Abnormal tau induces cognitive impairment through two different mechanisms: synaptic dysfunction and neuronal loss

The hyperphosphorylated microtubule-associated protein tau is present in several neurodegenerative diseases, although the causal relationship remains elusive. Few mouse models used to study Alzheimer-like dementia target tau phosphorylation. We created an inducible pseudophosphorylated tau (Pathological Human Tau, PH-Tau) mouse model to study the effect of conformationally modified tau in vivo. Leaky expression resulted in two levels of PH-Tau: low basal level and higher upon induction (4% and 14% of the endogenous tau, respectively). Unexpectedly, low PH-Tau resulted in significant cognitive deficits, decrease in the number of synapses (seen by EM in the CA1 region), reduction of synaptic proteins, and localization to the nucleus. Induction of PH-Tau triggered neuronal death (60% in CA3), astrocytosis, and loss of the processes in CA1. These findings suggest, that phosphorylated tau is sufficient to induce neurodegeneration and that two different mechanisms can induce cognitive impairment depending on the levels of PH-Tau expression.

Preparation of soluble and insoluble fractions was performed as described with some modifications 2 . The forebrains were homogenized in 10 volumes, 300 l of the homogenates were centrifuged at 14,000 rpm for 20 min and the supernatants were collected as RAB fraction. tested in an open-square white arena, 60*60cm, 40cm high as previously described 5 . The following objects were used: two black metal cylinders, 6*7cm; an orange disk, 1.5*5cm. The task started with a habituation trial, during which the animals were placed in the empty arena for 10 min. The next day, mice were again place in the same arena containing two identical objects (familiarization phase). Exploration was recorded in a 10-min trial. Sniffing, touching, and stretching the head toward the object at a distance of no more than 2cm were scored as object investigation. Six hours later (test phase), mice were again placed in the arena containing two objects: one identical to one of the objects presented during the familiarization phase (familiar object), and a new, different object (novel object). The time spent exploring the two objects was recorded for 10 min. Memory was expressed as a discrimination index, namely (seconds on novel-seconds on familiar)/(seconds on novel+seconds on familiar) and was expressed as the percentage of time on each object.

Morris Water Maze.
Fifteen-month old mice were trained during a 4 day training period and were tested in four trials per day for their ability to locate a hidden platform aided by visual clues surrounding the water pool as previously described 6 . Mice were allowed free access to food and water and maintained at constant temperature (25°C). Spatial memory was measured by Morris water maze (142 cm in diameter, opaque water, 26±1°C, automated swim-path monitoring; HVS Image). Mice were subjected to a training period for 4 days and were then tested a month later for their ability to search for the platform in the same water pool with visual clues, but where the platform had been removed. Hidden platform training is a total of 20 trials/day, 11×11×39 cm hidden platform, placed 30 cm from the wall and 2 cm bellow the water level). At the end of the training, a 60-s probe trial (platform absent) was administered. Through these training sessions, mice acquired spatial memory about location of the safe platform. The time mice stay in the previous platform quadrant (quadrant time) and their swimming path were recorded by a video camera. Videos were subsequently analyzed by AmyMaze software. The longer a mouse stays in the previous platform-located quadrant, the better it scores spatial memory. The spatial memory of a mouse was expressed as quadrant time (%).
Passive avoidance task. Using passive avoidance paradigm, we tested the ability of mice to learn to avoid an electrical shock as previously described 7 . Mice received six trials a day for five days and measured their learning after repetitive training. The apparatus has a bright and a dark compartment with a computer-controlled door between them. The delivery of electric shocks and the raising and lowering of the door and the latencies at which the animals stepped into the dark from the bright compartment were controlled by the computer. Each animal was gently placed in the light compartment for 10 s, after which the guillotine door was raised and the time the animal waited before crossing to the dark (shock) compartment was recorded as the latency. The trial ended when an animal waited more than 180s to cross to the other side, or if it received an electrical shock in the dark side after crossing. Once the animal crossed with all four paws to the next compartment, the door was closed and a 1.5 mA foot shock was delivered for 5 s. Mice that showed immobility from the previous experiment were excluded from this test.
The chambers were cleaned carefully between animals with a tap water solution containing ethanol (5% v/v).
Training sessions: each mouse was trained by gently placing it in the light compartment then when it stepped through the dark compartment putting the 4 paws on the grid floor, the door automatically closed and an electric shock of 0.5 mA was delivered for 3s. Repeat the training 6 times per day, with an upper cut-off time of 180s.
Retention session: each mouse was introduced to the light compartment and the latency to step-through to the dark compartment was recorded as a passive avoidance behavior indicating memory retention, with an upper cut-off time of 180s. No electric shock was delivered during retention session.
Statistical Analysis. All data are presented as average ± SEM. Statistical analysis was performed using STATISTICA software. N = 3-11 mice/group were used for each experiment.
P-values less than or equal to 0.05 were considered statistically significant.