Adenylate Cyclase Toxin from Bordetella pertussis - The Journal of ...

THEJOURNAL

OF

Vol. 264, No. 32. Issue of November 15, pp. 19379-19384,1989 Printed in U.S.A.

BIOLOGICAL CHEMISTRY

0 1989 by The American Society for Biochemistry and Molecular Biology, Inc.

Adenylate Cyclase Toxinfrom Bordetella pertussis IDENTIFICATION AND PURIFICATION OF THE HOLOTOXIN MOLECULE* (Received for publication, May 8, 1989)

Erik L. HewlettSQ, Valery M. Gordon$$, John D. McCaffery$ll, William M. Sutherlandll, and Mary C. Gray$ From the Departments of *Medicine, §Pharmacology,and IlAnatomy and Cell Biology, University of Virginia Schoolof Medicine, Charlottesuille; Virginia22908

Bordetella pertussis adenylate cyclase (AC) toxin is a calmodulin-activated adenylate cyclase enzyme which has the capacity to enter eukaryotic targetcells and catalyze the conversion of endogenous ATP into cyclic AMP. In this work, the AC holotoxin molecule is identified and isolated. It is a single polypeptide of apparent 216 kDa as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. MonoAC activity clonal antibodies which immunoprecipitate from extracts of wild type B. pertussis(BP338) react with this 216-kDa band on Western blots, and it is absent from a transposon Tn5 mutant (BP348) specifically lacking AC toxin. Isolation of the 216-kDa protein to >85% purity by hydrophobic chromatography, preparative sucrose gradient centrifugation,and affinity chromatographyusing either calmodulin-sepharose or monoclonal antibody coupled to Sepharose4B yields stepwise increases in AC toxin potency, to a maximum of88.3 wmol of cAMP/mg of target cell protein/mg of toxin. Electroelution of the 216-kDa band following sodium dodecyl sulfate-polyacrylamide gel electrophoresis yields a preparation with both AC enzyme andtoxin activities. These data indicate that this protein represents theAC holotoxin molecule.

least three other species of calmodulin-activatable AC (50 kDa, 45 kDa, and 43 kDa). Theseenzymes appear to be related as they can be cleaved proteolytically to yield the 43-kDa protein (5), but they have no AC toxin activity. Weiss et al. (6) andHanskiand Farfel (7), on theother hand, have demonstrated that in crude bacterial extracts the AC toxin activity is associated with proteins of 190-200 kDa. Rogel et al. (8) showed that polyclonal antiserum raised against the smaller (47-kDa) form of adenylate cyclase cross-reacts with a species of 200 kDa. Glaser et al. (9) have cloned the putative AC gene from B. pertussis and expressed it in Escherichia coli. The gene encodes for a proteinof 1706 amino acids (calculated 177,316 Da), which is considerably larger than the AC enzymes previously identified, but corresponds to thesize of the protein with AC toxin activity described by Rogel et al. (8) and in thiswork. In the present study, extracts of B. pertussis are used for identification and isolation of the B. pertussis AC toxin, a protein with apparent mass of 216 kDa. This single protein, which is immunologically related tothe 43-kDa catalytic fragment, possesses both enzyme and toxin activities and is the AC holotoxin. EXPERIMENTALPROCEDURES

Bordetella pertussis, the bacterium that causes whooping cough, produces a novel adenylate cyclase (AC)’ toxin which is able toenter host cells, whereupon it is activated by calmodulin to produce cAMP from endogenous ATP (1, 2). This AC toxin is an importantvirulence factor for B. pertussis, chiefly by virtue of the inhibitory effect of accumulated cAMP on host phagocyte function (1, 3). Adenylate cyclase of B. pertussis was first isolated from supernatant culture medium by Hewlett and Wolff and shown to have an apparent mass of 70 kDa but no AC toxin activity (the ability to increase cAMP levels in mammalian target cells) (4).’More recently, Ladant et al. (5) have isolated at

* This work wascarried out with the support of National Institutes of Health Grant RO-1 AI18000 (to E. L. H.) and Grant DK28942 from the University of Virginia Diabetes Center. 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. B Present address: Vanderbilt University School of Medicine, Nashville, TN. ’ The abbreviations used are: AC, adenylate cyclase; PBS, phosphate-buffered saline; PVDF, polyvinylidene difluoride; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; ELISA, enzyme-linked immunosorbent assay; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis. E. L. Hewlett and J. Wolff, unpublished data.

