Aug 25, 2015 - 255, 10710-10716). In this report, we describe experiments involving nuclear Overhauser enhancement NMR spectroscopy which were under-.
THE JOURNAL OF BIOLOGICAL CHEMISTRY
Vol. 256. No. 16, Issue of August 25. pp. 8579-8581. 1981 Printed in U.S.A.
Diphtheria Toxin SITE AND CONFIGURATION OF ADP-RIBOSYLATION OF
DIPHTHAMIDE IN ELONGATION FACTOR 2* (Received for publication, April 6, 1981)
Norman J. Oppenheimertg and James W. Bodleyl +From the Departmentof Pharmaceutical Chemistry, University of California, San Francisco, California 94143 and The YDepartment of Biochemistry, University of Minnesota, Minneapolis, Minnesota 55455
Diphtheria toxin inactivates protein synthesis elon- Here, too, the exact site and configuration of ADP-ribosylagation factor 2by catalyzing the ADP-ribosylation of a tion is unknown. novel derivative of histidine, diphthamide, in the proCholera toxin functions to perturbCAMPlevels as theresult tein (Van Ness,B. G., Howard, J. B., and Bodley,J. w. of ADP-ribosylation of amembrane-boundprotein which (1980) J. Biol. Chem. 255, 10710-10716). In this report, regulates adenylate cyclase (3). The amino acid residue in the we describe experimentsinvolving nuclear Overhauser regulatory subunit which is ADP-ribosylated by cholera toxin enhancement NMR spectroscopywhichwereunderis unknown; however, cholera toxin will catalyze the ADPtaken to elucidatethe site ofADP-ribosylationof ribosylation of the free amino acid, arginine, as well as other diphthamideandtheconfigurationof the glycosidic guanidino compounds. In this case, the ADP-ribose is atbond formed by the toxin. The essential result of these tached to one of the guanidino nitrogens of arginine via an experiments is that, in ribosyl-diphthamide obtained a-linkage ( 6 ) .The Escherichia coli heat labile toxin is strucby enzymatic digestion of ADP-ribosyl-elongationfactor-2, the H-5 imidazole proton is near the R-4 proton turally distinct from, but in other ways functionally similar of ribose. Thisresult and others areconsistent with the to, cholera toxin (7). In thisreport, we describe nuclear Overhauser enhancement interpretationthatdiphtheriatoxincovalently attaches ADP-ribose to the imidazole N-1 of diphthamideNMR spectral experiments with ribosyl-diphthamide which were designed to elucidate the location and configuration of via an a-glycosidic linkage. the glycosidic linkage formed by diphtheria toxin. The results of theseexperiments establish that diphtheria toxin, like cholera toxin, catalyzes ADP-ribosylation with inversion of DiDhtheria toxin is one of several toxic proteins which configuration at theribose. interferes with eukaryotic cell metabolism by catalyzing the ADP-ribosylation of an essential intracellular protein (2, 3). EXPERIMENTALPROCEDURES Protein synthesisis inhibited by diphtheria toxin as theresult Preparation of Samples-Ribosyl-diphthamide was obtained by of ADP-ribosylation of EF-2.’ Inthis reaction, the toxin sequential enzymatic digestion of ADP-ribosyl-EF-2 as described exhibits unusual specificity in that it will ADP-ribosylate all previously (8). For ’H NMR spectroscopy, approximately 5 pmol of E F - ~ s but , under the usual reaction conditions, it will not ribosyl-diphthamide was exchanged twice with 99.8% D20, dissolved modify any other protein (2). Recent experiments have indi- in 0.35 ml of 100% D20 (low paramagnetic content; Aldrich Chemical cated that at least a portionof this specificity results from the Co.), vacuum degassed three times, and then kept under drynitrogen. attachment of ADP-ribose to EF-2via a glycosidic linkage to The pH of the sample was approximately 2.0. NMR Spectroscopy-The experiments were conducted on a Nicoa novel derivative of histidine, diphthamide, not previously let NT-360 NMR spectrometer equipped with a Nicolet 1180 comobserved in proteins (4). Experiments with ribosyl-diphtham- puter and computer-controlled decoupling accessory. Spectra were ide obtained by enzymatic digestion of ADP-ribosyl-EF-2 obtained in the Fourier transform mode using quadrature detection suggested that theADP-ribose is attached via one of the two and a 60’ pulse angle. Spectral width was 2400Hz and 8K of data imidazole nitrogens of diphthamide in an unknown configu- points were accumulated, and after using a 2 Hz exponential line ration (4). Pseudomonas toxin, although immunologically dis- broadening, a 32K transform was performed. Difference spectra were generated by subtraction of the I6K “real” portionsof the irradiated tinct from diphtheria toxin, appears to function in an essen- and control spectra. The decoupler was under computer control and tially identical manner through the ADP-ribosylation of the gated off during data acquisition. The duration of the pre-irradiation same amino acid residue in EF-2 (Ref. 5 and Footnote 2). pulse was three times the spectral acquisition time and decoupler * This work was supported by National Institutes of Health Grants GM-22982 (to N. J. 0.) and GM-23276 and GM-26832 (to J. W. B.). The experiments were conducted at the University of California, Davis, NMR Facility which is supported by aNational Science Foundation Grant CHE 79-04832 to the Department of Chemistry, University of California, Davis. This is paper XXIX in the series “Studies on Translocation.” The preceding paper in the series is Ref. 1. The costs of publication of this article were defrayed in part by the payment of page charges. This articlemusttherefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 8 National Institutes of Health ResearchCareer Development Awardee, 1979-1984. The abbreviations used are: EF-2, elongation factor 2; NOE, nuclear Overhauser enhancement. B. G. Van Ness, B. H. Iglewski, and J. W. Bodley, unpublished experiments.
