THE JOURNAL OF BIOLOGICAL CHEMISTRV 0 1987 by The American Society of Biological Chernista, Inc.

Val. 262, No. 18, Issue of June 25. pp. 8707-8711 1987 Printed in irS.A.

Active Site of Pseudomonas aeruginosa Exotoxin A GLUTAMIC ACID 553 IS PHOTOLABELED BY NAD AND SHOWS FUNCTIONAL HOMOLOGY WITH GLUTAMIC ACID 148 OF DIPHTHERIA TOXIN*

(Received for publication, December 1, 1986)

Photoaffinity labeling with native NAD, a method employed earlier with diphtheria toxin (DT), was used to identify an active site residue of Pseudomonas aeru- ginosa exotoxin A (ETA). An enzymically active frag- ment (M, 27,000), derived by partial digestion of ETA with thermolysin, was irradiated with ultraviolet light (254 nm) in the presence of various radiolabeled prep- arations of NAD. Label from the nicotinamide moiety was efficiently transferred to the protein (maximally 0.79 mol/mol), and the label was exclusively located at position 553. This position, like that photolabeled in DT (position 148), corresponds to glutamic acid in the native protein. Chromatographically identical photo- products were generated at these positions in the two toxins. Glu-553 lies in a cleft in domain I11 that is believed to represent the active site of ETA, and other evidence supports the notion that Glu-553 of ETA and Glu-148 of DT are directly involved in catalysis. When Glu-553 of ETA was aligned with Glu-148 of DT, we found similarities of local primary structure not de- tected earlier. These results suggest that the catalyti- cally active domains of ETA and DT may be evolution- arily related, and they provide information that should prove useful for preparing vaccines against ETA by recombinant DNA methods.

Pseudomonas aeruginosa exotoxin A (ETA)‘ and diphtheria toxin (DT) represent two members of an expanding class of bacterial exotoxins that act by ADP-ribosylation mechanisms (1). Such exotoxins catalytically transfer the ADP-ribosyl moiety of NAD into covalent linkage with specific target proteins of eucaryotic cells, resulting in alterations of cellular metabolism. For both ETA and DT, the target protein is elongation factor 2 (EF-2). ADP-ribosylation inactivates EF- 2, blocking protein synthesis and causing cell death. Other examples include cholera toxin and pertussis toxin, which ADP-ribosylate separate regulatory subunits of adenylate cy- clase and thereby modulate intracellular production of CAMP.

The evolutionary relationships between ETA and DT re- main uncertain. With regard to enzymatic properties these two toxins are remarkably similar. Both toxins are highly

* This work was supported by National Institutes of Health Grants AI22848, AI22021, and CA39217, and by the Shipley Institute of Medicine. 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.

~~

.# To whom correspondence should be addressed. ’ The abbreviations used are: ETA, Pseudomonas aeruginosa exo-

toxin A, DT, diphtheria toxin; RP-HPLC, reverse-phase high per- formance liquid chromatography.

specific for EF-2 and ADP-ribosylate the protein at the same position, a post-translationally modified histidine residue termed diphthamide (2,3). The stereochemistry of transfer is identical (3, 4), and the kinetic constants are similar for the two toxins (5). Structurally, however, the two proteins show major differences. The order of functional sites is different, with the catalytic domain being amino-terminal in DT and carboxyl-terminal in ETA. The amino acid sequences report- edly show little homology by conventional methods of com- parison (6); the number and arrangement of disulfide bridges are different; and immunological cross-reactivity is demon- strable only by the most sensitive methods of detection (7).

Using a novel photoaffinity labeling reaction, we recently identified a glutamic acid residue within the catalytic center of DT (8). Irradiating mixtures of DT fragment A and NAD with UV light induced efficient transfer of the nicotinamide moiety to a specific residue, Glu-148. Results from site-di- rected mutagenesis have supported the notion that this resi- due is crucial for catalytic activity; substitution of Glu-148 with aspartic acid reduced enzymatic activity by more than 100-fold, apparently with little effect on the binding of NAD (9). Also, deletion of Glu-148 was found to cause a dramatic decrease in enzymic activity (10).

In the present report we have used the same photoaffinity labeling method to probe the active site of ETA and have explored similarities in primary structure between the two toxins after aligning the photolabeled residues.

