Abstract
Synaptic dysfunction and neuronal excitatory/inhibitory imbalance have been implicated in Alzheimer's disease (AD) pathogenesis. Although intensive studies have been focused on the excitatory synaptic system, much less is known concerning the mechanisms mediating inhibitory synaptic dysfunction in AD. We reported previously that protocadherin-γC5 (Pcdh-γC5), a member of clustered Pcdh-γ subfamily of cadherin-type synaptic adhesion proteins, functions to promote GABAergic synaptic transmission. We reveal here that Pcdh-γC5 is enriched in vesicular GABA transporter-positive synaptic puncta and its expression levels are increased in neuronal hyperexcitation conditions, upon β-amyloid (Aβ) treatment, and in amyloid precursor protein (APP)/presenilin-1 (PS1)-transgenic mice of both sexes. This is associated with elevated levels of GABAergic proteins and enhanced synaptic inhibition. Genetic knock-down experiments showed that Pcdh-γC5 modulates spontaneous synaptic currents and Aβ-induced synaptic alterations directly. Our results support a model in which Pcdh-γC5 senses neuronal hyperexcitation to augment GABAergic inhibition. This adaptive mechanism may be dysregulated under chronic excitation conditions such as AD, leading to aberrant Pcdh-γC5 expression and associated synaptic dysfunction.
SIGNIFICANCE STATEMENT Synaptic dysfunction is causal for Alzheimer's disease (AD). Here, we reveal a novel pathway that contributes GABAergic synaptic dysfunction in AD mediated by protocadherin-γC5. Our study not only identifies a new mechanism mediating excitatory/inhibitory balance in AD, but may also offer a new target for potential therapeutic intervention.
Introduction
Alzheimer's disease (AD) is characterized by the progressive decline of memory and cognition, with complicated pathological mechanisms. The two hallmarks of AD are the extracellular deposition of β-amyloid (Aβ) plaques and the intracellular accumulation of neurofibrillary tangles. In addition, the alteration of synaptic function, disruption of neuronal excitatory and inhibitory balance, and eventual neuronal loss are widely recognized pathogenic events occurring in AD (Palop and Mucke, 2010a, 2010b). It has been reported that the risk of unprovoked seizures is significantly higher in AD patients compared with nondemented controls (Hauser et al., 1986; Mendez et al., 1994; Hesdorffer et al., 1996; Lozsadi and Larner, 2006). Particularly, in AD patients with early onset dementia, the occurrence risk of epilepsy is 87-fold higher than control population during early stages (Amatniek et al., 2006), implicating neuronal hyperexcitability in disease pathogenesis. This is further supported by aberrant epileptiform activity and seizures in animal models of AD (Palop et al., 2007; Minkeviciene et al., 2009). Much research has been focused on glutamatergic excitatory synaptic system, whereas the inhibitory GABAergic synaptic system has received less attention.
In the vertebrate brain, GABA is the major inhibitory neurotransmitter that conveys its inhibitory effect through GABAA receptors (Moore, 1993; Lüscher and Keller, 2004). Emerging evidence indicates that the GABAergic synaptic system undergoes pathological changes. In a mouse model of AD, aberrant epileptic activity was also found associated with sprouting of inhibitory GABAergic synaptic terminals and enhanced synaptic inhibition, suggesting a compensatory response to the neuronal hyperexcitability. Dysregulated inhibition may interfere with synaptic function and contribute to learning and memory deficit in AD. Indeed, application of a GABAA receptor blocker has been shown to restore the cognitive deficient in animal models of AD (Marczynski, 1998; Yoshiike et al., 2008). However, the underlying mechanism for this inhibitory compensation is unclear.
Protocadherin-γC5 (Pcdh-γC5) is a member of clustered Pcdh-γ subfamily, which belongs to the well known cadherin family of synaptic adhesion molecules. Like cadherins, Pcdh members are hypothesized to engage homophilic or heterophilic interactions between adjacent cells. Pcdh-γs are predominantly expressed in the nervous system and have been reported to play important roles in neuronal connectivity, neurite self-avoidance, and neuron–astrocyte interactions, among others (Stoeckli, 2014; Keeler et al., 2015a, 2015b). We reported previously that Pcdh-γC5 interacts with GABAA receptors to promote its trafficking and GABAergic synaptic stabilization (Li et al., 2012). In the present study, we observed a significant increase of Pcdh-γC5 in amyloid precursor protein (APP)/presenilin-1 (PS1) mice and upon Aβ treatment in vitro and this was associated with elevated levels of GABAergic proteins and inhibitory synaptic function. Interestingly, Pcdh-γC5 can be induced under neuronal hyperexcitation conditions and knocking down endogenous Pcdh-γC5 rescued Aβ-induced synaptic dysfunction, supporting a critical role of Pcdh-γC5 in mediating synaptic function in AD pathogenesis.
