Infections frequently cause or complicate illnesses associated with hypoxaemia and local tissue hypoxia. The influence of hypoxia on the interaction between host and pathogen is therefore of considerable interest. S. aureus is a major pathogen in critical care where patients may have profound hypoxaemia and at a tissue level, S. aureus frequently infects ischaemic wounds. We investigated the effect of systemic hypoxia on host-pathogen interactions using a subcutaneous S. aureus infection model in mice.
C57BL/6 mice were shaved, injected with a low dose of S. aureus (5x107 SH1000) and placed in a hypoxic chamber (10% O2) or left in room air. At 6 or 12 hours mice were assessed clinically and rectal temperature recorded. Clinical assessment of mouse sickness behaviour was made by two independent observers blinded as to which oxygen tension the mice had been exposed. Mice were anaesthetised and tissue samples (blood, skin, lung, spleen, kidney, liver and brain) obtained for analysis.
Mice injected with live bacteria and placed in hypoxia developed a phenotype of sickness behaviour and hypothermia. Infected hypoxic mice had significantly higher sickness scores and lower body temperature than infected normoxic mice or hypoxic mice injected with PBS (rectal temperature at 12 hours: hypoxic 33.4˚C ± 0.74, normoxic 37.7˚C ± 0.24, hypoxic PBS-injected 38.9˚C ± 0.26, p < 0.0001). Surprisingly, we found no evidence of bacteraemia, enhanced cytokine production, vascular leak or lung injury in the hypoxic infected mice. However, these animals had significant circulatory dysfunction, with hypotension, bradycardia and echocardiographic evidence of impaired left ventricular function. Interestingly, myeloid-cell deficiency of either HIF-1α or HIF-2α protected mice from the adverse systemic phenotype in this model, implicating the host innate immune response in the pathogenesis of the phenotype.
These findings imply that hypoxia may adversely alter the host response to a minor bacterial challenge, leading to profound systemic illness and that in such a setting modulation of the HIF pathway may be a possible therapeutic option.