C. psittaci modulates virulence by alteration of host immunity, which is conferred by small differences in the chromosome.

Genetic mapping studies may provide association between sequence variants and disease susceptibility that can, with further experimental and computational analysis, lead to discovery of causal mechanisms and effective intervention. We have previously demonstrated that polymorphisms in immunity-related GTPases (IRG) confer a significant difference in susceptibility to Chlamydia psittaci infection in BXD recombinant mice. Here we combine genetic mapping and network modeling to identify causal pathways underlying this association. We infected a large panel of BXD strains with C. psittaci and assessed host genotype, IRG protein polymorphisms, pathogen load, expression of 32 cytokines, inflammatory cell populations, and weight change. Proinflammatory cytokines correlated with each other and were controlled by a novel genetic locus on chromosome 1, but did not affect disease status, as quantified by weight change 6 days after infection In contrast, weight change correlated strongly with levels of inflammatory cell populations and pathogen load that were controlled by an IRG encoding genetic locus (Ctrq3) on chromosome 11. These data provided content to generate a predictive model of infection using a Bayesian framework incorporating genotypes, immune system parameters, and weight change as a measure of disease severity. Two predictions derived from the model were tested and confirmed in a second round of experiments. First, strains with the susceptible IRG haplotype lost weight as a function of pathogen load whereas strains with the resistant haplotype were almost completely unaffected over a very wide range of pathogen load. Second, we predicted that macrophage activation by Ctrq3 would be central in conferring pathogen tolerance. We demonstrated that macrophage depletion in strains with the resistant haplotype led to neutrophil influx and greater weight loss despite a lower pathogen burden. Our results show that genetic mapping and network modeling can be combined to identify causal pathways underlying chlamydial disease susceptibility.