’
power was adjusted to give an 80-90s saturation of the irradiated resonance with a minimum effect on the intensity of adjacent resonances. The NOE was measured as the ratio of peak heights in a control off-resonance spectrum to those in the difference spectrum obtained by subtracting the control spectrum from the decoupled spectrum. Since only qualitative information on the spatial proximity of protons is needed to establish the anomeric configuration, no attempt was made to optimize conditions for a quantitative measurement of the full intrinsic NOE. RESULTS
A number of NMR criteria have been used for the assignment of the anomeric configuration of glycoside linkages in nucleosides including: nuclear Overhauser enhancements (9), the relaxation time of the anomeric proton (10, l l ) , and the chemical shift of the anomeric proton (12). With the exception
8579
8580
ADP-ribosylation of Elongation Factor2 TABLEI Nuclear Overhauser effects for ribosyl-diphthamide Irradiated proton
H-5 R-1 H*N--C=O
R-2 R-3 R-4
Nuclear Overhauser enhancement"
H-5
R-1
R-2
R-3
%
%
%
%
2.3 2.1 1 k 0.5 0 4.6
5.8 0 0
0
0
8.4
0 8.3
6.0 0
R-4 %
4.4 0 0 1
2.0
* 0.5
Measured as the ratio of peak height in the difference spectrato that in the reference spectrum obtained under identical conditions of decoupler power. TABLEI1 Stereochemistry of enzymes cleaving the nicotinamide-ribosyl bond Acceptor
a-Specific Poly(ADP-ribose) synthase(17) Cholera toxin (6) E . coli enterotoxin (19) Turkey erythrocyte enzyme (20) Diphtheria toxin P-Specific Calf spleen NADase (21) NADase brainPig (22)
1
Adenosine 2'-OH Guanidinium Guanidinium Guanidinium Diphthamide (inEF-2) CH30H and nucleophilic heterocycles
the data are summarized in Table I. Note that a NOE is observed between the H-5 and R-4 protons but not between I HDO I the H-5 and the R-3 protons. Furthermore, the R-1 protons shows a large NOE with the R-2 but not the R-4 proton. These results establish the spatial proximity of the imidazole H-5 proton toR-1 and R-4, and theR-1 proton toR-2 but not R-4. This patternof enhancements canonly beaccommodated N-1 nitrogen by an a-glycosidic linkageand substitution on the 1 1 ( 1 1 1 ' I . ~ " I ~ ' ' ' 7.5 6.5 6.0 5.0 4.5 4.0 of the imidazole ring as shown in Fig. 2. The absence of a detectable NOE between R-4 the and R-1 protons is consistent PPM with a n a-glycosidic linkage, since these protons would be FIG. 1. Difference spectrafromnuclearOverhauser en- trans, whereas in the P-glycosidic linkage, they would be cis hancement experiments on ribosyl-diphthamide D,O in at 360 and thusa significant NOE would be expected (9, 13). SubstiMHz. The control spectrum withits amplitude reducedby a factor of tution on the more hindered H-3 position can be excluded, 10 for comparative purposesis shown at the bottom of the figure. The because theH-5 proton would be too distant togive a detectfive resonances of interest are identified immediately beneath this spectrum in relation to the proposed structure of ribosyl-diphthamide, able NOE with the ribosyl protons. A small but significant shown at the top of the figure. The upper five are differencespectra, NOE (-1 k 0.5%)was also observed between the R-1 and Cwith the decoupledprotonindicatedin the left column and the 1 resonances and between theH-5 and H-P resonances (data decoupled frequency indicatedby the bars in the spectra. not shown). These effects establish the proximity of the respective protons and are only consistent with the substitution of the NOE method,however, they require thecomparison of pattern shown in Fig. 2. spectral properties of both the a- and p-anomers. The NOE is DISCUSSION measured by a double resonance experiment in which irradiaThe enzymes which cleave the nicotinamide-glycosyl bond tionwith aweak radiofrequency pulse of oneresonance causes a change in the intensity of other resonances in the of NAD+ have beencategorized as either NAD+-glycohydrolases (NADases) or ADP-ribosyl transferases. This is not an molecule. The intensitychange will beobservedonlyfor protons that havea significant dipole-dipole interaction with entirely clear distinction, however, because various transferthe irradiated proton. More importantly, this interactionoc- ases, including diphtheria toxin (2), can also hydrolyze NAD' curs through space and has a strong, l/r6, dependence on the in the