Amyloid-β Fosters p35/CDK5 Signaling Contributing to Changes of Inhibitory Synapses in Early Stages of Cerebral Amyloidosis

Abstract

Early changes in inhibitory synapse connectivities are thought to contribute to the excitation/inhibition imbalance preceding neurodegeneration in Alzheimer’s disease (AD). Recently, we reported a robust increase in the level of different key-proteins of inhibitory synapses in hippocampal subregions of pre-symptomatic APPswe-PS1 mice, a model of cerebral amyloidosis. Besides increased inhibitory synaptic clusters on parvalbumin-positive projections in CA1 and CA3, we observed impaired communication between these two hippocampal areas of young APP-PS1 mice. Interestingly, the phosphorylation of gephyrin, a major organizer of inhibitory synapses, was also increased. Here, we demonstrate that the protein levels of CDK5, a kinase involved in the phosphorylation of gephyrin, and its regulatory protein p35 are also significantly increased in hippocampal subregions of young APP-PS1 mice. Consistently, the expression of hAPP-swe in cultured hippocampal neurons resulted in higher p35-protein levels, indicating a possible molecular link between increased Aβ-production and the elevated p35/CDK5 levels seen in vivo. Further, a shRNA mediated downregulation of p35-expression in hippocampal neurons correlated with a decrease in gephyrin phosphorylation and in a reduced density of synaptic γ2-GABA_A-receptor clusters. These findings, together with the detection of gephyrin colocalization with CDK5 and p35 by immunostaining and proximity ligation experiments in vivo and in vitro, are supporting our hypothesis that Aβ has a profound impact on inhibitory network properties, likely mediated at least in part by p35/CDK5 signaling. This further underscores the impact of altered inhibitory synaptic transmission in AD.

Keywords

Alzheimer’s disease APP-PS1 CDK5 CDK5r1 inhibitory synapse p35

INTRODUCTION

Patients as well as different mouse models carrying mutations in genes that cause early-onset familial Alzheimer’s disease (AD) such as amyloid-β protein precursor (AβPP) and presenilin-1 (PS1) exhibit impairment of hippocampal synaptic plasticity long before the onset of clinical symptoms [1, 2]. It is largely acknowledged that in addition to the excitatory system, inhibitory GABA(γ-aminobutyric acid)ergic neurotransmission also undergoes significant changes in the AD brain, already in early stages of disease pathogenesis [3, 4]. Different components of the inhibitory system seem to be affected including GABA levels, GABA dependent network oscillations, expression levels, morphological and functional characteristics of GABA_A receptors and regulatory proteins [5, 6]. Previously, we reported an increase in the protein levels of both gephyrin, the main organizer of inhibitory synapses, and the postsynaptic γ2-GABA(A)-receptor subunit in hippocampal subregions of young pre-symptomatic double transgenic APP_swe/PS1_L166P mice (APP-PS1) (1–3 months) [7]. Subsequently, we demonstrated an increased number of presynaptic parvalbumin-positive projections in the hippocampal CA1 and CA3 subregions, which correlated with an increased number of parvalbumin-related GABAergic synapses in particular in the CA1 subregion of the hippocampus. Analysis of sharp-wave ripples (SWRs) and γ-oscillations revealed changes in different parameters such as amplitudes and frequencies when comparing CA1 and CA3 subregions suggesting an impaired communication between networks of these two areas of hippocampus in 3-month-old APP-PS1 mice [8]. In addition, we encountered a steady elevation in the phosphorylation level of gephyrin in the hippocampus in this early stage of cerebral amyloidosis.

Phosphorylation of synaptic proteins by protein kinases represents one essential pathway regulating structural and functional properties underlying synaptic plasticity [9]. Cyclin-dependent kinase 5 (CDK5) is a serine/threonine kinase that play important roles during central nervous system development [10, 11]. Emerging evidence suggests that CDK5 is involved also in a huge number of neuronal functions in the adult brain encompassing synaptic plasticity and homeostasis [12 –14]. Especially well documented is the decisive role of the CDK5-dependent signaling in regulating different aspects of synaptic plasticity at excitatory synapses [12, 15] involving neurotransmitter release [16], dendritic spine density [17, 18], postsynaptic N-methyl-D-aspartate (NMDA] receptor clustering [19, 20], and even expression of neurotransmitter receptors by nuclear signaling [21]. Phosphorylation of various synaptic components including the prototypic excitatory synaptic scaffold protein, PSD-95 (postsynaptic density protein 95) seems to represent a major molecular mechanism in mediating these regulatory functions [13 , 22–25].

