Pseudomonas aeruginosa Exotoxin A Interaction with Eucaryotic ...

T H EJOURNALOF BIOLOGICAL CHEMISTRY cs 1992 by The American Society for Biochemistry and Molecular B ~ o l o a Inc. ,

Vol. 267, No. 27, Issue of September 25, pp. 19107-19111,1992 Prrnted In U.S.A.

Pseudomonas aeruginosa Exotoxin A Interaction with Eucaryotic Elongation Factor 2 ROLEOFTHEHis426RESIDUE* (Received for publication, March 25, 1992)

Sean P. KesslerS and Darrell R. Galloways From the Department of Microbiology, Columbus, Ohio 43210-1292

Pseudomonasaeruginosa exotoxin A (ETA) catalyzes the transfer of the ADP-ribose moiety of NAD+ onto eucaryotic elongation factor 2 (EF-2). To study the ETA site of interaction with EF-2, animmobilized EF-2 binding assay was developed. This assay demonstrates that ETA, in the presence of NAD+, binds to immobilized EF-2. Additionally, diphtheria toxin was also found to bind to the immobilized EF-2 in the presence of NAD’. Comparative analysiswas performed with a mutated form of ETA (CRM 66) in which a histidine residue at position 426 has been replaced with a tyrosine residue. This immunologically crossreactive, ADP-ribosyl transferase-deficient toxindoes not bind to immobilized EF-2, thus explaining its lack of ADPRT activity. ETA bound to immobilized EF-2 cannot bind the monoclonal antibody TC- 1 which specifically recognizes the ETA epitope containing Immunoprecipitation of native ETA by mAb TC-1 is only achieved by incubating ETA in the presence of NAD+.Diethyl pyrocarbonate modification of the residue blocks ETA binding to EF-2 and prevents the binding of the TC-1 antibody. Analogs of NAD+ containing a reduced nicotinamide ring or modified adenine moieties cannot substitute for NAD+ in the immobilized binding assay. Collectively,these data support our proposal that the site of ETA interaction with EF-2includes His426and that a molecule of NAD+ is required for stable interaction.

covalently modifies eucaryotic elongation factor 2 (EF-2) so that it nolonger functions in protein synthesis(2). Due to its enzymaticnature,justoneinternalizedETA molecule is sufficient to kill an ETA-sensitive cell. The x-raycrystallographic structure of ETA at 3.0-A resolution has been determined (3), and the structural gene for ETA (tox A) has been cloned and sequenced (4, 5). These studies indicate that ETA possesses three structuraldomains. Domain I is composed of subdomain Ia (1-252), postulated to play arole in ETA binding ato specific cell surface receptor(s), and subdomain Ib(365-404) with no known function. Following receptor-mediated endocytosis, domain I1 (253-364) is believed to aid in translocationof ETA or an ETA fragment from the endocytic vesicle into thecytoplasmic compartment. The ADPRT active region (catalytic domain) of the 66,583dalton protein, lies in domain I11 (405-613) which contains the NAD+ binding site (active site) and asignificantcleft structure bordered by two helices (3, 6). Current research in several laboratorieshas focused on activity. The domain I11 structure as it relates to ADPRT active site of ETA has been correlated