Materials and Methods
Animals.
The APPswe/PSEN1dE9 (APP/PS1) mice carrying mutations on both human APP and PS1 proteins were purchased from The Jackson Laboratory. The transgenic mice were backcrossed with C57BL/6 mice. Both female and male APP/PS1-transgenic mice and littermate wild-type (WT) mice were used in this study. For comparing the expression levels of Pcdh-γC5 and GABAergic markers between WT and APP/PS1 mice, both sexes of 9-month-old mice were used. For the immunofluorescence experiment, both sexes of 10- 13-month-old WT and APP/PS1 mice were used. Images were taken and quantified from two to three brain sections of each mouse. For electrophysiological recording with brain slices, both sexes of WT and APP/PS1 mice at ages as indicated were used. For the kainic acid (KA) injection experiment, male WT mice 2–4 months of age were used. The numbers of mice used for each experiment are specified in the figures and figure legends.
All experiments were performed 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.
Antibodies and reagents.
The rabbit anti-Pcdh-γC5 antibody was made in-house by immunization against aa 30–43 of mouse Pcdh-γC5 (GenBank accession no. NM_033583.3). The rabbit anti-APP antibody (Ru369) was developed as described previously. The mouse anti-vesicular GABA transporter (vGAT) antibody was from Santa Cruz Biotechnology (catalog #sc-393373). The rabbit anti-glutamate decarboxylase (GAD)65/67 antibody was from Millipore (catalog #AB1511). The mouse anti-β actin (catalog #ab8226) and rabbit anti-GAPDH (catalog #ab181602) antibodies were from Abcam. The mouse anti-vesicular glutamate transporter 1 (vGluT1) antibody was from Synaptic Systems (catalog #135511). All Alexa Flour-conjugated secondary antibodies for immunofluorescence were purchased from Invitrogen. The HRP-conjugated secondary antibodies for Western blotting were from Thermo Fisher Scientific. All reagents used for electrophysiological recording were from Sigma-Aldrich.
Western blotting.
Briefly, the cortex was dissected from WT or APP/PS1 mice, followed by lysis with RIPA buffer (containing 10 mm Tris-HCl, 137 mm NaCl, 0.1% SDS, 1% Triton X-100, and 1% sodium deoxycholate, pH 7.4) and protease inhibitor. The homogenate was then centrifuged at 12,000 rpm for 20 min. The collected protein samples were separated with SDS-PAGE and transferred to PVDF membrane. The blots were then incubated with indicated primary antibodies at 4°C overnight and HRP-conjugated secondary antibodies at room temperature for 2 h, followed by detection with enhanced chemiluminescence (Millipore, catalog #WBKLSO500). The intensity of blotting bands was measured and quantified using ImageJ software.
RNA isolation and quantitative RT-PCR.
Total RNA from mouse cortex or hippocampus was extracted using TRIzol reagent (Invitrogen, catalog #15596-018) and reverse transcribed to cDNA using GoScript Reverse transcription Kit (Promega, catalog #A5001). Quantitative RT-PCR was conducted using the Applied Biosystems (ABI) 7500.
The primers used for mouse gene expressions were as follows: vGAT forward: 5′-CATTCATTATCAGCGCGGCG-3′; vGAT reverse: 5′-AGCACGAACATGCCCTGA-3′; GAD65 forward: 5′-TCGGATCTGAAGATGGCTCTGC-3′; GAD65 reverse: 5′-TAGAGCAGAGCGCACAGCTT-3′; GAD67 forward: 5′-TTCCACTCCTTCGCCTGCA-3′; and GAD67 reverse: 5′-AGAAGCCACAGATCTTCAGGCC-3′.
Cell cultures and transfection.
