View Full Text | Abstract |
Article as PDF | Print this Article |
Pubmed | PMC |
PubReader | Export to Citation |
Email Alerts | Open Access |
Exp Neurobiol 2023; 32(1): 42-55
Published online February 28, 2023
https://doi.org/10.5607/en22044
© The Korean Society for Brain and Neural Sciences
Hyun-ju Lee1*, Jin-Hee Park1,2, Justin H. Trotter3, James N. Maher4, Kathleen E. Keenoy4, You Mi Jang1, Youngeun Lee5, Jae-Ick Kim5, Edwin J. Weeber3 and Hyang-Sook Hoe1,2,4*
1Department of Neural Development and Disease, Korea Brain Research Institute (KBRI), Daegu 41062, 2Department of Brain and Cognitive Sciences, Daegu Gyeongbuk Institute of Science & Technology (DGIST), Daegu 42988, Korea, 3Department of Molecular Pharmacology and Physiology, USF Health Byrd Alzheimer’s Institute, University of South Florida, Tampa, FL 33613, 4Department of Neuroscience, Georgetown University Medical Center, Washington, DC 20057, USA, 5Department of Biological Sciences, Ulsan National Institute of Science and Technology (UNIST), Ulsan 44919, Korea
Correspondence to: *To whom correspondence should be addressed.
Hyang-Sook Hoe, TEL: 82-53-980-8310, FAX: 82-53-980-8309
e-mail: sookhoe72@kbri.re.kr
Hyun-ju Lee, TEL: 82-53-980-8313, FAX: 82-53-980-8309
e-mail: hjlee@kbri.re.kr
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Amyloid precursor protein (APP) plays an important role in the pathogenesis of Alzheimer’s disease (AD), but the normal function of APP at synapses is poorly understood. We and others have found that APP interacts with Reelin and that each protein is individually important for dendritic spine formation, which is associated with learning and memory,
Keywords: APP, Reelin, Dendritic spine, Alzheimer’s disease, Ras signaling
Many Alzheimer’s disease (AD) studies have focused on the synaptotoxic effects of the amyloid β (Aβ) peptide and neglected the possibility that amyloid precursor protein (APP) itself is important for synapse formation and function. We previously demonstrated that APP is highly localized in synapses and is involved in dendritic spine formation. For example, in primary hippocampal neurons
Recruitment of APP to both pre- and post-synaptic sites is required for synapse formation [3]. In addition, APP is involved in presynaptic vesicle release and postsynaptic N-methyl-D-aspartate receptor (NMDAR) trafficking, indicating a role in synaptic connectivity [4]. Furthermore, overexpression of human APP increases synaptic density, while expression of familial Alzheimer’s disease (FAD)-mutated APP has no effect [5]. Overexpression of sAPP, a soluble N-terminal fragment liberated by α-secretases, improves long-term potentiation (LTP) and enhances spatial memory [6]. Thus, the synaptic functions of full-length APP and sAPPα appear to directly oppose the neurotoxic effects of Aβ (and perhaps sAPPβ), the accumulation of which leads to impaired synaptic physiology and synapse loss [7, 8]. Despite the neurotoxic properties of Aβ, enhancement of long-term potentiation (LTP) and spatial memory by APP overexpression requires beta-site APP cleaving enzyme 1 (BACE1), suggesting that the APP intercellular domain (AICD) is also critical for the effect of APP on synaptic function [9].
The extracellular matrix glycoprotein Reelin is required for neuronal migration during embryonic brain development [10] and has been implicated in the pathoprogression of neurodevelopmental diseases such as attention deficit and hyperactivity disorder (ADHD) and autism spectrum disorder (ASD) and in neurodegenerative diseases, including AD. A genome-wide association study (GWAS) revealed that single nucleotide polymorphisms of Reelin are associated with an increased risk of ADHD [11], and patients with ASD have higher plasma levels of Reelin than healthy controls [12]. In addition, Reelin-positive neuronal cells are decreased in a mouse model of AD, whereas Reelin levels in cerebrospinal fluid (CSF) are increased in AD patients [13, 14]. Furthermore, genetic ablation of the
How is Reelin associated with cognitive dysfunction in neurodevelopmental and neurodegenerative diseases? Similar to APP, several studies have reported that Reelin plays a critical role in regulating synaptic function. Specifically, Reelin regulates dendritic spine formation in the early, postnatal hippocampus [16], and intraventricular administration of Reelin enhances dendritic spine density, synaptic plasticity, and spatial learning in the hippocampus in adult wild-type mice [17]. A role for Reelin in regulating synaptic function in the adult brain is further supported by the reduction in LTP in heterozygous
Ras/ERK signaling upregulates dendritic spine plasticity, and Ras signaling facilitates α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR) trafficking in cultured hippocampal slices, thereby enhancing LTP [21, 22]. Reelin promotes ERK/p90RSK signaling in mouse cortical neuronal culture [23]. In addition, Reelin regulates Ras-PI3K signaling to suppress cancer cell migration [24]. However, the molecular mechanisms by which Reelin alters dendritic spine formation are unclear. We and others recently reported that Reelin and APP individually alter dendritic spine formation [25], and in this study, we further investigated the physiological function of Reelin and the interaction between Reelin and APP at synapses. We found that Reelin treatment increased dendritic spine formation by regulating Ras/ERK/CREB signaling in mature primary hippocampal neurons. Importantly, Reelin injection did not affect cortical and hippocampal dendritic spine numbers in APP KO mice, suggesting that Reelin is required for the regulation of dendritic spine formation by APP. Taken together, our results indicate that Reelin interacts with APP to facilitate postsynaptic AMPAR trafficking and subsequent activation of Ras signaling to regulate dendritic spine formation.
