The vast surface area of the distal lung, sparsely populated with resident immune cells, is the often the site of emerging (COVID-19) or established (Tuberculosis) infection diseases. For example, severe COVID-19 result in microvascular thrombosis in the lung and systemic endothelialitis, but the underlying dynamics of damage to the vasculature and whether it is a direct consequence of endothelial infection or an indirect consequence of immune cell mediated cytokine storms is still unclear. In contrast, the minimum infectious dose in tuberculosis can be as low as one, which makes the early stages of infection difficult to study even in animal models. Yet heterogenous outcomes in the early stages of infection significantly alter the course of infection and may explain why the proportion of exposed individuals who develop clinical tuberculosis is low. In my talk, I will describe efforts to develop the lung-on-chip as model systems for respiratory infectious diseases and showcase the ability of these systems to enable new experiments that would not be possible in vivo.

In the case of COVID-19, we used a vascularised human lung-on-chip model [1] where we find that rapid infection of the underlying endothelial layer leads to the generation of clusters of endothelial cells with low or no CD31 expression, a progressive loss of endothelial barrier integrity, and a pro-coagulatory microenvironment. These morphological changes do not occur if endothelial cells are exposed to SARS-CoV-2 apically. Viral RNA persisted in individual cells, which generated a response skewed towards NF-KB mediated inflammation and an antiviral interferon response which was transient in epithelial cells but persistent in endothelial cells. Perfusion with Tocilizumab, an inhibitor of trans IL-6 signalling slows the loss of barrier integrity but does not prevent the formation of endothelial cell clusters with reduced CD31 expression. Further work is ongoing to understand the role of cell-cell communication in the inflammation observed.

Prior to this, we used a a murine lung-on-chip model for tuberculosis to recreate a level of pulmonary surfactant deficiency that would be lethal in vivo [2]. Using long-term time-lapse imaging, we measured the growth rates of small bacterial microcolonies in the induvial host cells at the air-liquid interface and showed that pulmonary surfactant secreted by epithelial cells dramatically reduced bacterial growth in both epithelial cells and macrophages, whereas deficient levels of surfactant led to unimpeded growth. These insights suggest a greater role for alveolar epithelial cells in early tuberculosis than previously assumed.