CDK5 is activated through direct binding by its neuron specific activators, p35 (CDK5R1), p39 (CDK5R2), or their calpain-mediated cleavage products p25 or p29, each of them with varying activities in different subcellular localizations [26, 27]. Since CDK5 and p35 transgenic and knockout mice exhibit morphological and functional synaptic changes which result in severe spatial learning and memory deficits, physiological CDK5 activation is considered essential for maintaining homeostatic synaptic plasticity and hippocampus-dependent learning in adult animals [10 , 28]. Deregulation, in particular hyperactivation, of CDK5 seems to be a main contributor to the pathogenesis of some neurodegenerative diseases including AD, primarily by causing aberrant hyperphosphorylation of various substrates like tau and neurofilament proteins as well as APP and PS1 [29 –31]. Studies on postmortem AD brains revealed significantly higher activity of CDK5 [32] and colocalization of CDK5 with early stage neurofibrillary tangles in neurons [33]. In mouse models, intracerebroventricular injection of the amyloid-β (1–40) peptide led to CDK5 hyperactivation with subsequent tau hyperphosphorylation, synaptotoxicity, and neuronal cell death [34, 35].

In contrast to the numerous studies demonstrating the importance of CDK5 for the glutamatergic neurotransmission [29 –35], data about a potential role of CDK5 at inhibitory synapses are sparse. Even more, despite the broad literature documenting dysregulated CDK5 signaling in AD and pointing toward CDK5 as potential target for AD therapy [36], no data are yet available about putative CDK5-dependent changes of the GABAergic synaptic system in this neurodegenerative disease.

Recently, Li and coworkers reported that pharmacological inhibition of CDK5 reduced the expression of the major GABA synthesizing enzymes, glutamate decarboxylase (GAD) 67 and GAD65, and impaired mIPSC (miniature inhibitory postsynaptic current) frequency, thus suppressing local GABAergic circuits in the visual cortex of adult mice [37]. This is in line with our earlier report in which we could show that shRNA-mediated knockdown or pharmacological inhibition of CDK5 in cultured hippocampal neurons result in reduced phosphorylation of gephyrin and a lower number of γ2-containing GABA_A receptor clusters [38]. Altogether, these findings suggest an important additional role for CDK5 in also regulating the plasticity of inhibitory synapses. Gephyrin is the major scaffold protein at inhibitory synapses and has similar roles as PSD-95 at the excitatory synapses in clustering of glutamatergic receptors: it is essential for clustering of subtypes of GABA_A- and glycine receptors at inhibitory postsynaptic membrane specializations in forebrain and spinal cord, respectively [39, 40]. Gephyrin is substrate for various posttranslational modifications, which are thought to induce conformational changes thereby altering its clustering, trafficking, and binding properties [41, 42]. Specifically, Flores et al. demonstrated that activity-dependent rearrangement of perisomatic inhibitory synapses in the hippocampus involves the phosphorylation of gephyrin [43]. Battaglia et al. reported activity-dependent changes within gephyrin clusters involving phosphorylation, which are assumed to modify the affinity of gephyrin for synaptic GABA_ARs and resulting in increased lateral motility of receptor molecules at postsynaptic membrane specializations [44].

