Early Long-Term Memory Impairment and Changes in the Expression of Synaptic Plasticity-Associated Genes, in the McGill-R-Thy1-APP Rat Model of Alzheimer's-Like Brain Amyloidosis

Abstract

Accruing evidence supports the hypothesis that memory deficits in early Alzheimer Disease (AD) might be due to synaptic failure caused by accumulation of intracellular amyloid beta (Aβ) oligomers, then secreted to the extracellular media. Transgenic mouse AD models provide valuable information on AD pathology. However, the failure to translate these findings to humans calls for models that better recapitulate the human pathology. McGill-R-Thy1-APP transgenic (Tg) rat expresses the human amyloid precursor protein (APP751) with the Swedish and Indiana mutations (of familial AD), leading to an AD-like slow-progressing brain amyloid pathology. Therefore, it offers a unique opportunity to investigate learning and memory abilities at early stages of AD, when Aβ accumulation is restricted to the intracellular compartment, prior to plaque deposition. Our goal was to further investigate early deficits in memory, particularly long-term memory in McGill-R-Thy1-APP heterozygous (Tg+/-) rats. Short-term- and long-term habituation to an open field were preserved in 3-, 4-, and 6-month-old (Tg+/-). However, long-term memory of inhibitory avoidance to a foot-shock, novel object-recognition and social approaching behavior were seriously impaired in 4-month-old (Tg+/-) male rats, suggesting that they are unable to either consolidate and/or evoke such associative and discriminative memories with aversive, emotional and spatial components. The long-term memory deficits were accompanied by increased transcript levels of genes relevant to synaptic plasticity, learning and memory processing in the hippocampus, such as Grin2b, Dlg4, Camk2b, and Syn1. Our findings indicate that in addition to the previously well-documented deficits in learning and memory, McGill-R-Thy1-APP rats display particular long-term-memory deficits and deep social behavior alterations at pre-plaque early stages of the pathology. This highlights the importance of Aβ oligomers and emphasizes the validity of the model to study AD-like early processes, with potentially predictive value.

Keywords: Alzheimer disease; Camk2b; Grin2b; amyloid beta-precursor protein; cognitive dysfunction; long-term memory; neuronal plasticity; social behavior alterations.

Figure 1

Exploration and habituation to an OF. McGill-R-Thy1-APP Tg+/– male rats (dark gray bars) and their wt littermates (white bars) were trained and tested for short-term (ST, intra-session) habituation to an OF at (A) 3, (B) 4, and (C) 6-months of age. Bar diagram shows median and interquartile ranges (25;75) for the number of crossings recorded during each minute of a 5 min session (training, Tr). Significant differences were assessed by Wilcoxon matched pairs signed rank test; *P < 0.05, **P < 0.01, ***P < 0.001 (3 mo: N_wt = 8, N_Tg+/− = 11; 4 mo: N_wt = 8, N_Tg+/− = 11; 6 mo: N_wt = 11, N_Tg+/− = 10). (D) Long-term (LT) habituation to an OF (LTM). Bar diagram shows median and interquartile ranges (25;75) for the total number of crossings recorded during a 5 min training session (Tr, plain bars) and during the test session (Te, stripped bars) which was performed 24 h later (schematic representation of the task steps on top). No significant differences were observed between any selected pair of training sessions amongst the wt or Tg+/– animals (Unmatched Kruskal-Wallis test with Dunn's multiple comparisons). Total crossings during test session were significantly lower than during training session for the three different ages, either for wt or Tg animals. Statistically significant differences were assessed by Wilcoxon matched pairs signed rank test; *P < 0.05, **P < 0.01 (3 mo: N_wt = 8, N_Tg+/− = 8; 4 mo: N_wt = 8, N_Tg+/− = 9; 6 mo: N_wt = 11, N_Tg+/− = 8).

