Abstract
Accumulation of astrocytes around β-amyloid (Aβ) plaques is one of the earliest neuropathological changes in Alzheimer's disease (AD), but the underlying mechanisms and significance remain unclear. Cell adhesion molecule protocadherin-γC5 (Pcdh-γC5) has been reported to implicate in AD. Here, we find elevated expression levels of Pcdh-γC5 in the brain of 5×FAD mice and Aβ-treated astrocytes and further reveal that Pcdh-γC5 deficiency leads to exacerbated Aβ deposition in 5×FAD mice. Deletion of Pcdh-γC5 impairs astrocyte migration, astrocytic response to Aβ signaling, and Aβ phagocytosis in both cultured astrocytes in vitro and 5×FAD mice in vivo. Both male and female mice were used in this study. Our findings support a model in which increased expression level of Pcdh-γC5 promotes astrocyte migration in response to Aβ signaling and engulfment and phagocytosis of neurotoxic Aβ plaques, therefore exerting a critical neuroprotective function in AD.
Significance Statement
Astrocytes play a pivotal role in AD pathogenesis. Recent evidence identified Pcdh-γC5 as a potential early-stage biomarker in the CSF of AD patients. Here, by using Pcdhgc5 gene knock-out astrocytes, we unravel the essential function of Pcdh-γC5 in promoting astrocyte migration and Aβ phagocytosis via the PI3K-Akt-actin signaling pathway. By crossing Pcdhgc5 knock-out mice and 5×FAD mice or downregulating astrocytic Pcdh-γC5 expression with shRNA, we demonstrate the neuroprotective role of Pcdh-γC5 in ameliorating Aβ plaque pathogenesis by regulating astrocyte activity. Our investigation provides new clues in clarifying the underlying mechanisms of AD pathogenesis, especially the critical role of astrocytes in AD.
Introduction
Accumulation of β-amyloid (Aβ) in the brain is one of the hallmarks of Alzheimer's disease (AD) and plays a key role in initiating and propagating AD pathogenesis (Querfurth and LaFerla, 2010; Gomez-Arboledas et al., 2018). Besides neuronal dysfunction, emerging evidence has given emphasis on the importance of nonneuronal cells, such as astrocytes and microglia, in the underlying mechanisms of AD (Liddelow and Barres, 2017; Hansen et al., 2018). Astrocytes are the most abundant cell type and constitute 20–40% of the total number of cells in the mammalian central nervous system (Khakh and Sofroniew, 2015; Ding et al., 2021). In physiological conditions, astrocytes are responsible for a wide variety of critical functions such as secreting neurotrophic molecules, modulating synaptic function, composing blood–brain barrier, and maintaining microenvironment homeostasis (Farhy-Tselnicker and Allen, 2018; Verkhratsky and Nedergaard, 2018). While in pathological conditions like AD, remarkable proliferation and activation of astrocytes were observed. Specifically, the accumulation of reactive astrocytes surrounding Aβ plaques is one of the earliest pathological features in AD (Wyss-Coray et al., 2003; Carter et al., 2012). Traditionally, reactive astrocytes were considered a passive component in the microglia-mediated neuroinflammatory cascade with neurotoxic consequences, contributing to AD pathogenesis (Wyss-Coray, 2006; Heneka et al., 2015). However, it's increasingly appreciated that reactive astrocytes also play a neuroprotective role by constituting a barrier around Aβ plaques and suppressing plaque progression in AD. In addition, reactive astrocytes have been reported to phagocytose and degrade Aβ, reducing Aβ plaque burden in the brain (Wyss-Coray et al., 2003; Koistinaho et al., 2004; Kraft et al., 2013; Gomez-Arboledas et al., 2018; Preman et al., 2021), but the underlying molecular mechanisms remain incompletely understood.
Protocadherin (Pcdh) is a family of cell adhesion molecules containing clustered Pcdh-α, Pcdh-β, and Pcdh-γ members and some nonclustered Pcdhs (Wu et al., 2001). Clustered Pcdhs express prominently in the central nervous system and play critical roles in many neuronal processes including neuronal development, dendrite arborization, synaptogenesis, and cell survival (Junghans et al., 2005; Chen et al., 2012; Jia and Wu, 2020). Pcdh molecules are Type I transmembrane proteins with six extracellular cadherin domains. The extracellular domain of various Pcdhs on cell membrane interacts with each other and forms a bridge between adjacent cells, mediating recognition and cell adhesion (Thu et al., 2014; Brasch et al., 2019). The intracellular domain of Pcdhs has been reported to bind focal adhesion kinase or proline-rich tyrosine kinase 2 (Pyk2), modulating actin filament activity (Chen et al., 2009; Garrett et al., 2012; Keeler et al., 2015).
In a large-scale cerebrospinal fluid (CSF) proteomic analysis in AD patients, Pcdh-γC5, a member of Pcdh-γs, was revealed to correlate with the level of both Aβ42 and phosphorylated tau (p-tau181), the two most prominent hallmarks of AD. The expression level of Pcdh-γC5 increased significantly in the CSF of AD patients, highlighting Pcdh-γC5 as a potential early-stage biomarker of AD (Tao et al., 2024). The gene expression alteration of Pcdh-γC5 has also been detected in the brain of AD patients and animal models of AD, as well as neural cells generated from human iPSCs that express APOE4, a prevalent genetic AD risk factor (Nativio et al., 2018; Meyer et al., 2019). In a protein array–based interactome screen, some clustered Pcdhs including Pcdh-γC5 were found significantly bound to Aβ (Oláh et al., 2011), further implicating Pcdh-γC5 in AD pathogenesis. Therefore, here we generated a Pcdhgc5 gene knock-out mouse and sought to investigate the role of Pcdh-γC5 in the neuropathology of AD.
Materials and Methods
Animals
The 5×FAD transgenic (34840-JAX) mice were obtained from The Jackson Laboratory. The Pcdhgc5 knock-out mice were generated by Cyagen Biosciences. Wild-type C57BL/6J mice were obtained from Xiamen University Laboratory Animal Center. Mice were housed in a 12 h light/dark cycle under specific pathogen-free conditions, with free access to water and food. No more than four adult mice were kept per cage. The room temperature was kept at 22–28°C, and the humidity was kept at 45–55%. All animal experiments were conducted in accordance with the National Institutes of Health's Guidelines for the Care and Use of Laboratory Animals and were approved by the Animal Ethics Committee of Xiamen University (XMULAC20200164 and XMULAC20180058). Both male and female mice were used in this study. For determining the expression of Pcdh-γC5, 8-month-old male WT and 5×FAD mice were used. For stereotactic injection of FAM-Aβ, 2-month-old male WT mice were used. For immunoblotting and immunofluorescence experiments in WT, 5×FAD and 5×FAD/Pcdhgc5+/− mice, both male and female mice of 6–8 months old, were used. For stereotactic injection of adeno-associated virus (AAV), 6-month-old male 5×FAD mice were used. The number of mice used for each experiment is specified in the figure legends. Mice were anesthetized with isoflurane, and anesthesia was verified by toe pinch.
