Biochemical analysis of CRM 66. A nonfunctional Pseudomonas aeruginosa exotoxin A

Pseudomonas aeruginosa exotoxin A (ETA) is an ADP-ribosyltransferase which inactivates protein synthesis by covalently attaching the ADP-ribose portion of NAD+ onto eucaryotic elongation factor 2 (EF-2). A direct biochemical comparison has been made between ETA and a nonenzymatically active mutant toxin (CRM 66) using highly purified preparations of each protein. The loss of ADP-ribosyltransferase activity and subsequent cytotoxicity have been correlated with the presence of a tyrosine residue in place of a histidine at position 426 in CRM 66. In the native conformation, CRM 66 demonstrated a limited ability (by a factor or at least 100,000) to modify EF-2 covalently and lacked in vitro and in vivo cytotoxicity, yet CRM 66 appeared to be normal with respect to NAD+ binding. Upon activation with urea and dithiothreitol, CRM 66 lost ADP-ribosyltransferase activity entirely yet CRM 66 retained the ability to bind NAD+. Replacement of Tyr-426 with histidine in CRM 66 completely restored cytotoxicity and ADP-ribosyltransferase activity. These results support previous findings from this laboratory (Wozniak, D. J., Hsu, L.-Y., and Galloway, D. R. (1988) Proc. Natl. Acad. Sci. U. S. A. 85, 8880-8884) which suggest that the His-426 residue of ETA is not involved in NAD+ binding but appears to be associated with the interaction between ETA and EF-2.     

Alkylation of cysteine 41, but not cysteine 200, decreases the ADP-ribosyltransferase activity of the S1 subunit of pertussis toxin

Sulfhydryl-alkylating reagents are known to inactivate the NAD glycohydrolase and ADP-ribosyltransferase activities of the S1 subunit of pertussis toxin, a protein which contains two cysteines at positions 41 and 200. It has been proposed that NAD can retard alkylation of one of the two cysteines of this protein (Kaslow, H.R., and Lesikar, D.D. (1987) Biochemistry 26, 4397-4402). We now report that NAD retards the ability of these alkylating reagents to inactivate the S1 subunit. In order to determine which cysteine is protected by NAD, we used site-directed mutagenesis to construct analogs of the toxin with serines at positions 41 and/or 200. Sulfhydryl-alkylating reagents reduced the ADP-ribosyltransferase activity of the analog with a single cysteine at position 41; NAD retarded this inactivation. In contrast, sulfhydryl-alkylating reagents did not inactivate analogs with serine at position 41. An analog with alanine at position 41 possessed substantial ADP-ribosyltransferase activity. We conclude that alkylation of cysteine 41, and not cysteine 200, inactivates the S1 subunit of pertussis toxin, but that the sulfhydryl group of cysteine 41 is not essential for the ADP-ribosyltransferase activity of the toxin. These results suggest that the region near cysteine 41 contributes to features of the S1 subunit important for ADP-ribosyltransferase activity. Using site-directed mutagenesis, we found that changing aspartate 34 to asparagine, arginine 39 to lysine, and glutamine 42 to glutamate had little effect on ADP-ribosyltransferase activity. However, substituting an asparagine for the histidine at position 35 markedly decreased, but did not eliminate, ADP-ribosyltransferase activity. Chou-Fasman analysis predicted no significant modifications in secondary structure of the S1 peptide with the change of histidine 35 to asparagine. Thus, histidine 35 may interact with a substrate of the S1 subunit without being essential for catalysis.     

The cytotoxic effect of diphtheria toxin on the actin cytoskeleton

Diphtheria toxin (DT) and its N-terminal fragment A (FA) catalyse the transfer of the ADP-ribose moiety of nicotinamide adenine dinucleotide (NAD) into a covalent linkage with eukaryotic elongation factor 2 (eEF2). DT-induced cytotoxicity is versatile, and it includes DNA cleavage and the depolymerisation of actin filaments. The inhibition of the ADP-ribosyltransferase (ADPrT) activity of FA did not affect the deoxyribonuclease activity of FA or its interaction with actin. The toxin entry rate into cells (HUVEC) was determined by measuring the ADP-ribosyltransferase activity. DT uptake was nearly 80% after 30 min. The efficiency was determined as K_m = 2.2 nM; V_max = 0.25 pmol.min⁻¹. The nuclease activity was tested with hyperchromicity experiments, and it was concluded that G-actin has an inhibitory effect on DT nuclease activity. In thepresence of DT and mutant of diphtheria toxin (CRM197), F-actin depolymerisation was determined with gel filtration, WB and fluorescence techniques. In the presence of DT and CRM197, 60–65% F-actin depolymerisation was observed. An in vitro FA-actin interaction and F-actin depolymerisation were reported in our previous paper. The present study thus confirms the depolymerisation of actin cytoskeleton in vivo.     

