Presenilin-1 familial Alzheimer's disease mutation alters hippocampal neurogenesis and memory function in CCL2 null mice

Abstract

Aberrations in hippocampal neurogenesis are associated with learning and memory, synaptic plasticity and neurodegeneration in Alzheimer's disease (AD). However, the linkage between them, β-amyloidosis and neuroinflammation is not well understood. To this end, we generated a mouse overexpressing familial AD (FAD) mutant human presenilin-1 (PS1) crossed with a knockout (KO) of the CC-chemokine ligand 2 (CCL2) gene. The PS1/CCL2KO mice developed robust age-dependent deficits in hippocampal neurogenesis associated with impairments in learning and memory, synaptic plasticity and long-term potentiation. Neurogliogenesis gene profiling supported β-amyloid independent pathways for FAD-associated deficits in hippocampal neurogenesis. We conclude that these PS1/CCL2KO mice are suitable for studies linking host genetics, immunity and hippocampal function.

Keywords: Chemokine; Hippocampus; Long-term potentiation; Morris water maze; Neurogliogenesis.

Fig. 1

Overexpression of a PS1 FAD mutant in CCL2-null mice results in fewer newborn neurons in the DG of the hippocampus. (A, B) Immunohistochemical detection of Dcx-labeled cells in the DG from 4-month-old (A) and 7-month-old (B) mice. Scale bar, 100 µm. (C, D) Quantification of the number of Dcx-labeled cells in the DG from 4-month-old (C) and 7-month-old (D) mice (n = 7 mice per group, 12 sections per mouse, P = 0.6728 (C) and 0.0016 (D)). Data are presented as mean ± SEM ^aa,bbP < 0.01 vs non-Tg or PS1, ^cP < 0.05 vs CCL2KO, one-way ANOVA, Newman-Keuls post hoc test.

Fig. 2

Overexpression of a PS1 FAD mutant in CCL2-null mice results in reduced adult neurogenesis in the dentate SGZ. (A) Immunofluorescent detection of BrdU and NeuN double-labeled cells in the SGL from 7-month-old mice. Scale bar, 50 µm. (B) Quantification of the number of BrdU and NeuN double-labeled cells in the SGL from 7-month-old mice after three-weeks BrdU labeling (n = 7 mice per group, 12 sections per mouse, P = 0.0051). (C) No difference in BrdU incorporation between the transgenic lines. NPCs were isolated from all animal groups and cultured with BrdU. Immunofluorescence using FITC-conjugated anti-BrdU antibody was performed. Quantification of BrdU fluorescent density shows consistent proliferation rate between all lines (n = 3 cultures per each group, P = 0.7462 as determined by two-way ANOVA, Bonferroni post hoc test). (D, E) Immunofluorescence of MAP2/GFAP (D) or TuJ1/s100β (E)-positive cells differentiated from NPCs. Scale bars, 100 µm. (F–I) Quantification of MAP2 (F), GFAP (G), TuJ1 (H) or s100β (I)-positive cells. Data are presented as the percentage of total cells in in vitro cultivation (n = 3 cultures per each group from 3 independent experiments, P = 0.0138 (F), 0.0137 (G), 0.0386 (H) or 0.0056 (I)). (J) Immunofluorescent detection of BrdU and s100β double-labeled cells in the SGL from 7-month-old mice. Scale bar, 50 µm. (K) Quantification of the number of BrdU and s100β double-labeled cells in the DG from 7-month-old mice after three-weeks BrdU labeling (n = 7 mice per group, 12 sections per mouse, P = 0.0015). Data are presented as mean ± SEM ^a,b,cP < 0.05 vs non-Tg, PS1 or CCL2KO, ^{aa, bb,cc}P < 0.01 vs non-Tg, PS1 or CCL2KO, one-way ANOVA, Newman-Keuls post hoc test.