Antibodies
The rabbit anti-Pcdh-γC5 antibody was developed in-house against the QLRYSVVEESEPGT epitope of mouse Pcdh-γC5 (GenBank accession no. NM_033583.3, 30–43 aa), and its specificity for Pcdh-γC5 has been verified previously (Li et al., 2010, 2012; Su et al., 2024). The other antibodies used were as follows: mouse anti-β-amyloid (Abcam, catalog #ab126649; RRID: AB_3095985), rabbit anti-NeuN (Cell Signaling Technology, catalog #12943; RRID:AB_2630395), mouse anti-GFAP (Cell Signaling Technology, catalog #3670; RRID: AB_561049), rabbit anti-presenilin 1 (ABclonal Technology, catalog #A2187; RRID: AB_2764205), rabbit anti-Iba1 (Cell Signaling Technology, catalog #17198; RRID: AB_2820254), rabbit anti-GAPDH (Abcam, catalog #ab181602; RRID:AB_2630358), mouse anti-APP (Merck Millipore, catalog #MAB348; RRID:AB_94882), mouse anti-s-100β (Sigma-Aldrich, catalog #S2657; RRID:AB_261477), rabbit anti-PI3K (Cell Signaling Technology, catalog #4292; RRID: AB_329869), rabbit anti-phospho-PI3K (Cell Signaling Technology, catalog #4228; RRID: AB_659940), rabbit anti-Akt (Cell Signaling Technology, catalog #9272; RRID: AB_329827), rabbit anti-phospho-Akt (Cell Signaling Technology, catalog #9271; RRID: AB_329825), and mouse anti-PSD-95 (Millipore, catalog #MAB1596, RRID: AB_2092365). The Alexa Flour–conjugated secondary antibodies for immunofluorescence were purchased from Jackson ImmunoResearch. The HRP-conjugated secondary antibodies for Western blotting were from Boster Biological Technology.
Primary astrocyte cultures
After anesthetized with isoflurane, the cortices of P0 mouse pups for regular neonatal astrocytes or 2-month-old male mice for mature astrocytes were dissected quickly in ice-cold HBSS after decapitation. The meninges were carefully removed, and the tissue was digested with 0.25% trypsin at 37°C for 20 min. Trypsin inhibitor was added to stop the reaction, followed by centrifugation at 3,000 rpm for 2 min. The solution was replaced with a culture medium (DMEM containing 10% FBS and penicillin–streptomycin) with DNase. The astrocytes were isolated by pipetting several times. After centrifugation, the cells were resuspended in a fresh culture medium and seeded with a desired density. The cultures were then kept in a 37°C incubator.
Oligomeric Aβ preparation
Oligomeric Aβ (1–42) peptide was prepared as previously described (Li et al., 2017). Briefly, 1 mg human Aβ (1–42) peptide (Invitrogen, 03-112) was dissolved in 500 μl of HFIP and incubated at room temperature for 30 min. For the 25 μl Aβ aliquots, HFIP was evaporated with gentle N2 air blowing. Before use, Aβ in each aliquot was dissolved with 10 μl of DMSO, then diluted with 100 μl of ice-cold F12 medium, followed by incubation at 4°C for 24–36 h. The prepared oligomeric Aβ was then diluted with astrocyte culture medium to a final concentration of 1 μM and was incubated with cells for various periods.
Determination of Aβ40 and Aβ42 levels
Briefly, mouse cortices were homogenized in TBS buffer containing protease inhibitor followed by centrifugation at 20,000 × g, 4°C for 2 h. The supernatants were collected as TBS fraction and pellets were further extracted with TBSX (TBS with 1% Triton X-100) containing protease inhibitor at 4°C for 30 min. After centrifugation at 20,000 × g, 4°C for 2 h, the supernatants were collected as TBSX fraction. The concentrations of human Aβ40 and Aβ42 in TBS and TBSX fractions were determined with ELISA kits (Ruixin Biotech, catalog #RX106222H and catalog #RX100431H) following the manufacturer's instructions.
Wound healing assay
Astrocytes cultured on coverslips were allowed to grow to 80–90% confluence. A scratching wound was built across the cell monolayer using a 200 μl pipette tip. After rinsed with PBS, cells were returned to a 37°C incubator and maintained in a serum-free medium. Then the cells were immunostained with phalloidin-iFluor 594 (Abcam, catalog #ab176757) at various time points after scratching and visualized under a microscope. The astrocytic migration rate at each time point was determined as (Initial scratch width – Current scratch width) / Initial scratch width.
Transwell migration assay
Primary astrocytes suspended in serum-free DMEM medium were added to the upper transwell chamber with an 8.0 μm polycarbonate membrane (Costar, 3422). After 30 min, the medium in the bottom chamber was replaced from serum-free medium to regular DMEM medium containing 10% FBS with or without 1 μM Aβ. After incubation at 37°C for 36 h, cells that remained on the upper surface of the membrane were removed with a cotton swab. Cells migrated through the membrane were then stained with hematoxylin–eosin (Solarbio Life Sciences, G1120) and visualized under a microscope.
Immunoblot analysis
Protein levels were determined with immunoblot analysis as described previously (Su et al., 2024). In brief, mouse brain tissue or cultured cells were lysed in RIPA buffer containing protease inhibitor at 4°C for 30 min followed by centrifugation at 12,000 rpm for 15 min. An equal amount of protein samples was subjected to 10% SDS-PAGE and transferred onto PVDF membranes. Separated proteins were detected by incubating with primary antibodies at 4°C overnight and HRP-conjugated secondary antibodies at room temperature for 1 h. The intensity of blotting bands was quantified by using ImageJ software.
Immunofluorescence staining
For brain slices, anesthetized mice were perfused through the ascending aorta with PBS and then 4% paraformaldehyde for fixation. Dissected mouse brains were postfixed in 4% paraformaldehyde at 4°C overnight and then transferred to 30% sucrose for cryoprotection. Frozen brains were sectioned in a 15 μm thickness with a freezing microtome. For cultured astrocytes, cells were fixed with 4% paraformaldehyde for fixation. Mouse brain slices or fixed cell cultures on coverslips were incubated with 2% goat normal serum and 0.25% Triton X-100 at room temperature for 1 h, followed by primary antibodies at 4°C overnight. After washing, the brain slices or cells were incubated with fluorophore-conjugated secondary antibodies at room temperature for 1 h. Images of mounted brain sections or cells were captured by using Nikon A1R laser confocal microscopy and analyzed with ImageJ software.
