Age-dependent neuroinflammation and cognitive decline in a novel Ala152Thr-Tau transgenic mouse model of PSP and AD

Introduction Mutations of Tau are associated with several neurodegenerative disorders. Recently, the Tau mutation A152T was described as a novel risk factor for frontotemporal dementia spectrum disorders and Alzheimer disease. In vitro Tau-A152T shows a decreased binding to microtubules and a reduced tendency to form abnormal fibers. Results To study the effects of this mutation we generated a mouse model expressing human full-length Tau with this mutation (hTau40AT). At young age (2–3 months) immunohistological analysis reveals pathological Tau conformation and Tau-hyperphosphorylation combined with Tau missorting into the somatodendritic compartment of neurons. With increasing age there is Tau aggregation including co-aggregates of endogenous mouse Tau and exogenous human Tau, accompanied by loss of synapses (especially presynaptic failure) and neurons. From ~10 months onwards the mice show a prominent neuroinflammatory response as judged by activation of microglia and astrocytes. This progressive neuroinflammation becomes visible by in vivo bioluminescence imaging after crossbreeding of hTau40AT mice and Gfap-luciferase reporter mice. In contrast to other Tau-transgenic models and Alzheimer disease patients with reduced protein clearance, hTau40AT mice show a strong induction of autophagy. Although Tau-hyperphosphorylation and aggregation is also present in spinal cord and motor cortex (due to the Thy1.2 promoter), neuromotor performance is not affected. Deficits in spatial reference memory are manifest at ~16 months and are accompanied by neuronal death. Conclusions The hTau40AT mice mimic pathological hallmarks of tauopathies including a cognitive phenotype combined with pronounced neuroinflammation visible by bioluminescence. Thus the mice are suitable for mechanistic studies of Tau induced toxicity and in vivo validation of neuroprotective compounds. Electronic supplementary material The online version of this article (doi:10.1186/s40478-016-0281-z) contains supplementary material, which is available to authorized users.


CatWalk test:
A detailed gait analysis was performed using the CatWalk XT 10.5 system (Noldus). Transgenic hTau40 AT mice (n=14, mixed genders) were compared to age-matched control littermates (mixed genders, n=14) at the age of 11 months. Briefly, the CatWalk system consists of an enclosed walkway on a glass plate that is traversed by the mouse. Green light enters at the long edge of the plate and is completely internally reflected. In case a mouse's paw contacts the glass surface, light is scattered downwards resulting in a sharp and bright digital image of a paw print. The whole run is tracked by a video system placed under the glass plate. Mice were trained to cross the glass plate in three consecutive trials. The number of compliant runs was n=3 with a minimum duration of 0.5s and a maximum duration of 3s. The following parameters were analyzed: (a) intensity of the paw print (arbitrary units, a.u.), expressing the mean pressure of each paw during the floor contact; (b) stride length (cm), describing the distance between two consecutive paw placements of a particular paw; (c) swing (s),measuring the time between one step and the next for a particular paw; (d) stance (s), indicating the amount of time a particular paw rests on the walkway; (e) step cycle (s), exhibiting the time between two consecutive paw placements (stance + swing duration); (f) duty cycle (%) showing the ratio between stance duration and step cycle duration; (g) base of support (cm) representing the distance between the front or the hind paws; (h) regularity index (RI, %) as a measure of interlimb coordination defined by the percentage of normal step sequence pattern in relation to the total number of paw placements.

Fig. S2: No spine loss in hTau40
AT mice at 10 months of age. (a) Dendritic spine density analyzed by Golgi-staining. Representative images of apical dendrites of CA1-hippocampal neurons of 10 mo old WT and hTau40 AT mice. Scale bar: 3µm. (b) Quantification of (a). No significant differences of the CA1 apical dendritic spine density are detected between 10 months old hTau40 AT mice (red bar) and age-matched WT mice (grey bars) by two-sided t test. Bars show mean ± SEM. n.s, not significant.