Immunochemical identification of the ADP-ribosyltransferase in botulinum C1 neurotoxin as C3 exoenzyme-like molecule

Botulinum C1 neurotoxin and C3 exoenzyme were purified to apparent homogeneity from the culture filtrate of Clostridium botulinum type C strain 003-9. Both preparations catalyzed ADP-ribosylation of the same substrate, the Mr 22,000 rho gene product (Gb). When the light and heavy chains of C1 toxin were separated, ADP-ribosyltransferase activity in the toxin was quantitatively recovered in the light chain fraction. Anti-C1 toxin antiserum precipitated the ADP-ribosyltransferase activity and the neurotoxicity of C1 toxin in parallel, whereas it had no effect on C3 exoenzyme. On the other hand, anti-C3 exoenzyme antiserum precipitated the ADP-ribosyltransferase activities of both C3 exoenzyme and C1 toxin. This antibody, however, did not precipitate the neurotoxicity of C1 toxin. The ADP-ribosyltransferase in C1 toxin was quantitatively adsorbed onto the anti-C3 antibody column and separated from the majority of C1 toxin protein. The enzyme was then eluted with acidic urea and Western blotting analysis of this eluate revealed the appearance of a protein band positively stained with anti-C3 antibody at a position similar to that of C3 exoenzyme. Quantitative determination by enzyme-linked immunosorbent assay showed that the C3-like immunoreactivity is present in the C1 toxin molecules at the molecular ratio of 1 to 1,000. These results suggest that the ADP-ribosyltransferase activity in C1 toxin is expressed by a C3-like molecule which is present in a small amount in the toxin preparation and appears to bind to the toxin component(s). The above results also indicate that the ADP-ribosyltransferase in C1 toxin is not related to its neurotoxin action.     

Identification of a single amino acid substitution in the diphtheria toxin A chain of CRM 228 responsible for the loss of enzymatic activity.

CRM 228 (T. Uchida, A. M. Pappenheimer, and R. Greany, J. Biol. Chem. 248:3838-3844, 1973), a mutant form of diphtheria toxin which completely lacks ADP-ribosyltransferase activity, contains five amino acid substitutions. The two amino acid changes that fall within the A chain of the toxin (G79D and E162K) were separately analyzed by substituting a variety of other amino acids at these sites. The substitution at position 79 (G79D) singularly appears to account for the loss of enzymatic activity found in CRM 228.     

Restoration of enzymic activity and cytotoxicity of mutant, E553C, Pseudomonas aeruginosa exotoxin A by reaction with iodoacetic acid

Pseudomonas aeruginosa exotoxin A (ETA) is inactivated greater than 1,000-fold when an active site glutamic acid, E553, is mutated to aspartic acid (Douglas, C.M., and Collier, R. J. (1987) J. Bacteriol. 169, 4967-4971). To test the effect of creating a carboxyl-containing side chain at position 553 longer than that of glutamic acid, we first replaced Glu-553 with cysteine by site-directed mutagenesis of cloned ETA and then carboxymethylated the cysteine side chain with iodoacetic acid. The E553C mutation reduced ADP-ribosyltransferase and cytotoxic activities greater than 10,000-fold. Reaction of the mutant with iodoacetic acid enhanced enzymic activity 2,500-fold, to a level approximately one-sixth that of wild type toxin, and restored cytotoxicity to a slightly lesser extent. Iodoacetamide did not activate the mutant, and neither iodoacetic acid nor iodoacetamide affected the activity of wild type toxin. These results show that the carboxyl group of Glu-553 is important for ADP-ribosylation activity and imply flexibility in the enzyme-substrate complex in accommodating the slightly longer S-carboxymethylcysteine side chain. This general approach may have applications in protein engineering as well as in studying carboxyl side chain functions in enzymes.     

Enzymic activity of cholera toxin. II. Relationships to proteolytic processing, disulfide bond reduction, and subunit composition.