Materials-Phosphate-buffered saline (PBS), RPMI 1640, penicillin-streptomycin, fetal bovine serum, Hanks’ balanced salt solution, Bordet-Gengou, and Stainer-Scholte medium werefromGIBCO (Grand Island, NY). Phenyl-Sepharose CL-4B and calmodulin-Sepharose 4Bwere obtained from Pharmacia LKB Biotechnology Inc. Polyvinylidene difluoride (PVDF) membranes were obtained from Millipore Corp. Peroxidase-conjugated goat anti-mouse IgG heavy and light chains were obtained from Jackson ImmunoResearch (West Grove, PA). All other reagents were obtained from Sigma, unless otherwise indicated. Culture of Organisms and Purification of Adenylate Cyclase ToxinB. pertussis (strain BP338) was maintained at -20 “C in skim milk. The organisms were plated on Bordet-Gengou agar containing 20% sheep blood and then grown in modified Stainer-Scholte liquid medium as previously described (10) for 16-20 h to an ODsmof 0.6-0.8. Bacteria from 7.2 liters were concentrated by centrifugation (14,000 X g X 40 min) and resuspended in culture medium. Crystalline urea was added to a final concentration of 4 M, and the suspension was stirred for 16 h a t 4 “C.The bacteria were then removed by centrifugation, and the supernatant was diluted 1:l with Buffer A (10 mM Tricine, pH 8, 0.5 mM EGTA, 0.5 mM EDTA) to reduce the urea concentration to 2 M. This material was then passed over a 35-ml phenyl-Sepharose CL-4B column. The column was washed with Buffer A containing 2 M urea and eluted with Buffer A containing 8 M urea. This material was then concentrated and simultaneously dialyzed against Buffer A using a collodion membrane (75-kDa exclusion, Schleicher & Schuell). The concentrate was loaded onto a gradient of 5-20% sucrose prepared in Buffer A and centrifuged a t 200,000 X g for 16-18 h. The gradient was fractionated into 0.5-ml aliquots and assayed for adenylate cyclase enzymatic activity, and peak fractions

