Cognitive Decline and BPSD Are Concomitant with Autophagic and Synaptic Deficits Associated with G9a Alterations in Aged SAMP8 Mice

Abstract

Behavioural and psychological symptoms of dementia (BPSD) are presented in 95% of Alzheimer's Disease (AD) patients and are also associated with neurotrophin deficits. The molecular mechanisms leading to age-related diseases are still unclear; however, emerging evidence has suggested that epigenetic modulation is a key pathophysiological basis of ageing and neurodegeneration. In particular, it has been suggested that G9a methyltransferase and its repressive histone mark (H3K9me2) are important in shaping learning and memory by modulating autophagic activity and synaptic plasticity. This work deepens our understanding of the epigenetic mechanisms underlying the loss of cognitive function and BPSD in AD. For this purpose, several tasks were performed to evaluate the parameters of sociability (three-chamber test), aggressiveness (resident intruder), anxiety (elevated plus maze and open field) and memory (novel object recognition test) in mice, followed by the evaluation of epigenetic, autophagy and synaptic plasticity markers at the molecular level. The behavioural alterations presented by senescence-accelerated mice prone 8 (SAMP8) of 12 months of age compared with their senescence-accelerated mouse resistant mice (SAMR1), the healthy control strain was accompanied by age-related cognitive deficits and alterations in epigenetic markers. Increased levels of G9a are concomitant to the dysregulation of the JNK pathway in aged SAMP8, driving a failure in autophagosome formation. Furthermore, lower expression of the genes involved in the memory-consolidation process modulated by ERK was observed in the aged male SAMP8 model, suggesting the implication of G9a. In any case, two of the most important neurotrophins, namely brain-derived neurotrophic factor (Bdnf) and neurotrophin-3 (NT3), were found to be reduced, along with a decrease in the levels of dendritic branching and spine density presented by SAMP8 mice. Thus, the present study characterizes and provides information regarding the non-cognitive and cognitive states, as well as molecular alterations, in aged SAMP8, demonstrating the AD-like symptoms presented by this model. In any case, our results indicate that higher levels of G9a are associated with autophagic deficits and alterations in synaptic plasticity, which could further explain the BPSD and cognitive decline exhibited by the model.

Keywords: BPSD; G9a; SAMP8; aggressivity; autophagy; cognitive decline; epigenetics.

Figure 1

Results of the elevated plus maze (EPM) and open field (OF) tests in male SAMR1 and SAMP8 mice at 12 months of age. For EPM: time spent in closed arms (A), time spent in centre (B) and number of rearings (C). For OF: total distance (D), number of rearings (E) and time spent in the border zone (F). Results of the three-chamber test (TCT) in aged SAMR1 and SAMP8 male mice. Habituation (G), preference for chambers (H) and exploration time (I). Bars show the mean ± SEM (for EPM and OF: n = 11–12 mice per group; for TCT n = 13–10 mice per group). Statistical tests: t-Student analysis (A–G,I); two-way ANOVA followed by Šídák’s post hoc test (H). * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001.

Figure 2

Results of the resident–intruder test (RI) in male SAMR1 and SAMP8 mice at 12 months of age. Attack latency (A), number of attacks (B), number of offensive uprights (C), number of lateral threats (D), time in aggression (E) and number of rearings (F). Bars show the mean ± SEM (For RI: n = 11–12 mice per group). Statistical test: t-Student analysis. ** p < 0.01; *** p < 0.001; **** p < 0.0001. Polygonal graph presenting a summary of the relevant behavioural parameters (G).

Figure 3

Summary of NORT for short-term memory (A). Bars show the mean ± SEM (for NORT: n = 12 mice per group). Statistical test: t-Student analysis. *** p < 0.001. Correlations between the time exploring mice indicating sociability (B), attack latency (C), distance travelled in the central zone (D) and time spent in open arms (E) and the discrimination index (DI). Pearson’s correlations were calculated for the time exploring mice (sociability), distance travelled in the central zone and time spent in open arms, and Spearman’s correlation was calculated for attack latency in both groups (n = 10–12 mice per group). r and p-values are indicated on the graphs.

Figure 4

Representative gene expression for Ehmt2 (A). Gene expression levels were determined by real-time PCR. Representative global H3 modification patterns’ quantification (B). Bars show the mean ± SEM (n = 4–6 mice per group); * p < 0.05; ** p < 0.01. Representative WBs (C) and quantification for the ratio of p-JNK/JNK total (D), Beclin-1 (E), p-62 (F) and the ratio of LC3I/LC3II (G). Values in the bar graphs are adjusted to 100% for the protein levels of the SAMR1 group. Bars show the mean ± SEM (n = 3–5 mice per group). Statistical test: t-Student analysis. * p < 0.05; ** p < 0.01.

Figure 5

Representative WB (A) and quantification for the ratio of p-ERK/ERK total (B). Values in the bar graphs are adjusted to 100% for the protein levels of the SAMR1 group. Bars show the mean ± SEM; * p < 0.05. Representative gene expression levels of Dnmt3a (C), Comt (D), Zif-268 (E) and cFos (F). Gene expression levels were determined by real-time PCR. Bars show the mean ± SEM (n = 4–8 mice per group). Statistical test: t-Student analysis. * p < 0.05; **** p < 0.0001.

Figure 6

Representative gene expressions for NT3 (A), Bdnf (B), Ngf (C), TrkA (D) and TrkB (E). The gene expression levels were determined by real-time PCR. Bars show the mean ± SEM (n = 4–8 mice per group). Statistical test: t-Student analysis. * p < 0.05; *** p < 0.001. Representative images and tracings of Golgi-stained cortical neurons from the aged SAMR1 male mice group (top) and the age SAMP8 male mice group (bottom) (scale bar = 100 µm) (F). Aged SAMP8 male mice showed a reduction in the number of neuronal intersections compared with the SAMR1 control group through a Sholl analysis (G). Representative images of the spine density of the aged SAMR1 male mice group (top) and the aged SAMP8 male mice group (bottom) by Golgi staining (H) and quantification (I). Bars show the mean ± SEM (n= 28–30 neurons from 6 different mice per group). Statistical test: t-Student analysis. **** p < 0.0001.