Cell Counting Kit-8 assay
Cell proliferation was examined using the Cell Counting Kit-8 (CCK-8) assay. Cells were plated in a 96-well culture plate and maintained in a 37°C incubator overnight. In each well, 10 μl of CCK-8 reagent (US Everbright, C6005L) was to cells. The cell optical density was determined at various time points by measuring absorbance at 450 nm with an enzyme immunoassay instrument (Bio-Rad Laboratories).
Stereotactic injection of FAM-labeled Aβ
FAM-labeled Aβ (AnaSpec, AS-23526-01) was dissolved in DMSO and diluted with F12 medium to a concentration of 100 μM. Then the FAM-Aβ solution was incubated at 37°C for 2 h, followed by 4°C overnight. Prepared oligomeric FAM-Aβ was stereotaxically injected into anesthetized mouse brain bilaterally (ML, ±2.0 mm; AP, −2.0 mm; DV, 1.8 mm) by using an automated stereotaxic injection apparatus (RWD Life Science). After injection, the syringe was left in place for 10 min and then withdrawn slowly. Mouse brain slices were collected 16 h later, immunostained, and analyzed.
Aβ phagocytosis assays
The 100 μM oligomeric FAM-Aβ stock was diluted with serum-free DMEM medium to a concentration of 0.5 μM FAM-Aβ and was added to primary cultured astrocytes. After incubation for 24 h, cells were washed with DMEM medium, followed by immunofluorescence staining. After incubation with 0.5 μM FAM-Aβ, cells were washed with ice-cold PBS and trypsinized for detection of FAM fluorescence using fluorescence-activated cell sorting (FACS).
Stereotaxic injection of AAV
Pcdh-γC5 shRNA (5′-GGATAATGGTGACCCTTCA-3′) or scrambled negative control (5′-GAAGTCGTGAGAAGTAGAA-3′) was cloned into the pAAV-GfaABC1D-mCherry-WPRE vector (AAV 2/5, 1.0 × 1012 vg/ml; OBiO Technology). For infection, AAV was stereotaxically injected into the cortex (AP, −2.0 mm; ML, ±1.8 mm; DV, 0.8 mm) and hippocampus (AP, −2.0 mm; ML, ±1.8 mm; DV, 2.0 mm) of anesthetized mice as described previously (Su et al., 2024). Behavioral tests were performed 4 weeks after virus injection. Then the mice were killed for immunofluorescence.
Morris water maze test
Mice were subjected to training and probe test in the Morris water maze as described previously (Su et al., 2024). In brief, mice were placed into a circular water tank at a random position, and escape latency to find the underwater platform was recorded. The mouse will be guided to the platform if it fails to find the platform within 60 s. After 6 d of training, the probe test was conducted. The platform was removed, and the number of mice that crossed the platform region was quantified.
Experimental design and statistical analysis
In this study, all experiments were blinded and randomized. The experimental design, handling of mice, and data collecting and analyzing were identical across experiments.
Microsoft Excel and GraphPad Prism software were used for statistical analyses. Data were presented as means ± SEM. For comparisons between two groups, two-tailed unpaired Student's t tests were used. For multiple comparisons, one-way or two-way ANOVA with post hoc Tukey's tests were used. Statistical significance was defined as *p < 0.05, **p < 0.01, and ***p < 0.001, and p > 0.05 was considered nonsignificant.
Results
The expression level of Pcdh-γC5 increases in the brain of 5×FAD mice and Aβ-treated astrocytes
Cell adhesion molecule Pcdh-γC5 has been shown to express specifically in the central nervous system and play important roles in the brain (Li et al., 2012, 2017; Su et al., 2024). To investigate the function of Pcdh-γC5, an antibody against mouse Pcdh-γC5 was generated, and the specificity was validated by using the cortical lysate of Pcdhgc5−/− mouse (Fig. 1A). Here, we determined the expression of Pcdh-γC5 in neurons, astrocytes, and microglia, the three major cell types in central nervous system. Pcdh-γC5 was found mainly expressed in neurons and astrocytes, with the highest expression level in astrocytes (Fig. 1B,C). Pcdh-γC5 has been implicated in AD pathogenesis (Nativio et al., 2018; Tao et al., 2024), prompting us to examine its expression in 5×FAD mice, a commonly used animal model of AD. Interestingly, the expression level of Pcdh-γC5 increased significantly in the cortex and hippocampus of 5×FAD mice (Fig. 1D,E), the most vulnerable brain regions of AD. The specificity of Pcdh-γC5 antibody was also tested with immunofluorescence on primary cultured neurons, astrocytes, and brain slices from Pcdhgc5−/− mice (Fig. 1F–H). Immunofluorescence staining showed increased Pcdh-γC5 clusters in the brain of 5×FAD mice, with some clusters colocalizing with neuronal marker MAP2 or astrocyte marker GFAP, especially around Aβ plaques (Fig. 1I–L). Pcdh-γC5 belongs to the cell adhesion molecule protocadherin (Pcdh) family, which is important in synaptogenesis and maintenance of synaptic function. In our previous study, we examined the critical role of Pcdh-γC5 in mediating synaptic dysfunction and Aβ-induced neuronal hyperactivity in AD (Li et al., 2017; Su et al., 2024). However, very few studies about Pcdhs in astrocytes have been reported. Given the extensive expression of Pcdh-γC5 in astrocytes and its enhanced expression level in 5×FAD mice, here we focus on the potential function of astrocytic Pcdh-γC5 in AD pathogenesis.
Increased expression level of Pcdh-γC5 in the brain of 5×FAD mice and Aβ-treated astrocytes. A, By using lysates from the cortex of WT and Pcdhgc5−/− mouse, Pcdh-γC5 antibody was validated to detect a specific band at ∼120 kD (arrow). B, C, The expression of Pcdh-γC5 was determined in primary cultured neurons, astrocytes, and microglia. n = 7 independent experiments. D, E, The protein level of Pcdh-γC5 was quantified in the cortex and hippocampus of WT and 5×FAD mice. n = 8 mice. F–H, The specificity of Pcdh-γC5 antibody was validated with immunofluorescence in primary cultured neurons identified by MAP2 (F), astrocytes identified by GFAP (G), and brain sections (H) from WT and Pcdhgc5−/− mice. I, Brain sections of WT and 5×FAD mice were immunostained with thioflavin-S (Thio-S) for Aβ plaques, MAP2 for neurons, and Pcdh-γC5. The arrows indicate colocalization of Pcdh-γC5 and MAP2. J, Quantification of Pcdh-γC5 puncta colocalized with MAP2 in the brain of WT and 5×FAD mice. n = 20 images from four mice. K, Brain sections of WT and 5×FAD mice were immunostained with Thio-S, GFAP for astrocytes, and Pcdh-γC5. The arrows indicate colocalization of Pcdh-γC5 with GFAP. L, Quantification of Pcdh-γC5 puncta colocalized with GFAP in the brain of WT and 5×FAD mice. n = 20 images from four mice. M, N, The protein level of Pcdh-γC5 was determined in cultured astrocytes in the absence or presence of 1 μM oligomeric Aβ. n = 8 independent experiments. O, P, The protein level of Pcdh-γC5 was quantified in astrocytes treated with 1 μM oligomeric Aβ in the absence or presence of 1 mM EGTA. n = 3 independent experiments. Scale bars: F–H, 20 μm; I, K, 10 μm. One-way ANOVA with post hoc Tukey's test was used for comparisons in C, N, and P. Two-tailed unpaired Student's t test was used for comparisons in E, J, and L. Data were presented as means ± SEM. *p < 0.05, ***p < 0.001.