19379

19380

Adenylate Cyclase Toxin Purification

were pooled. The procedure to this point was employed in previous studies ( 3 ) . Sucrose gradient pool with added Tween 20 (0.1%) and CaCI2 (5 mM) was passed over either a calmodulin-Sepharose 4R or an immunoaffinity column of Sepharose 4B coupled with the monoclonal antibody 9D4. The column was washed with 3 column volumes of 10 mM Tricine, pH 8, 1 mM CaCI,, 0.5 M NaCI, 0.1% Tween 20, and eluted with Buffer A containing 8 M urea, 0.1% Tween 20. Monoclonal Antihodv Preparation and Characterization-BALB/c mice were immunized with sucrose gradient-purified adenylate cyclase emulsified in complete Freund's adjuvant, by injecting 100 pg of protein divided between subcutaneous and intraperitoneal sites. The mice were boosted4 weeks later with similarinjections of immunogen in incomplete Freund's adjuvant. Sixweeks following the initial injection, sera from the mice had high titers of specific antibody as demonstrated by an indirect ELISA (11) using sucrose the gradient pool coated at 1 pg per well to microtiter plates. Eight weeks after the secondary injection, the mousewith the highest circulating antibody titerwas given an intrasplenic injection (12,13) of 6 p g and an intraperitonealinjection of 20 pg of the sucrose gradient pool in PRS. Four days later, the mouse was killed, and its spleencells were fused with the SpZ/O-Ag 14myeloma (14) by established procedures (15). Culturescontaininghybridomas (16) were first screened for specific antibodyproduction by indirect ELISA. Thoseculturesdemonstrating a positiveresponse in the ELISA were subsequently screened for their ability toimmunoprecipitate adenylatecyclase enzymatic activity with the use of the following procedure. Sucrose gradient-purified adenylate cyclase was incubated for 1618 h at 4 "C with relevant or irrelevant monoclonal antibody culture supernatants in a final volume of 100 pl. T o each tube, 100 pl of PBS containing 5 mg/ml insoluble protein A (Sigma) was addedand incubated for 2h at 2.5 "C. After centrifugation in an Eppendorf centrifuge for 5 min, the pellet was washed twicein PBS,resuspended in 50 pl of 60 mM Tricine, pH 8, with 0.15 BSA, and assayed for adenylate cyclase enzyme activity. The control, towhich no antibody was added, had activity of 1877 pmol/l0 min/l0 pl. The protein A pellet-associated AC enzymaticactivities with control value subtracted were as follows: 1) irrelevant monoclonal, 91 pmol/l0 min/lO pl; 2) monoclonal 9D4,28,362 pmol/l0 min/l0 pl; and 3)monoclonal 1H6, 29,764 pmol/l0 min/l0 pl. Hybridoma cultures that demonstrated a positive antibody response in both the ELISA and the immunoprecipitation assay were cloned twice by limiting dilution and cryopreserved. Primary and secondary clones were also selected on the basis of their ability to give both a positive response against the sucrose gradient pool in the ELISA assay and to immunoprecipitate adenylatecyclase enzymatic activity with protein A. Ascitcs Purification and Affinity Column Preparation-Large amountsof antibodyfrom cloned hybridomas 9D4 (1gGz8),1H6 (IgGI), and MHS-5 (IgCI) were produced as ascites by injecting 2.0 X lo6 cells per mouse intraperitoneally into BALB/c mice that. had been previously primed hy injecting 0.5 ml of Pristane (2,6,10,14-tetramethvlpentadecane; Sigma). TheIgG fraction was purified and coupled to activated CH-Sepharose4R (I'harmacia) for use in immunoaffinity purification of the AC toxin. Briefly, an equal volume of saturated ammonium sulfate was added to ascites, and the mixturewas stirred a t 4 "C for 3 h. Following centrifugation (7000 X g) for 20 min, the precipitate was dissolved in phosphate-buffered saline,pH 7.2,0.02%sodium azide (PRS/NaNa). This solutionwas then dialyzed sequentially against PRS/NaNR for 48 h and 0.01 M Tris, pH 8, for 16 h a t 4 "C. The resulting material was passed over a 50-ml DEAE-cellulose column, and theIgG fraction eluted with a linear gradient of 0-0.2 M KCI. Purified IgC (75 mg) was then coupled to 1 g of activated CHSepharose 4H (I'harmacia) according to manufacturer's instructions. .SDS-I'A(;I.:. Wcsten Blot, and Electroelution-Electrophoresis of protein samples was performed according to Laemmli (17) on 1.5mm-thick SIX-polyacrylamidegels. Proteins were separated on SDSpolyacrylamide gels, transferred to nitrocellulose sheets, and incuhated with monoclonal antihodies as described by Towhin et al. (18). Electroelution of the 216-kDa hand was performed by the technique of Hunkapiller et al. (19), using an ELU-40 electroelution apparatus (C.R.S. Scientific Co., San Diego, CA). Briefly, SDS-polyacrylamide gels were stained with 0.5% Coomassie Blue for 15-20 min a t 25 "C, destained in 3 0 5 methanol, 10%acetic acid for 2-3 h, and thedesired hand was excised. The gel piece was washed in 10 mM Tris, pH 8, for 2 h, diced into I-mm cuhes, and loaded into the elution cell. The gel cuhes were covered with soaking buffer (2% SDS, 0.2 M Tris acet,ate,

pH 7.8), overlaid with elution buffer (0.1% SDS, 0.05 M Tris acetate, pH 7.8), and soaked for 3-5 h. Electroelution was carried out at 4 OC for 20-24 h a t 100 V. The protein was then precipitated using 40% trichloroacetic acid and redissolved in 50 mM Tris acetate with 0.1% SDS at pH 7.8. Apparent molecular mass of the holotoxin molecule was determined using 5% SDS-PAGE with myosin (200 kDa), microtubule-associated protein-2 (MAP-2) (290 kDa), and thyroglobulin

A

116 92 -

200

66

-

45

-

31

-

21

A

B

C

B

C

B 200

-

116 -

92 66

-

45

-

31

-

21

A

FIG.1. Monoclonal antibodies that immunoprecipitate AC enzyme activity recognize a216-kDaprotein by Western blot. Panel a, sucrose gradient pool (210 pg/ml) was separated by 12.5% SDS-PAGE, transferredont.0 a PVDF membrane, and incubated with monoclonal antihodies 9D4 (lane A), 1H6 (lane R), and MHS-5 (lane C ) . Panel b, urea extracts (25 pg each) of BP338, wild type strain (lower A ) , and two T n 5 mutants, BP348 (lane R ) and RP347 (lane C ) , both lacking AC, were fractionated by 12.5% SDS-PAGE, transferred to PVDF, and incubatedwith monoclonal antibody 9D4.

