INFECTION AND IMMUNITY, May 1999, p. 2090–2095 0019-9567/99/$04.0010 Copyright © 1999, American Society for Microbiology. All Rights Reserved.
Vol. 67, No. 5
Epitope Mapping of Monoclonal Antibodies against Bordetella pertussis Adenylate Cyclase Toxin S.-J. LEE,1 M. C. GRAY,2 L. GUO,2† P. SEBO,3 1
AND
E. L. HEWLETT1,2*
2
Departments of Pharmacology and Medicine, University of Virginia, Charlottesville, Virginia, and Institute for Microbiology, Czech Academy of Sciences, Prague, Czech Republic3 Received 28 October 1998/Returned for modification 16 December 1998/Accepted 26 January 1999
Adenylate cyclase (AC) toxin from Bordetella pertussis is a 177-kDa repeats-in-toxin (RTX) family protein that consists of four principal domains; the catalytic domain, the hydrophobic domain, the glycine/aspartaterich repeat domain, and the secretion signal domain. Epitope mapping of 12 monoclonal antibodies (MAbs) directed against AC toxin was conducted to identify regions important for the functional activities of this toxin. A previously developed panel of in-frame deletion mutants of AC toxin was used to localize MAb-specific epitopes on the toxin. The epitopes of these 12 MAbs were located throughout the toxin molecule, recognizing all major domains. Two MAbs recognized a single epitope on the distal portion of the catalytic domain, two reacted with the C-terminal 217 amino acids, one bound to the hydrophobic domain, and one bound to either the hydrophobic domain or the functionally unidentified region adjacent to it. The remaining six MAbs recognized the glycine/aspartate-rich repeat region. To localize these six MAbs, different peptides derived from the repeat region were constructed. Two of the six MAbs appeared to react with the repetitive motif and exhibited cross-reactivity with Escherichia coli hemolysin. The remaining four MAbs appeared to interact with unique epitopes within the repeat region. To evaluate the roles of these epitopes on toxin function, each MAb was screened for its effect on intoxication (cyclic AMP accumulation) and hemolytic activity. The two MAbs recognizing the distal portion of the catalytic domain blocked intoxication of Jurkat cells by AC toxin but had no effect on hemolysis. On the other hand, a MAb directed against a portion of the repeat region caused partial inhibition of AC toxin-induced hemolysis without affecting intoxication. In addition, the MAb recognizing either the hydrophobic domain or the unidentified region adjacent to it inhibited both intoxication and hemolytic activity of AC toxin. These findings extend our understanding of the regions necessary for the complex events required for the biological activities of AC toxin and provide a set of reagents for further study of this novel virulence factor. Bordetella pertussis, a gram-negative bacterium which causes whooping cough, produces several essential virulence factors (37, 38). One of these is adenylate cyclase (AC) toxin, which invades eukaryotic cells, catalyzing the conversion of ATP into cyclic AMP (cAMP) after activation by host calmodulin (4, 13, 22, 24, 25). The consequences of this intoxication include inhibition of host immune cell function (9, 30) and macrophage death through apoptosis (29). AC toxin has also been demonstrated to elicit K1 efflux from sheep erythrocytes in a process that is thought to represent an antecedent event to osmotic lysis of sheep erythrocytes (16). Very little is known, however, about the mechanism by which AC toxin penetrates membranes. AC toxin is synthesized and secreted as a single polypeptide of 177 kDa and consists of four principal domains (13, 28). (i) The N-terminal catalytic domain (amino acids 1 to 400) is activated by calmodulin to convert endogenous ATP into cAMP (4). (ii) The hydrophobic region (amino acids 500 to 700) is hypothesized to include as many as four membranespanning a-helices and may contribute to channel formation in membranes (3, 35). (iii) The repeat region (amino acids 1000 to 1600), which contains 38 copies of the glycine/aspartate-rich motif GGXGXDXLX, is involved in Ca21 binding (27, 33).
Such a tandem arrangement of glycine/aspartate-rich repeats characterizes the RTX (repeats in toxin) family (10, 39). The secondary structure of the repetitive motif is predicted to be similar to that of the alkaline protease of Pseudomonas aeruginosa (2). One of the proposed functions of this repeat region is targeting AC toxin to the cytoplasmic membrane of target cells; however, no specific cell surface receptor has been identified (15, 28). (iv) The C-terminal domain (amino acids 1600 to 1706) contains the secretion signal and seems to play a structural role, since a deletion mutant lacking the secretion signal has no biological activity (14, 28). Over the past 10 years, we have prepared a number of hybridoma cells secreting monoclonal antibodies (MAb) directed against AC toxin, two of which have been described previously. MAb 9D4 and 1H6 were used for Western blotting in the initial purification of AC toxin and identification of the holotoxin molecule (26). In addition, MAb 1H6 was used to characterize the conformational change, which occurs after Ca21 is bound to AC toxin (27). To help identify functionally important regions of AC toxin, we have localized epitopes of a set of MAbs by using a panel of in-frame deletion mutants of AC toxin. In addition, each MAb has been evaluated for its effect on AC toxin-induced hemolysis and intracellular cAMP accumulation, to begin to elucidate the relationships of the structure and function of AC toxin.