The HEK293T cells were maintained in high-glucose DMEM supplemented with 10% fetal bovine serum. For primary cortical neurons, the cortices of postnatal day 0 pups were isolated and diced to small pieces with forceps in ice-cold HBSS. Tissue was then transferred to fresh HBSS with 0.25% trypsin (Invitrogen, catalog #15090046) for digestion at 37°C for 15 min. Trypsin inhibitor (Invitrogen, catalog #17075029) was added to stop the reaction and tissue was collected after centrifugation at 3000 rpm for 2 min. Neuronal culture medium (Neurobasal A medium with 2% B27 supplement, 0.5 mm l-glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycine) containing 0.1 mg/ml DNase (Sigma-Aldrich, catalog #D5025) was added to collected tissue and neurons were isolated by pipetting until there were no visible tissue clumps, centrifuging again, and resuspending the precipitate with fresh culture medium. Neurons were then seeded to poly-L-lysine-pretreated coverslips with the desired density and maintained in a 37°C incubator.
Cells were transfected with various vectors using Lipofectamine 2000 transfection reagents (Invitrogen,catalog #11668027) and following the product protocol.
DNA constructs and RNA interference.
The full-length mouse Pcdh-γC5 cDNA (GenBank accession no. NM_033583.3) was subcloned to pcDNA3.1/myc-His A vector with myc and His tags at the C terminus. The two small hairpin RNAs (shRNAs) of Pcdh-γC5 were synthesized and subcloned to LentiLox 3.7 (pLL3.7) vector containing an EGFP tag. The specific targeting region in Pcdh-γC5 is 5′-GGATAATGGTGACCCTTCA-3′.
Aβ oligomer preparation and treatment.
Five hundred microliters of 1,1,1,3,3,3,-hexafluoro-2-propanol (HFIP) was added to 1 mg of lyophilized Aβ1-42 peptide (Invitrogen, catalog #03-112), followed by 30 min of incubation at room temperature. The solutions were aliquoted into a 25 μl volume. HFIP was allowed to evaporate in open tubes in a fume hood and any remaining traces of HFIP were removed with gentle N2 air blowing. The dried peptide was stored over desiccant at −80°C until use. Then, 10 μl of DMSO was added to each Aβ aliquot and pipetted thoroughly at room temperature. The freshly resuspended Aβ was diluted in 90 μl of ice-cold F12 cell culture medium to 100 μm concentration and transferred to 4°C for 24–36 h. The appropriate volume of prepared Aβ oligomer was added to the medium of cultured neurons to a final concentration of 1 μm and incubated for 2 h. The neurons were then subjected to electrophysiological recording or protein extraction.
Electrophysiological recording.
For recordings on brain slices, WT or APP/PS1 mice were anesthetized and decapitated. The brains were quickly excised and cut into 400-μm-thick coronal slices with vibrating microtome (Leica VT1200S) in ice-cold ACSF containing the following (in mm): 126 NaCl, 18 NaHCO3, 2.5 KCl, 1.2 NaH2PO4, 1.2 CaCl2, 2.4 MgCl2, and 11 glucose aerated with 95% O2–5% CO2. For mice older than 5 months, high-sucrose ACSF containing the following (in mm) 206 sucrose, 26 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 2 CaCl2, 2 MgSO4, and 10 glucose was used instead. The slices were then transferred to continuously aerated ACSF and incubated at 32°C for 30 min and then at room temperature for at least 30 min. For recording, the slices were transferred to a recording chamber continuously perfused with aerated ACSF. For whole-cell recording of sEPSCs and sIPSCs, pipette solution contained the following (in mm): 140 CsCH3SO3, 2 MgCl2, 5 TEA-Cl, 10 HEPES, 1 EGTA, 2.5 Mg-ATP, and 0.3 Na2-GTP adjusted to pH 7.2–7.4 with CsOH. Spontaneous EPSCs (sEPSCs) were recorded at a holding potential of −70 mV and spontaneous IPSCs (sIPSCs) were recorded at a holding potential of 0 mV. For evoked spikes recording, the brain slices were perfused with ACSF and the glass pipette was filled with solution contained the following (in mm): 140 K-gluconate, 2 MgCl2, 0.1 CaCl2, 10 HEPES, 1.1 EGTA, 0.3 Na2-GTP, and Na2-ATP adjusted to pH 7.25 with KOH.