AdenylatePurification Cyclase Toxin

A 3

8 2

-

P

A

80-

x c .-

.-

60-

19381

-

MHS-5 Column

0 40-

1oo.000

Enzyme Toxin

-

0

8 V

W1

W2

IOOx .= >

80 -

4 -

60-

.c

9D4 Column

17

0.001

W3 8M1 8M2 8M3

r-7

Enzyme Toxin

0.01

0.1

1

10

100

B. pertussis Protein (pg/rnl)

B

1oo,ooo

3

10,000

m

c

-

40-

1,000

$

20-

100

0

0-=--

V

c

8o

1

W1

W2 W3

CaM Column

IC,

8M1 8M2 8M3

n

10 0.001 ~

0.01

0.1

1

10

100

B. pertussis Protein (pg/rnl)

FIG. 3. AC toxin potency is increased by calmodulin or monoclonal affinity columns following partial purification. Two separate preparations ofAC toxin were purified as described under “Experimental Procedures.” Presented in both panels are: urea extract (O), phenyl-Sepharose (A),and sucrose gradient pool (a). Panel a shows activity eluted from 9D4 immunoaffinity column (0). Panel b shows activity eluted from the calmodulin affinity column (A). Intracellular cAMP was measured as described under “Experimental Procedures” and displayed on a 4-cycle log scale.

V

W1

W2

W3 8M1 8M2 8M3

FIG. 2. Monoclonal antibody 9D4 and calmodulin affinity columns both purify AC enzyme and toxin activities. Sucrose gradient pool with added Tween 20 (0.1%) and CaCI2 (5 mM) was passed over an irrelevant antibody (MHS-5) column (panel a), 9D4 immunoaffinity column (panel b), or calmodulin affinity column @anel C ) , and the void ( V ) was collected. The column was washed with 3 column volumes of 10 mM Tricine, pH 8, containing 1 mM CaCI2, 0.5 M NaC1, 0.1% Tween 20 ( W l , W2, W3), and eluted in three fractions with buffer A containing 8 M urea and 0.1% Tween 20 (8M1, 8M2, 8M3). Aliquots (10 pl) of each fraction were assayed for AC enzyme (solid bars) and AC toxin (open bars) activities as described under “Experimental Procedures.” (330 kDa) as standards.Laser densitometry of stained SDS-polyacrylamide gels, using a BioImage Visage 2000, was used to determine the level of purity of gel-fractionated samples. Adenylate Cyclase Enzymatic Actiuity-Enzymatic activity was measured by conversion of [w3’P]ATP to [32P]cAMPin a cell-free assay as described previously (20). Briefly, the reaction was carried out at30 “C andcontinued for 10 min in a final volume of 60 rl. Each assay contained 60 mM Tricine, pH 8.0, 10 mM MgCI,, 1 mM ATP (with 2 X 10’ to 5 X lo5 cpm of [w3’P]ATP), and 1 PM calmodulin. The reaction was terminated by the addition of 100 pl of a solution containing 1%sodium dodecyl sulfate (SDS), 20 mM ATP, and 6.25 mM cAMP (including 15,000-20,000 cpm of [3H]cAMP per tube for calculation of recovery). The cAMP formed was isolated by the double column method of Salomon et al. (21). Adenylate Cyclase Toxin Activity-Adenylate cyclase toxin activity was determined by quantitation of intracellular cAMP accumulation in 5774 cells treated with preparations containing the toxin. 5774 cells were maintained in culture flasks in RPMI 1640 supplemented with 10% fetal bovine serum and 50 units/ml penicillin, 50 pg/ml streptomycin in 5% CO,. For intracellular cAMP determinations, cells were plated in 24-well tissue culture plates and grown to 80-

100% confluency. Immediately before assay, medium was removed by aspiration and replaced with Hanks’ balanced salt solution at 37 ‘C. AC toxin was then added to cells and incubated for 1 h a t 37 “C. Intracellular cAMP was extracted with 0.1 N HC1 at 25 “C for 30 min and measured by automated radioimmunoassay (22). Protein was extracted with 0.2 N NaOH and quantitated using the method of Lowry et al. (23). RESULTS