To explore the expression of Pcdh-γC5 in response to Aβ, primary cultured astrocytes were treated with 1 μM oligomeric Aβ for various time periods. Compared to control, the expression level of Pcdh-γC5 in astrocytes increased significantly after Aβ treatment for 16 and 24 h, in a time-dependent manner (Fig. 1M,N). The Pcdh family is Ca2+-dependent (Frank and Kemler, 2002; Junghans et al., 2005; Chen and Maniatis, 2013; Jia and Wu, 2020). Cellular Ca2+ homeostasis is known to play pivotal roles in many facets of the nervous system. Dysregulated Ca2+ signaling contributes to pathophysiological conditions including AD. Elevated Ca2+ influx induced by Aβ has been reported in both neurons and astrocytes (Bosson et al., 2017; Tong et al., 2018; Shah et al., 2022). Therefore, the effect of Ca2+ signaling on Aβ-induced elevation of Pcdh-γC5 expression was examined. In the presence of EGTA, the chelator of Ca2+, Aβ treatment lost the ability to enhance Pcdh-γC5 expression in astrocytes (Fig. 1O,P), suggesting that Aβ increased the expression level of Pcdh-γC5 in a Ca2+-dependent manner.
Pcdh-γC5 deficiency accelerates Aβ plaque pathogenesis in 5×FAD mice
Aggregation of Aβ plaques in the brain is one of the neuropathological hallmarks of AD, highly contributing to the neurodegenerative process (Hardy and Higgins, 1992). Given the increased expression level of Pcdh-γC5 in the brain of 5×FAD mice and Aβ-treated astrocytes, to investigate the role of Pcdh-γC5 in AD, Pcdhgc5 gene knock-out mice were crossed with 5×FAD mice, a wildly used AD mouse model that develops aggressive Aβ pathology starting at 1.5 months of age (Richard et al., 2015). Then Aβ plaques in the brains of 5×FAD and 5×FAD;Pcdhgc5+/− mice were quantified. Intriguingly, compared with 5×FAD mice, the number and area of Aβ plaques were significantly elevated in both the cortex and hippocampus of 5×FAD;Pcdhgc5+/− mice (Fig. 2A–C), suggesting that elevated expression level of Pcdh-γC5 in 5×FAD mice plays a critical role in the control of pathological Aβ accumulation. Pcdh-γC5 deficiency significantly accelerated the neurotoxic Aβ plaque pathogenesis in 5×FAD mice.
Pcdh-γC5 deficiency accelerates Aβ plaque pathogenesis in 5×FAD mice. A, Immunofluorescence staining for Aβ plaques with antibody against Aβ in the brain of 5×FAD and 5×FAD;Pcdhgc5+/− mice. B, C, Quantification of Aβ plaques in the cortex and hippocampus, respectively. The quantified region of the hippocampus in this figure and Figures 6–8 includes CA1, CA2, CA3, and dentate gyrus, as indicated by the line. n = 5 mice. D, The mRNA levels of IL-1β, IL-6, and TNF-α in the cortex of WT, 5×FAD, and 5×FAD;Pcdhgc5+/− mice were quantified by RT-PCR. E, Brain sections were immunostained with Aβ for plaques and Iba1 for microglia. F, Quantification of microglia in mice with different genotypes. n = 10 images from three mice. G, H, The protein levels of Pcdh-γC5, APP, and PS1 were determined in the cortex of WT, 5×FAD, and 5×FAD;Pcdhgc5+/− mice. n = 6 mice. I, J, Cortices from 5×FAD and 5×FAD;Pcdhgc5+/− mice were extracted in TBS and then in TBSX. Concentrations of Aβ40 and Aβ42 in TBS (I) and TBSX (J) fractions were measured with ELISA. n = 6 mice. K, L, Brain sections of WT, 5×FAD, and 5×FAD;Pcdhgc5+/− mice were immunostained with neuronal marker NeuN, and the number of neurons in the cortex was quantified. n = 10 images from three mice. Scale bars: A, 400 μm; E, 10 μm; K, 50 μm. Two-tailed unpaired Student's t test was used for comparisons in B, C, I, and J. One-way ANOVA with post hoc Tukey's test was used for comparisons in D, F, H, and L. Data were presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001.
To determine whether the increased Aβ plaques are mediated by neuroinflammation via microglia, proinflammatory cytokines in the brain of WT, 5×FAD, and 5×FAD;Pcdhgc5+/− mice were measured with RT-PCR. The mRNA levels of IL-1β, IL-6, and TNF-α were significantly enhanced in 5×FAD mice but unchanged with Pcdh-γC5 deficiency (Fig. 2D). Consistently, the microgliosis in 5×FAD and 5×FAD;Pcdhgc5+/− mice was not significantly different either (Fig. 2E,F). We further examined whether Pcdh-γC5 deficiency has a potential effect on the expression of amyloid precursor protein (APP) and presenilin 1 (PS1), which are overexpressed in 5×FAD mice and vital for Aβ production. The protein levels of both APP and PS1 were significantly increased in 5×FAD mice, and Pcdh-γC5 deficiency didn't alter the expression of APP and PS1 (Fig. 2G,H). In addition, the levels of soluble Aβ40 and Aβ42 peptides were quantified with ELISA, and no significant difference was detected between 5×FAD and 5×FAD;Pcdhgc5+/− mice (Fig. 2I,J). Neuronal loss is another pathological characteristic of AD. Therefore, we examined the number of neurons in the brain of WT, 5×FAD, and 5×FAD;Pcdhgc5+/− mice. Compared to WT mice, a significant decrease of neurons was observed in 5×FAD and 5×FAD;Pcdhgc5+/− mice, but with no significant difference between them (Fig. 2K,L). These results suggest that the exacerbated Aβ plaque pathology induced by Pcdh-γC5 deficiency is not through alteration of microglia activation, neuronal Aβ production, or neuronal loss.
