Search

Haemophilus ducreyi express two porin proteins, termed OmpP2A and OmpP2B. To test whether expression of OmpP2A and OmpP2B was necessary for virulence in humans, eight volunteers were experimentally infected with the parent (35000HP) in one arm and a double OmpP2A OmpP2B mutant (35000HP::P2AB) in the other arm. The pustule formation rates were 58.3% (95% CI, 33.2-83.5%) for the parent and 41.7% (95% CI, 19.3-64.0%) for the mutant (P=0.25). Biopsy of 35000HP and 35000HP::P2AB-infected sites yielded similar amounts of bacteria in quantitative culture. These results indicate that expression of OmpP2A and OmpP2B is not necessary to initiate disease or to progress to pustule formation in humans.

The cytolethal distending toxin of Haemophilus ducreyi (HdCDT) is a three-component toxin that induces the arrest of the mammalian cell cycle in the G2 phase. All of the individual gene products, CdtA, CdtB and CdtC, are required for toxic activity on cultured mammalian cells. The CdtB component alone exerts nuclease activity. The individual HdCDT components were purified by affinity chromatography or ion-exchange chromatography followed by gel-filtration. HdCDT was reconstituted and purified by the immobilization of a GST-CdtB fusion on a GSTrap column and the subsequent addition of cell sonicates from Escherichia coli recombinants that produced CdtA and CdtC. The purified HdCDT preparation contained all three CDT proteins, as detected by immuno-blotting, and had high cytotoxic activity (10(6)CPU/ml). Immunization of rabbits with the HdCDT complex and with the individual CdtA, CdtB and CdtC proteins elicited high titres of antibodies, as detected by ELISA. All of the immune sera had toxin-neutralizing activities. The pathological effects of the HdCDT complex were investigated in rabbits, since the proliferation of two rabbit cell lines, SIRC and RK-13, was inhibited by HdCDT. Intradermal injection of HdCDT (1, 10, 50 and 100microg protein) into naive rabbits resulted in dose-dependent skin reactions (erythema) about 24h after injection. Similar effects were not observed when the individual HdCDT proteins were injected. HdCDT injection into immune rabbits resulted in dose-dependent skin responses that were characterized by both erythema and oedema. Histological evaluation of the 24-h lesions in naive rabbits that were injected with HdCDT, revealed moderate levels of inflammatory cells, which were mainly granulocytes and macrophages, and dilatation of blood vessels. The skin reactions in HdCDT-injected immunized rabbits showed pronounced vascular changes and extensive infiltration of inflammatory cells, including eosinophils. All of the pathological changes healed after 3 days. In conclusion, purified HdCDT holotoxin is a complex of all three CDT proteins and all three components induce neutralizing antibodies when injected in rabbits. HdCDT causes dose-dependent pathologic skin reactions in both naive and immune rabbits, which is characterized by increased inflammatory responsiveness after each immunization.

Haemophilus ducreyi is the causative agent of the genital ulcer disease Chancroid. Chancroid has been shown to increase the risk of heterosexual transmission of HIV. Little is known regarding the attachment or localization of this organism to human cells in either the dermal or epidermal layer. In this study the attachment of H. ducreyi to human foreskin fibroblast (HFF) cells was further characterized. Attachment was mediated by more than one mechanism. Proteinase K treatment but not trypsinization of H. ducreyi significantly reduced attachment suggesting protein involvement. In addition, purified lipooligosaccharide (LOS) was able to inhibit attachment in a dose dependent manner. It appeared that the organism binds to fibronectin in the extracellular matrix of HFF cells, since competition studies using fibronectin showed that it was able to significantly reduce attachment in a dose dependent manner whereas collagen did not. We hypothesize that the attachment of H. ducreyi involves both a protein mediator of attachment (likely pili) as well as LOS and that one or both of these bacterial components interacts with fibronectin in the extracellular matrix to mediate attachment to HFF cells.

The response to heat shock in Haemophilus ducreyi has been investigated by labelling newly synthesized proteins with a pulse of 35S-methionine at a range of temperatures between 30-40 degrees C. Proteins were then separated by SDS-PAGE and analysed by autoradiography of the resultant profiles. Four heat shock proteins (HSP) were identified of apparent molecular weight < 14, 58.5, 74 and 78 kDa. The three larger proteins appeared to be independently controlled and inducible in a range of H. ducreyi strains. The protein profiles were also probed by Western blot with serum from an immunized rabbit, serum from a rabbit infected intradermally with H. ducreyi, mouse monoclonal antibodies (MAb) and a pool of human sera from patients with chancroid. The rabbit sera, the human serum pool and MAb (BB11) reacted with epitopes on the HSP of 58.5 kDa, whereas only the immunized rabbit serum and another MAb (CC11) reacted with epitopes on the HSP of 78 kDa. The HSP of 58.5 kDa appears to be immunogenic and may play a part in the host response to H. ducreyi and the pathogenesis of chancroid.

