Role of intrachain disulfides in the activities of the CdtA and CdtC subunits of the cytolethal distending toxin of Actinobacillus actinomycetemcomitans

Abstract

The cytolethal distending toxin (Cdt) of Actinobacillus actinomycetemcomitans is an atypical A-B-type toxin consisting of a heterotrimer composed of the cdtA, cdtB, and cdtC gene products. The CdtA and CdtC subunits form two heterogeneous ricin-like lectin domains which bind the holotoxin to the target cell. Point mutations were used to study CdtC structure and function. One (mutC216(F97C)) of eight single-amino-acid replacement mutants identified yielded a gene product that failed to form biologically active holotoxin. Based on the possibility that the F97C mutation destabilized a predicted disulfide, targeted mutagenesis was used to examine the contribution of each of four cysteine residues, in two predicted disulfides (C96/C107 and C135/C149), to CdtC activities. Cysteine replacement mutations in two predicted disulfides (C136/C149 and C178/C197) in CdtA were also characterized. Flow cytometry and CHO cell proliferation assays showed that changing either C96 or C149 in CdtC to alanine abolished the biological activity of holotoxin complexes. However, replacing C107 or C135 in CdtC and any of the four cysteines in CdtA with alanine or serine resulted in only partial or no loss of holotoxin activity. Changes in the biological activities of the mutant holotoxins correlated with altered subunit binding. In contrast to elimination of the B chain of ricin, the elimination of intrachain disulfides in CdtC and CdtA by genetic replacement of cysteines destabilizes these subunit proteins but not to the extent that cytotoxicity is lost. Reduction of the wild-type holotoxin did not affect cytotoxicity, and the reduced form of wild-type CdtA exhibited a statistically significant increase in binding to ligand. A diminished role for intrachain disulfides in stabilizing CdtA and CdtC may have clinical relevance for the A. actinomycetemcomitans Cdt. The cdt gene products secreted by this pathogen assemble and bind to target cells in periodontally involved sites, which are decidedly reduced environments in the human oral cavity.

FIG. 1.

Effect of point mutations in CdtC on the cytotoxic activities of reconstituted heterotoxin. (A) Effects on CHO cell proliferation. Affinity-purified recombinant wild-type CdtA-His₆, CdtB-His₆, and CdtC-His₆ or mutant CdtC-His₆ proteins were preincubated in folding buffer, and the mixture (5 μg/ml of culture medium) was added to CHO cell cultures. Cell colonies were stained and counted after 6 days of culturing, and the data are expressed as numbers of CFU. Amino acid substitutions are shown for each mutant. CdtC designates holotoxin reconstituted with wild-type CdtC-His₆. All samples were run in triplicate. Statistically significant differences between the numbers of CFU of untreated CHO cell cultures and of those treated with wild-type or mutant heterotoxins are marked with an asterisk (P < 0.000005). (B) Assessment of DNA damage by PFGE. Reconstituted wild-type or mutant heterotrimers (10 μg of total protein/ml) were added to HeLa cells, and the cultures were incubated for 36 h. The cells were then prepared for PFGE as described in Materials and Methods. The numbers represent the relative percentages of DNA retained in the well versus that in the gel. (C) Assessment of heterotoxin assembly by differential dialysis. Heterotrimers composed of preincubated mixtures of either wild-type or mutant CdtC-His₆ proteins were dialyzed for 48 h. The protein composition of the material remaining after dialysis was examined by Western blotting as described in Materials and Methods. CdtA′ represents the truncated form (17 to 18 kDa) of CdtA-His₆. Results of all experiments were typical of a minimum of three trials.

FIG. 2.

Effect of point mutations on the binding activities of CdtC. (A) Saturation curve of the binding of wild-type recombinant CdtC-His₆ to CdtA-His₆ on thyroglobulin-coated wells from an ELISA. Bound Cdt protein was detected with anti-His · Tag monoclonal antibody (1:3,000 dilution) and antimouse immunoglobulin G horseradish peroxidase conjugate (1:3,000 dilution) as described in Materials and Methods. The dashed line marks the absorbance value of bound CdtA-His₆ at an input concentration of 4 μg/well. All CdtC-His₆ concentrations were run in triplicate. (B) Affinity-purified recombinant CdtC-His₆ and the mutant proteins (10 μg/well) were added to ELISA plate wells coated with thyroglobulin (white bars) or thyroglobulin and 4 μg/well CdtA-His₆ (black bars). mutC162 has a single nucleotide change (T46C) but no corresponding amino acid change. Bound protein was detected as in the experiment whose results are shown in panel A, and absorbance values were compared to those of wells containing only 4 μg/well CdtA-His₆ (middle dotted line). All samples were run in triplicate. Statistically significant differences between the absorbance values for the mutant and wild-type CdtC-His₆ proteins bound to wild-type CdtA-His₆-coated thyroglobulin are marked with an asterisk (P < 0.005). Results of all experiments were typical of three trials.

FIG. 3.