In this study we investigated if p35/CDK5 signaling could be involved in the upregulation of the perisomatic gephyrin phosphorylation and GABAergic inhibition in hippocampal subregions of young, presymptomatic APP-PS1 mice described previously [38, 45]. For this purpose, primarily, we investigated if increased heterologous expression of mutated hAPP in cultured hippocampal neurons affects the expression of proteins related to CDK5 signaling and inhibitory synapses. Secondly, we examined the expression of CDK5 and its regulatory subunit p35 in the hippocampus of 3-month-old APP-PS1 mice by immunoblotting and immunofluorescence and compared the subregional and subcellular distribution of these proteins at inhibitory synapses with that of age-matched wild-type mice. Thirdly, we analyzed if the shRNA-mediated knockdown of p35 influences the formation of inhibitory synaptic structures with focus on gephyrin phosphorylation and GABA_A receptor clustering in hippocampal neurons. Our results show an early, presumably Aβ-fostered upregulation of p35/CDK5 signaling which affects inhibitory synapse organization and plasticity in the hippocampus of pre-plaque-stage APP-PS1 mice.

MATERIALS AND METHODS

Hippocampal cell cultures and lentiviral infection

Construction of expression plasmids

DNA fragments encoding the complete coding sequence of human APP695 wild-type and human APPswedish (APP695 KM670/671NL) (APPswe) mutant, which have been tagged with a N-terminal myc-epitope coding sequence cloned into the Kpnl site of an APP cDNA [46] (Eggert S, unpublished] were amplified by PCR, creating BamH1 and EcoRI flanking restriction sites. These PCR-fragments were subsequently inserted into the BamHI and EcorRI restriction sites of the lentiviral vectors pFUGW [47] and pFSGW [48] replacing the corresponding EGFP encoding BamH1 and EcoRI fragments of these plasmids. A ubiquitin promoter (U) in the vector pFUGW or a human synapsin promoter (S) in pFSGW drove the expression of the described APP-constructs.

Construction of p35-knock-down vectors (Supplementary Figure 1)

The following sequences coding for shRNAs directed toward p35 were cloned into pFSGW using the human synapsin promoter to drive EGFP expression as described in [48]. The p35-shRNA-coding sequences: p35rat shRNA1 : 5^′-TTTgcaagaacgccaaggacaagaGTGAAGCCACAGATGtcttgtccttggcgttcttgcTTTTT-3’ and p35rat shRNA2 : 5^′-TTTgcctctcagttatcaacctcaGTGAAGCCACAGATGtgaggttgataactgagaggcTTTTT-3^′ were first cloned into the BstBI and BbsI restriction sites of the plasmid pCMV-U6 delBbsI allowing the subsequent recloning of NheI-BstBI restriction fragments, containing the described shRNA sequences under the control of the U6 promotor, into the NheI and BstBI sites of pFSGW.

Lentivirus preparation

Recombinant lentiviral particles were produced as described previously [47]. In brief, HEK293LTV cells (Cell Biolabs) were transfected with equimolar amounts of pFSGW, pΔ8.9, and pVSVg using polyethyleneimine (Sigma-Aldrich). Virus particles were harvested 2 days after transfection by collecting the cell culture supernatant and concentrated by ultracentrifugation (90 min at 75,000 g (25,000 rpm, SW32Ti rotor, Beckman Coulter)). Virus containing solution was aliquoted, shock frozen in liquid nitrogen and stored at –80°C for further use.

Cell culture

Primary cultures of rat hippocampal neurons were prepared from E19 embryos plated at a density of 60,000 cells/cm² and transfected as described previously [48] into 24-well plates or 6-well plates (for protein extracts and Western analysis). Infection with lentivirus dilutions was performed two days after plating (days in vitro/div2) with I. lentiviruses expressing either humanAPP-wild-type, humanAPP-Swedish mutation or EGFP or II. for p35 knockdown experiments, pFSGW based lentiviruses expressing EGFP only, EGFP with a scrambled shRNA as control and two knock-down lentiviruses expressing EGFP with two different p35 directed shRNAs. Cells were harvested for immunoblotting or fixed for immunocytochemistry at div 14-15 or 21.

Animals and tissue preparation

Three-month-old male APP-PS1 mice and age-matched non-transgenic littermates (WT) were used. The double transgenic mice on a C57BL/6J genetic background coexpress the KM670/671NL “Swedish” mutated amyloid precursor protein (APP) and the aggressive L166P-mutation carrying human presenilin 1 (PS1) under the control of a neuron-specific Thy1 promoter element. Through the elevated secretion of human Aβ peptide APP-PS1 mice model aspects of cerebral amyloidosis with amyloid plaque deposition in the hippocampus starting at 4-5 months of age [49]. Cognitive impairment, including deficits in the Morris Water maze test and impairments of LTP in the hippocampal subregion CA1 were reported starting at seven-eight months of age [50, 51].