* Corresponding author. Mailing address: Box 419, School of Medicine, University of Virginia, Charlottesville, VA 22908. Phone: (804) 924-5945. Fax: (804) 982-3830. E-mail:
[email protected]. † Present address: Department of Protein Chemistry, Immunnex Corp., Seattle, WA 98101.
MATERIALS AND METHODS Bacterial strains, plasmids, and recombinant DNA techniques. Escherichia coli XL1-Blue (Stratagene) was used to overexpress wild-type toxin and the
2090
VOL. 67, 1999
MAbs AGAINST B. PERTUSSIS ADENYLATE CYCLASE TOXIN
deletion mutant proteins. E. coli M15/pREP4 (Qiagen) (Nals Strs Rif s lac ara gal mtl F2 recA1 uvr1) was used for the production of small peptides of the repeat region which were poorly expressed in XL1-Blue. Plasmid pREP4 contains the lacI gene to give the host bacterium 10-fold-higher levels of lac repressor. All the plasmids in this study were transformed into E. coli competent cells by the calcium chloride cold-shock method. Deletion constructs (see Fig. 1) were described previously (28, 34). To enhance the resolution of mapping, small peptides of the repeat region in Fig. 3 (left) were constructed. An insert introducing six histidine residues at the N terminus (pR1) was obtained by subcloning the SacI-SmaI fragments of pT7ACT1 into the SacI-SmaI sites of pQE30 (Qiagen). Plasmid pT7CACT1 was derived from pCACT3 by placing cyaC under control of both lacZp and T7p promoters and was used for preparation of AC toxin in earlier studies (5, 16). To make pR2, an insert was prepared by digesting pT7ACT1 with SphI, creating blunt-ended fragments with S1 nuclease (Boehringer Mannheim), and further digesting the linearized pT7ACT1 with SacI. This insert was subcloned into the SacI-SmaI sites of pQE30. Plasmid pR3 was constructed by taking a DNA fragment encoding amino acids 1156 to 1319 and amplifying it by PCR with oligonucleotide primers 5-CAT GCG AGC TCT GGG GCC-3 and 5-TCC CCC GGG CCC CCC GTA-3. The amplified fragment was double digested with SacI and SmaI and ligated into the same sites of the pQE30 vector. Plasmid pR4 contains the DNA insert encoding amino acids 1320 to 1489. This DNA fragment was amplified by PCR with oligonucleotides 5-CGC CCA TCC GGG GGG CTG GGC GAC-3 and 5-GTC GAC CCG GGC CGC TGA-3 and cloned into the BamHI and SmaI sites of pQE30. To construct pR5, pT7CACT1 was digested with ClaI and BlpI and the ends were filled in with Klenow polymerase (New England BioLabs) and then ligated into the pGEX2T (Pharmacia) SmaI site. The direction of the inserted fragment was confirmed by restriction mapping. Production of MAbs against AC toxin. Two groups of MAbs, (i) 9D4, 7C7, 2F5, 1H6, 2B12, and 4H2 and (ii) 3D1, 5D1, 10A8, 6E1, 10A8, and 2A12, were derived from separate fusions. Hybridoma cells in the first group were prepared from the fusion of myeloma cells and splenocytes of BALB/c mice immunized by standard procedures (8) with AC toxin purified by sucrose gradient centrifugation (26). Screening for hybridoma cells producing MAbs directed against AC toxin was based upon immunoprecipitation of the enzymatic activity of AC toxin from a preparation comparable to that used for immunization. At a later date, when AC toxin had been purified and characterized, the second panel of antibodies was derived from BALB/c mice immunized with a palmitoylated synthetic peptide of the acylation region, surrounding Lys983, and boosted with holotoxin. By using purified holotoxin (1 mg/well) as the antigen, the culture supernatants containing MAbs directed against AC toxin were identified by indirect ELISA. From each hybridoma, ascites was prepared as previously described (26) and all the MAbs were purified from the ascites by affinity chromatography on protein A-Sepharose. The titer of each MAb was defined as the dilution of purified MAb eliciting an optical density of greater than 1.00 in an indirect ELISA. The isotype of each MAb was determined with an Isostrip MAb isotyping kit (Boehringer Mannheim) as specified by the manufacturer. Preparation of AC toxin and mutant proteins. E. coli XL1-Blue, transformed with the plasmid encoding wild-type AC toxin or its derivative, was grown in 23 YT (1.6% Bacto Tryptone, 1% Bacto Yeast, 85 mM NaCl) (Difco) to optical density at 600 nm of 0.2, induced with 1 mM IPTG (Boehringer Mannheim), and grown for another 4 h at 37°C. The bacteria were sonicated and extracted in 8 M urea–50 mM Tris-HCl (pH 8.0)–150 mM NaCl. Soluble proteins were separated from cell debris by centrifugation. Holotoxin and mutant proteins were further purified by affinity chromatography on calmodulin-Sepharose 4B (Pharmacia) as described previously (27, 34). Urea extract of the DAC deletion mutant was used in this study because this deletion mutant cannot be purified by calmodulinSepharose. His-tagged proteins pR1, pR2, pR3, and pR4, were purified by Ni1-agarose affinity chromatography (Qiagen) as specified by the manufacturer. The GST fusion protein pR5 was purified by