A brain-penetrant triazolopyrimidine enhances microtubule-stability, reduces axonal dysfunction and decreases tau pathology in a mouse tauopathy model

Abstract

Background: Alzheimer's disease (AD) and related tauopathies are neurodegenerative diseases that are characterized by the presence of insoluble inclusions of the protein tau within brain neurons and often glia. Tau is normally found associated with axonal microtubules (MTs) in the brain, and in tauopathies this MT binding is diminished due to tau hyperphosphorylation. As MTs play a critical role in the movement of cellular constituents within neurons via axonal transport, it is likely that the dissociation of tau from MTs alters MT structure and axonal transport, and there is evidence of this in tauopathy mouse models as well as in AD brain. We previously demonstrated that different natural products which stabilize MTs by interacting with β-tubulin at the taxane binding site provide significant benefit in transgenic mouse models of tauopathy. More recently, we have reported on a series of MT-stabilizing triazolopyrimidines (TPDs), which interact with β-tubulin at the vinblastine binding site, that exhibit favorable properties including brain penetration and oral bioavailability. Here, we have examined a prototype TPD example, CNDR-51657, in a secondary prevention study utilizing aged tau transgenic mice.

Methods: 9-Month old female PS19 mice with a low amount of existing tau pathology received twice-weekly administration of vehicle, or 3 or 10 mg/kg of CNDR-51657, for 3 months. Mice were examined in the Barnes maze at the end of the dosing period, and brain tissue and optic nerves were examined immunohistochemically or biochemically for changes in MT density, axonal dystrophy, and tau pathology. Mice were also assessed for changes in organ weights and blood cell numbers.

Results: CNDR-51657 caused a significant amelioration of the MT deficit and axonal dystrophy observed in vehicle-treated aged PS19 mice. Moreover, PS19 mice receiving CNDR-51657 had significantly lower tau pathology, with a trend toward improved Barnes maze performance. Importantly, no adverse effects were observed in the compound-treated mice, including no change in white blood cell counts as is often observed in cancer patients receiving high doses of MT-stabilizing drugs.

Conclusions: A brain-penetrant MT-stabilizing TPD can safely correct MT and axonal deficits in an established mouse model of tauopathy, resulting in reduced tau pathology.

Keywords: Alzheimer’s disease; Microtubule; Tauopathy; Therapeutic.

Fig. 1

PK and pharmacodynamic profiling of 51657. a Plasma and brain levels of 51657 (structure inset) after a 5 mg/kg i.p. dose. Brain levels exceed plasma levels at all times, with terminal plasma and brain T_1/2 values of 1–1.5 h. Error bars represent SEM, with n = 3 per time point. b Levels of brain AcTub in WT mice 3 days after the last dose of vehicle or 10 mg/kg of 51657. Error bars represent SEM of n = 3 per treatment, with an unpaired t-test used to determine significance of compound effect

Fig. 2

PS19 mouse blood cells were unaffected by 12 weeks of 51657 dosing. Total blood cell counts were determined for WT mice receiving vehicle (n = 10) or PS19 mice receiving vehicle (n = 6), 3 mg/kg of 51657 (n = 7) or 10 mg/kg of 51657 (n = 8). No differences in (a) total white cell counts, (b) red cell counts or (c) neutrophil counts were observed between the vehicle- and 51657-treated mice as determined by one-way ANOVA. Error bars represent SEM