APP-PS1 mice were obtained from Prof. Dr. M. Jucker (University of Tübingen, Germany) and bred in the animal unit of the University of Kaiserslautern. Offspring was genotyped as described elsewhere [49, 51]. All animal procedures have been carried out in accordance with the European Communities Council Directive (86/609/EEC) and were approved by the responsible regional state authorities of Baden-Württemberg (T-65/15 and G-72/17).

Mice (4–6 animals/group) were deeply anesthetized with isoflurane and perfused transcardially with phosphate buffered saline (PBS, pH 7.4) (Sigma Aldrich) (∼10 ml) followed by 4% paraformaldehyde (PFA, ∼50 ml) (Roth-Histofix 4%). The animals were decapitated and the brains were prepared. Brain hemispheres were separated, transferred to 4% (w/v) PFA for 1 h (at 4°C) rinsed in PBS and cryoprotected by immersion in 10% (w/v, 30 min), 20% (w/v, 1 h), and 30% (w/v) sucrose (overnight). The samples were mounted in OCT embedding compound (VWR Chemicals) and snap-frozen on an absolute ethanol-dry ice mixture. For biochemical analysis, fresh hippocampus was dissected from unfixed brain hemispheres and immediately frozen in liquid nitrogen. Tissue samples were stored at –80°C until use.

Protein extracts and immunoblot analysis

Cultured hippocampal cells (6-well plates) were washed once with ice-cold PBS, harvested and homogenized in 350 μl ice-cold lysis buffer. Extracts were passed through a cannula (0.4×19 mm, tenfold) and centrifuged (15 min, 13,000 g) before total lysates were used for protein determination using a BCA-assay kit (Thermo Scientific). Frozen hippocampus tissue was homogenized as described elsewhere [7]. Proteins were blotted on polyvinylidene difluoride membranes (Millipore) according to the manufacturer’s instructions. Membranes were probed with primary antibodies presented in Table 1. The corresponding horseradish peroxidase-conjugated secondary antibodies (Bio-Rad) were detected using ECL Prime detection kit (Amersham Biosciences). After exposure to hyperfilms (Amersham Bioscience) pixel intensities of the bands of interest were analyzed using ImageJ. Average band intensities were averaged for three independent experiments (hippocampal cell culture extracts) or from five animals (hippocampal tissue)/group.

Immunolabeling

For brain sections

Coronal cryostat sections (8 μm) cut from fresh-frozen brains were mounted on SuperFrost Plus slides (Menzel GmbH) and stored at –20°C. They were then thawed, fixed with 4% (w/v) PFA (Roth-Histofix 4%) for 8 min and preincubated for 20 min at 95°C in sodium citrate buffer (10 mM, 0.05% Tween-20, pH 6.0) for antigen retrieval. After three washes in PBS (pH 7.4) and preincubation in blocking solution (5% normal horse serum, 5% bovine serum albumin, 0.2% TritonX-100) (Roth) for 60 min, sections were incubated overnight at 4°C with primary antibodies diluted in blocking solution without TritonX-100. Individual sections were incubated with up to three primary antibodies (for colocalization of antigens), detected with corresponding secondary antibodies conjugated to fluorophores (Vector Laboratories, Invitrogen, Jackson Immunoresearch Laboratories) and mounted with mounting medium Mowiol/Dabco (Roth). To reduce autofluorescence tissue sections were treated with Autofluorescence Eliminator Reagent (Merck Millipore) according to the manufacturer‘s recommendations. Controls omitting the primary antibodies were included. Serial sections from APP-PS1 and WT mice were labelled simultaneously using the same batches of solutions to avoid differences in immunolabeling conditions. The primary antibodies and concentrations used are presented in Table 1.