glutathione-Sepharose affinity chromatography (Pharmacia) as specified by the manufacturer. E. coli hemolysin was prepared as described previously (11). Immunoblotting. Prepared proteins were subjected to sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (10% polyacrylamide) by the method of Laemmli (29a). After SDS-polyacrylamide gel electrophoresis, the proteins were transferred to Immobilon filters (PVDF; Millipore), blocked with 1% bovine serum albumin dissolved in TSB (50 mM Tris [pH 7.5], 200 mM NaCl), and incubated for 2 h at room temperature with MAb at 1:1,000 dilution. The membrane was washed three times with TSB and incubated with peroxidaseconjugated anti-mouse immunoglobulin G for 1 h at room temperature. The membrane was washed three times with TSB and exposed to 0.5 mg of chloronaphthol per ml in TSB–0.01% hydrogen peroxide for 5 min at room temperature to allow visualization. Proteins pR1 through pR5 were dotted onto a nitrocellulose membrane by using a minifold apparatus (Schleicher & Schuell). This membrane was allowed to air dry and tested for MAb reactivity as described above. Intoxication by AC toxin. Intoxication was determined by measuring intracellular cAMP accumulation in Jurkat cells exposed to AC toxin. For determination of the inhibitory effect of antibodies, toxin (2.5 mg/ml) was incubated with each MAb (10 mg/ml) at room temperature for 5 min. Jurkat cells (106/ml) in Hanks balanced salt solution were added to the mixture, which was then incubated at 37°C for 30 min. At the end of the incubation, the cells were centrifuged for 7
2091
TABLE 1. Isotypes and titers of MAbs used in this study MAb
Isotype
ELISA titer
3D1 10A1 5D1 2A12 10A8 6E1 9D4 7C7 2F5 2B12 1H6 4H2
IgG1 IgG1 IgG1 IgG2a IgG2b IgG1 IgG2a IgG2a IgG1 IgG1 IgG1 IgG1
1:1 3 105 1:5 3 104 1:1 3 105 1:5 3 104 1:5 3 104 1:1 3 105 1:2 3 104 1:5 3 104 1:2 3 104 1:5 3 104 1:1 3 105 1:5 3 104
min at 2,200 3 g, and medium was removed by aspiration. The cells were resuspended in 1 ml of 0.1 N HCl for extraction of intracellular cAMP. After centrifugation, the supernatants were carefully transferred to tubes for cAMP measurement by automated radioimmunoassay (7, 16). Determination of hemolytic activity. Hemolytic activity was determined by measuring the amount of hemoglobin released into the culture medium. AC toxin (5 mg/ml) was incubated with each MAb (20 mg/ml) for 5 min at room temperature. Washed sheep erythrocytes at 5 3 108/ml in Hanks balanced salt solution were added, and the mixture was incubated for 5 h at 37°C. Nonlysed sheep erythrocytes were pelleted by centrifugation at 2,200 3 g for 5 min, and hemoglobin released into the supernate was quantitated by measurement of the optical density at 541 nm. Background hemoglobin release was subtracted from experimental values.
RESULTS Screening MAbs against AC toxin. Hybridomas that were found to secrete MAbs against AC toxin were used to induce ascites fluid in BALB/c mice, and the antibodies were purified on a protein A affinity column. The isotypes of all MAbs used in this study were determined by immunodiffusion and are listed in Table 1. In addition, ELISA titers for each purified MAb were measured as described in Materials and Methods and are also listed in Table 1. Epitope mapping of anti-AC toxin MAbs. The epitope recognized by each anti-AC toxin MAb was determined by the pattern of immunoreactivity against a set of in-frame deletion mutants of AC toxin. The toxin derivatives used in this study, which are schematically presented in Fig. 1, were constructed so that some or all of each major domain of the toxin was deleted (28, 34). These deletion mutant proteins were separated on an SDS-polyacrylamide gel, transferred to PVDF, and tested for their reactivity with each MAb by Western blotting. The results of this type of evaluation for all the MAbs tested are summarized in Table 2. As an example, the reactivity of each MAb with DR, the deletion mutant lacking a major part of the repeat region (amino acids 1009 to 1489), is shown in Fig. 2. Of the 12 MAbs, 5, i.e., 6E1, 9D4, 4H2, 2B12, and 2F5, did not react with this deletion mutant, suggesting that the deleted segment contains the epitopes of these MAbs. One possible explanation for the prevalence of MAbs to this region is that some or all are directed against 1 of the 38 glycine/ aspartate-rich repeat motifs. To test this hypothesis, peptides of the repeat region were constructed and are shown schematically in Fig. 3 (left). Interestingly, these five MAbs reveal several different immunoblotting patterns, as shown in Fig. 3 (right). MAbs 9D4 and 2F5 seem to recognize repetitive epitopes in the repeat region because they are reactive to all derivatives of the protein containing amino acids 1156 to 1489, including pR3 and pR4 segments, which do not overlap. Of 12 MAbs tested for cross-reactivity with E. coli hemolysin, only 9D4 and 2F5 showed cross-reactivity (Fig. 4). MAb 9D4 was
2092
LEE ET AL.