Fig. 3

PS19 mice treated with 51657 had significantly increased ON MT density and reduced axonal dystrophy. ON sections from vehicle-treated WT mice (n = 10) and PS19 mice treated with vehicle (n = 9) or 3 mg/kg (n = 8) or 10 mg/kg (n = 10) of 51657 were imaged by EM, and the number of MTs and dystrophic axons within treatment-masked images were counted as previously described [25]. a Quantification of MT density in ON sections demonstrates that vehicle-treated PS19 mice have a MT deficit relative to vehicle-treated WT mice, and treatment of PS19 mice with 3 mg/kg or 10 mg/kg of 51657 increases MT density to a level comparable to that of WT mice. b Quantification of ON EM images reveals a significant reduction in axonal dystrophy in PS19 mice receiving either 3 mg/kg or 10 mg/kg of 51657 compared to vehicle-treated PS19 mice. After quantification, a Grubb’s test determined there was an extreme outlier within the 10 mg/kg 51657 group that was not used for quantification. Analyses consisted of a one-way ANOVA with Tukey’s post-hoc analysis of between group differences. Error bars represent SEM. c Representative ON images from a vehicle-treated WT and PS19 mouse, with example MTs indicated by arrows. As depicted in the PS19 vehicle image, hexagonal fields of 0.035 μm² were overlaid on the ON images, with MTs counted within the hexagon and on three of the six borders to avoid repeat counting of MTs of MTs on adjacent hexagons (see also [27]). Scale bar represents 0.5 μM. d Representative ON images from a vehicle-treated WT and PS19 mouse, as well as a PS19 mouse that received a twice-weekly dose of 10 mg/kg of 51657. Vehicle-treated PS19 mice have greater axonal dystrophy, as evidenced by fewer intact axons and more axons that are demyelinated or debris-filled, than vehicle-treated WT mice. ONs of PS19 mice treated with 51657 more closely resembled those of vehicle-treated WT mice. Scale bar = 2 μm

Fig. 4

IHC staining and quantification of tau pathology. Three bregma-matched brain sections from each PS19 mice receiving vehicle (n = 12), or 3 mg/kg (n = 11) or 10 mg/kg (n = 12) of 51657, were stained with AT8 antibody to visualize tau pathology. a Representative images from a vehicle-treated PS19 mouse with an average amount of tau pathology (Bregma − 2.5). Regions of stained sections encompassing the hippocampus and entorhinal cortex were imaged and a fixed threshold was applied to distinguish AT8-positive staining from background (images on right; yellow, low AT8 signal; orange, moderate AT8 signal; red, high AT8 signal), followed by quantification of AT8-positive tau pathology. b A plot of the combined AT8-positive pathology from the entorhinal cortex and hippocampus from PS19 mice in each treatment group reveals a trend toward reduced tau pathology in the 51657-treated mice

Fig. 5

PS19 mice treated with 51657 show a reduction of insoluble pathological forms of tau. Brains of PS19 mice treated with vehicle (n = 12), or 3 mg/kg (n = 11) or 10 mg/kg (n = 12) of 51657 were sequentially extracted to remove high salt- and RIPA-soluble proteins. The remaining insoluble fraction was solubilized in 2% SDS and analyzed by immunoblotting, utilizing antibodies that recognize (a) total tau (17025 antibody), (b) phospho-tau (AT8 antibody) and (c) tau acetylated at residue K280 (AcTau; tau K280 antibody). The lower dose of 51657 caused a significant reduction of all three forms of insoluble tau, and the higher dose of 51657 dose resulted in a reduction of all insoluble tau species, with a significant reduction of AcTau. After quantification, a Grubb’s test determined there were extreme outliers within some treatment groups, resulting in the removal of a sample (B2) from the 10 mg/kg 51657 group from the AcTub antibody immunoblot (see Additional file 1: Figure S6), and a sample from each treatment group in the 17025 antibody immunoblot. d Soluble tau levels as measured by ELISA within the high salt fractions were unaffected by 51657-treatment. All comparisons consisted on one-way ANOVA with Tukey’s post-hoc analysis of between group differences. Error bars represent SEM

Fig. 6

Barnes maze testing of WT and PS19 mice treated with vehicle or 51657. The vehicle-treated PS19 mice had a modest deficit relative to vehicle-treated WT mice in successfully identify the escape compartment in the maze during the first two days of testing, and the PS19 mice receiving 51657 showed a non-significant trend toward improvement on day 1 and 2 compared to the vehicle group. Because of the modest amount of tau pathology and absence of overt neuron loss in the 12-month female PS19 mice, the behavioral deficits were mild and all treatment groups showed nearly 100% performance by the third and fourth days of testing