with NAD’ binding, and several critical residues have been identified within this site (7), e.g. Glu553.Our previous work has indicated that the ~ i ~ 4 2residue 6 also plays a key role in ADPRT activity (8, 9) and is located within a major cleft which may be associated with EF-2 binding (9, 10). Comparative studies utilizinga mutated ADPRT-deficient toxin designated CRM 66, which contains a tyrosinein place of histidine a t position 426, revealed theimportance of the residue in ADP-ribosylation of EF-2 (8). This mutation nearly abolishes ADPRT Pseudomonasaeruginosa exotoxinA (ETA)’isthebest activity butdoes not affect NAD’ binding (8).Further analycharacterized extracellular virulence factor associatedwith sis was performed usingthe monoclonal antibody TC-1,which this opportunistic pathogen. ETA, one of several bacterial recognizes an epitopewithin the majorcleftincluding the toxins known as mono-ADP-ribosyltransferases (ADPRT), residue. Exposure of this cleft epitope occurs following catalyzes the transfer of the ADP-ribosemoiety of NAD+ onto treatment with urea and dithiothreitol. TC-1 is able to bind a specific target protein (1). Like diphtheria toxin (DT), ETA ETA and inhibit ADPRT activity of the ADPRTactive form of toxin without diminishing NAD+ glycohydrolase activity * This work was supported by National Institutes of Health Grant (10). Therelationship between exposure by conformaAI123429-04 and by Research Career Development Award AI0087602. This work was partially funded by a grant from Sigma Xi, the tional change and ADPRT activity was investigated using a Research Society. The costs of publication of this article were de- 36-kDa toxinfragment (PE 36) which containstheETA frayed in part by the payment of page charges. This article must catalytic domain. This fragment has full enzymatic activity therefore be hereby marked “aduertisement” in accordance with 18 and bindsmAb TC-1 without the requirement for prior urea/ U.S.C. Section 1734 solely to indicate this fact. $Presentaddress:The ClevelandClinic Foundation,Dept. of dithiothreitol treatment toeffect any conformationalchange. Monoclonal antibody TC-1 binding toP E 36 and subsequent Molecular Biology, 9500 Euclid Ave., Cleveland, OH 44195. $! To whom correspondence and reprint requests should be ad- inhibition of its ADPRTactivity confirms that PE36 isin the dressed Dept. of Microbiology, The Ohio State University,484 West conformation necessaryfor ADPRT activity and that the 12th Ave., Columbus, OH 43210-1292. Tel.: 614-292-3761; Fax: 614- major cleft is in the open figuration (10). These findings led 292-1538. to ourproposal that His42fiis involved in the interactionwith ’ The abbreviations used are: ETA, P. aeruginosa exotoxin A; DT, EF-2 and that this binding event is related to arequired C. diphtheriae diphtheria toxin; EF-2, elongation factor 2; ADPRT, ADP-ribosyltransferase; mAb, monoclonal antibody; ELISA,enzyme- conformational change in thetoxin molecule. linked immunosorbent assay; DEPC, diethyl pyrocarbonate. The data presented in this report more precisely defines

