The Neuroprotective Effect of Neural Cell Adhesion Molecule L1 in the Hippocampus of Aged Alzheimer's Disease Model Mice

Abstract

Alzheimer's disease (AD) is a severe neurodegenerative disorder and the most common form of dementia, causing the loss of cognitive function. Our previous study has shown, using a doubly mutated mouse model of AD (APP/PS1), that the neural adhesion molecule L1 directly binds amyloid peptides and decreases plaque load and gliosis when injected as an adeno-associated virus construct (AAV-L1) into APP/PS1 mice. In this study, we microinjected AAV-L1, using a Hamilton syringe, directly into the 3-month-old APP/PS1 mouse hippocampus and waited for a year until significant neurodegeneration developed. We stereologically counted the principal neurons and parvalbumin-positive interneurons in the hippocampus, estimated the density of inhibitory synapses around principal cells, and compared the AAV-L1 injection models with control injections of green fluorescent protein (AAV-GFP) and the wild-type hippocampus. Our results show that there is a significant loss of granule cells in the dentate gyrus of the APP/PS1 mice, which was improved by AAV-L1 injection, compared with the AAV-GFP controls (p < 0.05). There is also a generalized loss of parvalbumin-positive interneurons in the hippocampus of APP/PS1 mice, which is ameliorated by AAV-L1 injection, compared with the AAV-GFP controls (p < 0.05). Additionally, AAV-L1 injection promotes the survival of inhibitory synapses around the principal cells compared with AAV-GFP controls in all three hippocampal subfields (p < 0.01). Our results indicate that L1 promotes neuronal survival and protects the synapses in an AD mouse model, which could have therapeutic implications.

Keywords: APP/PS1 mice; Alzheimer’s disease; GABAergic interneurons; adhesion molecule L1; hippocampus; synapses.

Figure 1

(A) Schematic depiction of the experimental design. (B) Representative image of the APP/PS1 mouse hippocampus injected with AAV-GFP (green) and immunofluorescently stained for NeuN (red). The overlay (yellow) highlights the transduced cells. (C) Higher magnification of the CA1 hippocampal subfield, with GFP (green) transduced pyramidal cells. (D,E). Immunostaining with L1 555 antibody (green) in wild-type (D) and AAV-L1 injected (E) hippocampus cells. Note the robust transduction of the CA1 pyramidal cells with AAV-L1. CA—cornu ammonis, DG—dentate gyrus, pyr—stratum pyramidale, rad—stratum radiatum. Scale bars: 200 µm (B); 25 µm (C–E).

Figure 2

Injection of AAV-L1 reduces the loss of hippocampal NeuN-positive granule neurons in the dentate gyrus of APP/PS1 mice. (A,C,E) Representative images of NeuN-immunostained (NeuN+) neurons in the CA1 (A), CA3 (C), and DG (E) subfields of the hippocampus. Or—stratum oriens, pyr—stratum pyramidale, rad—stratum radiatum, gr—stratum granulosum, hil—hilus of the dentate gyrus. Scale bar: 50 µm. (B,D,F) Densities of the hippocampal NeuN-positive neurons in the pyramidal layer of the CA1 (B), CA3 (D), and granule cells in the DG (F) in wild-type (WT), AAV-L1, and AAV-GFP-injected APP/PS1 mice. Data are shown as mean + standard deviation. Asterisks indicate the differences between treatments, while the hashtag indicates a difference from the wild-type control; one-way ANOVA with Holm–Sidak post hoc, p < 0.05; n = 5 mice/group.

Figure 3

Injection of AAV-L1 reduces the loss of hippocampal parvalbumin-positive interneurons in APP/PS1 mice. (A,C,E) Representative images of parvalbumin-immunostained (PV+) interneurons in the CA1 (A), CA3 (C), and DG (E) subregions of the hippocampus. Or—stratum oriens, pyr—stratum pyramidale, rad—stratum radiatum, gr—stratum granulosum, mol—stratum moleculare, hil—hilus of the dentate gyrus. Scale bar: 25 µm. (B,D,F) Densities of PV+ neurons in the CA1 (B), CA3 (D), and DG (F) in the wild-type (WT) and AAV-L1 or AAV-GFP-injected APP/PS1 mice. Data are shown as mean + standard deviation. Asterisks indicate the difference between treatments, the hashtag indicates a difference from the wild-type control; one-way ANOVA with Holm–Sidak post hoc, p < 0.05; n = 5 mice/group.

Figure 4

The loss of perisomatic inhibitory synapses on principal neuron cell bodies in the hippocampus of APP/PS1 mice is reduced by AAV-L1 injection. (A,C,E) Representative confocal micrographs of VGAT-(red) and parvalbumin (PV, green)-immunostained perisomatic terminals around CA1 (A), CA3 (C), pyramidal neurons (pyr) and DG granule cells (gr) (E). Scale bar: 10 µm. (B,D,F) Diagrams represent the number of parvalbumin-positive/VGAT-positive (PV+) and parvalbumin-negative/VGAT-positive (PV-) perisomatic terminals per unit length (mm) in the CA1 (B), CA3 (D), and DG (F) of wild-type (WT) mice, and APP/PS1 mice injected with either AAV-L1 or AAV-GFP. Data are shown as mean + SD. Asterisks indicate a difference between the injections (AAV-L1 or AAV-GFP), hashtags indicate a difference between APP/PS1 mice and the wild-type control, in a two-way ANOVA with the factors “parvalbumin expression” and “viral injection”, followed by the Holm–Sidak post hoc method, p < 0.05; n = 5 mice/group.