Figure 2

Long-term memory (LTM) of inhibitory avoidance (IA) in 3-, 4-, and 6-month-old McGill-R-Thy1-APP Tg+/– rats. (A) IA task: Training and Test Latencies. Tg+/– (dark gray bars) and wt littermates (white bars) were trained in a step through IA to a mild foot-shock (0.5 mA) and tested 24 h later. Latencies to enter the dark compartment (from a lighted one) were recorded. The bar diagram represents median of latencies and interquartile ranges (25;75). Test latencies (Te, stripped bars) to enter the dark compartment were significantly higher than training latencies (Tr, plain bars) for 3, 4, and 6-month-old wt rats. However, only 3-month-old Tg+/– rats showed a significant difference between Te and Tr latencies. Neither 4- nor 6-month-old Tg rats showed differences between Te and Tr latencies. Significant differences were assessed by Wilcoxon matched pairs signed rank test; *P < 0.05, **P < 0.01, ***P < 0.001. (B). IA task: (Te – Tr) Latencies Difference. Bar diagram represents median of (Te-Tr) difference of latencies, with interquartile ranges (25;75) for the same groups as in A. Differences were significantly lower in Tg+/– (stripped bars) than in wt (plain bars) rats of 4 (light gray) and 6 (dark gray) months, whilst there were no statistically significant differences among the difference of latencies by genotype in 3-month-old rats (Unmatched Kruskal-Wallis test with Dunn's multiple comparisons. *P < 0.05, ***P < 0.001). Tt-Tr latency difference of 3-month-old Tg+/– rats was significantly different form zero, while latency differences for 4- and 6-month-old Tg+/– rats were not different from zero. Unmatched Kruskal-Wallis test with Dunn's multiple comparisons. *P < 0.05) (3 mo: N_wt = 8, N_Tg+/− = 7; 4 mo: N_wt = 7, N_Tg+/− = 11; 6 mo: N_wt = 11, N_Tg+/− = 13). (C) Tail-Flick latency. Tail withdrawal latency from a hot (51 ± 1°C) water bath is represented in the scatter plot. No significant differences were observed between any selected pair of animals (Tg or wt) of a given age (Unpaired Mann-Whitney test).

Figure 3

Novel object recognition (NOR) task performance by 4-month-old McGill-R-Thy1-APP Tg+/– rats (A) Schematic representation of the task steps. Habituation: Rats were left to freely explore an OF for habituation to the environment (10 min/session, along 3 days). Training (Tr): On the 4th day, each rat was left to re-explore the OF in the presence of two identical objects (A–A′), for 5 min. Test (Te): On the 5th day, trained rats were split in two groups: Rats from one group were exposed for 5 min to the same two objects (A–A′) (as internal control at 24 h), while rats from the other group, to a familiar and a novel object (A, B). (B) Training. Latency to start exploring an object. Bars represent latencies as mean ± SEM for each group. There were no significant differences between Tg+/– (dark gray bars, N_Tg+/− = 19) and wt (white bars, N_wt = 16) rats (Student's t-Test). (C) Training. Total time spent exploring objects (A+A′) by Tg/+- and wt rats during the 5 min Tr session. No significant differences were found between groups (Student's t-Test). (D) Training. Time spent exploring each object (A or A′). There were no significant differences when comparing the time spent exploring A and A′ (Paired-t-test). (E) Test. On the 5th day, when tested in the presence of a novel (B) and a familiar (A) object (A-B, striped-bars), wt rats showed a significantly higher Discrimination Index in the Test session: [DI_Te (A-B) = (time exploring B - time exploring A)/(time exploring B + time exploring A)], compared to the Discrimination Index of Training [DI_Tr (A-A′) = (time exploring A′ - time exploring A)/(time exploring A′ + time exploring A)] (plain bars) or against zero. At variance, DI_Te (A-B) for Tg+/– rats was similar to DI_Tr (A-A′)—and it was not significantly different from zero. The two other groups of Tg+/– and wt rats were tested with same objects (A-A′) used in the Tr session, as an internal control after 24 h of Tr. In this case, the DI_Te (A-A′) for Tg+/– as well as that for wt rats were similar to the respective DI_Tr. One-way ANOVA, Dunnett post-hoc test; ***P < 0.001 [N_wt Te (A-A′) = 8; N_wt Te (A-B) = 8; N_Tg+/− Te (A-A′) = 9; N_Tg+/− Te (A-B) = 10].

