Central nervous system tuberculosis (CNS TB) is a debilitating disease, affecting young children disproportionately, and leading to severe neurological sequalae. Little is known, however, regarding the pathogenesis and mechanism(s) by which M. tuberculosis invade the CNS. In addition, though multiple clinical reports have shown the association of distinct M. tuberculosis strains with CNS disease, the microbial virulence factors required have not been described previously. In order to create a system for the study of CNS TB pathogenesis, murine and guinea pig models for CNS invasion were developed and characterized. Using bacterial mutant screens in our animal models of TB, we identified M. tuberculosis pknD (Rv0931c) as a key microbial factor required for CNS disease. These findings are CNS tissue- and host-cell (brain endothelia)-specific, and not observed in lung tissue, macrophages, lung epithelia or other endothelia. M. tuberculosis pknD encodes a "eukaryotic-like" serine-threonine protein kinase with an intracellular kinase, and an extracellular (sensor) domain. The sensor localizes to the bacterial cell wall and is homologous to proteins that interact with the host extracellular matrix protein, laminin. We have shown that M. tuberculosis pknD sensor binds to laminin, and is sufficient to trigger host-cell invasion. We propose a model in which interaction of the pknD sensor domain with host laminin triggers bacterial signaling via the kinase, enhancing the invasive and survival capacity of the bacilli These studies have identified a novel role for M. tuberculosis pknD, which may represent an important mechanism for bacterial invasion and subsequent CNS disease.