For cultured hippocampal neurons

The protocol was previously described in [38, 45]. Briefly, coverslips containing cultured neurons were washed once with PBS and subsequently fixed with 4% (w/v) PFA for 10 min at RT. After blocking with 5% horse-serum and 1% BSA solution containing 0.1% TritonX-100 (Roth) for 30 minutes cells were incubated with up to three primary antibodies (Table 1) overnight at 4°C followed on the next day by incubation with appropriate secondary antibodies conjugated to fluorophores (Vector Laboratories, Invitrogen, Jackson Immunoresearch Laboratories) for 30 min at room temperature. Both primary and secondary antibodies were diluted in blocking solution without TritonX-100. After further 3x washes in PBS, samples were rinsed in demineralized water and the coverslips mounted onto glass slides using Mowiol/Dabco (Roth). The slides were kept at 4°C in the dark until imaging.

Confocal laser scanning microscopy and quantitative immunofluorescence analysis

Confocal microscopy was performed with a Leica TCS SP8 microscope (Leica Microsystems CMS GmbH, Mannheim, Germany) using a HC PL APO CS2 63.0×1.40 oil objective.

For brain sections

The protocol was described in detail previously [7, 8]. Briefly, the emission filter settings were 490–540 nm for PMT2 (green), 555–620 nm for PMT3 (red), and 630–665 nm for PMT4 (cyan). The images were acquired in sequential mode with a frame average of 4. Stacks of 8 optical sections (1024×1024 pixels) spaced by 500 nm were recorded. Laser power and settings were identical for all samples in an experiment. Four randomly chosen fields within CA1 and three fields within CA3 and dentate gyrus (DG) of each hippocampus (n = 4–6 brains/group) were recorded for quantitative analysis. In each field, rectangular areas of 400×120 μm within the pyramidal cell layer (PCL) and stratum radiatum (SR) of CA1 and CA3 and the granular cell layer (GCL) of DG were randomly selected and quantified for mean fluorescence intensities using NIH’s Fiji (pacific.mpi-cbg.de/wiki/index.php/fiji). Mean fluorescence intensity of a region of interest was calculated from maximal intensity projection of 8 optical sections. To analyze fluorescence intensities at overlapping sites for p35, pGeph, and GABA_AR-γ2 in the triple labeled brain sections an ImageJ/Fiji macro was developed (CellNetworks Math-Clinic Core Facility, University of Heidelberg, Germany) as described [8]. This first semi-automatically segmented the immunofluorescent puncta using the threshold method and then automatically processed the generated binary masks to find overlapping signals between the three different confocal channels. The number of overlapping puncta and corresponding fluorescence intensities for each channel were then counted by the ImageJ/Fiji macro and a customized summary table was created in the output directory for each processed image for further validation and statistical analysis. Mean values calculated for each animal were used for final statistics. WTs were set to 1, and data are given as mean±SEM [7, 8].

For cultured hippocampal neurons

To measure p35 expression upon hAPP-swe expression images from myc-tagged hAPP-swe, hAPPwt expressing neurons as well as non-infected controls were acquired using a Zeiss Axio Observer fluorescence microscope (63×/1.46 oil objective). Mean fluorescence intensities of p35 immunoreactivities at the somata of infected (myc+) and control (identified by DAPI) neurons were measured using the NIH’s Fiji software. For quantification of synaptic Geph, pGeph and GABA_AR-γ2 clusters in p35 deficient hippocampal neurons GFP-positive pyramidal cells were acquired as 4-channel, sequential Z-stacks. All settings, including laser power, were kept constant between all recordings. Particle analysis was performed using NIH’s Fiji. In short, regions of interest (ROIs) containing all clusters within the most proximal part of the apical dendrite or of the entire circumference of the neuron’s soma were manually drawn on the Z-projections. Within the selected ROI, the Geph, pGeph or GABA_AR-γ2 signal was thresholded (thresholds were kept constant throughout all recordings) and all particles of at least 0.01 μm² were measured for size, average and maximal grey intensities. For each particle total fluorescence intensity was calculated as cluster size×(pixel fluorescence intensities - average cytoplasmic background fluorescence). VGAT positivity was used to subdivide clusters as synaptic or extrasynaptic. For the analysis of p35 immunofluorescence intensities, ROIs were manually drawn for the pyramidal cell soma. Average grey values within each area were obtained. Four different experiments with five till ten pyramidal neurons/experiment were analyzed if not other specified. Statistics were performed as one-way ANOVA.