Fig. 3

Morris water maze (MWM) tasks reveal impairments in memory acquisition and retention in PS1/CCL2KO mice. (A–D) MWM task at 2 months of age shows no difference in learning and memory between the transgenic lines at both MWM acquisition phase (A) where escape latency during a 1-minute trial was measured, and MWM retention phase (B, C). During a one-time probe trial after memory acquisition, number of annulus crossings (B) and duration spent in the goal quadrant between groups (C) were analyzed. (D) Measurement of average swimming speed of animals. Average swimming speeds are unchanged between groups. (E) MWM acquisition phase at 4 months of age (n = 11 (non-Tg), 10 (CCL2KO), 12 (PS1, PS1/CCL2KO)). PS1/CCL2KO mice show an initial trend to increase escape latencies as compared to other groups after day 6. (F, G) MWM retention phase. Number of annulus crossings is fewer in PS1/CCL2KO mice as compared to other groups (F), but no significant difference in duration spent in the goal quadrant between groups (G). (H) Measurement of average swimming speed of animals at 4 months of age. (I) MWM acquisition phase at 7 months of age (n = 10 (non-Tg, PS1, PS1/CCL2KO), 8 (CCL2KO)). PS1/CCL2KO mice show statistically higher escape latencies as compared to other groups. (J, K) MWM retention phase at 7 months of age. Both number of annulus crossings (J) and percent search time (K) are statistically fewer as compared to other groups. (L) Measurement of average swimming speed of animals at 7 months of age. Data are presented as mean ± SEM ^a,b,cP < 0.05, ^aa,bb,ccP < 0.01, ^aaa,bbbP < 0.001, ^a,aa,aaa vs non-Tg, ^b,bb,bbb vs PS1, ^c,cc vs CCL2KO, two-way ANOVA, Bonferroni post hoc (A, E, I) or one-way ANOVA, Newman-Keuls post hoc (F, J, K) test.

Fig. 4

Overexpression of a PS1 FAD mutant in CCL2-null mice leads to changes in hippocampal gene expression related to neurogenesis and synaptic function. (A) Heat map depicting fold changes in neurogenesis-related genes in the hippocampus between non-Tg and PS1/CCL2KO mice (n = 4 mice per group). Shades of green and magenta show down-regulated and up-regulated genes, respectively. (B–D) A conventional RT2-qPCR was performed to measure Nog (B), Fgf2 (C) and Neurod1 (D) expression using primer sets and synthesized cDNA with total RNA isolated from the hippocampus of each mouse (n = 4 mice per group, P = 0.0392 (B), 0.0007 (C) and 0.0029 (D)). Data are presented as mean ± SEM ^a,b,cP < 0.05, ^aa,bb,ccP < 0.01, ^aaaP < 0.001, ^a,aa,aaa vs non-Tg, ^b,bb vs PS1, ^c,cc vs CCL2KO, one-way ANOVA, Newman-Keuls post hoc test.

Fig. 5

Overexpression of a PS1 FAD mutant in CCL2-null mice affects expression of learning and memory-related molecules. (A) Immunohistochemical detection of c-fos-labeled cells in the dentate GCL at 7 months of age. Scale bar, 100 µm. (B) Calbindin expression in the DG of the hippocampus at 7 months of age. Scale bar, 200 µm. (C, D) Quantification of the number of c-fos-labeled cells (C, n = 7 mice per group, 12 sections per mouse, P = 0.0092) and calbindin expression levels (D, n = 7 mice per group, 10 sections per mouse, P = 0.0046). Data are presented as mean ± SEM ^a,b,cP < 0.05, ^aaP < 0.01, ^a,aa vs non-Tg, ^b vs PS1, ^c vs CCL2KO, one-way ANOVA, Newman-Keuls post hoc test.

Fig. 6

LTP measures in PS1/CCL2KO mice. (A) Time courses and average magnitudes of LTP recorded in the CA1 region of hippocampal slices prepared from non-Tg (□), PS1 (▲), CCL2KO (∇) and PS1/CCL2KO (●) animals. The graph plots the initial slope of the evoked fEPSPs in response to constant current stimuli. HFS (100 Hz, 500ms × 2) was delivered at the time indicated by an arrow. Each point in this graph represents an average of 6 slices. The representative fEPSPs taken at 10 minutes before and at 50 minutes after HFS (as indicated by letters a and b) in one of the non-Tg (left) and one of the PS1/CCL2KO (right) slices are shown below the LTP time course. Note that HFS produced a significant increase of fEPSP in non-Tg slice, but not in PS1/CCL2KO slice. (B) A bar graph showing the average LTP magnitudes measured from 50 to 60 min after HFS. Note a significant reduction of LTP magnitude in the hippocampal slices prepared from PS1/CCL2KO animals. Data are presented as mean ± SEM *P < 0.05 vs non-Tg, two-tailed t-test.