For sEPSCs and sIPSCs recording on cultured cortical neurons, coverslips seeded with neurons were transferred to the recording chamber perfused with ACSF and the same recording protocol described above was used. Recording data were collected with a Multi-Clamp 700B amplifier (Molecular Devices) using Clampex10 acquisition software (Molecular Devices), filtered at 1 kHz, sampled, and digitized with a Digidata 1440A (Molecular Devices). Raw data were analyzed offline with pClamp10.3 software. The frequency and amplitude of sEPSCs and sIPSCs were measured with Mini analysis software.
Immunofluorescence and amyloid plaque staining.
Mice were anesthetized, followed by perfusion through ascending aorta with 0.9% saline and then 4% paraformaldehyde for fixation. Brains were excised and postfixed overnight in 4% paraformaldehyde at 4°C, then transferred to 30% sucrose for cryoprotection. Frozen brains were sectioned with a freezing microtome. Parasagittal brain sections of 15 μm thickness were incubated with PBS containing 2% normal goat serum and 0.3% Triton X-100 at room temperature for 1 h, followed by incubation with primary antibodies. After washing with PBS, the brain sections were incubated with a mixture of secondary antibodies in PBS for 1.5 h, followed by washes. If needed, sections were subjected to DAPI (Sigma-Aldrich, catalog #D9542) staining for 10 min. To visualize amyloid plaques, brain sections were incubated with 1 mm thioflavine-S (Sigma-Aldrich, catalog #T1892) in 50% ethanol for 8 min, rinsed with 80% ethanol, and washed with PBS. After the staining, brain sections were mounted and analyzed by Nikon A1R laser confocal microscopy. Confocal microscopic images were analyzed using the ImageJ program. The original images were processed and converted into binary files and then the immunostained particles were analyzed, with puncta number, average puncta size, and total puncta area measured. The values were recorded and compared using Microsoft Excel and GraphPad Prism software.
Statistical analysis.
Microsoft Excel or GraphPad Prism software was used for the statistical analyses. The data are presented as means ± SEM. When two groups were compared, raw data were analyzed by Student's t test. For comparison of various intensities of current-evoked spikes between WT and APP/PS1 mice, two-way ANOVA was used. p < 0.05 was considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001). p > 0.05 was considered nonsignificant.
Results
Increased expression of Pcdh-γC5 and GABAergic synaptic markers in APP/PS1 mice
Pcdh-γC5 has been shown to play important roles in the brain, particularly in GABAergic inhibitory synaptic function and maintenance. Consistent with this view, we found that Pcdh-γC5 was highly expressed in various regions of the brain including cortex, hippocampus and cerebellum, whereas its expression was negligible in non-neuronal tissues (Fig. 1A). Furthermore, double immunofluorescence staining using anti-Pcdh-γC5 and anti-vGAT antibodies revealed a high degree (∼80%) of colocalization (Fig. 1B), indicating that Pcdh-γC5 is highly expressed in GABAergic neurons, where it is enriched in vesicular vGAT-positive presynaptic terminals.
Alteration of GABAergic transmission has been implicated in AD pathogenesis (Bareggi et al., 1982; Zimmer et al., 1984; Garcia-Marin et al., 2009; Gueli and Taibi, 2013; Ramos-Miguel et al., 2015), prompting us to examine Pcdh-γC5 expression in APP/PS1 mice, a commonly used animal model of AD. Interestingly, we detected significantly higher levels of Pcdh-γC5 in forebrain homogenates of 9-month-old APP/PS1 mice compared with WT littermate controls (Fig. 2A,B). Immunofluorescence staining of the cortex of 10- to 13-month-old APP/PS1 mice (Fig. 2C) revealed that, whereas the number of Pcdh-γC5-positive puncta was not affected (Fig. 2D), both the average puncta size (Fig. 2E) and total puncta area (Fig. 2F) were significantly increased in APP/PS1 mice. This was particularly the case in areas surrounding Aβ plaques (see enlarged views in Fig. 2C).