All experiments were approved and performed in accordance with protocols approved by the Animal Welfare and Use Committee of Georgetown University and the Institutional Animal Care and Use Committee of the University of South Florida (approval no. R3336).
APP KO mice (B6.129S7-APPtm1DBo/J) were obtained from Jackson Laboratory (Bar Harbor, ME, USA). Animals were housed under a standard 12-h light-dark cycle and fed normal chow
Recombinant Reelin and mock-conditioned medium (Mock medium) were produced using HEK293 cells as previously described [26-28]. Briefly, HEK293 cells were transfected with the plasmid pCrl, which carries the entire open reading frame of mouse Reelin (Addgene plasmid #122443, a gift from Tom Curran), or a control construct. The cells were then incubated in low-glucose DMEM with 0.2% BSA for 2 days, and the medium was collected, filtered, and concentrated by using Centricon® plus-80 centrifugal filter units (Millipore, Bedford, MA, USA). Purified recombinant Reelin was validated via immunoreactivity to an anti-Reelin (G20) antibody (Santa Cruz Biotechnology, Dallas, TX, USA).
Primary hippocampal neurons from E18–E19 Sprague Dawley rats were cultured at 150 cells/mm2 as described previously [29]. Primary hippocampal neurons (DIV 12 or DIV 19) were transfected with GFP plasmid using Lipofectamine 2000 for 24 h and subsequently treated with purified Reelin (0.7 or 1.4 μM) or Mock medium on DIV 13 or DIV 20 for 24 h. Then, immunocytochemistry was conducted on DIV 14 or DIV 21, and secondary dendritic spine density was analyzed to measure the dendritic spine number or puncta number/intensity of synaptic protein levels. In brief, dendritic spines were defined as protrusions ranging in length from 0.2 μm to 2.0 μm with a head and neck (mushroom shaped or thin) or without a neck (stubby) [30]. Immature filopodia were excluded in the analysis of dendritic spine density in primary hippocampal neurons. In addition, we measured the number of dendritic spines located on secondary/tertiary dendritic branches only to minimize the bias/variability caused by the hierarchy of dendrites as previously described [30]. Dendritic spine density was measured by dividing the number of manually quantified dendritic spines by the dendritic segment length, which ranged from 10 to 20 μm.
For
Twenty-four hours after GFP plasmid transfection, primary hippocampal neurons were treated with 1.4 μM Reelin or Mock medium for 24 h. Then, the cells were fixed with 4% paraformaldehyde (to measure the dendritic spine number) or cold methanol (to measure synaptic protein puncta number along the dendritic neuronal processes and dendritic spines) for 10 min, washed 3 times with 1× PBS for 5 min, and incubated with primary antibodies overnight. The next day, the cells were washed 3 times with 1× PBS for 5 min and incubated with secondary antibodies for 1 h at room temperature. Finally, the cells were washed with 1× PBS for 5 min and mounted. Fluorescence intensity and puncta number along the dendritic neuronal processes and dendritic spines were quantified by using the NeuronJ plugin (NeuronJ, NIH, Bethesda, MD, USA) and SynaptoCount. For intensity quantification, regions of interest (ROIs) were acquired by tracing GFP signals; then, the average intensity of the target protein puncta in the dendrites was measured. Table 1 provides the details of the antibodies used in the present study.