EXPERIMENTAL PROCEDURES AND RESULTS*

DISCUSSION

The studies reported here indicate that ETA, like DT, contains a glutamic acid residue within the nicotinamide subsite that is efficiently photolabeled in the presence of NAD. By chromatographic criteria the photoproduct formed at position 553 of ETA is the same as that identified at position 148 of photolabeled DT (13), namely a-amino-y-(6- nicotinamidy1)butyric acid. This photoproduct results from decarboxylation of the glutamic acid side chain, formation of a new bond between the y-methylene carbon of the truncated side chain and the N6 carbon of the nicatinamide ring, and elimination of the ADP-ribose moiety of NAD. Such a struc-

Portions of this paper (including “Experimental Procedures,” “Results,” Figs. 1-6, and Table I) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are available from the Journal of Biological Chemistry, 9650 Rockville Pike, Bethesda, MD 20814. Request Document No. 86M-4093, cite the authors, and include a check or money order for $5.20 per set of photocopies. Full size photocopies are also included in the microfilm edition of the Journal that is available from Waverly Press.

8707

8708 Active Site of P. aeruginosa Exotoxin A

ture implies proximity of the Glu-553 side chain to the nico- tinamide ring of bound NAD in the enzyme-substrate complex and is consistent with a model in which the Glu-553 side chain carboxylate is very close to the nicotinamide-ribose bond ruptured during transfer of ADP-ribose to EF-2.

Other evidence supports the notion that Glu-553 is at the active site and may be directly involved in catalysis. First, in the three-dimensional structure of ETA determined by x-ray crystallography (17), the Glu-553 side chain projects into a cleft in domain 111, the catalytic domain. This cleft is known to bind adenosine3 and may well represent the NAD binding site. Second, the photolabeled ETA fragment was virtually devoid of enzymic activity. Third, substitution of aspartic acid for glutamic acid at position 553 in a cloned ETA fragment has been found to decrease catalytic activity by more than 1000-fold? This result is similar to that obtained when DT Glu-148 was replaced with aspartic acid (9), suggesting that the precise spatial location of the carboxylate group of ETA Glu-553 or DT Glu-148 is important for enzymic activity and that these residues perform similar functions in the ADP- ribosylation reaction.

Although it has been reported that ETA shows no signifi- cant homology with DT (6), the apparent functional identity between ETA Glu-553 and DT Glu-148 provided a new ref- erence point for sequence comparisons. When these residues were aligned in the primary structure and the sequences surrounding these positions were compared, a region of local homology was observed. Within segments of 13 residues con- taining the aligned glutamic acids, 4 residues were identical (31%), and another 7 residues (54%) represented substitutions that are common in evolutionarily related proteins (18):

Exotoxin A (546-558) EEEGGRLETILGW

Diphther ia tox in (141-153) AEGSSSVEYINNW

The same alignment was produced if the active-site segment from ETA was scanned against the entire sequence of DT fragment A (193 amino acids) using the protein sequence similarity program FASTP (19). The reciprocal scan of ETA domain I11 (about 208 amino acids) with the DT segment gave identical results.

It is noteworthy that one of the residues common to both sequences is a tryptophan five positions C-terminal to the aligned glutamic acids, since independent evidence exists that Trp-153 is at or near the catalytic center of DT. Derivitization of this residue with 2-hydroxy-5-nitrobenzyl bromide (20) or substitution of a threonine residue at this position5 inacti- vated the enzyme. Also, binding of NAD to DT fragment A strongly quenches intrinsic tryptophan fluorescence and in- duces a broad weak absorbance band (X,,, 360 nm) that has been postulated to represent a charge-transfer complex (12). The intrinsic tryptophan fluorescence of catalytically active ETA fragments is similarly affected by NAD (5).'j

That these similarities of primary structure within the active sites of ETA and DT reflect broader structural homol- ogy is supported by direct comparison of their complete amino acid sequences. When domain I11 of ETA was used to search (19) the NBRF protein data base (4257 sequences), the most

........ . .. . . .

D. McKay, personal communication. ' C. Douglas and R. J. Collier, manuscript in preparation. ' R. K. Tweten and R. J. Collier, unpublished data.