Deletion of Pcdh-γC5 impairs astrocyte migration and Aβ-induced recruitment of astrocytes
Due to the influence of Pcdh-γC5 on Aβ plaque deposition, we assessed the function of Pcdh-γC5 in astrocytes. Migration of astrocytes was determined by using a wound healing method. The astrocytes cultured from WT or Pcdhgc5−/− mice were scratch-wounded, and their migration abilities were compared. At each tested time point until 48 h after scratching, the migration rate of Pcdhgc5−/− astrocytes was significantly lower than that of WT astrocytes (Fig. 3A,B), exhibiting a reduced capability to close the wound. To confirm that the decreased wound healing ability of Pcdhgc5−/− astrocytes was due to impaired migration, the proliferation of WT and Pcdhgc5−/− astrocytes was determined by CCK-8 assay. No significant difference was observed within 48 h (Fig. 3C), confirming that the reduced wound healing ability of Pcdhgc5−/− astrocytes resulted from impaired migration, but not decreased proliferation. WT and Pcdhgc5−/− astrocytes were also subjected to the transwell migration assay. Deletion of Pcdh-γC5 led to a significantly decreased number of astrocytes migrated across the transwell membrane. It has been reported that Aβ could recruit astrocytes and promote astrocyte migration (Wyss-Coray et al., 2003; Kraft et al., 2013). As expected, treatment with oligomeric Aβ significantly enhanced the migration of WT astrocytes through the transwell membrane. However, Aβ was incapable of promoting migration of Pcdhgc5−/− astrocytes, losing the ability to recruit astrocytes after deletion of Pcdh-γC5 (Fig. 3D,E). To further test whether Pcdh-γC5 modulates astrocytic response to Aβ-induced recruitment in vivo, oligomeric fluorescent FAM-labeled Aβ (FAM-Aβ) was stereotaxically injected into the brain of WT and Pcdhgc5−/− mice. In response to Aβ, astrocytes were found accumulated in the vicinity of FAM-Aβ in the WT mouse brain. While in the brain of Pcdhgc5−/− mice, the ability of astrocytes to respond to Aβ recruitment was compromised (Fig. 3F,G). It is known that increased GFAP expression is one of the characteristics of activated astrocytes (Ding et al., 2021; Escartin et al., 2021). To identify whether Pcdhgc5 gene knock-out modulates astrocytic GFAP expression around Aβ or astrocyte migration toward Aβ, another astrocyte marker S-100β was used to label astrocytes in the vicinity of injected FAM-Aβ. Similarly, S-100β-labeled astrocytes surrounding FAM-Aβ were significantly reduced in the brain of Pcdhgc5−/− mice (Fig. 2H,I). Taken together, these data reveal that Pcdh-γC5 is critical in mediating astrocyte migration and Aβ-induced recruitment both in vitro and in vivo.
Deletion of Pcdh-γC5 impairs astrocyte migration. A, Confluent astrocyte cultures were scratched, and subsequent wound healing by migrating cells was monitored for up to 48 h. Astrocytes were visualized by phalloidin staining. B, Astrocyte migration rate was determined at 6, 12, 24, 36, and 48 h postscratching. n = 3 independent experiments. C, The proliferation of WT and Pcdhgc5−/− astrocytes was measured using CCK-8 assay. n = 5 independent experiments. D, E, Primary cultured WT and Pcdhgc5−/− astrocytes were plated into transwell chamber inserts. Following 24 h incubation with a control medium or 1 μM oligomeric Aβ in the bottom chamber, astrocytes that migrated through the membrane were quantified. n = 25 images from five independent experiments. F, G, After stereotaxic injection of 100 μM oligomeric FAM-Aβ, brain sections from WT and Pcdhgc5−/− mice were immunostained with DAPI and GFAP. Astrocytes accumulated around Aβ (50 μm distant from Aβ signal as indicated by the dotted line) were quantified. n = 20 images from five mice. H, I, Brain sections from WT and Pcdhgc5−/− mice injected with FAM-Aβ were immunostained with DAPI and another astrocyte marker S-100β. Astrocytes surrounding Aβ signal were quantified. n = 8 images from three mice. Scale bars: A, D, 100 μm; F, H, 50 μm. Two-tailed unpaired Student's t test was used for comparisons in B, E, G, and I. Two-way ANOVA with post hoc Tukey's test was used in C. Data were presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001.
Pcdh-γC5 promotes astrocytic phagocytosis of Aβ
In the brain of AD patients and animal models of AD, accumulated reactive astrocytes have been commonly observed surrounding Aβ plaques, forming a barrier to limit the spreading of pathological Aβ in the brain parenchyma and clear Aβ deposits through phagocytosis, exerting neuroprotective function in AD (Verkhratsky et al., 2016; Preman et al., 2021; Konishi et al., 2022). It has been reported that the phagocytosis ability of astrocytes is dependent on mouse age. Only mature astrocytes cultured from adult, but not neonatal, mice was able to phagocytose and degrade Aβ, reducing Aβ plaques (Wyss-Coray et al., 2003; Koistinaho et al., 2004). To determine whether Pcdh-γC5 participates in astrocytic phagocytosis of Aβ, cultured WT and Pcdhgc5−/− mature astrocytes were incubated with oligomeric FAM-Aβ, and then the Aβ phagocytosed by astrocytes were assessed. Compared to WT astrocytes, deletion of Pcdh-γC5 significantly decreased the cellular FAM-Aβ fluorescence intensity, suggesting a reduced amount of FAM-Aβ phagocytized into cells (Fig. 4A,B). After incubation with FAM-Aβ, Aβ phagocytosis by WT and Pcdhgc5−/− mature astrocytes were also analyzed by flow cytometry. Both the proportion of FAM-Aβ+ astrocytes and the FAM-Aβ intensity were significantly reduced in Pcdhgc5−/− mature astrocytes (Fig. 4C,D), exhibiting compromised ability of Aβ phagocytosis. These results provide evidence that Pcdh-γC5 promotes the phagocytic capacity of astrocytes in response to Aβ.
Pcdh-γC5 is critical for Aβ uptake and phagocytosis in mature astrocytes. A, Cultured WT and Pcdhgc5−/− mature astrocytes were immunostained with DAPI and GFAP after incubation with 0.5 μM oligomeric FAM-Aβ for 24 h. Scale bar, 10 μm. B, FAM-Aβ phagocytosed by astrocytes was evaluated. n = 18 images from three independent experiments. C, Representative FACS dot plots of WT and Pcdhgc5−/− mature astrocytes after incubation with FAM-Aβ. D, Percentage of FAM-Aβ+ astrocytes and FAM-Aβ intensity of WT and Pcdhgc5−/− astrocytes were quantified. n = 6 independent experiments. Two-tailed unpaired Student's t test was used for comparison. Data were presented as means ± SEM. *p < 0.05, **p < 0.01.
