclude possible interference by charge repulsions. Desorbed by the ap- Recovered by the Nucleotides elution buffer plication buffer Adenine Adenosine 5‘-AMP 5‘-ADP 93.4 5”ADP-R NAD

B

%

97.3 4.2

0 89.0

3.0

91.3

1.2 2.0

88.0

A2 54 o.2

r

RETENTIONTIME (rnin) A254

x 1 0 3 cprn

I

(A)

Elutlon

0 ;

I 5

10

0.1

-4 T 10

2

x

6

8

I

12

14

16

RETENTIONTIME (mi.)

IO 3 cpm

II

4

(6)

Appllcation

FIG. 2. Criteria of purity of PR-AMP determined by high performance liquid chromatography. PR-AMP used for the experiment shown in Fig. 1A was adsorbed on a weak anion exchanger (Micropak, AX-IO, Varian Associates) and eluted with a linear gradient of potassium phosphate from 5 mM, pH 2.85, to 750 mM, pH 4.8, with a flow rate of 2 ml/min during a period of 15 min. A, authentic substrates: 1, adenine; 2, 5‘-AMP 3, 5’-ADP-R, 4, 5’-ADP. B, hydrolyzed product of poly(ADP-ribose) by snake venom phosphodiesterase: 5, PR-AMP.

The specificity of the AS-Gel601 was subsequently tested with a variety of adenine derivatives, as shown in Table I, and the affiiity column behaved as predicted with respect to its cis-diol specificity. Therefore the results obtained with PRFRACTION NUMBER AMP were not due to anartifact. FIG. 1. Chromatographic behaviorof PR-AMP ( A )and ADPR ( B )on boronate affinity resin (see “Materials and Methods”). The purity of enzymatically prepared PR-AMP was conf i i e d by high performance liquid chromatography (7), and Application buffer was 1M ammonium carbonate (pH 8.8) and elution as shown in Fig. 2B no contaminant could be detected. The buffer was 100 mM potassium phosphate (pH 4.0). The change of buffers is indicated by arrows. same chromatographic system completely separated adeno5

10

15

20

25

inInhibition Steric

11944

2'-(5"-Phosphoribosy1)-5'-AMP

sine, 5'-AMP, 5'-ADP-R, 5'-ADP, and PR-AMP (Fig. 2A). When PR-AMP was dephosphorylated by incubation with alkaline phosphatase, the product, ribosyladenosine, was not retained on the AG 1-X2 ion exchanger (Fig. 3B), whereas PR-AMP was adsorbed and eluted with a formic acid gradient (Fig. 3A). When the ribosyladenosine (tubes 1 and 2 in Fig. 3 B ) after freeze-drying was applied to anAffi-Gel601 column it was retained by the affinity columns at pH 8.8 and eluted as a single component with 100 m~ K phosphate, pH 4.0, as illustrated in Fig. 3C. It is apparent that dephosphorylation rendered the cis-diols of ribosyladenosine accessible to complexation with the boronate sites. The identity of ribosyladenosine and its separation from ADP-R, adenine, and adenosine was further established by reverse phase high performance liquid chromatography (19) as shown in Table 11. In order to confirm the availability of cis-diols in PR-AMP, quantitative periodate oxidation (14) was performed on t 9 s product as well as on standard nucleotides (Fig. 4).Resul?, which are in agreement with previous work of others (20,21), clearly show that the lack of complexation between cis-diols of PR-AMP and theboronate sites of Affi-Gel 601 is not due to some anomaly of ribose moieties of PR-AMP. It is known that nucleotides (e.g. 5'-AMP, 5'-ADP) do not bind to phenylboronatecolumns at low ionicstrength because

TABLE I1 Separation of adenine, adenosine, ADP-R, and ribosyladenosine by reuerse phase high performance liquid chromatography The substances listed were injected into a reverse phase column (Altex, Ultrasphere-ODS) and developed by isocratic elution. The solvent was 7 m~ ammonium formate (pH 5.3)/methanol (9/1, v/v), and the flow rate was 1.5ml/min. Nucleotide

Retention time

min

ADP-R Adenine Adenosine Ribosyladenosine

I

1

2.37 6.77

15.25 21.11

1 Adenine Adenosine

II

]

5'-AMP 5"ADP 5'-ADPR

1

NAD Product (PR-AMP)

100 n mol

each

1 I

I

100

200

periodote consumption ( n mol)

FIG. 4. Periodate titration of ribose-containing adenine derivatives. The periodate consumption was followed in l-ml volume containing 500 PM periodate, 10 mM acetate buffer, pH 4.0, and 100 of each ribose-containing substance.

