Indeed, overPSI-7977 expression of fractalkine, which would be predicted to inhibit microglial activation, has been shown to reverse behavioral deficits in rTg4510 animals. Furthermore, inflammatory markers could serve as invaluable pharmacodynamic endpoints in this model, because the downregulation of tau with doxycycline resulted in reduced inflammation. Therefore, inflammatory markers could be used as surrogate efficacy endpoints to monitor treatments directed at tau, such as tau phosphorylation, in addition to inflammatory treatments. Doxycycline reduced inflammatory markers in as little as one week when administered to animals that already displayed H-89 robust inflammation, suggesting the treatment was not just preventing further inflammation, but resolving extant inflammation. The possibility does remain, however, that doxycyline is reducing inflammation directly, since doxycycline has been reported to have anti-inflammatory properties in its own right. However, the observation that doxycycline appears to preferentially reduce human tau levels in a particular cortical layer and this is the same layer that shows the greatest reduction in GFAP staining suggests that at least some of the reduced inflammation observed with doxycycline treatment was a result of tau reduction. In summary, we have conducted a characterization of the natural history of tau-dependent changes in gene expression in the rTg4510 mice. The predominant class of gene expression changes were in inflammatory genes, which increased in an age-dependent manner, coinciding with increases in tau phosphorylation, decreases in markers of neuronal activity, and behavioral deficits. These inflammatory genes formed a network highly reminiscent of the immune function gene expression network perturbed in AD. Laser microdissection of hippocampal subfields revealed additional gene expression changes in AD-relevant synaptic transmission pathways. These perturbations may contribute to the observed behavioral deficits.Overall these data suggest that the rTg4510 recapitulates several aspects of human neurodegenerative disease, and could be used as a reasonable preclinical model to test diseaserelevant hypotheses and evaluate candidate therapeutic interventions.