To determine whether elevated Pcdh-γC5 levels in APP/PS1 mice were associated with altered expression of GABAergic inhibitory synapse genes, we examined the expression of the GABAergic synaptic markers GAD and vGAT (Fig. 3). We found that both the mRNA (Fig. 3A) and protein levels (Fig. 3B, quantified in 3C) of GAD65 and GAD67, two isoforms of GAD, and vGAT were significantly higher in APP/PS1 mice compared with WT controls. This was supported by immunofluorescence staining of brain slices of the cortex of 10- to 13-month-old APP/PS1 mice (Fig. 3D,E). Although the puncta numbers of GAD and vGAT (Fig. 3F) did not change significantly, both the average puncta size (Fig. 3G) and total puncta area (Fig. 3H) were increased significantly compared with WT controls.
Enhanced neuronal excitation and inhibition in APP/PS1 mice
To further investigate the changes of synaptic function in APP/PS1 mice, electrophysiological recording was performed in brain slices from WT and APP/PS1 mice at different ages (Fig. 4). General neuronal excitation was determined using current-clamp recording by injecting positively charged current of various intensities into individual forebrain neurons to induce depolarization. At both 4 and 10 months of age, the current-induced spikes were significantly higher in APP/PS1 mice compared with WT controls (Fig. 4A,B). This was associated with higher sEPSCs and sIPSCs as measured by voltage-clamp recording (Fig. 4C–G). The frequency of both sEPSCs and sIPSCs were increased significantly in APP/PS1 mice at all ages tested (Fig. 4D,E), indicating increased neurotransmitter release from both excitatory and inhibitory presynaptic terminals. In contrast, the amplitude of sEPSCs and sIPSCs were comparable between genotypes (Fig. 4F,G), suggesting no significant change of postsynaptic glutamate or GABA receptors.
Neuronal hyperexcitation induces changes of Pcdh-γC5 and synaptic markers
Because sEPSCs are increased in APP/PS1 mice, the elevated Pcdh-γC5 and GABAergic markers and sIPSCs could either be independent of or secondary to neuronal excitation. To clarify the relationship between neuronal hyperexcitation and enhanced GABAergic synaptic function, we examined Pcdh-γC5 and GABAergic markers in a widely used KA-induced neuronal hyperexcitation model (Dibué et al., 2014), which induces acute epileptic behaviors after intraperitoneal injection of KA. The protein levels of Pcdh-γC5, the inhibitory presynaptic markers GAD and vGAT, and the excitatory presynaptic marker vGluT1 were determined at various time points after KA injection (Fig. 5). We observed increased levels of Pcdh-γC5 and inhibitory and excitatory proteins 2–24 h after KA treatment, followed by a reduction 1 week after KA treatment. Because neuronal excitation is the initial trigger, time-dependent increases of Pcdh-γC5 and inhibitory proteins support the idea that enhanced GABAergic synaptic function serves as a compensatory mechanism to counter hyperexcitation. However, sustained enhanced neurotransmitter release might lead to the depletion of synaptic vesicles, resulting in the eventual reduction of Pcdh-γC5 and synaptic markers observed 1 week after KA injection (Fig. 5G,H).
Aβ treatment increases the expression of Pcdh-γC5 in cortical neurons
It has been reported that Aβ could induce Ca2+ influx and elicit epileptiform activity in animal models of AD (Minkeviciene et al., 2009; Parodi et al., 2010). To investigate whether Aβ itself could trigger the alteration of Pcdh-γC5, cultured cortical neurons were treated with 1 μm oligomeric Aβ for different times. At all of the time points tested (2, 8, and 24 h), the expression levels of Pcdh-γC5 increased significantly compared with vehicle-treated controls (Fig. 6A,B), although there were no significant differences across time points. The effect of Aβ doses on Pcdh-γC5 was then tested by incubating the cortical neurons with different concentrations of oligomeric Aβ for 8 h. The expression of Pcdh-γC5 increased at all Aβ concentrations tested (0.5, 1.0, and 1.5 μm) and a significant dose-dependent effect was detected at doses between 0.5 and 1.0 μm (Fig. 6C,D). The results support the notion that the increased Pcdh-γC5 observed in APP/PS1 mice is due to Aβ overproduction.