INFECT. IMMUN.
FIG. 1. Schematic diagrams of AC toxin-derivative deletion mutants. The lines correspond to the deleted portion of full-length AC toxin. The designation of each deletion mutant is given on the left, and the deleted segment is indicated on the right.
previously reported to react with E. coli hemolysin (HlyA) (11) and Neisseria meningitidis FrpA (36), which have a glycine/ aspartate-rich repeat domain characteristic of RTX-related molecules. On the other hand, MAb 6E1 reacts with all peptides except pR4 (amino acids 1320 to 1489), suggesting that it binds to a unique epitope at the proximal portion of the repeat region, namely, amino acids 1156 to 1319. Interestingly, MAb 7C7, which reacts with DR (containing amino acids 1 to 1008 and 1490 to 1706) and DHR3 (amino acids 1 to 384 and 1590 to 1706), also recognized pR1 (amino acids 1156 to 1489), pR4 (amino acids 1320 to 1489), and pR5 (amino acids 813 to 1627) as shown in Table 2 and Fig. 3 (right). This suggests that the epitope of 7C7 may be a repetitive motif located throughout amino acids 1320 to 1627. The last two MAbs, 2B12 and 4H2, react with pR5 (amino acids 813 to 1627), DH (amino acids 1 to 384, and 829 to 1706) and DCla (amino acids 1 to 827 and 888 to 1706) but not with DHR1 (lacking amino acids 385 to 1006). Therefore, they appear to recognize the region between amino acids 888 and 1006. However, 2B12 and 4H2 do not recognize DR, which includes amino acid 888 to 1006. These results suggest that the epitopes of both 2B12 and 4H2 are localized at amino acids 888 to 1006 but require additional amino acids distal to residue 1006 for their conformation.
TABLE 2. Immunoblot reactivity of MAbs against wild-type (CyaA) and mutant AC toxin proteins Constructa
CyaA DH DC1322 DR DAC DC1308 DHR1 DHR2 DHR3 DHR4 DC217 DC75 DBglII DClaI a b
MAbs 3D1 and 5D1 recognize DAC (amino acids 373 to 1706) and DC1308 (amino acids 1 to 399) but do not react with AC toxin deletion mutant DC1322 (amino acids 1 to 384). This would suggest that MAbs 3D1 and 5D1 recognize an epitope located between amino acids 385 and 399. A synthetic peptide consisting of amino acids 385 to 399 was synthesized; however, it was unable to elicit binding of 3D1 or 5D1 (data not shown). We interpret these data to indicate either that the epitope is not linear and these MAbs cannot bind to the synthetic peptide or that the epitope requires additional amino acids proximal to residue 385 for its conformation. Since we cannot rule out either of these possibilities and since both of these MAbs bind to DAC (amino acids 373 to 1706), we have provisionally assigned the epitope for these MAbs as amino acids 373 to 399. MAbs 10A1 and 2A12 do not react with deletion mutant DH, lacking the entire hydrophobic domain and functionally unidentified regions adjacent to the hydrophobic domain (Table 2). In addition, the deletion mutant DBgl (deletion between amino acids 624 and 780) has lost the epitope of MAb 10A1 but DCla (deletion between amino acids 828 and 887) has not. This result indicates that the epitope of 10A1 is likely to be located between amino acids 624 and 780. In contrast, MAb 2A12, which does not react with DH (amino acids 1 to 384 and 829 to 1706) or DC1308 (amino acids 1 to 398), recognized DBgl, suggesting that its epitope is localized at either end of the 624-to-780 deletion (between amino acids 399 and 623 or 781 and 828).
Reactivityb of MAb: 3D1 10A1 5D1 2A12 10A8 6E1 9D4 7C7 2F5 2B12 1H6 4H2
1 2 2 1 1 1 2 2 2 2 1 1 1 1
1 2 2 1 1 2 2 2 2 2 1 1 2 1
1 2 2 1 1 1 2 2 2 2 1 1 1 1
1 2 2 1 1 2 2 2 2 2 1 1 1 1
1 1 2 1 1 2 1 1 1 2 2 1 1 1
1 1 2 2 1 2 1 2 2 2 1 1 1 1
1 1 2 2 1 2 1 2 2 2 1 1 1 1
1 1 2 1 1 2 1 1 1 2 1 1 1 1
Deletion mutants of AC toxin are illustrated in Fig. 1. 1, immunoreactive; 2, not immunoreactive.