PM I

FRACTION NUMBER FIG. 3. Chromatographic behavior of PR-AMP and ribosyladenosine on AG 1-X2 and boronic acid (Affi-Gel 601) resins. A, PR-AMP was charged on AG 1-X2 formate resin and then eluted with a linear gradient from H 2 0 to 6 M formic acid. B, the eluted PR-AMP shown in A was lyophilized and dephosphorylated by alkaline phosphatase and was charged on AG 1-X2 formate resin. Ribosyladenosine was eluted without being adsorbed. C, ribosyladenosine from B was lyophilized and charged on boronic acid gel and eluted. AR-R, ribosyladenosine. Application and elution buffers in C were the same as describedFig. for 1.

of repulsion between negatively charged phosphate and boronate groups (4). However, this repulsion is effectively eliminated by the use of salt solutions of high ionic strength (4), as also demonstrated in experiments shown in Table I where the developing solvent was 1 M ammonium carbonate buffer. It follows that the absence of binding displayed by PR-AMP must be due to reasons other than generalized phosphateboronate repulsion. It was most reasonable to assume that a phosphate-dependent (see Fig. 3) steric effect was operative. In the absence of x-ray crystallographic data we analyzed the molecular structure of PR-AMP by a computergraphics model-building study utilizing currently available NMR data (23, 24). Only eleven conformational parameters are required to specify the conformation of PR-AMP with fixed bond angles and lengths. Nine are due to single bond rotations and two to sugar (ribose) puckering. Based on NMR analyses (22-24), the orientation of the phosphate group relating to the C3.-C4-C5.-O~ torsion angle in the AMP moiety of PR-AMP is gg and that of the adenine ring is anti, whereas in the 5'"phosphoribosyl moiety the orientation of the phosphate is tg (24). Input parameters for the sugars were taken from minimized coordinates (25) and the average bond lengths and angles of AMP and ribosylphosphate were obtained from previous reports (26-28). If the cis-diol of the ribose in the 5"-phosphoribosyl moiety were to form a boronate complex it would be constrained to an01endo or 01..ex0 conformation, and we have found by graphic computation only the 01-ex0 to be sterically favorable. In-the ribose of the AMP moiety, NMR data indicate equal

Inhibition Steric

in 2’-(5”-PhosphoribosyI)-5’-AMP

11945

REFERENCES

OZP

FIG. 5. Specific computer-derivedconformation of PR-AMP. In thisconformation steric interaction apparent is between the oxygen atoms of the phosphate groups of AMP (OP2, OP3)and the oxygen atoms of cis-diol (0,2, 0,3) present in the phosphoribosyl portion of the molecule. The dots depict the accessible surfaces of the interacting oxygen atoms. The numbering system reflects identificationof atoms by the computer program. The steric hindrance imposedby the above interaction inhibits the approach of a boronate group that has to be coplanar with the cis-diol.

proportionsof CZ. endo and Cy endo puckerings.Amajor observation in ourstudies is that the puckerings of the sugars significantlyalter the position of the phosphategroupsin relation to the rest of the molecule. With C3.endo puckering, the 5‘-phosphate of the AMP moiety is distantfrom the hydroxyl groups (cis-diol) of 5”-phosphoribosyl moiety, whereas with CZ.endo puckering the 5”phosphate is located near the cis-diol. Accordingly we built a model with C y endoexo-tg for the 5”-phosgg-anti for the AMP moiety and 01phoribosyl moiety. The contacts between the nonbonded atoms were checked and relieved by rotating about the sugarsugar bond. We found that only the phosphate of the AMP moiety can interact with the cis-diol group. This is shown in Fig. 5. If the pertinent phosphate groupis removed from the model and replaced by a hydroxyl group,as would happen in enzymaticdephosphorylation,thereissufficientfree space about the cis-diol to permit phenylboronate complexation,as is experimentally observed(Fig. 3). Acknowledgments-We wish to acknowledge with thanks thehelpful interest of Dr. R. Langridge, Director of the Computer Graphics Laboratory, supported by Research Resources Grant RR-1081 of the National Institutes of Health, United States Public Health Service, and we thank Clementina Moya Kun for typing the manuscript.

