Distinguishing multiple roles of T cell and macrophage involvement in determining lymph node fates during...

Tuberculosis (TB) is a disease of major public health concern with an estimated one-fourth of the world currently infected with M. tuberculosis (Mtb) bacilli. Mtb infection occurs after inhalation of Mtb, following which, highly structured immune structures called granulomas form within lungs to immunologically restrain and physically constrain spread of infection. Most lung granulomas are successful at controlling or even eliminating their bacterial loads, but others fail to control infection and promote disease. Granulomas also form within lung-draining lymph nodes (LNs), variably affecting immune function. Both lung and LN granulomas vary widely in ability to control infection, even within a single host, with outcomes ranging from bacterial clearance to uncontrolled bacterial growth. While lung granulomas are well-studied, data on LN granulomas are scarce; it is unknown what mechanisms drive LN Mtb infection progression and variability in severity. Recent data suggest that LN granulomas are niches for bacterial replication and can reduce control over lung infection. To identify mechanisms driving LN Mtb infection, we developed a multi-scale compartmental model that includes multiple lung-draining LNs, blood. We calibrated to data from a nonhuman primate TB model (one of the only models that parallels human TB infection). Our model predicts temporal trajectories for LN macrophage, T-cell, and Mtb populations during simulated Mtb infection. We also predict a clinically measurable infection feature from PET/CT imaging, FDG avidity. Using uncertainty and sensitivity analysis methods, we identify key mechanisms driving LN granuloma fate, T-cell efflux rates from LNs, and a role for LNs in pulmonary infection control.