Extraction of AC toxin from intact B. pertussis organisms with 4 M urea is the first step in purification (1, 3). Further fractionation by phenyl-Sepharose chromatography and preparative sucrose gradient centrifugation yields a partial purification of AC toxin (3). In order to purify the holotoxin molecule, however, additional steps arerequired. Two monoclonal antibodies, 9D4 and 1H6, selected for their ability to immunoprecipitate B. pertussis AC enzyme activity, react with a band of 216 kDa on Western blot (Fig. la). An irrelevant monoclonal antibody serves as a negative control and isunreactive in thisblot. Monoclonal 9D4, and toa lesser extent 1H6, also react with a family of smaller proteins which are proteolytic fragments. As demonstrated by Western blot, the 216-kDa protein and its degradation products are absent from extracts of two mutants, BP348 and BP347, which produce no AC toxin (Fig. lb) (24). BP348 is a mutant with no AC enzyme or toxin activities due to a single transposon Tn5 insertion near the amino-terminal end of the AC structural gene (24-26). BP347 is an avirulent mutant which produces no virulence-associated proteins (including AC toxin, pertussis toxin, and filamentous hemagglutinin) by a virtue of a single Tn5 insertion into the uir gene (24, 26). Further purification of the AC toxin is achieved with either

AdenylatePurification Cyclase Toxin

19382

TABLE I Purification of R. pertussis AC toxin cyclase

Adenylate Procedure

Urea extraction Phenyl-Sepharose chromatography Sucrose gradient 10.2 Calmodulin-Sepharose chromatography4.0 21.0 Electroelution of 216-kDa band

Protein Toxin Total Concentration

Enzyme

wlml

m.c

1.32 0.60 0.18 0.11 0.031

158 60

Specific activitv"

Totalh

Specific activitv'

TOtald

21.0 98.8 248 392 589

3318 5928 2530 1568

0.65 15.5 52.8 88.3

103 930 539 3.53

" Enzyme specific activity, pmol of cAMP/min/mg. Total enzyme activity, pmol of cAMP/min. ' Toxin specific activity, pmol of cAMP accumulated in target cells/mg of target cell protein/mg of toxin. dTotal toxin activity,pmol of cAMP accumulated in target cells/mg of target cell protein.

monoclonal antibody (9D4) immunoaffinity chromatography, or calmodulinaffinity chromatography. T o determine that the AC toxin was not bindingnonspecifically to the Sepharose matrix, a column prepared by coupling the irrelevant antibody, MHS-5, to Sepharose 4B was also tested. As shown in Fig. 2a, neither AC enzyme nor toxin activity from a sucrose gradient-purified preparation binds to the MHS-5-Sepharose column. In contrast, both the 9D4 monoclonalaffinity column and the calmodulin affinity column retain enzyme and toxin activity which is then eluted with>80% recovery by 8 M urea (Fig. 2, b and c ) . When separate urea extractsof B.pertussis organisms are processed through the entire purification procedure, with either immunoaffinity (Fig. 3a) or calmodulin 31affinity (Fig. 3b) chromatography as the final step, there are 1 " progressive increases in AC toxin activity. The relationshipof the 216-kDa protein to AC enzyme and 21 toxin activities is illustrated in Table I. Purification using calmodulin-Sepharose as the final step yields an 18.7-fold increase in AC enzyme activity anda 135-fold increase in AC A B C D E toxin activity. The differential enhancementof toxin activity over enzyme activity is the result of elimination of smaller FIG.4. The purification procedure isolates the 216-kDa molecular weight forms which possess enzyme, but not toxin protein. Urea extract, 13.2pg loaded (lane A ) , phenyl-Sepharose, 6.0 pg (lane R ) , sucrose gradient pool, 1.9 pg (lane C ) , and calmodulin activity. Analysis of the SDS-polyacrylamide gels shown in Fig. 4 by laser densitometry scanningreveals that theincrease affinity-purified material, 1.1 pg (lane D),were separated by 12.5% SDS-PAGE and stained with Coomassie Blue. Electroelution of the in toxin specific activity with each step is associated with 216-kDa band from the calmodulin affinity-purified reaction was increases only in the 216-kDa protein band. Furthermore, in performed as described under "Experimental Procedures," and the the calmodulin affinity-purified sample, the 216-kDa band is eluted fraction was separated by 12.5% SDS-PAGE, 0.5 pg loaded >85% pure. In order to determine, however, that themeasured (lane E). enzyme and toxin activities are due to the 216-kDa protein andnot residual smaller forms, the 216-kDa protein was Identification of 216-kDa holotoxin raises the question of electroeluted and shown by repeat electrophoresis to bea its relationship to the smaller species, such as the 43-kDa AC single band of unchanged mass (Fig. 4). Isolated as a single enzyme of Ladant et al. (5). Comparison of the enzymatic protein in this manner, 216-kDa the formhas bothAC enzyme activitiesis only appropriateon amolarbasis, sincethe and AC toxin activities (Table I). TheAC enzyme activity of holotoxin has a molecular mass of 5 times thatof the 43-kDa 589 pmol of cAMP/min/mg of protein or 128pmol of CAMP/ catalytic form. Ladant et al. (5) obtained a specific activity of min/nmol of enzyme represents a 1.5-fold increase over the 1600 pmol of cAMP/min/mg of protein or69.0 pmol of CAMP/ calmodulin affinity-purified material. The AC toxin activity min/pmol of enzyme, whereas the activity of the 216-kDa is, however, muchmorelabile than enzyme activity, with form is slightly greater (128 pmol/min/pmol of enzyme). >90% of toxin activity lost in 24 h of storage a t ambient The structural relatednessof the two proteins is addressed temperature during which time enzyme activity is reduced with the use of a polyclonal antiserum which was prepared 4 % (data not shown). It is not unexpected therefore that againstthe 43-kDa form (generouslyprovidedby Dr. A. the AC toxin activity of the isolated 216-kDa holotoxin (21 Ullmann, Institut Pasteur, Paris). When used in a Western pmol of cAMP/mg of target cell protein/pg of toxin) isreduced blot of sucrose gradient-purified material (Fig. 5), the rabbit following SDS-PAGE, electroelution, and trichloroacetic acid antiserum reacts with the216-kDa band identified by monoprecipitation. These data demonstrate unequivocally, how- clonal antibody 9D4. These data indicating theimmunologic ever, that the 216-kDa protein isolated from urea extraction relatedness of the 216-kDa and 43-kDa forms suggest, as of R. pertussis is the holotoxin and no other bacterial com- postulated by Rogel et al. (8) that the smaller catalyticform is derived from the 216-kDa holotoxin. ponents are necessary for toxin activity.