19107

Analysis of P. aeruginosa Exotoxin

19108

the relationship between ETA structure and its enzymatic function by analyzing the interaction between ETA and its protein target EF-2. The results support the proposal for a two-step process for the ADPRT reaction which states: 1) ETA mustfirstbind NAD'; 2) NAD+ binding induces a conformational change in ETAresulting in exposure of His426; 3) subsequent to NAD+ binding, the charged ETA. NAD+ complex is capable of interacting with EF-2; 4) the ADPribose moiety of NAD+ is transferred to diphthamide 715 on EF-2 concurrent with release of nicotinamide and H+. ETA must possess the His426residue and a bound molecule of NAD+ in order to bind EF-2 and carry out itstransferase activity. EXPERIMENTAL PROCEDURES

Purification of ETA, CRM 66, and EF-2-ETA and CRM 66 were purified from the culture supernatant fraction of P. aeruginosa strains PA103-29/pGW28 and PAO-PRl/pGW28as previously described (8, 11,12).Elongation factor 2 was purified from wheat germ as described by Breitenberger et al. (13, 14). Immobilized EF-2 Binding Assay-The procedure was performed as a modification of a previously published method (15). Immulon I1 microtiter plates(Dynatech Laboratories, Alexandria, VA)were coated with approximately 1.30 pmol of purified EF-2 in coating buffer (CB) (0.1 M sodium carbonate, 0.02% sodium azide (pH 9.6)) for 24 h at 37 'C. After washing with 40 mM borate buffered saline (pH 7.8), 0.1% sodium azide, 0.5% Tween, 0.5% bovine serum albumin fraction IV (W buffer),the plates were blocked for 1h with W buffer at 37 "C. Specified wells received 0-7.5 pmol of ETA or CRM 66 diluted in W buffer in the presence or absence of10 mM NAD+. Alternatively, EF-2-coated plates received 0-8.55 pmol of diphtheria toxin (DT) in the presence or absence of 10 mM NAD'. Following 2h incubation at 25 "C, the wells were washed with W buffer and received either ETA-specific polyclonal rabbit antiseradiluted 1:7500 in W buffer, monoclonal antibody TC-1-alkaline phosphatase conjugate (10) diluted 15000 in W buffer or polyclonal DT-specific rabbit antisera diluted 1:10,000 in W buffer. After 24-h incubation at 4 "C the wells were washed.Wells which receivedpolyclonal antisera were developed with subsequent 2-h incubations with goat anti-rabbit (1:2000) in W buffer and rabbit anti-goat alkaline phosphatase conjugate (1:ZOOO) in W buffer. All wells were then developed with pnitrophenyl phosphate in 44 mM sodium carbonate, 1mM MgClz (pH 9.8) (S buffer) and read at O.D. 410 nm. Triplicate well values were corrected for nonspecific binding and averaged. Inhibition Immobilized Binding Assay-Each well of an Immulon 11 microtiter plate was coated with 0.75 pmol of ETA in coating buffer for 24 h at 37 "C. The plate was washed and blocked as previously described. Designated wells receivedfemtomole quantities of purified EF-2, EF-2 10 mM NAD+, or NAD+ alone in W buffer for 20 h at 4 "C. The amount of EF-2 in each triplicate setof wells wasincreased by a factor of 10. After washing with W buffer, TC-1-alkaline phosphatase conjugate diluted 15000 in W buffer was applied to specified wells for 5 h at 25 "C. The plate was washed with W buffer and developed as previously described. All data reflects correction for nonspecific binding. Indirect Radioimmunoprecipitation Procedure-ETA labeled with lzaIwas precipitated as previously described with the following modifications (10, 16). Labeled ETA (36 ng) (specific activity, 5 X lo5 cpm/mg) was incubated in 50 mM Tris-HC1 (pH 8.0) containing 019 mM NAD+ (Sigma) for 30 min at 25 "C. Following initial incubation, each ETA. NAD+sample was incubated with 6 mg of monoclonal antibody TC-1 for 24 h at 4 "C. Addition of anti-species antisera and protein A-Sepharose beads (Sigma) aided collection of precipitated ETA. The immune complexes were washed and counted in aBeckman Instruments Gamma 4000 counter (Beckman, Arlington Heights, IL). A duplicate control sample for each NAD+ concentration was performed without the addition of the TC-1 antibody. Each data point represents the average of duplicate assays minus the value of the respective control samples. Toxin Binding to Immobilized EF-2 with NAD+Analogs-Individual wells of EF-2-coated (1.30 pmol/well) Immunlon I1 microtiter plates received 3 pg of ETA or DT incubated with NAD+ analogs or NAD' moieties a t concentrations between 20 and 0.001 mM in W buffer for 1.5 h at 25 "C. Toxin-specific polyclonal rabbit antisera was utilized as described above to observe bound toxin. Controls included wells receiving NAD+ analogs without toxin, toxins + NAD+,

+

A

and toxins alone. Duplicate data points were corrected for background and were averaged. Inhibition ELISA-DEPC-modified ETA-Picomole quantities of ETA were incubated in 10 mM NAD+/W buffer (pH 7.8) for 15 min at 25 "C asdescribed previously. Followingthis incubation, 34.5 nmol of DEPC was added to the mixture for 15 min at 25 "C. This entire mixture was then transferred to EF-2-coated Immunlon I1 microtiter plates prepared as previously described. The ELISA was developed as described previously with polyclonal rabbitantitoxinantisera. Controls included both ETA and ETA + NAD+ that had not been modified with DEPC. Data has been corrected for nonspecific binding. In another ELISA, ETA-coated wellsof an Immunlon I1 plate (prepared as described above) were treated with 10-fold increasing DEPC concentrations (3.2 X loo to 3.2 X lo6 pM) in 1.75% ETOH (final concentration) for 15 min at 25 "C. The wells were washed 10 times with W buffer prior to receiving either polyclonal rabbit antitoxin or TC-1 alkaline phosphatase conjugate as previously described. The wells were developed as noted previously. Triplicate wells were corrected for background and averaged. Controls included wells that did not receive DEPC and wells that received 1.75%ETOH. RESULTS AND DISCUSSION

