Model selection in drug testing: A cautionary tale from tauopathy research

View profile for Alexander Kertser

Head of Pipeline and Discovery at ImmunoBrain

On the Importance of Model Selection in Drug Testing For many human diseases, there are no true mouse equivalents. To still evaluate drug effectiveness in a complex, living organism, scientists have learned to engineer animal models that mimic selected aspects of human pathology. In essence, we take a mouse, deliberately disrupt one of its proteins, then “fix” it with a drug that restores what we just broke - and celebrate that we’ve cured the disease. Two major issues arise from this approach. First, such models often become benchmarks for drug testing despite capturing only a narrow mechanistic slice of the human condition. Treatments that look miraculous in mice frequently yield limited benefit in people. Second, these models are poorly suited for evaluating drugs that act through entirely different pathways and do not directly target the artificially overexpressed protein. In our recent bioRxiv preprint, ImmunoBrain tested an anti-PD-L1 blocking antibody in the PS19 mouse line - a tauopathy model that overexpresses mutated human tau (P301S). Unlike slower-progressing amyloid or tau models, PS19 mice develop rapidly advancing neurodegeneration with early mortality. Importantly, anti-PD-L1 does not target tau itself; it works by modulating immune activity in the periphery, which in turn supports natural repair processes within the CNS, regardless of the primary disease etiology. In previous studies using slower progressing mouse models of tauopathy, PD-L1 blockade was shown to ameliorate cognitive decline alongside sustained reductions in in pathological tau species. In contrast, in the aggressive PS19 model, while we observed a significant beneficial effect on cognitive performance, the effects on pathological tau were short-lived. This outcome aligns with the notion that a treatment that does not directly counteract the main driver of the model - in this case, massive tau overexpression – may be at a relative disadvantage in maintaining a long-lasting effect following a single dose. Our findings underscore how model kinetics can influence therapeutic outcomes and highlight the importance of calibrating treatment regimens accordingly. In humans, disease evolves over years rather than weeks, and this temporal difference is routinely considered during clinical translation. A similar awareness should guide comparisons between preclinical models. Careful interpretation is therefore essential - across models, not all “positive” or “negative” results are created equal. https://lnkd.in/dRgAYveA #Innovation #Biotech #HealthcareInnovation #TranslationalResearch #Neurodegeneration #DrugDiscovery #Neuroimmunology

To view or add a comment, sign in

Explore content categories