Proximity ligation assay (PLA) of hippocampal neurons

PLA, which allows to visualize individual protein interactions at a range of 30–40 nm, was performed according to the manufacturer’s guidelines (Duolink^® Orange Detection System, Olink Biosciences, Sweden) [52] with slight modifications: PFA 4% fixed GFP expressing neurons (see above: cell cultures) after blocking were incubated with a combination of two primary antibodies anti-gephyrin (3B11, Synaptic Systems) with anti-CDK5 (C8) (Santa Cruz) or anti-CDK5 (Novus Biologicals) or anti-p35 (Santa Cruz) antibodies overnight at 4°C. After washing and incubation with the corresponding DUO-link+/–secondary antibodies cells were newly incubated with an antibody against the pre-synaptic marker VGAT (Synaptic Systems) followed by a corresponding secondary antibody. After a new washing step, the subsequent ligation and polymerization reactions were performed subsequently for 30 and 100 min in dark at 37°C. After further washes, the coverslips were mounted onto glass slides using Mowiol/Dabco (Roth). To analyze the formation of PLA-signals, fluorescence images were acquired with a Leica confocal TCS SP8 microscope using a 63×Plan-Apochromat (1.4 NA) objective. Primary antibodies used are presented in Table 1.

Statistical analysis

Statistical evaluation was done in Prism (GraphPad Software Inc). Statistical significance of immunofluorescence and immunoblot data were determined using unpaired Student’s t-test or one-way ANOVAs with posthoc Tukeys multiple comparison tests: *p < 0.05, **p < 0.01, and ***p < 0.001. Numeric values are given as mean±standard error of the mean (SEM) or standard deviation (SD) as given.

RESULTS

hAPP-swe expression leads to both increased phospho-gephyrin and p35 protein levels in cultured hippocampal neurons

To test a putative link between initial amyloidogenesis, p35/CDK5 signaling and gephyrin phosphorylation, we expressed myc-tagged human APP carrying the Swedish mutation (hAPPswe), which is known to increase extracellular levels of Aβ [53] in cultured rat hippocampal neurons. Neurons transfected with a lentivirus construct expressing eGFP or myc-tagged wild-type human APP (hAPPwt) served as controls. Lentiviral expression constructs with either a ubiquitin (U) or a synapsin (S) promotor were used. The expression levels of p35, CDK5, GSK3β, gephyrin, and phosphorylated gephyrin (pGeph) were compared in homogenates of infected and control hippocampal cells (div15) by immunoblot analysis. Infection was assessed by using a hAPP specific antibody (WO2) (Fig. 1a).

Fig.1

Increased phospho-gephyrin and p35 protein levels in hAPPswe expressing cultured hippocampal neurons. a) Representative immunoblots of lysates obtained from cultured hippocampal neurons (div15): untreated control cultures (C), cultures expressing EGFP (GFP), cultures expressing mutated humanAPP (hAPPsw) and cultures expressing wild type human APP (hAPPwt) driven by a ubiquitin promotor (U) or a synapsin promotor (S). The presented blots were obtained from three different membranes: one membrane was probed subsequently with antibodies directed against phospho-gephyrin (pGeph, mAb7a, mouse), p25/35 (rabbit) and β-actin as loading control and after membrane stripping with an antibody against total gephyrin (3B11, mouse), a second membrane was probed subsequently with antibodies against CDK5, GSK3β and β-actin and a third membrane with antibodies against human APP (hAPP) and human + mouse APP. b) Quantification of p35-protein band intensities from three independent experiments (n = 3). Mean±SEM. *p < 0.05, **p < 0.01 hAPPswU/S versus control, ^§p < 0.05 hAPPswS versus hAPPwtS. Student’s t-test. c) Representative immunofluorescence (IF) images showing expression of p35 and hAPP in control cultures in comparison to hAPPsw and hAPPwt expressing cultured hippocampal neurons driven by a synapsin promoter. Confocal maximum intensity projection images (8 optical sections, 3.5- μm thick Z-stack); Scale bar: 10 μm. d) Quantification of mean IF intensities for p35-protein (see methods) (n = 30 cells/ group). Mean±SD; **p < 0.01 hAPPswe versus control, and *p < 0.05 hAPPSwt versus control. Student’s t-test.