Figure 4

Social approach behavior of McGill-R-Thy1-APP Tg+/– rats. (A) Schematic representation of the task steps. Tg+/– 4-month-old rats and their wt littermates were tested in an adapted social interaction-three chambered task. Step 1. Habituation: Erat was left to freely explore an OF for 5 min. Step 2. “Training”: Exploration of the OF with two empty cages added. Erat was left to explore the same OF now containing cages A and B, for 5 min, and the time Erat spent exploring each cage was recorded. Step 3. “Sociability”: An unknown novel rat (Nrat, same sex, and age as Erat) was placed inside one cage, and an unknown object (similar size and color as the Nrat) was placed in the other cage. The time Erat spent exploring either the Nrat or the object over 5 min was recorded. Step 4. “Preference for Social Novelty”: 10 min after Step 3, Erat was left to explore the OF with the already known Nrat in one cage and a novel N'rat (same sex and age of Nrat), in the other. Time Erat spent exploring either N or N'rat was recorded. (B) Latency to start exploring empty cages (A or B) (Step 2 of scheme in A) was similar for Tg+/– (dark gray bars, N_Tg+/− = 12) and wt (white bars, N_wt = 10) rats (bars represent mean latency ± SEM; Student t-test). (C) Time Erat spent exploring A or B empty cages. There were no significant differences between Tg+/– and wt rats (bars represent mean exploration time ± SEM; Paired t-test). (D) Number of visits either to A or B cages were not significantly different regardless the genotype (bars represent mean number of visits ± SEM; Paired t-test). (E) Preference Index (PI) for the three phases of the task: “Training” (control for location/cage (lack of) preference), “Sociability” and “Social Novelty.” “Training” (Step 2 of scheme in A). PI = A-B/A+B was not significantly different from zero for both genotypes, indicating that wt and Tg+/– Erats spent similar time exploring each cage, without any location/cage preference. 1st Test: “Sociability” (Step 3 of scheme in A). Both Tg+/– and wt Erats spent significantly more time exploring the Nrat than the object ('Sociability: PI_{Nrat−0bject} = Nrat-Object/Nrat+Object > 0). 2nd Test: “Preference for Social Novelty” (Step 4 in A). Wt rats explored significantly longer time the N'rat than the Nrat (“Preference,” ${\mathrm{PI}}_{wt\text{\_}{N}^{\prime }rat-Nrat}$ = N'rat-Nrat/N'rat+Nrat > 0), while Tg+/– rats spent similar time exploring the N'rat and the Nrat (“Preference,” ${\mathrm{PI}}_{Tg+/-\text{\_}{N}^{\prime }rat-Nrat}$ ≈ 0). ***P < 0.001, *P < 0.05, #P < 0.05; One-way ANOVA, Dunnett post-hoc test.

Figure 5

Social approach comparative behavior of McGill-R-Thy1-APP Homozygous Tg+/+ and Heterozygous Tg+/– rats. (A) Schematic representation of the task steps. 4-month-old Tg+/+, Tg+/– rats and their wt littermates were tested in an adapted version of the three-chambered task. Step 1: OF habituation. Step 2: Control for location/cages (lack of) preference. Erat explores the OF with A and B empty cages. Step 3: “Sociability” assessment. Erat explores the OF with a Nrat (novel rat, same sex, and age) in one cage, and an object (similar size and color than Nrat) in the other cage. (B) “Sociability”: Total exploration time of targets (Nrat or Object) during the 5 min session (Step 3 in A). Bar diagram represents the time (mean ± SEM) an Erat spent exploring an unknown Nrat, or an unknown object (O). Tg+/– animals (dark gray bars, N_Tg+/− = 13), Tg+/+ animals (light gray bars, N_Tg+/+ = 6), wt littermates (white bars, N_wt = 6). Object (O): stripped bars; Novel animal (Nrat): plain bars. Tg+/– and wt rats spent more time exploring a Nrat than the object; while Tg+/+ rats spent significantly more time exploring the object (***P < 0.001, **P < 0.01, Paired t-test). (C) Sociability Preference Index (PI = Nrat–O/Nrat+O) was calculated using the data shown in B. During the social task, wt (white bars) and Tg+/– (dark gray bars) Erats spent more time exploring the Nrat than the object (PI_N−0 > 0; plain bars), while Tg+/+ Erats spent more time exploring the object than the Nrat (PI_N−0 < 0; stripped bars). ****P < 0.0001, One-way ANOVA, Dunnett post- hoc test.

Figure 6

Expression of synaptic plasticity-associated genes. Total RNA was extracted from the hippocampus of (A) 3-month-old (N_wt = 7; N_Tg+/− = 4), (B) 4-month-old (N_wt = 4; N_Tg+/− = 4) and (C) 6-month-old (N_wt = 6; N_Tg+/− = 5) McGill-R-Thy1-APP Tg+/– male rats and their wt littermates. The relative expression of each transcript was calculated using the ΔΔCT method and standardized against the housekeeping genes 18S, GAPDH, HPRT, and β-actin. The name of the corresponding protein product is also indicated. Statistically significant differences were assessed by a two-tailed, unpaired t-test; *P < 0.05.