Cholera toxin containing intact A chain (Mr = 29,000) was isolated, and its enzymic properties were characterized. The "unnicked" form of the toxin, produced by a protease-deficient, hypertoxinogenic mutant of Vibrio cholerae 569B, had greatly reduced activity in catalyzing the NAD+-glycohydrolase and ADP-ribosyltransferase reactions as compared to the naturally nicked form commonly isolated. In the latter, the intact A chain has been cleaved by bacterial proteases to yield disulfide-linked A1 and A2 chains (Mr = 23,000 and 6,000, respectively). Digestion of unnicked toxin with trypsin or elastase yielded a nicked form similar to or identical with the naturally nicked toxin, but chymotryptic digestion did not. Disulfide bond reduction was necessary for expression of enzymic activity by naturally nicked or trypsin-nicked toxin, or the A1A2 protomer. Fractionation of thiol-treated, nicked cholera toxin by ion exchange, molecular exclusion, or affinity chromatography gave results suggesting that the reduced toxin displays enzymic activity while remaining structurally intact.     

Active-site mutations of diphtheria toxin: effects of replacing glutamic acid-148 with aspartic acid, glutamine, or serine

Glutamic acid-148, an active-site residue of diphtheria toxin identified by photoaffinity labeling with NAD, was replaced with aspartic acid, glutamine, or serine by directed mutagenesis of the F2 fragment of the toxin gene. Wild-type and mutant F2 proteins were synthesized in Escherichia coli, and the corresponding enzymic fragment A moieties (DTA) were derived, purified, and characterized. The Glu----Asp (E148D), Glu----Gln (E148Q), and Glu----Ser (E148S) mutations caused reductions in NAD:EF-2 ADP-ribosyltransferase activity of ca. 100-, 250-, and 300-fold, respectively, while causing only minimal changes in substrate affinity. The effects of the mutations on NAD-glycohydrolase activity were considerably different; only a 10-fold reduction in activity was observed for E148S, and the E148D and E148Q mutants actually exhibited a small but reproducible increase in NAD-glycohydrolytic activity. Photolabeling by nicotinamide-radiolabeled NAD was diminished ca. 8-fold in the E148D mutant and was undetectable in the other mutants. The results confirm that Glu-148 plays a crucial role in the ADP-ribosylation of EF-2 and imply an important function for the side-chain carboxyl group in catalysis. The carboxyl group is also important for photochemical labeling by NAD but not for NAD-glycohydrolase activity. The pH dependence of the catalytic parameters for the ADP-ribosyltransferase reaction revealed a group in DTA-wt that titrates with an apparent pKa of 6.2-6.3 and is in the protonated state in the rate-determining step.(ABSTRACT TRUNCATED AT 250 WORDS)     

Separation of the 24 kDa substrate for botulinum C3 ADP-ribosyltransferase and the cholera toxin ADP-ribosylation factor

Botulinum C3 ADP-ribosyltransferase modifies a approximately 24 kDa membrane protein believed to bind guanine nucleotides. Cholera toxin ADP-ribosylation factors are approximately 19 kDa GTP-binding proteins that directly activate the toxin. To evaluate a possible relationship between C3 ADP-ribosyltransferase substrate and ADP-ribosylation factor, they were partially purified from bovine brain. ADP-ribosylation factor, but not C3 ADP-ribosyltransferase substrate, stimulated auto-ADP-ribosylation of the choleragen A1 subunit whereas C3 ADP-ribosyltransferase substrate, but not ADP-ribosylation factor, was ADP-ribosylated by C3 ADP-ribosyltransferase. Thus, although both may be GTP-binding proteins, no functional similarity between ADP-ribosylation factor and C3 ADP-ribosyltransferase substrate was found.     

Structure-function analysis of the Rho-ADP-ribosylating exoenzyme C3stau2 from Staphylococcus aureus

