No abstract

Previous work in 3 subjects infected for 2 weeks indicated that experimental infection with Haemophilus ducreyi recruits CD4 cells to the skin at the pustular stage of disease. In order to describe the kinetics of the host response, 23 subjects were infected at 2 sites with a standardized dose of H. ducreyi. Subjects were biopsied 1 or 4 days after inoculation or when they developed a painful pustular lesion (days 7-14). Papules and pustules contained a predominant T cell infiltrate that consisted of CD45RO and CD4 cells of the alpha beta lineage. Both papules and pustules contained mixed or T helper 1 type cytokine mRNA and interleukin-8 and tumor necrosis factor-alpha mRNA. Although the subjects had no history of chancroid, their immune responses resembled delayed-type hypersensitivity reactions that occurred within 24 h of inoculation and persisted throughout the course of experimental infection.

We previously identified a heat- and protease-labile haemolytic activity expressed by Haemophilus ducreyi. In order to characterize the haemolysin at the molecular level, genomic DNA from H. ducreyi was probed with haemolysin genes from other Gram-negative organisms. The haemolysin genes of Proteus mirabilis hybridized to H. ducreyi DNA suggesting that the haemolysin of H. ducreyi is related to the Proteus/Serratia pore-forming family of haemolysins. Tn916 mutagenesis was employed to isolate haemolysin-deficient mutants. Approximately 5000 Tn916 transposon mutants were screened for the loss of haemolytic activity and two mutants were identified. One mutant, designated 35,000-1, was further characterized. Sequences flanking the Tn916 element in strain 35,000-1 were employed to identify clones from a lambda DASHII library of H. ducreyi strain 35,000 DNA. A 13 kb insert from one lambda clone was selected for further study. This 13 kb fragment was able to both confer haemolytic activity to Escherichia coli and complement the haemolysin deficiency in strain 35,000-1. The haemolysin gene cluster was cloned from this 13 kb insert and two genes, designated hhdA and hhdB, were identified. The derived amino acid sequence of these genes demonstrated homology to the haemolysin and activation/secretion proteins of P. mirabilis and Serratia marcescens.

The haemolysin of Haemophilus ducreyi is the newest member of the Proteus/Serratia family of pore-forming toxins. In order to assess the role of the haemolysin in virulence, we constructed an isogenic haemolysin-deficient mutant of H. ducreyi strain 35000 This strain, designated 35000-3, lacks detectable haemolytic activity. We tested H. ducreyi strains 35000 and 35000-3 for their cytopathic activity against human foreskin fibroblasts (HFFs). We observed strong cytopathic activity when strain 35000 was co-cultured with HFFs. In contrast, cytopathic activity was not observed when strain 35000-3 was co-cultured with HFF cells. We also analysed the isogenic pair of H. ducreyi strains for cytopathic activity against HeLa cells and the keratinocyte cell line HaCaT. Strains 35000 and 35000-3 were strongly cytotoxic when co-cultured with HeLa cells. HaCaT monolayers were slightly damaged by cocultivation with strain 35000-3 but this damage was much less than that observed when HaCaT cells were cocultured with strain 35000. These results indicate that the H. ducreyi haemolysin is responsible for the previously observed cytotoxic activity against HFF cells and is partially responsible for the activity observed with HaCaT cells. The haemolysin, however, is not responsible for the activity observed with HeLa cells.

Haemophilus ducreyi causes the genital ulcerative disease chancroid. One putative virulence factor of H. ducreyi is a pore-forming hemolysin that displays toxicity against human fibroblasts and keratinocytes. In order to test the role of the hemolysin in pathogenesis, an isogenic hemolysin-deficient mutant was constructed, designated 35000HP-RSM1. The lipooligosaccharide, outer membrane protein patterns, and growth attributes of 35000HP-RSM1 were identical to its parent, 35000HP. Human subjects were challenged on the upper arm with the isogenic isolates in a double-blinded, randomized, escalating dose-response study. Pustules developed at a similar rate at sites inoculated with the mutant or parent. The cellular infiltrate and bacterial load in lesions were also similar. These results indicate the hemolysin does not play a role in pustule formation. Due to the limitations of this model, the role of the hemolysin at later stages of infection could not be determined.