absence of the normal acceptor. Also, the membranedistance between the interacting nuclei (13).The observation bound NADases such as thoseisolated from pig brain or calf of a NOE between two protons indicates their spatial prox- spleen, besides hydrolyzing NAD+, also catalyze the exchange imity and thus can be used as a criterion for assigning the of the nicotinamide moiety with exogenous pyridines (14) and configuration of nucleosides. For a /3-anomer of a nucleoside, imidazoles (15). The NADC-cleaving enzymes can also be classified on the the diagnostic enhancements are between the R-1 and R-4 protons of the sugar and between the protons of the base and basis of the stereochemistryof their reaction products (Table will show a strong NOE 11).The two NADases which have been examined both yield the R-3 or R-5 protons. The a-anomer R-4) and reaction products with retention of configuration. The reacbetween theR-1 and R-2 protons (but none with the tions catalyzed byboth of these enzymes involvethe formation the base protonswill show a strong NOE to theR-4 proton. The difference spectra for pre-irradiation of the ribosyl and of ADP-ribosyl-enzyme intermediatesand proceed via an as the fist H-5 protons of ribosyl-diphthamide are shown in Fig. 1, and ordered Uni Bi mechanism with nictoinamide
-
ADP-ribosylation of Elongation Factor2
8581
FIG. 2. A Dreiding model of l-a~-ribofuranosyl-2-[3-carboxyamido3-(trimethylammonio)propyl]histidine, the proposed structure of ribosyl-diphthamide (4). The two-dimensional representation is parallel with the plane of the imidazole ring, with the H-5 and R-4 protons shown at theirclosest point of approach (-1.3 A). Carbon, nitrogen, and oxygen atoms are shown as black,white, and stippled, respectively.
7. Dallas, W. S., and Falkow, S. (1980)Nature 288,499-500 8. Van Ness, B. G., Howard, J. B., and Bodley, J . W. (1980)J.Biol. Chem. 255, 10717-10720 9. Cushley, R. J., Blitzer, B. L., and Lipsky, S. R. (1972)Biochem. Biophys. Res. Commun. 48, 1482-1488 10. Preston, C.-M., and Hall, L. D. (1974)Carbohydr. Res. 37, 267282 11. Tran-Dinh, S., Neumann, J.-M., Thiery, J.-M., Huynh-Dinh, T., Igolen, J., and Guschlbauer, W. (1977)J. Am. Chem. SOC.99, 3267-3273 12. Christl, M., Reich, M. J., and Roberts, J. D. (1971)J . Am. Chem. SOC.93,3463-3468 13. Noggle, J . H., and Schirmer, R. E. (1971)The Nuclear Ouerhauser Effect, Academic Press, New York 14. Zatman, L. J., Kaplan, N. O., Colwick, S. P., and Ciotti, M. M. (1954)J. Biol. Chem. 209, 453-466 15. Alivisatos, S. G. A,, Ungar, F., Lukacs,L., and LaMantia,L. (1960) J. Biol. Chem. 235, 1742-1750 REFERENCES 16. Schuber, F., Travo, P., and Pascal, M. (1976)Eur. J. Biochem. 1. Van Ness,B. G., Barrowclough, B., and Bodley, J . W. (1980) 69,593-602 17. Ferro, A. M., and Oppenheimer, N. J. (1978)Proc. Natl. Acad. FEBS Lett. 120, 4-6 2. Pappenheimer, A. M., Jr. (1977)Annu. Reu. Biochem. 46,69-94 Sci. U. S. A . 75,809-813 3. Moss, J., and Vaughan, M., (1979)Annu. Reu. Biochem. 48,581- 18. Deleted in proof. 600 19. Moss, J., Garrison, S., Oppenheimer, N. J., and Richardson, S. H. 4. Van Ness, B. G., Howard, J. B., and Bodley, J . W. (1980)J. Biol. (1979)J.Bwl. Chem. 254,6270-6272 Chem. 255, 10710-10716 20. Moss, J., Stanley, S. J., and Oppenheimer, N. J . (1979)J. Biol. 5. Iglewski, B. H., and Kabat, D. (1975)Proc. Natl. Acad. Sci. U. S. Chem. 254,8891-8894 A . 72, 2284-2288 21. Pascal, M., and Schuber, F. (1976)FEBS Lett. 66, 107-109 6. Oppenheimer, N. J . (1978)J.Biol. Chem. 253, 4907-4910 22. Oppenheimer, N. J . (1978) FEBS Lett. 94, 368-370
released product (16). On the other hand, none of the transferases including diphtheria toxin (2) are known to form ADPribosyl-enzyme intermediates, and all five of these enzymes listed in Table I1 lead to inversion of configuration. Mechanistic conclusions on the basis of present evidence would be premature. It is nonetheless reasonable to suggest that thetransferases, including diphtheria toxin, which operate with inversion of configuration may sharea common mechanism. This mechanism is undoubtedly distinct from the mechanism(s) of the NADases which operates with retention of configuration. Whether or not diphtheria toxin shares a mechanism in common with the other transferases, it is important to note that inversion of configuration does not necessarily imply an S i 2 reaction mechanism.