Exoenzyme C3stau2 from Staphylococcus aureus is a new member of the family of C3-like ADP-ribosyltransferases that ADP-ribosylates RhoA, -B, and -C. Additionally, it modifies RhoE and Rnd3. Here we report on studies of the structure-function relationship of recombinant C3stau2 by site-directed mutagenesis. Exchange of Glu(180) with leucine caused a complete loss of both ADP-ribosyltransferase and NAD glycohydrolase activity. By contrast, exchange of the glutamine residue two positions upstream (Gln(178)) with lysine blocked ADP-ribosyltransferase activity without major changes in NAD glycohydrolase activity. NAD and substrate binding of this mutant protein was comparable to that of the recombinant wild type. Exchange of amino acid Tyr(175), which is part of the recently described "ADP-ribosylating toxin turn-turn" (ARTT) motif [Han, S., Arvai, A. S., Clancy, S. B., and Tainer, J. A. (2001) J. Mol.Biol. 305, 95-107], with alanine, lysine, or threonine caused a loss of or a decrease in ADP-ribosyltransferase activity but an increase in NAD glycohydrolase activity. Recombinant C3stau2 Tyr175Ala and Tyr175Lys were not precipitated by matrix-bound Rho, supporting a role of Tyr(175) in protein substrate recognition. Exchange of Arg(48) and/or Arg(85) resulted in a 100-fold reduced transferase activity, while the recombinant C3stau2 double mutant R48K/R85K was totally inactive. The data indicate that amino acid residues Arg(48), Arg(85), Tyr(175), Gln(178), and Glu(180) are essential for ADP-ribosyltransferase activity of recombinant C3stau2 and support the role of the ARTT motif in substrate recognition of RhoA by C3-like ADP-ribosyltransferases.     

Pseudomonas aeruginosa exotoxin A: alterations of biological and biochemical properties resulting from mutation of glutamic acid 553 to aspartic acid

Glutamic acid 553 of Pseudomonas aeruginosa exotoxin A (ETA) was identified earlier as a putative active-site residue by photoaffinity labeling with NAD. Here ETA-E553D, a cloned form of the toxin in which Glu-553 has been replaced by aspartic acid, was purified from Escherichia coli extracts and characterized. Cytotoxicity of the mutant toxin for mouse L-M cells was less than 1/400,000 that of the wild type. The mutation caused a 3200-fold reduction in NAD:elongation factor 2 ADP-ribosyltransferase activity, as estimated by assays with an active fragment derived from the toxin by digestion with thermolysin. NAD glycohydrolase activity was reduced somewhat less, by a factor of 50, and photoaffinity labeling with NAD by a factor of 2. We detected less than 2-fold change in the values of KM for NAD or elongation factor 2 and no change in KD for NAD, as determined by quenching of protein fluorescence. The drastic reduction of ADP-ribosyltransferase activity therefore results primarily from an effect of the mutation on kcat, implying that Glu-553 plays an important and possibly direct role in catalyzing this reaction. The effects of the E553D mutation are similar to those of the E148D mutation in diphtheria toxin, supporting the notion that these two Glu residues perform the same function in their respective toxins.     

Artificial hybrid protein containing a toxic protein fragment and a cell membrane receptor-binding moiety in a disulfide conjugate. II. Biochemical and biologic properties of diphtheria toxin fragment A-S-S-human placental lactogen.

The biochemical and biologic properties of a purified disulfide conjugate of diphtheria toxin fragment A and human placental lactogen (toxin A-hPL) have been studied by (a) assaying the ADP-ribosyltransferase activity of the intact conjugate, (b) assaying the binding of the intact conjugate to mammary gland plasma membrane lactogenic receptors, and (c) assaying the effect of the conjugate on the rate of protein synthesis in rabbit mammary gland explants maintained in organ culture. The toxin A-hPL conjugate retains one-third of the NAD+:EF-2 ADP-ribosyltransferase activity of toxin A, and 26% of the hPL-binding activity to lactogenic receptors. Binding activity was demonstrated by radioreceptor assay and by assaying toxin A activity bound to membranes which was competitively displaced by excess hPL. Since the toxin A-hPL conjugate retained activities of its separate subunits, it could be regarded as a structural analogue of nicked diphtheria toxin with replacement of the original membrane-binding chain by another binding chain that is specific for lactogenic receptor. However, the conjugate failed to inhibit protein synthesis in organ-cultured mammary gland explants, although these were sensitive to native diphtheria toxin and could bind hPL. It is concluded from these results that the toxin A-hPL conjugate does not act as a functional analogue of diphtheria toxin with altered receptor specificity, and that the hPL receptor cannot mediate the entry of toxin A or toxin A-hPL from membrane-bound conjugate into the cytosol site of action of toxin A.     

Conformational integrity of a recombinant toxoid of Pseudomonas aeruginosa exotoxin A containing a deletion of glutamic acid-553.

A mutant form of Pseudomonas aeruginosa exotoxin A (ETA) carrying a deletion of glutamic acid-553, an important active-site residue, was expressed in an ETA-negative strain of P. aeruginosa and shown to be exported from the cells as efficiently as wild-type ETA. The mutant protein, purified from the culture medium, was devoid of ADP-ribosyltransferase activity. Protein conformation was barely perturbed by the deletion, as determined by a number of measures, including affinity for substrate NAD, proteinase sensitivity, absorbance and fluorescence spectroscopy, and differential scanning calorimetry. The conformational integrity and stability of the mutant toxin are consistent with potential use of the protein in vaccines or as a carrier in preparing conjugate vaccines.     