Pcdh-γC5 modulates actin cytoskeletal remodeling in astrocytes
Both cell migration and phagocytosis are processes dependent on actin remodeling and extension (Cowin, 2005; Giralt et al., 2016). Since Pcdhs have been reported to regulate actin filament activity (Chen et al., 2009; Garrett et al., 2012; Keeler et al., 2015), we further examined whether Pcdh-γC5 regulates astrocyte migration and Aβ phagocytosis by modulating actin dynamics. WT and Pcdhgc5−/− astrocytes were stained with phalloidin to show the actin cytoskeleton and cell morphology. Then the cellular perimeter and area of astrocytes were measured. Compared to WT astrocytes, deletion of Pcdh-γC5 didn't alter the astrocyte area but significantly decreased the perimeter and the ratio of perimeter to area (Fig. 5A–D), implying reduced cellular protrusions and impaired actin cytoskeleton extension. To further determine the effect of Pcdh-γC5 deficiency on actin dynamics, WT and Pcdhgc5−/− astrocytes were treated with latrunculin B, a drug to sequester G-actin and prevent F-actin assembly (Tomás et al., 2003; Giralt et al., 2016). After drug withdrawal, the disrupted actin cytoskeleton reassembled gradually, and its recovery ability was determined by measuring cell covered area in the absence or presence of Aβ at different time points. For WT astrocytes, Aβ significantly accelerated actin reassembling, with cells that covered a significantly bigger area. While after latrunculin B treatment, Pcdhgc5−/− astrocytes showed deficient actin reassembly ability. In addition, Aβ lost the ability to promote cell recovery (Fig. 5E,F). These results indicate that Pcdh-γC5 might regulate astrocyte migration and phagocytosis of Aβ via modulating actin cytoskeletal remodeling.
Pcdh-γC5 modulates actin cytoskeletal remodeling in astrocytes. A, WT and Pcdhgc5−/− astrocytes were stained with phalloidin to label actin filament. B–D, Astrocyte perimeter (B), area (C), and their ratio (D) were determined. n = 75 cells. E, WT and Pcdhgc5−/− astrocyte cultures were treated with 1 μM latrunculin B for 45 min and then washed with fresh culture medium in the absence or presence of 1 μM Aβ. Images represent phalloidin-labeled astrocytes before latrunculin B treatment (−45 min), at the end of latrunculin B application (0 min), and 15 and 60 min after washing. F, Actin repolymerization rate was quantified by measuring astrocyte covered area before and after latrunculin B treatment. n = 25 images from five independent experiments. G–I, The protein levels of Pcdh-γC5, PI3K, p-PI3K, Akt, and p-Akt were determined in WT and Pcdhgc5−/− astrocytes treated with 1 μM oligomeric Aβ or control. n = 6 independent experiments. Scale bar: A, E, 10 μm. Two-tailed unpaired Student's t tests were used for comparisons in B–D, H, and I. Two-way ANOVA with post hoc Tukey's test was used in F. Data were presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001.
It has been suggested that PI3K-Akt signaling can regulate actin activity and is important in astrocyte proliferation and migration (Zhao et al., 2017; Liao et al., 2022; Pérez-Núñez et al., 2023). To investigate whether Aβ-induced actin dynamics change is dependent on the PI3K-Akt–related pathway, we stimulated WT and Pcdhgc5−/− astrocytes with oligomeric Aβ and then evaluated the active phosphorylated PI3K (p-PI3K) and Akt (p-Akt). We found that Aβ treatment significantly increased the protein level ratios of both p-PI3K/PI3K and p-Akt/Akt in WT astrocytes, but not Pcdhgc5−/− astrocytes (Fig. 5G–I), implying that Pcdh-γC5 participates in Aβ-induced astrocytic processes through PI3K-Akt signaling pathway. Together, we revealed that Pcdh-γC5 is essential in astrocyte migration and phagocytosis of Aβ by regulating actin cytoskeletal dynamics, possibly through a downstream PI3K-Akt signaling pathway.
Pcdh-γC5 deficiency impairs Aβ-induced astrocyte activation in 5×FAD mice
In the brains of AD patients and animal models, especially in the vicinity of Aβ plaques, accumulation of reactive astrocytes was observed. The reactive astrocytes exhibit hypertrophic characteristics with thick and long processes and increased expression level of glial fibrillary acidic protein (GFAP; Liddelow and Barres, 2017; Escartin et al., 2021). To ascertain whether Pcdh-γC5 affects Aβ-induced astrocyte activation, GFAP-labeled reactive astrocytes were quantified in the brain of WT, 5×FAD, and 5×FAD;Pcdhgc5+/− mice. Compared with WT mice, the GFAP intensity was elevated significantly in the brain of 5×FAD and 5×FAD;Pcdhgc5+/− mice, showing a high level of astrocyte activation. Pcdh-γC5 deficiency didn't significantly change the general GFAP intensity in 5×FAD mice (Fig. 6A,B). We then evaluated the morphology of reactive astrocytes in the brain of 5xFAD and 5×FAD;Pcdhgc5+/− mice. For Aβ plaque-associated astrocytes, the number of branch intersections and total length of branches were significantly reduced in the brain of Pcdhgc5+/− mice (Fig. 6C–E), indicating less capable of activation and hypertrophy with Pcdh-γC5 deficiency. However, the morphology of astrocytes distant from Aβ plaques was indistinguishable between the two genotypes (Fig. 6F–H). Taken together, these data suggest that Pcdh-γC5 plays a critical role in coordinating astrocytes to take on a morphologically activated state in response to Aβ pathology, especially around Aβ plaques.
Pcdh-γC5 deficiency impairs astrocyte activation. A, Brain sections of WT, 5×FAD, and 5×FAD;Pcdhgc5+/− mice were immunostained with DAPI and GFAP. B, Quantification of activated astrocytes. n = 25 images in the cortex and 15 images in the hippocampus from five mice. C, Representative images and three-dimensional reconstruction of Aβ plaque-associated astrocytes in the brain of 5×FAD and 5×FAD;Pcdhgc5+/− mice. D, E, The number of branch intersections (D) and total branch length (E) of plaque-associated astrocytes were quantified. n = 50 cells. F, Representative images and three-dimensional reconstruction of astrocytes distant from Aβ plaques in the brain of 5×FAD and 5×FAD;Pcdhgc5+/− mice. G, H, The number of branch intersections (G) and total branch length (H) of astrocytes distant from Aβ plaques were quantified. n = 50 cells. Scale bars: A, 100 μm; C, F, 5 μm. One-way ANOVA with post hoc Tukey's test was used for comparisons in B. Two-tailed unpaired Student's t tests were used for comparisons in D, E, G, and H. Data were presented as means ± SEM. ***p < 0.001.