Current research in several laboratories working with P. aeruginosa exotoxin A has focused on domain I11 structure as it relates to enzymatic activity. The precise mechanism of ETA function, transfer of the ADP-ribose moiety of NAD+ onto the target protein elongation factor 2, has yet to be determined. The amino acid G ~ has u been ~ shown ~ ~ to be an active site residue critical for ADPRT activity (7), while we have proposed thatthe residue is required for ETA interaction with EF-2 (8-10). Because these sites are separated by considerable distance as revealed by inspection of the three-dimensional structure of ETA, it is proposed that there are two principle subsites within domain 111, namely one site associated with NAD' binding and another site for EF-2 interaction. The data presented in this report indicates thatbothand a bound molecule of NAD' are required for ETA interaction with EF-2. ETA Interaction with EF-2 Requires NAD+-The interaction between ETA and its protein target EF-2 was studied using a solid-phase immobilized binding assay as described under "Experimental Procedures." Picomole quantities of ETA or CRM 66 toxins suspended in 40 mM borate-buffered saline (pH 7.8) (W buffer) containing NAD+ (10 mM) were incubated inmicrotiter wells coated with purified wheat germ EF-2. Binding of ETA to theimmobilized EF-2 was detected using ETA-specific rabbit antisera. This binding assay revealed that ETA, previously incubated in the presence of NAD', bound to the EF-2 coated wells. ETA alone, or CRM 66 in the presence or absence of NAD', does not bind to the immobilized EF-2 (Fig. 1). Binding to EF-2 increases with increasing ETA concentration in the presence of a constant concentration of NAD+. Saturable binding to EF-2 occurred only in the presence of 10 mMNAD'. When 0.1% sodium dodecyl sulfate was included in the ETA NAD' preincubation mixture, binding to EF-2 was eliminated (data not shown). The CRM 66 toxin, which contains a substitution (His to Tyr) atposition 426, does not bind t o the immobilized EF-2, even in the presence ofNAD'. These results suggest that incubation with NAD+ alters theconformation of ETA allowing it to bind EF-2. Similar binding experiments were carried out using purified DT and DT-specific antisera (kindly provided by Dr. Leon Eidels, University of Texas, Dallas, TX). Maximal binding of DT to the immobilized EF-2 (1.30 pmol) occurred between 7.0 and 8.5 pmol of DT in the presence of 10 mM NAD' (data not shown). Because EF-2 is immobilized, the ADPRT reaction does not go to completion. Under the conditions of the assay the

Analysis of P. aeruginosa Exotoxin 0.5

r

0.4

0.3 0.2 0.1

0 0

a

6 2 4 PICOMOLESTOXIN +I- 1OrnM NAD+

FIG. 1. ELISA demonstrating ETA binding to immobilized EF-2 in the presence of 10 mM NAD'. Approximately 1.30pmol of EF-2 were coated per well of a microtiter plate. Indicated picomole quantities of ETA in 10 mM NAD+ (O), ETA or CRM alone (A), or CRM in 10 mM NAD+ (0)were incubated in the wells for 2 h at 25 "C. Bound toxin was detected with polyclonal ETA-specificrabbit antisera PI0 diluted 1:7500 in W buffer and goat anti-rabbit alkaline phosphatase conjugate. Data points represent an average of triplicate wells each corrected for background. TABLEI Substitution of NAD+ analogs in the immobilized EF-2 binding assay Binding activities for all assays were determined with 3 pmol of each toxin in the presence of 8 mMNAD' analog in W buffer. Data reflects triplicate assay wells corrected for background and averaged. Percent binding was relative to ETA or DT binding to 1.30 pmol immobilized EF-2 in the presence of 10 mM NAD+. The standard deviation for each value is 52%. Binding

NAD Analog

DT

ETA ?h

acid-AD+

0-NAD' (3-NADH @-NADP+ 0-NADPH Thionicotinamide-AD+ 100 Nicotinic 3-Pyridinealdehyde-AD+ Deamino-NAD+ 3-Pyridinealdehyde hypoxanthine dinucleotide ADP-ribose ATP ADP AMP

100 0 100 0 100 100 90 13 2 100 68 93 3

100 0 100 0 100 100 19 6 100 100 100 14

normal ADPRT reaction appears to be arrested at the point where the ETA-NAD+(or DT.NAD+) complex binds to the purified EF-2. CRM 66 toxin, which is capable of binding NAD+,isunabletobindtothe immobilized EF-2inthe presence of NAD+ under these same assay conditions, thus providing a good control for any nonspecific binding. This suggeststhatthe residue is requiredforefficient interaction between toxin and the EF-2 target protein. In an extension of these experiments,several NAD+ analogs were substituted for NAD+ in this ELISA assay in order to determine theireffect upon the abilityof ETA or DT tobind to the immobilized EF-2. Analogswith modification to the C3 carbon of the nicotinamide ring function in this assay with either ETA or DT. However,NAD+ analogs withreduced nicotinamide rings or modifications to the adenine moiety cannotsubstitute forNAD+ (Table I). Theseresultsare consistent withprevious findings which analyzed the binding of NAD+ analogs to D T (17) and demonstrate thespecificity