1 1 2 2 1 2 1 2 2 2 1 1 1 1
1 1 2 2 1 2 2 2 2 2 1 1 1 1
1 1 2 1 1 2 1 1 1 2 2 1 1 1
1 1 2 2 1 2 2 2 2 2 1 1 1 1
FIG. 2. Immunoblot analysis of the DR deletion mutant protein by using each MAb. DR protein from AC toxin was purified from a calmodulin affinity column, subjected to electrophoresis on a 10% polyacrylamide gel, transferred to a PVDF membrane, and immunoblotted. Molecular mass standards are indicated by arrows.
MAbs AGAINST B. PERTUSSIS ADENYLATE CYCLASE TOXIN
VOL. 67, 1999
2093
FIG. 3. Immuno-dot-blot analysis with MAbs 9D4, 2F5, 6E1, 7C7, 4H2, and 2B12, using different peptides of repeat region. (Left) Schematic representation of the peptides derived from the repeat region. (Right) The peptides of the repeat region were dotted onto nitrocellulose membrane and probed with each MAb as described in Materials and Methods.
There are two MAbs, 10A8 and 1H6, which recognize the C-terminal secretion signal region. These MAbs react with both DHR3, which consists of both N-terminal amino acids 1 to 384 and C-terminal amino acids 1590 to 1706, and DC75, lacking the C-terminal 75 amino acids. However, they do not bind DHR4, consisting of both the catalytic domain (amino acids 1 to 384) and C-terminal amino acids 1632 to 1706, or DC217, lacking the C-terminal 217 amino acids, suggesting that its epitope is located within amino acids 1590 to 1631. Inhibition of biological activities of AC toxin. To address the predicted functional role of each domain or epitope, we screened each MAb for its ability to affect the enzymatic activity required to convert ATP to cAMP in a cell-free system, the toxin activity required to enter cells, and the hemolytic activity. None of the MAbs had an effect on the enzymatic activity of AC toxin required to convert ATP to cAMP in a cell-free system (data not shown). The results of inhibition of intoxication and hemolytic activities are summarized in Table 3. When 3D1 or 5D1 was allowed to bind to AC toxin prior to its addition to cells, AC toxin-induced cAMP accumulation was inhibited by more than 95% in Jurkat cells. On the other hand, neither of these MAbs had any effect on the hemolytic activity of AC toxin, implying that they did not impair the binding of the toxin to cells as required for hemolysis. These results suggest that the N-terminal amino acids 373 to 399 recognized by 3D1 and 5D1 are important for delivery of the catalytic domain into the cell interior. MAb 2A12, which binds to amino acids 399 to 623 or 781 to 828, partially blocked cAMP accumulation in Jurkat cells but also interfered with the lysis of sheep erythrocytes. This suggests that this MAb inhibits the binding of AC toxin to cells or affects some other event common to these activities. MAb 6E1 strongly delayed the onset of the lysis of sheep erythrocytes but did not impair AC toxin-induced cAMP accumulation (Table 3). These results suggest that MAb 6E1 may block one of the steps required only for hemolysis.
DISCUSSION AC toxin is an essential virulence factor for B. pertussis (37, 38). This toxin is immunogenic and has been shown to elicit antibody responses in patients infected with B. pertussis and recipients of whole-cell pertussis vaccine (1, 6, 12). These observations led to the hypothesis that this immune response to AC toxin could be protective; indeed, this has been confirmed in a series of studies. AC holotoxin or a fragment which contains only adenylate cyclase activity has been shown to function as a protective antigen when used in immunizations (17, 18, 19). Subsequently, however, Betsou et al. (5) demonstrated that posttranslational acylation of AC toxin is required for protective activity. In addition, they observed that native AC toxin from B. pertussis is a more potent protective antigen than is recombinant AC toxin expressed in E. coli. This difference could be because native AC toxin is expressed with a different acylation pattern from recombinant AC toxin (20, 21). Furthermore, Betsou et al. localized the protective epitope of the toxin, demonstrating that the repeat region of AC toxin is necessary for protection and that sera from immunized infants recognized this region (6). This result is in contrast to previous results by Guiso et al. (19) indicating that the catalytic domain is sufficient for protective activity. The insufficient characterization of AC toxin as a protective antigen at the time when acellular vaccines were formulated precluded its being considered as a component in those products (23). The studies described above, however, establish a theoretical basis for inclusion of AC toxin in future acellular vaccines. Prior attempts to map MAbs to AC toxin were limited since a panel of deletion mutants was not available. Therefore, map-
TABLE 3. Inhibitory effects on AC toxin activities AC toxin activity
Inhibitory effect of MAba: 3D1 10A1 5D1 2A12 10A8 6E1 9D4 7C7 2F5 2B12 1H6 4H2
cAMP accu- 11 mulation Hemolysis 2 a
2
11 11
2
2
2
2
2
2
2
2
2
2
2
11 2
2
2
2
2
2
1
11, greater than 60% inhibition; 1, 10 to 30% inhibition; 2, less than 5% inhibition.