Haemophilus ducreyi cytotoxin-positive and -negative strains as well as bacterial sonicates and lipooligosaccharides (LOS) from such strains were evaluated for the capacity to produce dermonecrotic lesions, especially ulcers, after intradermal injections to rabbits and to different mouse strains, including nude mice. Dermonecrotic lesions of the ulcerous type were observed within 4 days and they were developed in both rabbits and mice with about 10(7) colony forming units (cfu) of H. ducreyi. Viable bacteria were isolated from the lesions up to 9 days after inoculation. All lesions healed spontaneously within 2-3 weeks. Bacterial sonicate (heated and unheated) and LOS preparations caused mainly abscess formation in rabbits, while in mice, a superficial, haemorrhagic ulceration was observed. To obtain ulceration at all injection sites, about 200 micrograms of LOS was required. Histological examination of acute, dermonecrotic lesions caused by viable bacteria showed deep necrosis, infiltrate of inflammatory cells, especially granulocytes and dilatation of blood vessels. The same type of inflammatory cells as seen in lesions caused by bacteria, were involved in the mouse lesions caused by bacterial sonicate and LOS preparations. The results indicate that LOS/endotoxin, probably in combination with other bacterial polysaccharides, can play a role in ulceration caused by H. ducreyi in animals; however, a relatively high amount of LOS preparation was necessary to cause dermal ulceration at all injection sites in the mouse model. The development of ulcers correlated with the endotoxin activity in bacterial sonicate and in LOS preparations. The model may therefore be useful to study the role of LOS components in development of ulceration. There was no significant difference in lesions caused by cytotoxin producing, respectively, non-producing H. ducreyi strains and cell-free preparations from such strains. The bacterial sonicates, cytotoxic for human cell lines, failed to kill animal cell lines, indicating that animal models do not adequately reflect the cytotoxin activity in experimental H. ducreyi infection. Antibodies to H. ducreyi sonicate and LOS, tested by means of ELISA, were found in pre-immune sera from both rabbits and mice. There was a significant antibody response to homologous cell sonicate and LOS, after primary and secondary infections with bacteria. Still, there was no clear difference between primary and secondary lesions in animals. Since animal lesions are mainly due to endotoxin activity, this may indicate that antibodies are of minor importance for protection in animal models.

Ten isolates of Haemophilus ducreyi, including seven clinical isolates and three laboratory reference strains, were assessed for their ability to produce extracellular enzymes which might contribute to their pathogenicity. Protease, elastase, lecithinase, lipase or collagenase enzyme activity were not detected in culture filtrates from any of the isolates tested using either plate or spectrophotometrical assays. Furthermore, cell-free culture filtrates did not exhibit cytotoxic activity in vitro when tested using either an established tissue culture cell line (Vero) or a primary cell culture of human keratinocytes. These results suggest that extracellular products do not play a role in host invasion or production of tissue damage by H. ducreyi.

Twelve strains of Haemophilus ducreyi isolated primarily from chancroid outbreaks in North America were examined for the presence of pili by transmission electron microscopy. We identified piliated cells in 10 of the 12 strains. Pilin extracts were prepared from the mechanically sheared cells of the 12 H. ducreyi strains as well as the stably piliated H. influenzae strain R890 and its non-piliated parent R906. Pili were present in 12 out of 12 H. ducreyi extracts and in the R890 extract but not in the R906 preparation. Pili were purified by cycles of differential pH solubilization and crystallization. In SDS-PAGE, the preparation consisted predominantly of a protein whose apparent relative molecular mass was 24,000 (24 k), and an electron micrograph showed that the preparation contained pili. Three H. ducreyi strains were passed 52 times on agar plates, and extracts prepared from these strains contained pili. There was no evidence of binding of erythrocytes obtained from nine mammalian and avian species to colonies of one of the stably piliated H. ducreyi strains. We conclude that H. ducreyi expressed pili, that the relative molecular mass of the pilin monomer was 24 k, that pilus expression was not readily lost in passage and that H. ducreyi pili may not bind to an erythrocyte receptor.