1. Onak, T. (1975) Organoborane Chemistry, Academic Press, New York 2. Steinberg, H. (1964) Organoborane Chemistry,Vol. I, John Wiley & Sons, New York 3. Weith, H. L., Wiebers, J. L., and Gilham, P. T. (1970) Biochemistry 9, 4396-4401 4. Rosenberg, M., Wiebers, J. L., and Gilham, P. T. (1972) Biochemistry 11,3623-3628 5. Schott, H., Rudloff, E., Schmidt, P., Roychoudhury, R., and Kossel, H. (1973) Biochemistry 12,932-938 6. Okayama, H., Ueda, K., and Hayaishi, 0.(1978) Proc. Natl. Acad. Sei. U. S. A . 75, 1111-1115 7. Romaschin, A. D., Kirsten, E., Jackowski, G., and Kun, E. (1981) J. Biol. Chem. 256, 7800-7805 8. Kun, E., Romaschin, A. D., Blaisdell, R. J., and Jackowski, G. (1981) in Metabolic Znterconuersion of Enzymes (Holzer, H., ed) pp. 280-293, Springer Publications, Heidelberg 9. Jackowski, G., and Kun, E. (1981) J. Biol. Chem. 256,3667-3670 10. Shima, T., Fujimura,S., Hasegawa, S., Shimizu, Y., and Sugimura, T. (1970) J. Biol. Chem. 245, 1327-1330 11. Minaga, T., Romaschin, A. D., Kirsten, E., and Kun, E. (1979) J. Biol. Chem. 254,9663-9668 12. Sulkowski, E., and Laskowski, M., Sr. (1971) Biochim. Biophys. Acta 240,443-447 13. Miwa, M., Tanaka, M., Matsuchima, T., and Sugimura, T. (1974) J. Biol. Chem. 249,3475-3482 14. Marinetti, G. V., and Rouser, G. (1955) J . Am. Chem. SOC.77, 5345-5349 15. Ferro, A. M., and Kun, E. (1976) Biochem. Biophys. Res. Commun. 71, 150-154 16. Hageman, J. H., and Kuehn, G. D. (1977) Anal. Biochem. 80, 547-554 17. Davis, C. W., and Daly, J. W. (1979) J . Cyclic Nucleotide Res. 5, 65-74 18. Moore, E. C., Peterson, D., Yang, L. Y., Yeung, C. Y., and Neft, N. F. (1974) Biochemistry 13,2904-2907 19. Sims, J. L., Juarez-Salinas, H., and Jacobson, M. K. (1980) Anal. Biochem. 106,296-396 20. Hasagawa, S., Fujimura, S., Shimiju, Y., and Sugimura, T. (1967) Biochim. Biophys. Acta 149, 369-376 21. Reeder, R. H., Ueda, K., Honjo, T., Nishizuku, Y., and Hayaishi, 0. (1967) J. Biol. Chem. 242,3172-3179 22. Miwa, M., Saito, H., Sakura, H., Saikawa, N., Watanabe, F., Matsushima, T., and Sugimura, T. (1977) Nucleic Acid. Res.4, 3997-4005 23. Ferro, A.M., and Oppenheimer, N. J. (1978) Proc. Natl. Acad. Sci. U. S. A. 75,809-813 24. Inagaki, F., Miyazawa, T., Miwa, M., Saito, H., and Sugimura, T. (1978) Biochem. Biophys. Res. Commun. 85, 415-420 25. Pattabiraman, N. (1979) Ph.D. dissertation, Indian Institute of Science, Bangalore, India 26. Sasisekharan, V., and Pattabiraman, N. (1978) Nature(Lond.) 275, 159-162 27. Prusiner,P., and Sundaralingam, M. (1976) Acta Crystallogr. Sect. B Struct. Crystallogr. Cryst. Chem. 32, 161-169 28. Dunita, J . D., Hawley, D. M., Miklos, D., White, D. N. J., Berlin, Y., Marusie, R., and Prelog, V. (1971) Helu. Chim. Acta 54, 1709-1713