FIG. 4. Western immunoblot analysis of the reactivity of each MAb with E. coli hemolysin. Approximately 5 mg of E. coli hemolysin was subjected to SDSpolyacrylamide gel electrophoresis, transferred to a PVDF membrane, and incubated as described in Materials and Methods.
2094
LEE ET AL.
INFECT. IMMUN.
FIG. 5. Proposed epitope map of MAbs against AC toxin. The linear sequence of AC toxin is shown along with its four major domains. Solid bars indicate the proposed epitope of each MAb.
ping of those MAbs had to be done by using cross-reactivity with mutants of the related RTX toxins (38a). For example, MAb 9D4 reacts with E. coli hemolysin (11) and N. meningitidis FrpA (36), suggesting that it binds an epitope in the glycine/ aspartate-rich repeat region which is the common structural feature among RTX toxins. In this report, we present the epitope mapping of 12 MAbs against AC toxin by using immunoblotting of toxin deletion mutants as summarized in Fig. 5. In addition, each MAb was screened for the ability to inhibit biological activities of AC toxin (Table 3). All these results were used to evaluate the structures of epitopes in parallel with functions previously assigned to major domains. For example, MAbs 3D1 and 5D1 block intoxication but have no effects on the hemolytic activity. Amino acids 373 to 399 comprise the region connecting the catalytic domain and putative membrane-spanning domain. Therefore, the reactivity of these MAbs with this region probably prevents the delivery of a portion of AC toxin to the inside of target cells. Another MAb, 6E1, recognizes the repeat region and reduces hemolytic activity but has no effect on AC toxin-induced cAMP accumulation. These data suggest that the epitope of 6E1 in the repeat region of AC toxin is involved in one of the steps hypothesized to be necessary for lysis of sheep erythrocytes, such as the formation of an oligomeric structure, but is less important for delivery of the catalytic domain to the target cell interior. In contrast, MAb 2A12, which recognize amino acids 399 to 623 or 781 to 828, may interfere with the insertion of AC toxin into the cell membrane, resulting in partial reduction of both activities of AC toxin. The remaining MAbs do not interfere with toxin or hemolytic activity, suggesting either that the interaction between MAb and AC toxin does not disrupt the conformational changes of AC toxin required for these activities or that the affinity of these MAbs for the toxin may be too low for them to remain bound to AC toxin when it interacts with the target cell membrane. The mapping data described in this study suggest that the repeat region may be immunodominant, since 6 of 12 MAbs recognize it. There seem to be two classes of MAbs that recognize the repeat region. Two of six MAbs, 9D4 and 2F5, clearly react with the repetitive motif located throughout the repeat region, since they bound to all the peptides of the repeat region constructed in this study. In parallel with these results, MAb 9D4 recognizes other RTX toxins (11, 36) and MAb 2F5, which was also demonstrated to react with E. coli hemolysin in this study, would be expected to react with other RTX toxins. The other MAbs, which recognize the repeat region, appear to have unique epitopes or nonconforming repetitive motifs in small portions of the repeat region.