Pcdh-γC5 is essential in astrocyte accumulation and phagocytosis of Aβ plaques
It has been suggested that reactive astrocytes accumulating around Aβ plaques constitute a neuroprotective barrier to prevent not only Aβ neurotoxicity spreading but also phagocytose Aβ plaques (Konishi et al., 2022). To gain insight into the impact of Pcdh-γC5 on Aβ plaque pathogenesis, we compared the amount of astrocytes surrounding Aβ plaques in the brain of 5×FAD and 5×FAD;Pcdhgc5+/− mice and found a significant decrease in both the cortex and hippocampus of 5×FAD;Pcdhgc5+/− mice (Fig. 7A–D). To better assess the astrocytic phagocytosis of Aβ, astrocytes associated with Aβ plaques were three-dimensionally reconstructed, and Aβ colocalized with astrocytes were considered phagocytosed Aβ. Consistent with their neuroprotective effect on Aβ plaque deposition, the processes of reactive astrocytes invaded deeply into Aβ plaques, colocalizing with Aβ, in the brain of 5×FAD mice. However, Pcdh-γC5 deficiency significantly decreased the amount of Aβ plaques colocalized with associated reactive astrocytes (Fig. 7E–H), showing a compromised phagocytic function of Aβ plaques. It is becoming increasingly clear that cross talk between astrocytes and neurons contributes to synaptic development, plasticity, and function. Disruption in neuron–astrocyte signaling results in synaptic and cognitive impairment in diseases (Ullian et al., 2001; Clarke and Barres, 2013; Chung et al., 2015). To know whether Pcdh-γC5 deficiency also affects astrocytic phagocytosis of synapses, PSD-95-labeled synapses colocalized with astrocytes were three-dimensionally reconstructed and quantified in the brain of 5×FAD and 5×FAD;Pcdhgc5+/− mice, and no significant difference was detected (Fig. 7I–L). Collectively, these results indicate that Pcdh-γC5 is essential in astrocyte accumulation and phagocytosis of Aβ plaques. Pcdh-γC5 deficiency not only significantly attenuated Aβ plaque-associated astrocytes but also reduced the amount of Aβ phagocytosed by reactive astrocytes.
Pcdh-γC5 is essential in astrocyte accumulation and phagocytosis of Aβ plaques. A, C, Representative images showing Aβ plaques associated with astrocytes in the cortex (A) and hippocampus (C) of 5×FAD and 5×FAD;Pcdhgc5+/− mice. B, D, Astrocytes associated with Aβ plaques in the cortex (B) and hippocampus (D) were quantified. n = 120 plaques in the cortex and 36 plaques in the hippocampus from four mice. E, G, Three-dimensional reconstruction of GFAP, Aβ, and Aβ colocalized with GFAP in the cortex (E) and hippocampus (G) of 5×FAD and 5×FAD;Pcdhgc5+/− mice. F, H, Quantification of Aβ colocalized with GFAP for each plaque. n = 12 plaques from four mice. I, K, Brain sections from 5×FAD and 5×FAD;Pcdhgc5+/− mice were immunostained with GFAP and PSD-95 for neuronal synapses, and three-dimensional reconstruction was performed in the cortex (I) and hippocampus (K). J, L, Quantification of PSD-95 colocalized with GFAP. n = 15 astrocytes from three mice. Scale bars: A, C, 20 μm; E, G, I, K, 5 μm. Two-tailed unpaired Student's t tests were used for comparisons. Data were presented as means ± SEM. *p < 0.05, **p < 0.01.
Downregulation of astrocytic Pcdh-γC5 expression exacerbates Aβ pathogenesis and cognitive impairment in 5×FAD mice
To better determine whether the aggravated Aβ plaques in 5×FAD;Pcdhgc5+/− mice are due to Pcdh-γC5 deficiency in astrocytes, we prepared an astrocyte-specific AAV containing Pcdh-γC5 shRNA, whose high efficiency has been validated previously (Li et al., 2012, 2017). Here, the knock-down efficiency of shRNA was confirmed again by infecting cultured astrocytes with AAV containing either scrambled negative control or Pcdh-γC5 shRNA (Fig. 8A,B). To downregulate astrocytic Pcdh-γC5 expression in vivo, AAV containing Pcdh-γC5 shRNA was stereotaxically injected into the cortex and hippocampus of 5×FAD mice. The infection of cells was indicated by the mCherry fluorescence (Fig. 8C). In the injected brain region, the mCherry signal of shRNA-AAV colocalized with astrocyte marker GFAP well, but not with neuronal marker NeuN, indicating its specific infection of astrocytes (Fig. 8D). The protein level of Pcdh-γC5 was determined in the brain of 5×FAD mice injected with either scrambled control-AAV (5×FAD-Ctrl) or Pcdh-γC5 shRNA-AAV (5×FAD-shRNA). A significant decrease of Pcdh-γC5 level was observed in the brain of 5×FAD-shRNA mice, further confirming the knock-down efficiency of Pcdh-γC5 shRNA in vivo (Fig. 8E,F). Then accumulation of Aβ plaques was examined. Compared to 5×FAD-Ctrl mice, a significant increase of Aβ plaques was observed in both the cortex and hippocampus of 5×FAD-shRNA mice (Fig. 8G–I). We further investigated the effect of astrocytic Pcdh-γC5 deficiency on cognitive behavior. In the Morris water maze, both 5×FAD-Ctrl and 5×FAD-shRNA mice showed impaired learning ability, with longer escape latency and swimming distance to find platform during training. Compared to 5×FAD-Ctrl mice, 5×FAD-shRNA exhibited even worse cognitive function, with significantly increased escape latency on Days 5 and 6 and longer swimming distance on Day 6 of training (Fig. 8J,K). A consistent result was obtained in the later probe test. The 5×FAD-shRNA mice crossed the platform region fewer times than WT and 5×FAD-Ctrl mice, aggravating the cognitive impairment of 5×FAD mice (Fig. 8L). As a control, the crossing numbers over the opposite quadrant were not significantly different among WT, 5×FAD-Ctrl, and 5×FAD-shRNA mice (Fig. 8M). Consistently, these results suggest that downregulation of astrocytic Pcdh-γC5 expression leads to exacerbated Aβ pathology and cognition impairment in 5×FAD mice.