A

19109

of this ELISA method in studying the interaction between ETA or DT and EF-2. NAD+-induced ETA Conformational Change-Recent investigations in this laboratory have suggested that EF-2 interaction with ETAoccurs following a conformational change in the toxin (10). In order to understand the relationship between NAD+ binding and EF-2 interaction with ETA, we were interested in determining whether NAD+ binding inducesa conformational change in ETA. Our experimental approach utilizedamonoclonal antibody designated TC-1 which has been shown to be specific for an epitope which includes the His4" residue (10). The TC-1 antibodyserves as aprobefora conformationalchange which results in the opening of the major cleft in domain I11 and exposure of the His4*' residue. Purified, nonactivated ETA (36 ng) labeled with '"1 was incubated with increasing quantities of NAD+ for 30 min followed by exposure to 6 pg of monoclonal antibody TC-1. The mixtures were incubated 24 h at 4 "C followed by subsequent incubation with rabbit anti-mouse antibody. The resulting immunecomplexes were isolated using Staphylococcus auremprotein Acovalently coupled toSepharosebeads (Sigma). After thorough washing, the beads containing the 12611-labeled immune complexes were collected and counted in a Beckman Gamma 4000 counter. The results in Fig. 2 indicate that bindingof NAD+ changes the conformation of ETA and results in the exposure of the Hisdz6 residue. Once the cleft structure opens, the monoclonal antibody TC-1 is able tobindatthe site. As NAD+ concentration increases, the numberof exposed His"' sites within the ETA population increases until saturation isreached a t approximately 10 mM NAD'. I t is this conformational change that facilitates ETA interaction withEF-2. An additional indication of a n NAD+-induced conformational change in ETA was obtained using native gel-shift assays. Radioiodinated ETA incubated in thepresence of 10 mM NAD' binds mAb TC-1 without prior activation of the ETA molecule. Analysis of thisimmune complex usinga nondenaturing gel system revealed a radioactive band with a molecular mass of 210 kDa corresponding to a complex composed of one ETA molecule bound by one TC-1 (IgG) molecule, however, these complexes only formed in the presence of NAD+ (data not shown). Presumably the same conformationalchange occurs in the

z

@Y

14

-

12

-

I

,

4

I

,

I

,

I

,

I

8 12 16 20 [NAD+]rnM FIG. 2. Immunoprecipitation of ETA bymonoclonal antibody TC-1 in the presence NAD+. Iodinated ETA (36 ng) was incubated in the indicated NAD+ concentration for 30 min at 25 "C prior to the addition of 6 mg of TC-1 antibody. Following 24-h incubation at 4 "C, the precipitated ETA was bound to protein ASepharose beads (Sigma),washed, and counted. Each data point was 0

corrected for nonspecific binding.

Analysis of P. aeruginosa Exotoxin

19110

A 0.8

z

0.6

0

0.4 /

d 0 0.2

0

0 1 2 3 4 5 18 X 1 Ox FEMTOMOLES EF-2 +/- 1 OmM NAD' FIG. 3. Inhibition of TC-1 binding to immobilized ETA by EF-2in thepresence of 10 mM NAD'. A microtiter plate coated with 0.75 picomol of ETA per well received the indicated femtomole quantities of EF-2 in the presence (0)or absence (0)of 10 mM NAD+ for 4 h at 25 "C. The x-axis indicates the 10-fold increase in EF-2 concentration per assay well. Followingplate washing, TC-1-alkaline phosphatase conjugate diluted 1:5000 in W buffer was incubated for 4 h at 25 "C in each well. Data points represent averages of triplicate wells and are corrected for background.

