Targeting hyperphosphorylated tau by immunotherapy is emerging as a promising approach to treat tauopathies such as Alzheimer’s disease and frontotemporal dementia. We have previously reported that active tau immunization clears tau aggregates from the brain and attenuates or prevents functional impairments in two different tangle mouse models. Here, we assessed the efficacy of passive immunization with the PHF1 antibody, which targets a phospho-epitope within one of our active immunogens. Homozygous female tangle mice (JNPL3, 2–3 months) were injected intraperitoneally once per week with PHF1 or pooled mouse IgG (250 μg/ 125 μL; n = 10 per group) for a total of 13 injections. Their behavior was assessed at 5–6 months of age and brain tissue was subsequently harvested for analyses of treatment efficacy. The treated mice performed better than controls on the traverse beam task (p < 0.03), and had 58% less tau pathology in the dentate gyrus of the hippocampus (p = 0.02). As assessed by western blots, the antibody therapy reduced the levels of insoluble pathological tau by 14–27% (PHF1, p < 0.05; PHF1/total tau, p < 0.0001) and 34–45% (CP13 or CP13/total tau, p < 0.05). Levels of soluble tau and sarkosyl soluble tau were unchanged, compared with controls, as well as total tau levels in all the fractions. Plasma levels of PHF1 correlated inversely with tau pathology in the brainstem (p < 0.01), with a strong trend in the motor cortex (p < 0.06) as well as with insoluble total tau levels (p < 0.02), indicating that higher dose of antibodies may have a greater therapeutic effect. Significant correlation was also observed between performance on the traverse beam task and PHF1 immunoreactivity in the dentate gyrus (p < 0.05) as well as with insoluble PHF1/total tau ratio on western blots (p < 0.04). These results show that passive immunization with tau antibodies can decrease tau pathology and functional impairments in the JNPL3 model. Future studies will determine the feasibility of this approach with other monoclonals and in different tangle models in which thorough cognitive assessment can be performed.

No abstract