AC toxin is a bifunctional protein exhibiting intoxication and hemolysis. MAbs characterized in this study aid in understanding the mechanism of each function by functional dissociation of AC toxin. Future studies will include the use of MAbs to visualize the assembly of AC toxin on the membrane by electron microscopy and to study the initial interaction of this toxin with target cells under various experimental conditions. ACKNOWLEDGMENTS We thank W. Sutherland for MAb production. This work was supported by National Institutes of Health grant AI18000 (to E.L.H.), grant P30CA44579 to the University of Virginia Cancer Center, and grant 310/98/0432 from the Grant Academy of the Czech Republic (to P.S.). REFERENCES 1. Arciniega, J. L., E. L. Hewlett, F. D. Johnson, A. Deforest, S. G. F. Wassilak, I. M. Onorato, C. R. Manclark, and D. L. Burns. 1991. Human serologic response to envelope-associated proteins and adenylate cyclase toxin of Bordetella pertussis. J. Infect. Dis. 163:135–142. 2. Baumann, U., S. Wu, K. M. Flaherty, and D. B. McKay. 1993. Threedimensional structure of the alkaline protease of Pseudomonas aeruginosa: a two-domain protein with a calcium binding parallel beta roll motif. EMBO J. 12:3357–3364. 3. Benz, R., E. Maier, D. Ladant, A. Ullmann, and P. Sebo. 1994. Adenylate cyclase toxin (CyaA) of Bordetella pertussis. J. Biol. Chem. 269:27231–27239. 4. Berkowitz, S. A., A. R. Goldhammer, E. L. Hewlett, and J. Wolff. 1980. Activation of prokaryotic adenylate cyclase by calmodulin. Ann. N. Y. Acad. Sci. 1:356–360. 5. Betsou, F., P. Sebo, and N. Guiso. 1993. CyaC-mediated activation is important not only for toxic but also for protective activities of Bordetella pertussis adenylate cyclase-hemolysin. Infect. Immun. 61:3583–3589. 6. Betsou, F., P. Sebo, and N. Guiso. 1995. The C-terminal domain is essential for protective activity of the Bordetella pertussis adenylate cyclase-hemolysin. Infect. Immun. 63:3309–3315. 7. Brooker, G., J. F. Harper, W. L. Terasaki, and R. D. Moylan. 1979. Radioimmunoassay of cyclic AMP and cyclic GMP. Adv. Cyclic Nucleotide Res. 10:1–33. 8. Chapman, M. D., W. M. Sutherland, and T. A. E. Platts-Mills. 1984. Recognition of two dermatophagoides pteronyssnus-specific epitopes on antigen P1 by using monoclonal antibodies: binding to each epitope can be inhibited by serum from dust mite-allergic patients. J. Immunol. 133:2488–2495. 9. Confer, D. L., and J. W. Eaton. 1982. Phagocyte impotence caused by an invasive bacterial adenylate cyclase. Science 217:948–950. 10. Coote, J. G. 1992. Structural and functional relationships among the RTX toxin determinants of gram-negative bacteria. FEMS Microbiol. Rev. 88: 137–162. 11. Ehrmann, I. E., M. C. Gray, V. M. Gordon, L. S. Gray, and E. L. Hewlett. 1991. Hemolytic activity of adenylate cyclase toxin from Bordetella pertussis. FEBS Lett. 278:79–83. 12. Farfel, Z., S. Konen, E. Wiertz, R. Klapmuts, P. A. Addy, and E. Hanski. 1990. Antibodies to Bordetella pertussis adenylate cyclase are produced in man during pertussis infection and after vaccination. J. Med. Microbiol. 32:173–177. 13. Glaser, P., D. Ladant, O. Sezer, F. Pichot, A. Ullmann, and A. Danchin. 1988. The calmodulin-sensitive adenylate cyclase of Bordetella pertussis: clon-
VOL. 67, 1999
MAbs AGAINST B. PERTUSSIS ADENYLATE CYCLASE TOXIN
ing and expression in Escherichia coli. Mol. Microbiol. 2:19–30. 14. Glaser, P., H. Sakamoto, J. Bellalou, A. Ullmann, and A. Danchin. 1988. Secretion of cyclolysin, the calmodulin-sensitive adenylate cyclase-haemolysin bifunctional protein of Bordetella pertussis. EMBO J. 7:3997–4004. 15. Gordon, V. M., W. W. Young, S. M. Lechler, M. C. Gray, S. H. Leppla, and E. L. Hewlett. 1989. Adenylate cyclase toxins from Bacillus anthracis and Bordetella pertussis. Different processes for interaction with and entry into target cells. J. Biol. Chem. 264:14792–14796. 16. Gray, M., G. Szabo, A. S. Otero, L. Gray, and E. Hewlett. 1998. Distinct mechanisms for K1 efflux, intoxication, and hemolysis by Bordetella pertussis AC toxin. J. Biol. Chem. 273:18260–18267. 17. Guiso, N., M. Rocancourt, S. Szatanih, and J. Alonso. 1989. Bordetella adenylate cyclase is a virulence associated factor and an immunoprotective antigen. Microb. Pathog. 7:373–380. 18. Guiso, N., M. Szatanik, and M. Rocancourt. 1990. Bordetella pertussis adenylate cyclase: a protective antigen against lethality and bacterial colonization in murine respiratory and intracerebral models, p. 207–211. In C. Manclark (ed.), Proceedings of the Sixth International Symposium on Pertussis. Department of Health and Human Services, Bethesda, Md. 19. Guiso, N., M. Szatanik, and M. Rocancourt. 