-

*

"

AdenylatePurification Cyclase Toxin

19383

protein toxins described by Gill (33), a portion of the 216(B) domainto kDaholotoxin must serve asthebinding promote the entry of the AC catalytic (A) domain into the 200 target cell, thereby conferring toxin activity. In light of the data from Glaser et al. (9), it is reasonable to consider that some part of the carboxyl-terminal 1300 amino acids consti116 tutes a binding domain. Analysis of the derived amino acid 92 sequence by Glaser et a1. (9) reveals a significant homology to the hlyA gene product(a-hemolysin) of E. coli (9). This 66 hemolysin is postulated to cause erythrocyte lysis by insertion into the target cell membrane with generation of transmembrane pores (34). The 216-kDa AC toxin isolated in this study has been shown to cause lysis of multilamellar dimyristoyl45 phosphatidylcholine/cholesterol/ganglioside liposomes (35) and hemolysis of sheep erythrocytes (36) providing further functional evidence that it t,he is product of the gene identified by Glaser et al. (9). The ability of this molecule to penetrate 31 an intact membranesuggests that the domainresponsible for this interaction may be functioning as the binding subunit which facilitates entry of the catalyticAC. The data presented here arein contrast to theconclusions reached by Masure et al. (37) who noted the presence of a 21 215-kDa protein with AC enzyme activity in extracts of B. pertussis, but stated that it had no AC toxin activity (37). A B They postulated that proteolytic release of the 43-kDa form FIG. 5. Antibodies directed against the 43-kDa catalytic resulted in generation ofAC toxin activity. The difference domain recognize the 216-kDa protein. Sucrose gradient-puri- between those observations and the data presented here are Iied material (210pg/ml) was separated by a 12.5% SDS-PAGE, likely due to the presence of components in the crude prepatransferredtoPVDFmembrane,andincubated withmonoclonal rations which inhibit toxin activity. antihody 9D4 ( h e A ) or polyclonal rabbit antiserum directed against Although the 216-kDa protein alone is sufficient for toxin the 43-kDa catalytic form of AC (lane R ) . activity, little is known of its processing during the course of intoxication. The toxinis cell-associated and does not appear DISCUSSION in the culture medium during growth (38). Hewlett et al. (38) Adenylate cyclase enzyme from B. pertussis has been stud- have demonstrated thatAC toxin is delivered by direct contact ied by a number of investigators and shown to exist in a between B. pertussis and target cells. It is hypothesized that variety of sizes (2, 5 , 6-9, 26-30). Previous studies from this the 43-kDa catalytic domain is the portion of the molecule laboratory and others have indicated AC thatactivity in crude that reaches the targetcell interior with the assistanceof the extracts can be separated into at least two fractions, onewith remainder of the holotoxin molecule. If this is thecase, there AC enzyme activity only (