We found elevated levels of pGeph and p35 in homogenates of both hAPPwt and hAPPswe expressing neurons compared to control cultures (Fig. 1a). By quantification of Western blot band intensities from three different experiments the increase in p35 protein in hAPPswe expressing cells proved to be significantly higher than in control extracts (Fig. 1b). By immunocytochemical analysis of transfected and control cultures the increase in p35 fluorescence intensities was significantly higher in both hAPPwt and hAPPswe neurons in comparison to controls (Fig. 1c, d). These results support our hypothesis that hAPPswe might trigger the increase of p35 protein levels thereby upregulating CDK5 activity. This in turn could contribute to elevated levels of pGeph in the APP-PS1 mouse model.

p35 and CDK5 are increased in the hippocampus of young APP-PS1 mice

To address the question, whether the expression of hAPP-swe in vivo can cause similar changes in p35 expression and consequently CDK5 activity, we compared the protein levels and distribution of these proteins in the hippocampus of APP-PS1 transgenic mice to age-matched controls. For our analysis we chose 3-month-old animals, as our earlier time course analyses (including 1-, 3-, 8-, and 12-month-old animals) revealed a significant increase in hippocampal pGeph levels at this age [7]. Our immunoblot analyses (Fig. 2a, Supplementary Figure 2) showed that the expression levels of both p35 and CDK5 were significantly higher in homogenates prepared from APP-PS1 hippocampus than in those prepared from WT mice. Quantification of the immunoblots corroborated that the changes are statistically significant (Fig. 2b). These results are mostly consistent with the findings of Pedros et al. who reported increased p35 (but not CDK5) protein levels in the hippocampus of 3- and 6-month-old APP-PS1 mice [54]. A systematic and detailed analysis of the distribution patterns of p35 and CDK5 in subregions of the hippocampus of mouse models for AD, which could help to clarify the functional roles of these proteins, is yet not available. Thus, next we performed single and double-labeled immunostainings of coronal brain sections and studied in detail the distribution and subcellular localization of these proteins in the hippocampus of APP-PS1 and control mice. The overall distribution pattern of both immunoreactivities was very similar in APP-PS1 and WT hippocampus (Fig. 2c). Both p35 and CDK5 immunoreactivities were mainly localized in the PCL of CA1 and CA3 as well as in the GCL of DG, suggestive for somatic localization. In the pyramidal cell soma both proteins showed pronounced but uneven immunoreactivity, with p35 in a more distinct, punctuated pattern, which also extended throughout the dendrites as seen in the SR (Fig. 2c). Corresponding to earlier findings [55, 56], p35 and CDK5 showed a similar subcellular distribution pattern, by double labeling being largely colocalized (Fig. 2c). Although the distribution of p35 and CDK5 immunoreactivities in the aforementioned subregions of the hippocampus was very similar between APP-PS1 and WT mice, the intensities of both p35 and CDK5 immunoreactivities were increased when compared to non-transgenic littermates. This difference was more prominent for CA1 and CA3 compared to the smaller extent for the DG of hippocampus (Fig. 2c). Indeed, quantification of the mean fluorescence intensities (see Methods) demonstrated significantly higher values especially for CA1 and CA3 in the APP-PS1 brain sections (p35: ∼ 45% increase in CA1 and CA3; CDK5: ∼25% increase in CA1 and CA3) as compared to WT hippocampi (Fig. 2d).