TABLE I1 Exotoxin A bound to immobilized EF-2 cannot bind monoclonal antibody TC-1 Native ETA (1.50 pmol) in the presence and absence of NAD+ (10 mM) was incubated in triplicate wells of a microtiter plate containing immobilized EF-2 (154 pmol). Bound toxin was detected with ETAspecific rabbit antisera or monoclonal antibody TC-1. Data reflects correction for background. Immobilized Primary

EF-2

ETA

EF-2

ETA

ETA

RaETA TC-1

ETA

NAD'

+ NAD+

Secondary

O.D. 410 nm

RaETA TC-1

0.003 f 0.001 0.002 f 0.001

RaETA TC-1

0.378 f 0.030 0.007 f 0.003 1.136 f 0.100 0.375 f 0.025

RaETA TC-1

1.189 f 0.100 0.435 f 0.030

CRM 66 toxin upon NAD' binding, but this cannotbe tested since mAb TC-1 does not recognize CRM 66 which has tyrosine in place of histidine at the 426 position (10). However, we have previously established that CRM 66 binds NAD+ with the same efficiency as wild type ETA (8). HiS426 Is Required for EF-2 Binding-Previously reported results from this laboratory suggest that the His426residue is involved in theinteraction between ETA and its protein target EF-2 (€410). In order to confirm this hypothesis, we reasoned that theinteraction between ETA and EF-2should block the binding of the monoclonal antibody TC-1tothe epitope. Using the ELISA assay described above, two experimental approaches were utilized to demonstrate that the binding of EF-2toETA occurs atthe site. First, we measured the ability of EF-2 tobind to ETAwhich had been immobilized to the wells of a microtiter plate. The resultsin Fig. 3 demonstrate that increasing amounts of purified EF-2, in the presence of NAD', inhibit the binding of the H i ~ ~ ~ ~ - s p e c i f i c antibody T C - 1 to theimmobilized ETA. In the absence of NAD', EF-2 binding does not occur which confirms the results reported above (see Fig. 1).In a separate set of experiments (Table 11), the binding of NAD+-treated ETA toimmobilized EF-2 is confirmed, and it is demonstrated that EF-2 binding blocks the TC-1 epitope. That is, binding

0

0.4

0.8

1.2

1.6

PICOMOLES ETA

FIG. 4. Inhibition of ETA+ NAD+ binding to immobilized EF-2 by DEPC. Picomole quantities of ETA were incubated in 10 mM NAD' for 15 min at 25 "C. This mixture received 34.5 nmol of DEPC for an additional 15 min at 25 "C. Data reflects indicated picomole quantities of ETA in 10 mM NAD+ (O),ETA in 10 mM NAD+ + 34.5 nmol DEPC (0),and ETA (A) incubated in wells of a microtiter plate containing 1.08 pmol of immobilized EF-2. Bound toxin was detected with ETA-specific rabbit antisera.Triplicate wells were corrected for background.

of ETA to immobilized EF-2 is revealed using polyclonal rabbit antitoxin, butunder the same conditions TC-1 binding is blocked when toxin is bound to theimmobilized EF-2. Since TC-1 binds tothe residue (lo), theseresults indicate that EF-2 binds at or near the His426site. This result also demonstrates that the interaction between toxin and EF-2 is specific under the assay conditions and not due to therandom binding of two proteins. Taken together, these results indicate that theHis426residue is required for EF-2 binding. Chemical Modificationof His426 Inhibits ADPRT FunctionDiethyl pyrocarbonate (DEPC), [C2H,0CO]20, reacts with histidyl residues to yield an N-carbethoxyhistidyl derivative (18). To further illustrate the importance of purified ETA was treated with DEPC and modification was followed by observing an increase in absorbance at 240 nm with no change in absorbance at 280 nm (data not shown). Modification of histidyl residues with 34.5 pmol of DEPC eliminates all ADPRT activity associated with 30 pmol of activated ETA (data not shown). Hydroxylamine was subsequently utilized to reverse the DEPC modification of histidine residues (18). When DEPC-modified ETA (2 pg) is treated with 1 mM hydroxylamine for 15min,restoration of ETA enzymatic activity is observed (data not shown). DEPC-modified ETA was analyzed for the ability to bind to immobilized EF-2 using the ELISA format previously described. Fig. 4 illustrates that ETA preincubated in 10 mM NAD+ then treated with DEPC does not bind to the immobilized EF-2. Furthermore, if the immobilizedtoxin is exposed to increasing concentrations of DEPC, TC-1binding is inhibited yet ETA-specific rabbit antitoxin is still able to bind (data notshown). This is confirmed by Western blot analysis of DEPC-treated toxin probed with mAb TC-1 andpolyclonal rabbit antitoxin (datanot shown). These results indicate that specific modification of the His426residue has occurred. In thisreport we have demonstrated that theHis426residue is required for the interaction between ETA and its protein substrate EF-2. This does not preclude the involvement of other residues in EF-2 binding. Interestingly, the interaction between ETA or DT andEF-2 requires the presence of NAD+. Furthermore, ETA could be quantitatively immunoprecipitated in the presence of increasing NAD' using Hi~~'~-specific mAb TC-1 (Fig. 2). This suggests that NAD' binding causes