Downregulation of astrocytic Pcdh-γC5 expression exacerbates Aβ pathogenesis and cognitive impairment in 5×FAD mice. A, B, The protein level of Pcdh-γC5 in cultured astrocytes infected with AAV containing scrambled negative control or Pcdh-γC5 shRNA was determined. n = 4 independent experiments. C, 5×FAD mice were stereotaxically injected with AAV containing Pcdh-γC5 shRNA. The mCherry fluorescence indicates AAV-infected brain regions. D, Brain sections from AAV-injected mice were immunostained with NeuN or GFAP. E, F, The protein level of Pcdh-γC5 was determined in the brain of 5×FAD-Ctrl and 5×FAD-shRNA mice. n = 5 mice. G–I, Aβ plaques were immunostained and quantified in brain sections of 5×FAD-Ctrl and 5×FAD-shRNA mice. n = 6 mice. J, K, WT, FAD-Ctrl, and FAD-shRNA mice were subjected to the Morris water maze. The escape latency (J) and swimming distance (K) during training period were quantified. The symbol * in blue represents comparison between WT and FAD-Ctrl mice. The symbol * in red represents comparison between WT and FAD-shRNA mice. The symbol # in black represents comparison between FAD-Ctrl and FAD-shRNA mice. n = 18 mice. L, M, The number of crossings over the platform region (L) and the opposite quadrant (M) during probe test was quantified. n = 18 mice. Scale bars: C, 1 mm; D, 10 μm; G, 500 μm. Two-tailed unpaired Student's t tests were used for comparisons in B, F, H, and I. Two-way ANOVA with post hoc Tukey's test was used in J and K. One-way ANOVA with post hoc Tukey's test was used in L and M. Data were presented as means ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001.
Discussion
In this study, we have identified Pcdh-γC5 as a pivotal intracellular regulator of neuroprotective astrocytic response in a mouse model of AD. Pcdh-γC5 regulates astrocyte migration and Aβ phagocytosis by controlling actin dynamics, via the PI3K-Akt signaling pathway. Pcdh-γC5 deficiency impairs Aβ-induced astrocyte accumulation and engulfment of Aβ plaques, leading to exacerbated plaque pathogenesis and cognitive impairment in 5×FAD mice.
Pcdh-γC5 belongs to the Ca2+-dependent cell adhesion molecule protocadherin (Pcdh), the largest subfamily of cadherin (Wu et al., 2001). Clustered Pcdhs abundantly are expressed in the central nervous system. The stochastic and combinatorial expression pattern of various Pcdh members encodes highly diverse neural identity codes on the cell membrane, which are central for cell recognition, interaction, synaptogenesis, axonal development, and neuronal circuit formation (Junghans et al., 2005; Chen et al., 2012; Jia and Wu, 2020). The critical role of Pcdhs during brain development has been well established. Dysfunction of Pcdhs has been reported in multiple neurological disorders including Down syndrome, autism, depression, bipolar disorder, schizophrenia, and AD (Herzberg et al., 2006; Iossifov et al., 2014; Krumm et al., 2015; El Hajj et al., 2016; Li et al., 2017; Almenar-Queralt et al., 2019; Shao et al., 2019; Jia and Wu, 2020; Su et al., 2024).
Although Pcdhs are expressed in both neurons and astrocytes, most researches on Pcdhs focus on their functions in neurons, and very few about astrocytes. Even though Pcdhs have been shown to regulate synaptogenesis and neuronal circuit development via neuron–astrocyte cross talk (Garrett and Weiner, 2009; Tan and Eroglu, 2021), their roles in astrocytes remain far to be elucidated. In this study, we examined whether Pcdh-γC5 contributes to astrocyte function of phagocytosing synapses. By labeling brain slices from 5×FAD and 5×FAD;Pcdhgc5+/− mice with PSD-95 and GFAP, the synapses phagocytosed by astrocytes were three-dimensionally reconstructed and quantified. No significant difference was detected, indicating that Pcdh-γC5 deficiency doesn't affect astrocytic phagocytosis of synapses. However, it's noticeable that GFAP mainly stains the major branches of astrocytes. It might not be good for staining the more distal and abundant terminals, which are synapse-proximal compartments of astrocytes. To better assess the astrocyte function of phagocytosing synapses, other antibodies that label astrocyte terminals could be used, or experiments in vitro by using cultured astrocytes and prepared cerebral synaptosome could be performed.
In addition to their neurological functions, Pcdhs are perhaps best known in oncogenesis (Pancho et al., 2020). Dysregulated expression of Pcdhs has been extensively associated with multiple types of cancer. DNA hypomethylation and altered expression level of PCDHGC5 have been reported in the brain samples of patients with pilocytic astrocytoma, significantly correlating with a decreased survival probability (Vega-Benedetti et al., 2019), further indicating the importance of Pcdh-γC5 in regulating astrocyte function.
In AD, accumulation of reactive astrocytes around Aβ plaques has been suggested to not only constitute a neuroprotective barrier but also phagocytose and degrade Aβ plaques (Konishi et al., 2022). However, not many researches focused on astrocytic phagocytosis of Aβ, one possible reason might be age dependence. It has been demonstrated that only mature astrocytes from adult mice, but not neonatal astrocytes, are capable of phagocytosing Aβ plaques (Wyss-Coray et al., 2003; Koistinaho et al., 2004). In our previous study, we showed that the expression of Pcdh-γC5 is also age-dependent (Li et al., 2010). We checked the expression level of Pcdh-γC5 in rat brains from E18 to P90 and found that Pcdh-γC5 was not expressed until the second postnatal week. A similar result was also reported in the mouse brain (Frank et al., 2005). Here, we examined the ability of both neonatal and mature astrocytes in phagocytosing oligomeric FAM-Aβ and obtained consistent results. Neonatal astrocytes didn't show any capability of Aβ phagocytosis (data not shown), while mature astrocytes engulf FAM-Aβ pretty well, exhibiting strong phagocytosis ability. These results indicate that the expression level of Pcdh-γC5 might be a reason to interpret the age dependence of astrocytic phagocytosis of Aβ in AD.
Our research reveals that AD-susceptible protein Pcdh-γC5 promotes astrocyte migration in response to Aβ signaling and phagocytosis of Aβ plaques, playing critical neuroprotective roles in AD pathogenesis. This study expanded the understanding of Pcdhs in regulating neuronal activity, especially their roles in neuronal disorders. It will be interesting to explore more function of Pcdh-γC5 and other Pcdhs in these processes.
Data Availability
Data needed to evaluate the conclusions are available from the corresponding author upon request.
Footnotes
This work was supported by the National Natural Science Foundation of China (82071195 and 81870845 to Y.L.; 82271219 to Z.H.).
The authors declare no competing financial interests.
- Correspondence should be addressed to Yanfang Li at yfli{at}xmu.edu.cn.