1991. Protective activity of Bordetella adenylate cyclase-hemolysin against bacterial colonization. Microb. Pathog. 11:423–431. 20. Hackett, M., L. Guo, J. Shabanowitz, D. F. Hunt, and E. L. Hewlett. 1994. Internal lysine palmitoylation in adenylate cyclase toxin from Bordetella pertussis. Science 266:433–435. 21. Hackett, M., C. B. Walker, L. Guo, M. C. Gray, S. Van Cuyk, A. Ullmann, J. Shabanowitz, D. F. Hunt, E. L. Hewlett, and P. Sebo. 1995. Hemolytic, but not cell-invasive, activity of adenylate cyclase toxin is selectively affected by differential fatty-acylation in Escherichia coli. J. Biol. Chem. 270:20250– 20253. 22. Hanski, E., and Z. Farfel. 1985. Bordetella pertussis invasive adenylate cyclase. Partial resolution and properties of its cellular penetration. J. Biol. Chem. 290:5526–5532. 23. Hewlett, E. L., and J. D. Cherry. 1997. New and improved vaccines against pertussis, p. 387–416. In M. M. Levine, G. C. Woodrow, J. B. Kaper, and G. S. Cobon (ed.), New generation vaccines, 2nd ed. Marcel Dekker, Inc., New York, N.Y. 24. Hewlett, E. L., I. E. Ehrmann, N. J. Maloney, P. R. Fremgen, E. M. Barry, A. A. Weiss, and M. C. Gray. 1992. Adenylate cyclase toxin from Bordetella pertussis: characterization of toxin-catalyzed intoxication and hemolysis, p. 241–248. In B. Witholt, J. E. Alouf, G. J. Boulnois, P. Cossart, B. W. Dijkstra, P. Falmagne, F. J. Fehrenbach, J. Freer, H. Niemann, R. Rappuoli, et al. (ed.), Bacterial protein toxins. Gustav Fischer Verlag, Stuttgart, Germany. 25. Hewlett, E. L., and V. M. Gordon. 1988. Adenylate cyclase toxin of Bordetella pertussis, p. 193–209. In A. C. Wardlaw and R. Parton (ed.), Pathogenesis
Editor: R. N. Moore
2095
and immunity in pertussis. John Wiley & Sons, Inc., New York, N.Y. 26. Hewlett, E. L., V. M. Gordon, J. D. McCaffery, W. M. Sutherland, and M. C. Gray. 1989. Adenylate cyclase toxin from Bordetella pertussis. Identification and purification of the holotoxin molecule. J. Biol. Chem. 264:19379–19384. 27. Hewlett, E. L., L. Gray, M. Allietta, I. Ehrmann, V. M. Gordon, and M. C. Gray. 1991. Adenylate cyclase toxin from Bordetella pertussis. Conformational change associated with toxin activity. J. Biol. Chem. 266:17503–17508. 28. Iwaki, M., A. Ullmann, and P. Sebo. 1995. Identification by in vitro complementation of regions required for cell-invasive activity of Bordetella pertussis adenylate cyclase toxin. Mol. Microbiol. 17:1015–1024. 29. Khelef, N., and N. Guiso. 1995. Induction of macrophage apoptosis by Bordetella pertussis adenylate cyclase-hemolysin. FEMS Microbiol Lett. 134: 27–32. 29a.Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685. 30. Pearson, R. D., P. Symes, M. Conboy, A. A. Weiss, and E. L. Hewlett. 1987. Inhibition of monocyte oxidative responses by Bordetella pertussis adenylate cyclase toxin. J. Immunol. 139:2749–2754. 31. Rogel, A., Z. Farfel, S. Goldschmidt, J. Shiloach, and E. Hanski. 1988. Bordetella pertussis adenylate cyclase. Identification of multiple forms of the enzyme by antibodies. J. Biol. Chem. 263:13310–13316. 32. Rogel, A., R. Meller, and E. Hanski. 1991. Adenylate cyclase toxin from Bordetella pertussis. The relationship between induction of cAMP and hemolysis. J. Biol. Chem. 266:3154–3161. 33. Rose, T., P. Sebo, J. Bellalou, and D. Ladant. 1995. Interaction of calcium with Bordetella pertussis adenylate cyclase toxin. J. Biol. Chem. 270:26370– 26376. 34. Sebo, P., and D. Ladant. 1993. Repeat sequences in the Bordetella pertussis adenylate cyclase toxin can be recognized as alternative carboxy-proximal secretion signals by the Escherichia coli alpha-haemolysin translocator. Mol. Microbiol. 9:999–1009. 35. Szabo, G., M. C. Gray, and E. L. Hewlett. 1994. Adenylate cyclase toxin from Bordetella pertussis produces ion conductance across artificial lipid bilayers in a calcium- and polarity-dependent manner. J. Biol. Chem. 269:22496–22499. 36. Thompson, S. A., L. L. Wang, A. West, and P. F. Sparling. 1993. Neisseria meningitidis produces iron-regulated proteins related to the RTX family of exoproteins. J. Bacteriol. 175:811–818. 37. Weiss, A. A., and E. L. Hewlett. 1986. Virulence factors of Bordetella pertussis. Annu. Rev. Microbiol. 40:661–686. 38. Weiss, A. A., E. L. Hewlett, G. A. Myers, and S. Falkow. 1983. Tn5-induced mutations affecting virulence factors of Bordetella pertussis. Infect. Immun. 42:33–41. 38a.Welch, R. Personal communication. 39. Welch, R. 1991. Pore-forming cytolysins of gram-negative bacteria. Mol. Microbiol. 5:521–528.