The ADP-ribosylating mosquitocidal toxin from Bacillus sphaericus: proteolytic activation, enzyme activity, and cytotoxic effects.

The mosquitocidal toxin (MTX) from Bacillus sphaericus SSII-1 is a approximately 97-kDa protein sharing sequence homology within the N terminus with the catalytic domains of various bacterial ADP-ribosyltransferases. Here we studied the proteolytic activation of the ADP-ribosyltransferase activity of MTX. Chymotrypsin treatment of the 97-kDa MTX holotoxin (MTX(30-870)) results in a 70-kDa putative binding component (MTX(265-870)) and a 27-kDa enzyme component (MTX(30-264)), possessing ADP-ribosyltransferase activity. Chymotryptic cleavage of an N-terminal 32-kDa fragment of MTX (MTX(30-308)) also yields MTX(30-264), but the resulting ADP-ribosyltransferase activity is much greater than that of the processed MTX(30-870). Kinetic studies revealed a K(m) NAD value of 45 microm for the processed 32-kDa MTX fragment, and a K(m) NAD value of 1300 microm for the processed holotoxin. Moreover, the k(cat) value for the activated MTX(30-308) fragment was about 10-fold higher than that for the activated holotoxin (MTX(30-870)). Precipitation analysis showed that the 70-kDa proteolytic fragment of MTX remains noncovalently bound to the N-terminal 27-kDa fragment, thereby inhibiting ADP-ribosyltransferase and NAD glycohydrolase activities. Glu(197) of MTX(30-264) was identified as the "catalytic" glutamate that is conserved in all ADP-ribosyltransferases. Whereas mutated MTX(30-264)E197Q has neither ADP-ribosyltransferase nor NAD glycohydrolase activity, mutated MTX(30-264)E195Q possesses glycohydrolase activity but not transferase activity. Transfection of HeLa cells with a vector encoding a fusion protein of MTX(30-264) with a green fluorescent protein led to cytotoxic effects characterized by cell rounding and formation of filopodia-like protrusions. These cytotoxic effects were not observed with the catalytically inactive MTX(30-264)E197Q mutant, indicating that the MTX enzyme activity is essential for the cytotoxicity in mammalian cells.     

Functional significance of active site residues in the enzymatic component of the Clostridium difficile binary toxin

Clostridium difficile binary toxin (CDT) is an ADP-ribosyltransferase which is linked to enhanced pathogenesis of C. difficile strains. CDT has dual function: domain a (CDTa) catalyses the ADP-ribosylation of actin (enzymatic component), whereas domain b (CDTb) transports CDTa into the cytosol (transport component). Understanding the molecular mechanism of CDT is necessary to assess its role in C. difficile infection. Identifying amino acids that are essential to CDTa function may aid drug inhibitor design to control the severity of C. difficile infections. Here we report mutations of key catalytic residues within CDTa and their effect on CDT cytotoxicity. Rather than an all-or-nothing response, activity of CDTa mutants vary with the type of amino acid substitution; S345A retains cytotoxicity whereas S345Y was sufficient to render CDT non-cytotoxic. Thus CDTa cytotoxicity levels are directly linked to ADP-ribosyltransferase activity.     

Diphtheria toxin translocation across cellular membranes is regulated by sphingolipids

Diphtheria toxin is translocated across cellular membranes when receptor-bound toxin is exposed to low pH. To study the role of sphingolipids for toxin translocation, both a mutant cell line lacking the first enzyme in de novo sphingolipid synthesis, serine palmitoyltransferase, and a specific inhibitor of the same enzyme, myriocin, were used. The serine palmitoyltransferase-deficient cell line (LY-B) was found to be 10-15 times more sensitive to diphtheria toxin than the genetically complemented cell line (LY-B/cLCB1) and the wild-type cell line (CHO-K1), both when toxin translocation directly across the plasma membrane was induced by exposing cells with surface-bound toxin to low pH, and when the toxin followed its normal route via acidified endosomes into the cytosol. Toxin binding was similar in these three cell lines. Furthermore, inhibition of serine palmitoyltransferase activity by addition of myriocin sensitized the two control cell lines (LY-B/cLCB1 and CHO-K1) to diphtheria toxin, whereas, as expected, no effect was observed in cells lacking serine palmitoyltransferase (LY-B). In conclusion, diphtheria toxin translocation is facilitated by depletion of membrane sphingolipids.     