Pcdh-γC5 modulates Aβ-induced synaptic dysfunction in cortical neurons
To provide mechanistic insights into Pcdh-γC5-mediated synaptic function, we designed a Pcdh-γC5 shRNA vector. The high efficiency knock-down was validated by transfecting a myc-tagged Pcdh-γC5 vector (Pcdh-γC5) or cotransfecting with Pcdh-γC5-myc and Pcdh-γC5 shRNA (Pcdh-γC5+sh) in HEK293 cells using pLL3.7 vector-cotransfected cells as a control (Fig. 7A,B). The effect of Pcdh-γC5 on synaptic function was investigated by whole-cell recording of cultured cortical neurons (Fig. 7C–Q). Knocking down Pcdh-γC5 expression decreased the frequency of both sEPSCs and sIPSCs significantly, but not the amplitude, indicating that Pcdh-γC5 modulates presynaptic neurotransmitter release directly (Fig. 7C–G). To determine whether Pcdh-γC5 could regulate Aβ-induced synaptic alteration, neurons were first transfected with Pcdh-γC5 shRNA or pLL3.7 control vector. Three days later, neurons were treated with 1 μm oligomeric Aβ for 2 h, followed by whole-cell recording (Fig. 7H–L). Consistent with the brain slices, Aβ treatment resulted in a higher frequency of sEPSCs and sIPSCs, but not amplitude. These effects were blunted by transfection with Pcdh-γC5 shRNA, suggesting that Pcdh-γC5 could regulate and rescue Aβ-induced synaptic changes efficiently.
We next investigated the effect of Pcdh-γC5 on cortical neurons from APP/PS1 mice, which generate more Aβ compared with WT mice. Neurons from APP/PS1 mice showed a significantly increased frequency of sEPSCs and sIPSCs. Similar to Aβ-treated neurons, knocking down Pcdh-γC5 efficiently rescued the frequency to normal levels, as recorded in WT neurons (Fig. 7M–Q), providing additional support that Aβ-induced synaptic dysfunction can be rescued by reducing Pcdh-γC5 expression.
Discussion
Compared with the pathological hallmarks of Aβ plaques and neurofibrillary tangles, synaptic dysfunction is known to best correlate with cognitive decline in AD (Palop and Mucke, 2016). In the past, considerable research has been focused on excitatory glutamatergic and cholinergic synaptic systems (Pozueta et al., 2013), whereas inhibitory GABAergic synapse systems received less attention because they were considered largely preserved in AD (Rossor et al., 1982; Reinikainen et al., 1988). However, emerging evidence implicates significant alteration and remodeling of GABAergic neurotransmission system in AD (Bareggi et al., 1982; Zimmer et al., 1984; Palop et al., 2007; Gueli and Taibi, 2013; Mitew et al., 2013; Jo et al., 2014; Wu et al., 2014; Zhan et al., 2014), which may contribute to overall network dysfunction. In this study, we show that the synaptic adhesion molecule Pcdh-γC5 is concentrated at GABAergic synaptic sites, where it mediates GABAergic synaptic transmission. Further, Pcdh-γC5 expression is elevated in APP/PS1 mice and in response to Aβ treatment and Pcdh-γC5 inhibition rescued the aberrant spontaneous synaptic currents induced by oligomeric Aβ and in neurons derived from APP/PS1 mice. The enhanced GABAergic synaptic function mediated by Pcdh-γC5 could act as a compensatory mechanism to neutralize the hyperexcitation induced by Aβ. However, sustained Pcdh-γC5 activation may result in the dysregulation of neuronal circuity and excitatory/inhibitory imbalance. Nevertheless, it is also possible that Pcdh-γC5, through its synaptic adhesion properties, modulates both excitatory and inhibitory activities in parallel. Regardless, our finding implicates a contribution of the Pcdh-γC5-dependent synaptic system in AD pathogenesis.
We have shown previously that Pcdh-γC5 interacts with GABAA receptors (Li et al., 2012). Our current results revealed that it colocalizes with presynaptic vGAT-positive puncta. This is in agreement with its characteristic as a synaptic adhesion molecule and the results combined support a model whereby presynaptic and postsynaptic Pcdh-γC5 interaction mediates synaptic transmission. Parallel increases in vGAT and GAD, which are specific GABAergic synaptic markers and responsible for presynaptic synthesis and release of neurotransmitter GABA, respectively, suggest that modulation of inhibitory synaptic function requires coordinated expression of multiple GABAergic proteins. In addition to the GABAergic neurons, Pcdh-γs was recently reported to interact with neuroligin-1 (Molumby et al., 2017), an adhesion protein mainly expressed in excitatory synapses. Therefore, it is conceivable that Pcdh-γC5 could modulate excitatory synaptic function independently or in corroboration with GABAergic neurons. In addition, Pcdhs have been documented to mediate the neuron–astrocyte interaction (Garrett and Weiner, 2009) and aberrant release of GABA from astrocytes has been reported to enhance neuronal tonic inhibition, leading to LTP deficit and cognitive impairment in AD-relevant settings (Mitew et al., 2013; Jo et al., 2014; Wu et al., 2014). Therefore, whether astrocytic Pcdh-γC5 also participates in GABAergic modulation is an interesting question and warrants further investigation.
