Back‐translation of clinical tau‐aggregation inhibitor trials to experimental models: negative interactions of dementia therapies

Background Alzheimer disease (AD) has attracted considerable interest in novel therapies. Two Phase III trials by TauRx Pharmaceuticals (TRx‐005, TRx‐015) failed to attain their endpoints. A significant number of recruits, however, remained on symptomatic treatment and a post‐hoc analysis confirmed this sub‐cohort was selectively resilient against the tau aggregation inhibitor hydromethylthionine (HMT; previously termed LMT). By contrast, drug‐naïve patients were highly sensitive to the novel therapy and showed little decline over 18‐month. We here back‐translated this study to the preclinic. Methods Experiments used female tau transgenic Line 1 mice (L1) aged 6‐8 months previously characterised for cognitive impairments, tau pathology, and sensitivity to the tau aggregation inhibitor HMT. Here, multiple groups and doses of drugs were chronically administered (up to 11 weeks) with pre‐exposure to the acetylcholine esterase inhibitor rivastigmine followed by add‐on of HMT. Drugs were also dosed singly. Results (i) Tested for spatial learning in the water maze, L1 mice were significantly impaired and benefitted from exposure to HMT. By contrast, combination therapy with rivastigmine precluded some of the benefits of HMT while rivastigmine alone was non‐efficacious. (ii) Chronic drug administration was followed by tissue harvest and histological preparation of sections through 6 brain regions (3 x cortical, 3x basal forebrain) and labelling of 6 presynaptic structural (3) and functional proteins (3. Reduced protein levels were found in cortex; they were heightened by HMT, but abnormally altered in the presence of rivastigmine (single and combination therapy). (iii) Chronic drug administration was followed by in vivo brain dialysis from hippocampus for acetylcholine at rest and under behavioural challenge. L1 showed a lowering of cholinergic neurones in the basal forebrain, but no difference in acetylcholine levels at rest; a heightening of choline levels was observed to HMT alone at rest and challenge, but not to combinations with rivastigmine. Conclusion We successfully back‐translate the clinical trials on HMT into the preclinic using L1 tau transgenic mice . Data suggest not only a reduced cognitive efficacy of HMT in subjects pre‐exposed to the choline esterase inhibitor rivastigmine, but also offer mechanistic explanations for potential interactions at presynaptic function and for transmitter release.