Mathematical modeling of Mycobacterium tuberculosis dissemination - Research Summary Tuberculosis (TB), a disease caused by the bacteria Mycobacterium tuberculosis (Mtb), remains a major infec- tious disease of humans in the world. After the initial local infection of one site in the lung Mtb somehow dissem- inates in the lung and often spreads beyond the lung. In fact, extrapulmonary TB is a hallmark of the disease in young children and immunocompromised adults that is difficult to diagnose and treat. Our understanding of Mtb dissemination, both within the lung and beyond, remains limited, however. In this proposal we assembled a team of scientists with expertise in computational biology (Ganusov, Aitchison, Duffy, Langston) and TB pathogenesis (Urdahl, Sherman, Behar) to provide quantitative understanding of mechanisms of Mtb dissemination the lung and systemically. To this end, we will be using a number of highly innovative techniques such as i) a novel animal model of TB: infection of mice with an ultra low dose (ULD, 1-3 colony forming units, CFU) of Mtb along with a set of 50 barcoded Mtb strains, ii) an Mtb strain H37Rv-pBP10 with the replication clock plasmid, allowing to estimate how quickly bacteria are eliminated in vivo, and iii) mRNA-based gene signatures predicting bacterial numbers in murine lungs and TB disease progression risk in humans. With three complementary specific aims we will pro- vide detailed, quantitative understanding of fundamental processes of how Mtb disseminates from the deposition in lung alveoli to the whole lung and systemically. In Aim 1 we will determine the pathway of Mtb dissemination within the lung using a novel model of ULD-infected mice that mimics better human infection than many other animal models. In particular, we will discriminate between alternative hypotheses of Mtb spread in the lungs such the “bubble model” (in which Mtb spreads locally between lung lobes) and the “reseeding model” (in which Mtb spreads hematogenously to different parts of the lung after disseminating systemically). In Aim 2 we will determine the contribution of different cell populations, including Mtb-specific CD4 T cell response, to kinetics of Mtb dissemination systemically in mice infected with conventional doses (CD, 150 CFU) of Mtb. To parameterize best fit models we will use data from experiments with Mtb H37Rv carrying the replication clock plasmid pBP10. Finally, in Aim 3 we will attempt to improve on our recently derived mRNA-based gene signatures predicting CFU in murine lungs using cutting-edge graph theory-based methods of data dimensionality reduction. We will also perform experiments and define a new signature predicting disseminated TB in mice, and test its accuracy using data from monkeys and humans. Taken together, by combining experimental data from highly innovative experiments involving novel techniques (ultra low dose infections, barcoded strains, replication clock plasmid, microarray-based gene signatures) we will provide a quantitative understanding of how Mtb disseminates in the lung and systemically in the body.