Analysis of P. aeruginosa Exotoxin A a conformational change in domain 111 of ETA resulting in the exposure of We therefore propose a more accurate model which defines the precise interactions between the twodomain 111 subsites of ETA, i.e. the sitefor substrate NAD' and the target protein elongation factor 2 site. The first step in the reaction requires NAD' binding at the NAD' binding site. It appears that the criticalpoint of contact involves the adenine moiety of NAD+ since modifications to this ring structure are not tolerated. Secondly, NAD' binding alters the conformation of ETA facilitating the interaction between the ETA.NAD' complex and EF-2' The ETA.NAD' interaction with EF-2 takes place at the EF-2 subsite and requires the residue. Subsequent t o EF-2 binding, the ADP-ribose moiety of NAD' is transferred onto EF-2 effecting release of nicotinamide and return of ETA to an NAD' uncharged state. Acknowle&ements-We thank D ~ L~~~ . Eidels for the polyclonal rabbit anti-DT and Dr. Gary Means for comments and discussion. REFERENCES 1. Liu P V., Sadoff, J. C., Iglewski, B. H., and Sokol, P. A. (1980) J. Infect. dis.'142,538-546

19111

2. Iglewski, B.H., Liu, P. v., andKabat, D.(1977) Infect.Immun. 15, 138144 3. Allured, V. S., Collier, R. J., Carroll, S. F., and McKay, D.B. (1986) Proc. Natl. Acad. Sci. U. S. A . 83,1320-1324 4. Gray, G. L., Smith, D.H., Baldridge, J. Harkins, R.N.,VasiLM. L. Chen, E. Y., and Heyneker, H. L. (1984) Proc. Natl. Acad. Sci. U. S. A. 8 1 , 2645-2649 5. Mozola, M. A., Wilson, R. B., Jordan, E. M., Draper, R. K., and Clowes, R. C. (1984) J. Bacteriol. 1 6 9 , 683-687 6. Hwang, J., Fitzgerald, D. J., Adhya, S., and Pastan, I. (1987) Cell 4 8 , 129136 7. Douglas, C.M., and Collier, R. J. (1987) J . Bacterid. 169,4967-4971 8. Galloway, D. R., Hedstrom, R. C., McGowan, J. L., Kessler, S. P., and Wozniak, D.J. (1989) J . Biol. Chem. 264,14869-14873 9. Wozniak, D.J., Hsu, L. Y., and Galloway, D.R. (1988) Proc. Natl. Acad. Sei. U. S. A. 77,7199-7203 10. McGowan, J. L., Kessler, S. P., Anderson, D. C., and Galloway, D.R. (1990) J. Bid. Chem. 266,4911-4916 11. Wozniak, D. J., Cram, D. C., Daniels, C. J., and Galloway, D. R. (1987) Nucleic Acids Res. 16,2123-2135 12, Leppla, s. (1976) Infect. Immun, 14, 1077-1086 13. Breitenberger, A., C. Moore, N., M. Russell, D. W., and Spremulli, L. L. (1979) A d . Biochem. 99,434-440 14. SPremuh L. L., Walthall, B. J., Lax, s. R., and Ravel, J. M. (1977) Arch. Biochem. Biophys. 178,565-575 15, Enpall, E., and Perlman, p, (1971) Immum&m&ry 8, 871-879 16. Galloway, D. R., Hedstrom, R.C., and Pavlovskis, 0.R. (1984) Infect. Immun. 4 4 , 262-269 17. Kandel, J., Collier, R. J., and Chung, D.W. (1974) J. B i d . Chem. 2 4 9 , 2088-2097 18. Miles, E. W. (1977) Enzymol. Methods 47,431-442

s.,