A touch screen-automated cognitive test battery reveals impaired attention, memory abnormalities, and increased response inhibition in the TgCRND8 mouse model of Alzheimer's disease

Abstract

Transgenic mouse models of Alzheimer's disease (AD) with abundant β-amyloid develop memory impairments. However, multiple nonmnemonic cognitive domains such as attention and executive control are also compromised early in AD individuals, but have not been routinely assessed in animal models. Here, we assessed the cognitive abilities of TgCRND8 mice-a widely used model of β-amyloid pathology-with a touch screen-based automated test battery. The test battery comprises highly translatable tests of multiple cognitive constructs impaired in human AD, such as memory, attention, and response control, as well as appropriate control tasks. We found that familial AD mutations affect not only memory, but also cause significant alterations of sustained attention and behavioral flexibility. Because changes in attention and response inhibition may affect performance on tests of other cognitive abilities including memory, our findings have important consequences for the assessment of disease mechanisms and therapeutics in animal models of AD. A more comprehensive phenotyping with specialized, multicomponent cognitive test batteries for mice might significantly advance translation from preclinical mouse studies to the clinic.

Fig. 1

Automated spontaneous object recognition. (a) Mean total touches to novel and familiar object during the choice phase for each genotype and delay. (b) Discrimination ratio based on object touches (difference between novel object touches and familiar object touches divided by total object exploration (novel object touches − familiar object touches)/(novel object touches + familiar object touches)). Thus, a score of 0 corresponds to no preference for the novel object, whereas a score of 1 corresponds to exploration of the novel object only. (c) Mean total approaches to novel and familiar object during the choice phase for each genotype and delay. (d) Discrimination ratio based on object approaches. The discrimination ratio was calculated from the total number of approaches as in (b): ((novel object approaches − familiar object approaches)/(novel object approaches + familiar object approaches)). Data are presented as mean ± standard error of the mean. * p < 0.05.

Fig. 2

Visual discrimination, reversal, and retention. (a) Task acquisition: sessions required to reach criterion (> 80% correct responses on 2 consecutive days). (b) Performance levels (choice accuracy) during 3 sessions of baseline testing (bl1–bl3), after reversing the task contingencies (the previous S+ became S−, and the previous S− the S+), and (c) after a 10 day, test-free interval (r1 and r2). Data are presented as mean ± standard error of the mean.

Fig. 3

Baseline performance on the 5-choice serial reaction time task (5-CSRTT) with a 2-second stimulus duration. (a) Response accuracy excluding omitted trials. (b) Percentage of omitted trials. (c) Percentage of premature responses before appearance of the stimulus. (d) Mean total perseverative responses per session. (e) Mean latency of response. (f) Mean latency of reward collection after a correct response. Data are presented as mean ± standard error of the mean.

Fig. 4

Five-choice serial reaction time task (5-CSRTT) performance on probe trials with 1.5-, 1-, 0.8-, and 0.6-second stimulus duration. (a) Response accuracy excluding omitted trials. (b) Percentage of omitted trials. (c) Percentage of premature responses before appearance of the stimulus. (d) Mean total perseverative responses per session. (e) Mean latency of response. (f) Mean latency of reward collection after a correct response. Data are presented as mean ± standard error of the mean. * Simple main effect, p < 0.05.

Fig. 5

Memory extinction. Percentage of stimuli mice touched during each session (responding no longer led to a food reward). Data are presented as mean ± standard error of the mean.