We observed neuronal hyperexcitation and enhanced presynaptic neurotransmitter release in APP/PS1 mice. Enhanced GABAergic inhibition has been proposed as a compensatory mechanism to restore disrupted balance between inhibitory and enhanced excitatory activities in AD progression (Palop et al., 2007). Our data showing that direct induction of hyperexcitation by KA leads to upregulation of Pcdh-γC5 and the inhibitory markers vGAT and GAD is in support of this idea. It is also consistent with reports that Aβ treatment in vitro increases both neuronal excitability (Minkeviciene et al., 2009) and enhanced levels of functional GABAA receptors (Zhan et al., 2014). Pcdh-γs have been implicated in multiple neuronal functions, including cell survival, dendrite arborization, neurite self-avoidance, the neuron–astrocyte interaction, and the formation of functional neural circuits (Garrett et al., 2012; Stoeckli, 2014; Keeler et al., 2015a, 2015b). Like classic cadherins, Pcdh-γ family members undergo proteolytic processing by metalloproteinases and the γ-secretase, which results in the release of the intracellular domain, followed by its nuclear translocation and signal transduction (Haas et al., 2005; Hambsch et al., 2005; Junghans et al., 2005). Interestingly, the generation of cadherin intracellular domain in neurons could be stimulated by membrane depolarization or direct activation of NMDA receptors (Junghans et al., 2005; Reiss et al., 2005). Therefore, Pcdh-γC5 may sense neuronal hyperexcitation and initiate transcriptional regulation in a similar manner. However, diminished perisomatic GABAergic synaptic terminals were observed in the cortex of AD patients and APP/PS1 mice, especially around amyloid plaques (Garcia-Marin et al., 2009; Ramos-Miguel et al., 2015). In addition, through EEG recording, reduced gamma oscillatory activity due to interneuron impairment was also observed in hAPP mice (Verret et al., 2012). These apparent contradictory findings may be caused by differences in the mouse models or the timing of the experiment. Indeed, by studying various AD mouse models at different ages, it was reported that excitatory glutamatergic and inhibitory GABAergic synaptic terminals were increased at early ages but declined at late stages (Bell and Claudio Cuello, 2006; Marttinen et al., 2015).
In addition to Aβ-induced changes of GABAergic transmission, abnormal activation of GABAergic system was also found to participate in the hyperphosphorylation of tau and the formation of neurofibrillary tangles (Nykänen et al., 2012; Nilsen et al., 2013). In addition, using various human ApoE knock-in mouse models, it was revealed that ApoE4 specifically released from GABAergic interneurons, but not pyramidal neurons or astrocytes, is responsible for its deleterious effect on neuronal loss and cognitive deficit (Li et al., 2009; Knoferle et al., 2014; Tong et al., 2014). Together, these findings support a widespread role of GABAergic dysfunction in AD pathogenesis. It will be interesting to explore whether Pcdh-γC5 is also involved in these processes.
Footnotes
This work was supported by Natural Science Foundation of Fujian Province of China (Grant 2017J01151 and 2016J01411), the National Natural Science Foundation of China (Grant 81301105 and 81771164), National Institutes of Health (Grant R01 AG032051), and the National Key Research and Development Program of China (Grant 2016 YFC1305903).
The authors declare no competing financial interests.
- Correspondence should be addressed to either Yanfang Li or Hui Zheng, Fujian Provincial Key Laboratory of Neurodegenerative Disease and Aging Research, Institute of Neuroscience, College of Medicine, Xiamen University, Xiamen, 361005 Fujian, China. yfli{at}xmu.edu.cn or huiz{at}bcm.edu