TPA induction of EL4 resistance to macrophage-released TNF: role of ADP-ribosylation in tumoricidal activities of TNF and other factors

Activated macrophages synthesize and release numerous tumoricidal soluble factors that can be divided into receptor- or nonreceptor-dependent agents. Tumor necrosis factor (TNF) would be an example of the former. In our experimental model the killing of EL4 thymoma cells by syngeneic activated macrophages involves, but not exclusively, TNF. Our results show that approximately 50% of the anti-EL4 activity expressed by macrophages can be specifically inhibited with rabbit anti-mouse TNF antibody. EL4 variants resistant to the lytic activity of TNF were still susceptible to macrophage-mediated lysis. A tumor-promoting phorbol ester, TPA, rendered TNF-sensitive and -insensitive EL4 cells resistant to M phi-mediated lysis. However, TPA down-regulated TNF-specific binding sites on both TNF-sensitive and -resistant cell surface membranes, suggesting that resistance to TNF involves postligand:receptor events. Tumor cell G-protein involvement (ADP-ribosylation), as a result of TNF-TNF receptor interactions, was investigated. The results showed that pertussis toxin was cytotoxic against TNF-sensitive and -resistant EL4 cells but not against TPA-treated target cells. Inhibitors of ADP-ribosyltransferase inhibited pertussis toxin cytotoxicity and macrophage-mediated lysis but did not interfere with recombinant TNF lytic activity.     

Evaluation of the proposed interaction of nucleic acid with Clostridium difficile toxins A and B and the effects of nucleases on cytotoxicity

Both DNA and RNA were found to co-purify with Clostridium difficile toxin B but not toxin A. DNAase treatment greatly reduced the cytotoxicity of toxin B but not of toxin A. RNAase had no effect on either toxin. The effects on toxin B were shown to be due to a contaminating protease and could be inhibited by the serine protease inhibitor phenylmethylsulphonyl fluoride.     

The 1.8 Å cholix toxin crystal structure in complex with NAD+ and evidence for a new kinetic model

Certain Vibrio cholerae strains produce cholix, a potent protein toxin that has diphthamide-specific ADP-ribosyltransferase activity against eukaryotic elongation factor 2. Here we present a 1.8 Å crystal structure of cholix in complex with its natural substrate, nicotinamide adenine dinucleotide (NAD⁺). We also substituted hallmark catalytic residues by site-directed mutagenesis and analyzed both NAD⁺ binding and ADP-ribosyltransferase activity using a fluorescence-based assay. These data are the basis for a new kinetic model of cholix toxin activity. Further, the new structural data serve as a reference for continuing inhibitor development for this toxin class.     

Biochemical and immunochemical studies of proteolytic fragments of exotoxin A from Pseudomonas aeruginosa

Limited proteolysis of Pseudomonas aeruginosa exotoxin A by four proteases (chymotrypsin, Staphylococcal serine proteinase, pepsin A and subtilisin) resulted in the formation of polypeptides having a molecular mass of approximately 25 kDa. They possessed both enzymatic activity and residual antigenicity. Their N-terminal sequence analysis showed that the different proteases cleaved exotoxin A in a very restricted area within domain Ib (amino acids 365-404). As a result, the polypeptides contained a large portion (13-34 amino acids) of domain Ib linked to the adjacent C-terminal domain III (amino acids 405-613). The major fragment derived from subtilisin cleavage, at a final yield of 35% (S-fragment; residues 392-613; 24201 Da; pI 4.7) possessed the same level of ADP-ribosyltransferase activity as uncleaved exotoxin A (by mass), and a 37-fold higher NAD-glycohydrolase activity. Polyclonal antibodies from rabbits against exotoxin A completely inhibited the ADP-ribosyltransferase activity of both exotoxin A and the S-fragment, but not the NAD-glycohydrolase activity of the S-fragment. Antibodies against the S-fragment neutralized the ADP-ribosyltransferase activity of exotoxin A. These data determine the primary proteolytic cleavage site of exotoxin A, suggest that some residues in the amino acid sequence 392-404 of exotoxin A seem to have a role in binding or positioning elongation factor 2 (EF-2) and show that antibodies recognize the EF-2-binding site but not the NAD(+)-binding site.