S. F. Carroll and R. J. Collier, unpublished data.

homologous sequence (aside from ETA itself) was DT frag- ment A. In this alignment, the photolabeled glutamic acids did not coincide, but extensive homology was found between other parts of the two proteins, including regions before and after the active site sequences. These findings will be dis- cussed in a separate communication.

Although other explanations exist, we believe that the functional similarity of ETA Glu-553 and DT Glu-148, to- gether with the apparent homology of both local and distant flanking sequences, indicates that the catalytic domains of ETA and DT represent highly divergent forms of an unknown ancestral domain. A more critical appraisal of evolutionary relationships between ETA and DT will be possible after elucidation of the x-ray crystallographic structure of DT (21), which will permit comparison of the three-dimensional struc- tures of the two proteins. In combination with in vitro muta- genesis of active site residues, such studies may provide a useful approach to the design of vaccines against these and other ADP-ribosylating exotoxins.

Acknowledgments-We thank William Lane for expertise in auto- mated amino acid sequencing and William R. Pearson and the Na- tional Biomedical Research Foundation Molecular Biology Computer Research Resource for providing the program FASTP.

REFERENCES 1. Collier, R. J., and Mekalanos, J. J. (1980) in Multifunctional

Proteim (Bisswanger, H., and Schmincke-Ott, E., eds) pp. 261- 291, John Wiley & Sons, New York

2. Van Ness, B. G., Howard, J. B., and Bodley, J. W. (1980) J. Bwl. Chem. 255,10717-10720

3. Iglewski, B. H., Liu, P. V., and Kabat, D. (1977) Infect. Immun. 15,138-144

4. Oppenheimer, N. J., and Bodley, J. W. (1981) J. Biol. Chem. 256,8579-8581

5. Chung, D. W., and Collier, R. J. (1977) Infect. Immun. 16, 832- 841

6. Gray, G. L., Smith, D. H., Baldridge, J. S., Harkins, R. N., Vasil, M. L., Chen, E. Y., and Heyneker, H. L. (1984) Proc. Natl. Acad. Sei. U. S. A. 81,2645-2649

7. Sadoff, J. C., Buck, G. A., Iglewski, B. H., Bjorn, M. J., and Groman, N. B. (1982) Infect. Zmmun. 37, 250-254

8. Carroll, S. F., and Collier, R. J. (1984) Proc. Natl. Acad. Sci

9. Tweten, R. K., Barbieri, J. T., and Collier, R. J. (1985) J. Biol.

10. Emerick, A., Greenfield, L., and Gates, C. (1985) DNA ( N Y ) 4,

11. Lory, S., and Collier, R. J. (1980) Infect. Immun. 28,494-501 12. Kandel, J., Collier, R. J., and Chung, D. W. (1974) J. Bioi. Chem.

13. Carroll, S. F., McCloskey, J. A., Crain, P. F., Oppenheimer, N. J., Marschner, T. M., and Collier, R. J. (1985) Proc. Natl. Acad. Sci. U. S. A. 82,7237-7241

14. Carroll, S. F., and Martinez, R. J. (1981) Biochemistry 20,5981- 5987

15. Bidlingmeyer, B. A., Cohen, S. A., and Tarvin, T. L. (1984) J. Chromutogr. 336,93-104

16. Leppla, S. H., Martin, 0. C., and Muehl, L. A. (1978) Biochem. Bwphys. Res. Commun. 81,532-538

17. Allured, V. S., Collier, R. J., Carroll, S. F., and McKay, D. B. (1986) Proc. Natl. Acad. Sci. U. S. A. 83, 1320-1324

18. Dayhoff, M. (1978) Atlas of Protein Sequence Anulysis, Vol. 5, Suppl. 3, p. 352, National Biomedical Research Foundation, Silver Spring, MD