Positions of the amino acid substitutions in CdtC and CdtA. (A) Alignment of the deduced amino acid sequences of A. actinomycetemcomitans recombinant CdtC-His₆ (5) and H. ducreyi CdtC (GenBank accession number U53215). Unique amino acid substitutions identified in the random-mutagenesis library and from the targeted cysteine mutagenesis are shown. Nonidentical residues are marked by shaded boxes. (B) Alignment of the deduced amino acid sequences of A. actinomycetemcomitans recombinant CdtA-His₆ (5) and H. ducreyi CdtA (GenBank accession number U53215). Amino acid substitutions for the four cysteine residues are shown. Nonidentical residues are marked by shaded boxes. The substitution C197S is in mutA65 from Cao et al. (5). (C) Alignment of the deduced amino acid sequences of A. actinomycetemcomitans recombinant CdtA-His₆ and CdtC-His₆, comparing the relative locations of the substituted cysteine residues. Identical amino acids are marked by shaded boxes. Cysteine residues in CdtA-His₆ and CdtC-His₆ not predicted to form disulfides are marked with an asterisk.

FIG. 4.

Locations of the amino acid substitutions in CdtC-His₆ and CdtA-His₆ in the crystal structure of the A. actinomycetemcomitans Cdt. The coordinates (Protein Data Bank accession number 2F2F) for the crystal structure were computer modeled. (A) The eight unique amino acid substitutions in the A. actinomycetemcomitans recombinant CdtC-His₆ mutants having a binding-deficient phenotype are shown at their corresponding positions in the A. actinomycetemcomitans crystal structure. Positions of the mutated residues within the structure are in orange. The position of the one substitution (F97C) that resulted in a noncytotoxic phenotype is underlined. The amino [N(E21)]- and carboxy [C(S186)]-terminal residues in CdtC are labeled. Only the backbone structure is shown. (B) Locations of predicted disulfide-forming cysteine residues in CdtC and CdtA in the A. actinomycetemcomitans crystal structure. There is one amino acid difference between the CdtA-His₆ shown in Fig. 3 (5) and that used to form the Cdt crystal structure (Protein Data Bank accession number 2F2F). Calculated distances between paired cysteines predicted to form disulfide bonds are shown in angstroms. Only the backbone structure is shown. Note that the structure in panel B is rotated for clarity relative to that in panel A. (C) Crystal structure of the A. actinomycetemcomitans Cdt showing the surface exposed residues in CdtC (blue) and CdtA (green). Exposed cysteines are yellow and are labeled. (D) Same structural model as that shown in panel C but rotated to show the exposed cysteines in CdtA.

FIG. 5.

Effect of cysteine mutant holotoxins on the proliferation of CHO cells. (A) Holotoxin was reconstituted with either wild-type CdtC-His₆ or CdtC-His₆ cysteine mutant proteins and added to CHO cell cultures as described in the legend to Fig. 1. Cell colonies were stained and counted after 6 days of growth, and the data were expressed as numbers of CFU. All samples were run in triplicate. Statistically significant differences between the numbers of CFU in untreated CHO cell cultures and in those treated with wild-type or mutant heterotoxin are marked with asterisks (*, P < 0.005; **, P < 0.05). Images of stained colonies from the untreated and mutCc107 holotoxin-treated cultures are shown in the inset. nc, normal colonies; sc, small colonies. (B) Same experiment as that shown in panel A except that holotoxins were reconstituted with the wild-type and cysteine mutant CdtA-His₆ proteins.

FIG. 6.

Effect of cysteine replacement on the ligand and subunit binding activities of CdtC-His₆ and CdtA-His₆ proteins. ELISA plates were coated with thyroglobulin. Wild-type CdtA-His₆ (4 μg/well) and wild-type CdtC-His₆ (3.5 μg/well) were added individually to the thyroglobulin-coated wells (white bars). Other wells received 4 μg/well of wild-type CdtA-His₆ followed by 3.5 μg/well of the individual wild-type and cysteine mutant CdtC-His₆ proteins (gray bars). Other thyroglobulin-coated wells received 10 μg/well of wild-type or cysteine mutant CdtA-His₆ proteins (black bars). The plates were developed, and the absorbance was determined as described in the legend to Fig. 2. All samples were run in triplicate. Statistically significant changes in the binding of the CdtC-His₆ and CdtA-His₆ cysteine mutant proteins from that of the corresponding wild-type proteins are marked by asterisks (*, P < 0.0001; **, P < 0.005). Absorbance ratios based on the binding of wild-type CdtA-His₆ to thyroglobulin were calculated as described in Materials and Methods. Dotted lines mark the absorbance values of wild-type CdtA-His₆ (4 and 10 μg/well) plus CdtC-His₆ (3.5 μg/well) bound to thyroglobulin. Results are representative of three experiments.

FIG. 7.

Effects of a reducing agent on wild-type holotoxin and CdtA-His₆ activities. (A) Purified recombinant wild-type CdtA-His₆, CdtB-His₆, and CdtC-His₆ were reconstituted in refolding buffer under reducing and nonreducing conditions as described in Materials and Methods. These preparations were then added to CHO cell cultures growing in medium with and without 10 mM DTT. Colonies were stained and counted after 6 days of growth and were expressed as numbers of CFU. Black bars, unreduced culture and unreduced heterotoxin; white bars, reduced culture and reduced heterotoxin. Statistical differences between the effects of the unreduced and reduced samples are shown as P values. red, reduced. (B) ELISA plates were coated with either reduced or nonreduced thyroglobulin (thyroglobulin). Reduced or unreduced wild-type CdtA-His₆ was then added to the wells in increasing concentrations. Bound CdtA-His₆ was detected as described in the legend to Fig. 2. Unfilled squares, thyroglobulin and reduced CdtA-His₆; filled squares, reduced thyroglobulin and reduced CdtA-His₆; open circles, thyroglobulin and CdtA-His₆; solid circles, reduced thyroglobulin and CdtA-His₆. All samples were run in triplicate a minimum of three times.