APP KO mice (3 months old; n=4 mice per genotype per treatment) underwent bilateral intraventricular injections of Reelin (210 nM) or control saline as described previously [17]. Briefly, anesthetized APP KO mice were positioned on a stereotaxic frame (Stoelting, Wood Dale, IL, USA). After making an incision in the scalp and muscle, two holes were drilled in the dura, and Reelin or saline was injected using a Hamilton syringe at coordinates AP -0.35 mm, ML ± 0.75 mm, and DV -2.5 mm from the bregma at a flow rate of 1 μl/min. After the surgery, the mice were allowed to recover and monitored. Five days after the injection, the brains of the mice were perfused with PBS and immersed in the solutions provided in the FD Rapid Golgi Stain kit (FD NeuroTechnologies, Ellicott City, MD) [25]. The dehydrated brains were sliced at a thickness of 150 μm using a vibratome (VT1000S, Leica, Wetzlar, Hessen, Germany) and mounted on glass slides. Images of dendrites in the hippocampal CA1 pyramidal neurons and in cortical layer II/III were obtained by bright-field microscopy (Axioplan 2. Zeiss, Oberkochen, Baden-Württemberg, Germany) at 63× magnification. Dendritic spines were manually counted using ImageJ software (NIH, Bethesda, MD, USA).
GraphPad Prism 7 software (GraphPad Software, San Diego, CA, USA) was used for graph generation and statistical analysis. Data are presented as individual data points and the mean±SEM. Student's t test was used for pairwise comparisons, and one-way analysis of variance (ANOVA) with Tukey’s multiple-comparisons test was used for multiple comparisons. Asterisks indicate significance: *p<0.05, **p<0.01, and ***p<0.001.
Dendritic spine density is reduced in organotypic hippocampal cultures from
We then examined whether Reelin regulates the number of excitatory synapses
A previous study found that Reelin modulates NMDAR subunit expression/trafficking during the peak of synaptogenesis [26]; however, the effects of Reelin on NMDAR expression in mature neurons has not been established. Therefore, we examined whether Reelin regulates NMDAR expression levels in mature neurons. Specifically, GFP plasmid-transfected primary hippocampal neurons (DIV 19) were treated with 1.4 μM Reelin or vehicle (Mock medium) for 24 h and immunostained with antibodies against the NMDAR subunits GluN1, GluN2A, or GluN2B. Interestingly, Reelin treatment did not alter the fluorescence intensity of total GluN1 and total GluN2A but significantly decreased the fluorescence intensity of total GluN2B (Fig. 2A~C). These data suggest that Reelin differentially regulates the total expression levels of GluN1/GluN2A and GluN2B in mature primary hippocampal neurons.
To examine whether Reelin treatment modulates the trafficking of the AMPAR GluA1 and GluA2 subunits during the peak of synaptogenesis
Ras/Rap regulator polo-like kinase 2 (Plk2) suppresses dendritic spine formation and AMPAR trafficking, leading to impaired memory formation [32]. Moreover , we demonstrated that APP increases dendritic spine formation via Ras signaling
We then examined the effects of Reelin on Ras downstream ERK/CREB signaling
Next, we tested whether Ras activity is required for the effects of Reelin on dendritic spinogenesis. Primary hippocampal neurons (DIV 19) were transfected with GFP plasmid and HA plasmid (control for RasN17) or GFP plasmid and RasN17 plasmid (Ras inactivator) and treated with Reelin (1.4 µM) or vehicle (Mock medium) for 24 h. Dendritic spine number was then measured in secondary/tertiary distal dendrites. Consistent with the findings in Fig. 1A, Reelin treatment significantly increased dendritic spine density in mature primary hippocampal neurons compared with Mock treatment (Fig. 4D). However, RasN17 treatment blocked the Reelin-mediated increase in dendritic spine number in mature primary hippocampal neurons (Fig. 4D). These data suggest that Reelin promotes dendritic spine formation via Ras signaling in mature primary hippocampal neurons.