Fig. S3: Morris water maze test shows learning and memory deficits of hTau40
AT mice at 16 months of age. (a) While hTau40 AT mice display unaltered learning behavior at 10 months of age, (b) 16 months old hTau40 AT mice exhibit significant learning impairments in comparison to age-matched controls as indicated by increased path length during Morris water maze acquisition (interaction effect, genotype x day, p=0.027, F (4,116) =2.88).

(c-d) Swimming speed does not differ between hTau40
AT and control mice (p>0.05 for all comparisons) at 10 months and 16 months of age, excluding motor disabilities as underlying cause for different learning rates. (a-d) Data shows mean path length (cm) or swim speed (cm/s) ± SEM. Statistics: two-way repeated measure analysis of variances (2-way ANOVA) with post hoc Fishers LSD multiple comparisons test. Asterisks indicate significant differences in path length between hTau40 AT mice and control group (interaction effect, genotype x day), *:p<0.05. (e-f) Bar diagrams show time spent in quadrants (%) for each probe trial of (e) 10 months and (f) 16 months old hTau40 AT mice in comparison to controls. The dotted line indicates the 25% chance level for the mice to choose the correct target region given by the presence of four quadrants (target, right, opposite, left). From probe trial 1 onwards, mice exhibit a highly significant preference for the target and avoidance for the opposite quadrant similar to control mice at 10 months of age (e), whereas the preference of the target quadrant is much less pronounced in hTau40 AT mice at 16 months of age (f), arguing for deficits in short-and long-term memory in old transgenic animals. Bars represent mean values ± SEM. Statistics: two-tailed one sample t-test against chance level of 25%; *:p<0.05; **:p<0.01; ***:p<0.001, ****:p<0.0001. MWM: Morris water maze, WT: wild-type animals; n: number of mice; mo: months; PT: probe trial; LTPT: long-term probe trial; opp.: opposite quadrant.

Fig. S4: Normal motor function upon expression and accumulation of hTau40
AT in brain and spinal cord. (a) A Rotarod test using a constant speed paradigm (16 rpm) and (b) an accelerating speed paradigm (4-40rpm) does not reveal any motor disabilities of hTau40 AT in comparison to control mice at 10 and 16 months of age as scored by the latency to fall from the rotating rod (s). Bars represent group mean values ± SEM. Statistics: unpaired two-tailed Student´s t-test between transgenic mice and controls, p>0.05 for all comparisons. WT: wild-type animals; mo: months of age; n: number of mice; T1: trial 1; T2: trial 2.

Fig. S5: No gait abnormalities due to expression and accumulation of hTau40
AT in brain and spinal cord. (a-h) A detailed Catwalk digital foot print analysis shows no gait abnormalities of hTau40 AT mice in comparison to controls at 12 months of age. No significant differences are detected between transgenic and control mice for parameters related to single paw or inter-limb coordination such as (a) the intensity of the paw print in arbitrary units (a.u.), expressing the mean pressure of each paw during floor contact; (b) the stride length in cm, describing the distance between two consecutive paw placements of a particular paw; (c) the swing in s, measuring the time between one step and the next for a particular paw; (d) the stance in s, indicating the amount of time a particular paw rests on the walkway; (e) the step cycle in s, exhibiting the time between two consecutive paw placements (stance + swing duration); (f) the duty cycle in %, showing the ratio between stance duration and step cycle duration; (g) the base of support in cm, representing the distance between the front or the hind paws; (h) the regularity index in % as a measure of inter-limb coordination defined by the percentage of normal step sequence pattern in relation to the total number of paw placements. Bars represent group mean values ± SEM. Statistics: unpaired two-tailed Student´s t-test between transgenic mice and controls, p>0.05 for all comparisons. WT: wild-type animals; mo: months of age; n: number of mice; FP: front paw; HP: hind paw. AT mice were stained with Iba1. Note phagocytic microglia in the hilus of aged hTau40 AT mice (b, red arrow) compared to resting microglial cells of age-matched WT mice (a). WT: wildtype; A152T: hTau40 AT transgenic mouse strain; mo: months; Scale bar: 50µm (a-b