19. Lipman, D. J., and Pearson, W. R. (1985) Science 2 2 7 , 1435- 1441

20. Michel, A., and Dirkx, J. (1977) Biochim. Eiophys. Acta 491 ,

21. Collier, R. J., Westbrook, E. M., McKay, D. B., and Eisenberg,

U. S. A. 81, 3307-3311

Chem. 260, 10392-10394

78

249,2088-2097

286-295

D. (1982) J. Bwl. Chem. 257 , 5283-5285

Active Site of P. aeruginosa Exotoxin A 8709

8710 Active Site of P. aeruginosa Exotoxin A

94 K+ 67K -. 43K+

30K-c

20.1 K-

1 2 3 4 5 6 7 8 9 1 0

0 1 5 10 15 20 25 50 Thermolysin, pg/ml

F ~ g u ~ e 1 Production of an enzymically a c I w ETA fragment by parltal dws l ian w l lh thermolyr8n

ETA (5 mg/ml) tn T E buffer was activated by cncub8ltm with urea-dilh~othre~tol. a3 dercribed under

Ex~eromenfal Procedures. and carboxymefhylafed with iodoacetic acid After exchange #"lo 50 mM

Trir-HCI, pH 8 0, and addition of NAD 10 1 mM, ETAac, (2 5 mg/ml) was digested with #mirearing

conccntrafionr of thermolyrm lor 30 min at 25'C The reacttons were terminated by addmg EDTA

to 5 mM, and aliquols were analyzed by eleclraphare3ir in 10% polyacrylam8de gels ~ o n l a n i n g sodium

dodecyl sulfate Lanes 1 and 10, molec~lsr wetght rlandardr The remainmg h e r conlam sampler

of ETA,,, dtgesled with the indicaled ~oncenlrations of lhcrmolyrin (MI 37.500) The predominate

lntermedmler ~n the degredalion of ETAac1 are fragments of MI 42.Mx). 34.000, and 27,000

.s I s % 3 06

\ F

3 04

Q.

-2 & si * 0 2

B \

4 ' 0 0 6 0 IO 20 30 40 50 6C

frradiolion Time, min

1 .:

? Q 1s I

4

$ O!

n ld 50 70 90 110 130

frucfion ~ & K Z 3 Fractionation of (carbonyl-14C)-labeled ETA peptides by ~~ze-exdus ion

chromatography The enzymically active ETA fragment was irradiated for 45 min 21 3 m W / m 2 ~n the prerenceof [ca,b~nyl-'~C]NAD, desalted io10 acetic acid. and lyaphylized After rcrwpenr8on ~n 0 2 M N-ethylmorpholine-acetale. 0 1 mM CaCI2, pH 8 2. lhe phololabeled fragment was digested

wlrh trypsin (2%. w/w) far 4 h 11 37% The reaction was terminated by addlng g l a ~ # d l a c e t ~

a~idro30%(v/v) ,andlhediger twarappl~edtoaO5~HOcm~olumnofSephade~G-50ruperf inc

equilibrated in the same solvent The flow rale was 0 8 ml/h, and 150 PI frxtions were

collected Aliquot$ of alternate frd~lions were removed and counted by lhquid rcinlillalion

Fractianrwerepwled arlollowr G50-I, 66-78. G50-11, 79-91

I1 I I I I I

0

b

m~rturer (50 mM Trir-HCI. pH 7 5 ) containing profem (201rM) and NAD (40 PM) raddabeled in the

4°C and Nrradmed on icewilh a G15-T8 germtcidal lamp (predominantly 253 7 nm) The cntenr#ly nlcofinamidc (0.a). adenylatephorphale (.),or Ihcadrnsne (A) moiety were prepared at

of UV hght received by the sampler was 3 mW/cm2 At 1nterva11 up to 60 min. aliquot5 were

removed and mired with carrier ovalbumin Acid-prccwilable radioactivity was then dctermmed

0 IO 20 30 40 50 Elufion rime, min

- 9 Purificationof the (ca,bonyl-14C]-labeled 650-11 peptide by reverse-phase HPLC

Fraclnonr Eorrerpondlng to the major radloacrive peak eluting from Sephadei G-50 (G50- 11) were

pooled. concentrated by lyaphilizalion. and chromalographed on a Waters C18 uBondapak column

The d u m n was eluted with a linear gradient of aceton#trde (16 to 43%. 42 min) tn

trtllumoacetic acid (0 I%, "/VI, lhe affluent was monitored far both absorbance at 214 nm

(panel 2 ) and for radioactivity (panel b)

Active Site of P. aeruginosa Exotoxin A 8711

I I r I I I

F O o 4 1 * ILe" -

1 Pro I l l