Previously, we found that APP enhances dendritic spine formation
Next, we investigated whether APP is required for Reelin-mediated dendritic spine formation
Previous studies have separately examined the roles of Reelin and APP in dendritic spine formation and synaptic plasticity
In a previous study, we investigated the effects of Reelin and APP on dendritic neurite outgrowth and dendritic branching in developing hippocampal neurons (DIV 14, the peak of synaptogenesis). Given that Reelin has a pivotal role in embryonic neuronal translocation/migration and postnatal dendritic growth/synaptic development [34], our previous findings implicated an interaction between Reelin and APP in the regulation of dendritic outgrowth during neuronal development [25]. Interestingly, a recent study demonstrated that organotypic hippocampal cultures from
Trafficking of NMDARs and AMPARs to the surface of excitatory postsynaptic neurons plays a critical role in LTP induction and memory consolidation [36]. In primary hippocampal neurons, blocking Reelin activity upregulates GluN2B-expressing dendritic spines [35]. However, Reelin treatment downregulates surface and total GluN2B levels but upregulates surface and total GluN2A levels in hippocampal slice cultures [26]. This NMDAR subunit switch from GluN2B to GluN2A and subsequent increase in GluA1 are accompanied by downregulation of silent synapses and synaptic transmission failure, implying that Reelin plays a critical role in synaptic plasticity by modulating synaptic NDMAR/AMPAR subunit composition [26]. Interestingly, a previous study demonstrated that NMDAR GluN1 subunit levels are reduced in primary hippocampal neurons from
Regarding the effects of Reelin on AMPAR trafficking and expression, several studies have confirmed that Reelin treatment enhances surface GluA1 expression and GluA2 phosphorylation in primary hippocampal neuronal culture [26, 41]. In addition, we previously reported that primary hippocampal neurons cultured from
NMDARs trigger Ras signaling via RasGRF, and the subsequent activation of the Ras-ERK cascade leads to CREB phosphorylation, which facilitates synaptic plasticity and memory formation [43, 44]. The Ras-ERK pathway has been implicated in dendritic complexity in hippocampal neuronal culture [45]. In addition, we and others have found that Reelin regulates NMDAR/AMPAR expression/trafficking to alter dendritic spine formation, but the underlying molecular mechanism remains to be elaborated. To address this gap, we examined the effects of Reelin on Ras signaling for the first time in the present study and found that Reelin treatment activated the Ras-ERK-CREB pathway and promoted dendritic spinogenesis in a Ras signaling-dependent manner (Fig. 4). Interestingly, an association of Reelin with neuronal maturation via enhancement of ERK signaling in a VLDLR/ApoER2-independent manner in cortical neuronal culture has been reported [23]. Given that Reelin binds to APP [25], we assume that Reelin interacts with APP and/or other receptors (e.g., VLDLR and ApoER2) to activate the Ras-ERK-CREB cascade and thereby promote dendritic spinogenesis
We and others have reported that APP also plays an important role in dendritic spinogenesis and synapse formation [1]. We previously demonstrated that APP interacts with Reelin extracellularly and that this interaction has a synergistic effect on neurite outgrowth [25]. The present study is the first to examine whether Reelin regulates dendritic spine formation, which is involved in cognitive function, by cooperating with APP
How do Reelin and APP alter dendritic spinogenesis? The mechanisms that Reelin utilizes to regulate synapse formation may not be mutually exclusive. Dendritic spines have high concentrations of actin fibers, and actin likely mediates the fluctuating sizes of dendritic spines over time. Interestingly, APP is highly enriched in the motile regions of neuronal growth cones [46] and affects cell movement
In conclusion, this study is the first to demonstrate that Reelin regulates the formation of dendritic spines in an APP-dependent manner. These data are further evidence of the important function of APP in dendritic spinogenesis and its distinction from the function of Aβ, which can result in synaptic dysfunction in AD. In particular, the present data suggest that APP is a downstream target of the Reelin signaling pathway in normal brain function and that alterations in this relationship may be relevant to the cognitive disturbances associated with AD.
This work was supported by the KBRI basic research program through KBRI funded by the Ministry of Science, ICT & Future Planning (grant numbers 22-BR-02-03, 22-BR-02-12, and 22-BR-05-02, H.S.H.), National Research Council of Science & Technology (NST) grant funded by the Korean government (grant number CCL22061-100, H.S.H.), and the National Research Foundation of Korea (grant number 2022R1F1A1074320, H.J.L.). The graphical abstract was created with BioRender.com.
List of antibodies used for ICC
Immunogen | Host species | Dilution | Manufacturer | Catalog no. |
GFP | Mouse | 1:200 | Novus Biologicals | 9F9.F9 |
GFP | Rabbit | 1:200 | Invitrogen | A11122 |
Synaptophysin | Mouse | 1:200 | Sigma Aldrich | S5768 |
PSD-95 | Mouse | 1:200 | NeuroMab | 07-028 |
GluA1 | Rabbit | 1:200 | Calbiochem | PC246 |
GluA2 | Mouse | 1:100 | BD Pharmingen | 556341 |
RasGRF1 | Rabbit | 1:200 | Santacruz Biotechnology | SC-224 |
pERK1/2 | Mouse | 1:200 | Invitrogen | 13-6200 |
pCREB | Rabbit | 1:200 | Millipore | 06-519 |
Immunogen | Host species | Dilution | Provider | Epitope |
GluN1 | Mouse | 1:200 | Dr. Barry Wolfe | Amino acid 656-811 |
GluN2A | Rabbit | 1:200 | Dr. Barry Wolfe | Amino acid 934-1142 |
GluN2B | Mouse | 1:200 | Dr. Barry Wolfe | Amino acid 934-1457 |