A Blood RNA Signature in a Novel Murine Model Predicts Human Tuberculosis Risk

Tuberculosis (TB) is a highly heterogeneous disease that develops in a subset of individuals infected with aerosolized Mycobacterium tuberculosis (Mtb). We sought to develop an improved experimental model by infecting mice with an ultra-low dose (ULD, 1-3 founding Mtb bacteria), which is believed to reflect a physiologic inoculum. ULD-infected mice exhibited bacterial burdens ranging from <10 to 106 CFUs within individual lungs. Furthermore, they exhibited well-circumscribed granulomas that shared features with human granulomas. To monitor outcomes in live mice, we identified a blood RNA signature that correlated with lung bacterial burdens. Remarkably, this mouse-derived signature predicted Mtb infection outcomes across species, including disease scores in non-human primates and risk of progression to active TB in humans. Overall, the ULD mouse model recapitulates key features of human TB and provides a new platform to study the biology of Mtb infection in a tractable experimental system.