Characterization of Haemophilus ducreyi cdtA, cdtB, and cdtC mutants in in vitro and in vivo systems

Abstract

Haemophilus ducreyi expresses a soluble cytolethal distending toxin (CDT) that is encoded by the cdtABC gene cluster and can be detected in culture supernatant fluid by its ability to kill HeLa cells. The cdtA, cdtB, and cdtC genes of H. ducreyi were cloned independently into plasmid vectors, and their encoded proteins expressed singly or in various combinations in an Escherichia coli background. All three gene products had to be expressed in order for E. coli-derived culture supernatant fluids to demonstrate cytotoxicity for HeLa cells. Isogenic H. ducreyi cdtA and cdtB mutants were constructed and used in combination with the wild-type parent strain and a previously described H. ducreyi cdtC mutant (M. K. Stevens, J. L. Latimer, S. R. Lumbley, C. K. Ward, L. D. Cope, T. Lagergard, and E. J. Hansen, Infect. Immun. 67:3900-3908, 1999) to determine the relative contributions of the CdtA, CdtB, and CdtC proteins to CDT activity. Expression of CdtA, CdtB, and CdtC appeared necessary for H. ducreyi-derived culture supernatant fluid to exhibit cytotoxicity for HeLa cells. Whole-cell sonicates and periplasmic extracts from the cdtB and cdtC mutants had no effect on HeLa cells, whereas these same fractions from a cdtA mutant had a very modest cytotoxic effect on these same human cells. CdtA appeared to be primarily associated with the H. ducreyi cell envelope, whereas both CdtB and CdtC were present primarily in the soluble fraction from sonicated cells. Both the cdtA mutant and the cdtB mutant were found to be fully virulent in the temperature-dependent rabbit model for experimental chancroid.

FIG. 1

Partial restriction map of the H. ducreyi 35000 chromosomal DNA insert in pJL300 and related plasmids. Restriction sites in parentheses indicate vector cloning sites. The arrows indicate the direction of transcription. Plasmid pJL300 was derived from the pBR322 vector; both it and pJL303 have been described (44). The kan1 cartridge was blunt-ended and ligated into the similarly treated RsrII site in cdtA to construct pJL301 and into the similarly treated MscI site of cdtB to produce pJL302. Plasmid pDL20-A is a modified pBR322 vector containing the 1.1-kb PstI- and EcoRI-digested PCR product with the cdtA gene. Plasmid pDL10-B is a modified pACYC184 vector containing the 1.2-kb SalI- and BamHI-digested PCR product with the cdtB gene. Plasmid pJL300-C is pLS88 containing the cdtC gene (44).

FIG. 2

Western blot-based detection of H. ducreyi CdtA, CdtB, and CdtC proteins expressed by wild-type and mutant H. ducreyi strains. Whole-cell sonicates (lanes 1 to 4) and periplasmic extracts (lanes 5 to 8) of these four strains were probed with the H. ducreyi CdtA-reactive MAb 1G8 (A), the H. ducreyi CdtB-reactive MAb 20B2 (B), the H. ducreyi CdtC-reactive MAb 8C9 (C), the ZnuA-reactive MAb 3F1 (D), and the MOMP-reactive MAb 3F12 (E). Equivalent amounts of whole-cell sonicate and periplasmic extract were loaded into each lane. Lanes: 1 and 5, wild-type H. ducreyi 35000; 2 and 6, H. ducreyi cdtA mutant 35000.301; 3 and 7, H. ducreyi cdtB mutant 35000.302; 4 and 8, H. ducreyi cdtC mutant 35000.303. Molecular size markers (in kilodaltons) are indicated on the right.

FIG. 3

Killing of HeLa cells by CDT in subcellular fractions and culture supernatant fluids from wild-type and mutant strains of H. ducreyi. Cytotoxic activity was assessed in culture supernatant fluids (A), periplasmic extracts (B), and whole-cell sonicates (C) obtained from the wild-type H. ducreyi strain 35000, the cdtA mutant 35000.301, the cdtB mutant 35000.302, and the cdtC mutant 35000.303. HeLa cells were exposed to filter-sterilized culture supernatant fluids, periplasmic extracts, or whole-cell sonicates as described in Materials and Methods. After 96 h, wells in which viable cells were present were detected by the use of the tetrazolium reduction-based assay described in Materials and Methods. An optical density at 492 nm (OD₄₉₂) greater than 1.8 indicates wells in which all of the HeLa cells were viable.

FIG. 4

Cytopathic effect of CDT from wild-type and mutant strains of H. ducreyi. HeLa cells were exposed to whole-cell sonicates from the wild-type strain 35000 (A), the H. ducreyi cdtA mutant 35000.301 (B), the H. ducreyi cdtB mutant 35000.302 (C), and the H. ducreyi cdtC mutant 35000.303 (D), and 96 h later the wells were photographed at an original magnification of ×40. (E) HeLa cells exposed to the buffer used for the preparation of the whole-cell sonicates were used as the negative control.

FIG. 5

Detection of Cdt proteins in soluble and insoluble fractions from H. ducreyi cells. Whole-cell sonicates were prepared from the wild-type strain 35000 as described in Materials and Methods by using centrifugation at 3,000 × g to remove whole cells and gross debris. Equivalent amounts of whole-cell sonicate (lane 1), soluble cell contents (lane 2), cell envelopes (lane 3), and Sarkosyl-insoluble material (lane 4) were probed by Western blot analysis using the CdtA-reactive MAb 1G8 (A), the CdtB-reactive MAb 20B2 (B), the CdtC-reactive MAb 8C9 (C), and the MOMP-reactive MAb 3F12 (D). Molecular size position markers (in kilodaltons) are on the right.

FIG. 6

Agarose gel electrophoresis of H. ducreyi multiplex RT-PCR products. The sizes of the predicted RT-PCR products were as follows: pal, 355 bp; cdtA, 132 bp; cdtB, 460 bp; cdtC, 271 bp. Templates included the following: lane 1, 100 ng of H. ducreyi 35000 genomic DNA (positive control); lane 2, no template (negative control); lanes 3 and 4, 2 μg of RNA isolated from H. ducreyi-infected rabbit lesions. The reaction mixture loaded in lane 4 was not subjected to the reverse transcription step of the RT-PCR protocol and served as a control to detect DNA contamination of the RNA template. Lane M, DNA size markers (in base pairs).