Understanding the host innate immune response to the histotoxic clostridia

The histotoxic clostridia, Clostridium perfringens and Clostridium septicum, are Gram positive, spore forming anaerobic bacilli that are the primary cause of gas gangrene in humans. They mediate disease via the production of numerous extracellular toxins and spreading factors and treatment to date relies on either debridement or amputation of the infected region in association with concurrent antibiotic treatment. However, this regime is not always successful and if left untreated, individuals succumb to disease with 24 to 48 post infection. The host innate immune response to these clostridial diseases is distinctive compared to that of other bacterial infections due to the paucity of polymorphonucleocytes (PMNs) within the infected regions. This leukostatic paucity is termed vascular leukostasis and is characterised by the sequestration of PMNs in the surrounding vasculature and their failure to enter into infected regions and successfully clear the infection. Remarkably, both C. perfringens and C. septicum infections share similar physiological manifestations, specifically tissue necrosis and lack of PMNs in the site of infection; however, what is now becoming apparent is that the underlying interactions between the two organisms and the host innate immune response differ significantly. Therefore, the aim of this thesis was to clarify the different mechanisms by which the histotoxic clostridia interact with, and modulate, the host innate immune response and to develop potential alternative treatment options for histotoxic clostridial myonecrosis. These studies showed that the opioid-based analgesics buprenorphine and morphine are able to perturb the progression and development of C. perfringens-mediated myonecrosis in a mouse model of infection. The results demonstrated that these compounds alter the vascular leukostatic phenotype in infected tissues, allowing neutrophils and macrophages to enter and clear the infection. qRT PCR analysis carried out on infected thigh tissues isolated from mice pretreated with buprenorphine showed higher levels of chemokine expression, which correlated with the increased numbers of PMNs found in these tissues and suggested that opioid-based analgesics are able to alter immune trafficking into a C. perfringens infected region by modulating the chemokine profile. Buprenorphine was also able to stop the progression of C. perfringens myonecrosis post-infection, with 60% of mice treated with buprenorphine 3.5 h post-infection surviving till the end of the experimental procedure. Accordingly opioid treatment has the potential to be both therapeutic and prophylactic. Notably, these protective effects did not extend to C. septicum-mediated myonecrosis, further highlighting that although both clostridial species are able to cause similar disease, the immunological differences that contribute to this pathology differ significantly. To further characterise the mechanism by C. septicum α-toxin modulates the innate immune response, the ability of α-toxin to interact with the MAPK signalling pathway was considered. This work has shown that α-toxin is able to activate all three arms of the MAPK pathway in a dose-dependent manner, that it requires both pore formation and calcium influx to do so and that it occurs in a c-Raf dependent and independent manner. Analysis of thigh tissues isolated from mice infected with an isogenic panel of strains showed that MAPK activation at the site of infection can occur independent of toxin, most likely by other uncharacterised virulence factors as well as bacterial cell wall components such as peptidoglycan and lipotechoic acid. Comparatively, activation at distal sites infection such as spleen appears to be primarily p38 and ERK dependent. These data offer a physiological mechanism by which α-toxin may initiate cellular oncosis and thus stop cellular trafficking into infected regions. Finally, the role of the C. perfringens cysteine protease α-clostripain was characterised, which had been proposed to be involved in toxin turnover and therefore to be involved in disease progression. A C. perfringens strain was constructed in which the α-clostripain gene was insertionally inactivated by using TargeTron technology, and an isogenic panel of strains constructed. We went on to complement this mutation in trans with a full-length copy of the α-clostripain gene, and went onto assess the panel of strains using the mouse myonecrosis model. Virulence testing showed that α-clostripain was not required for disease progression although the strain over-expressing α-clostripain showed reduced virulence in vivo; suggesting the potential for α-clostripain to be involved during infection.