Fig.2

Increased p35 and CDK5 protein levels in the hippocampus of young APP-PS1 mice. a) Representative immunoblots of hippocampus lysates obtained from 3-month-old WT and APP-PS1 mice. The presented blots were obtained from two membranes: one membrane was probed successively with antibodies against p25/p35, CDK5 and β-actin, a second membrane with anti-human APP (W02) and anti-β-actin antibodies. (Dotted lines separate different images from the same membrane.) b) Quantification of protein band intensities from five mice/group (n = 5). Mean±SEM; Student’s t-test, *p < 0.05. c) Representative IF images demonstrating increased expression of p35 and CDK5 within CA1, CA3, and dentate gyrus (DG) of the hippocampus of 3-month-old APP-PS1 mice in comparison with WT. Both p35 and CDK5 immunoreactivities are mainly localized in the PCL of CA1 and CA3 as well as in the GCL of DG, and both proteins show uneven immunoreactivity in the cell soma. Note the similar distribution pattern of p35 and CDK5. Confocal maximum intensity projection images (8 optical sections, 3.5- μm thick Z-stack); Scale bar: 50 μm. d) Quantification of mean IF intensities (see Methods) for p35 and CDK5 measured in the pyramidal cell layer (PCL) of CA1 and CA3 and the granular cell layer (GCL) of DG. n = 4–6 animals/group; Means±SEM; *p < 0.05, **p < 0.01; Student’s t-test.

The increased activity of CDK5 signaling observed in late stage AD-brains was primarily proposed to be the result of stress-induced activation of calpain and subsequent cleavage of p35 to p25, which lead to hyperactivation and subcellular mislocalization of CDK5 [57 , 59]. However, data concerning p25 levels in brains of AD patients and mouse models are conflicting, since increased as well as unchanged and even decreased levels being reported [60, 61]. Using an anti-p25/p35 antibody, we could not detect any relevant elevation in the p25 levels in the Western blots of hippocampal homogenates of APP-PS1 mice in comparison to WT (Fig. 2a). Similarly, we could not detect an obvious subcellular redistribution of p35/p25 and CDK5 immunoreactivities into the nucleus linked to neurodegeneration [57, 62] in the brain sections of APP-PS1 mice compared to WT. Altogether, our data suggest a specific increase of uncleaved p35 in young APP-PS1 hippocampus. Thus, these findings demonstrate that the content of CDK5 and its activator p35 is significantly increased in the hippocampus of 3-month-old APP-PS1 mice compared to controls without obvious differences in the subregional and subcellular distribution pattern of these proteins.

Increased p35/CDK5 colocalize and correlate with elevated phospho-gephyrin and GABA_AR-γ2 in the hippocampus of young APP-PS1 mice

Previously, we found that in addition to gephyrin, other components of the inhibitory synapse were also upregulated in the hippocampus of young (1–3-month-old) APP-PS1 mice, including the GABA_AR γ2-subunit [7]. Furthermore, we demonstrated an increased number of pGeph-positive synapses on the parvalbumin-positive projections contacting the perisomatic region of pyramidal cells in the CA1 and CA3 subregions of the APP-PS1 hippocampus [8].

To further test our hypothesis of a possible involvement of p35/CDK5-dependent phosphorylation of gephyrin in the organization and plasticity of inhibitory synapses in the APP-PS1 hippocampus we performed triple immunofluorescence staining by using antibodies against p35, phospho-gephyrin (mAb7a) and the γ2-subunit of GABA_AR, the latter being indicated as crucial for gephyrin-dependent clustering of GABA_ARs [63]. Interestingly, recent studies suggest that the role of the γ2-subunit in GABA_AR clustering might involve rather a complex formation with neurologin-2 and Lhfpl4 (lipoma HMGIC fusion partner-like 4), a putative auxiliary subunit of GABA_ARs than its direct binding to gephyrin [64 , 66]. As seen in Fig. 3, the analysis of immunofluorescence stained brain sections confirmed again the increased expression of p35 parallel to pGeph and GABA_AR-γ2 in APP-PS1 hippocampus in comparison to WT. The detailed morphological analyses provided evidence that p35 immunoreactivity frequently overlay with pGeph and GABA_AR-γ2 immunoreactivities especially on the soma of pyramidal cells of CA1 (Fig. 3, inset) and CA3 indicating their very close spatial relationship. Similarly, p35 overlapping with pGeph+ and GABA_AR-γ2 or VGAT positive signals could be observed in vitro on triple labeled cultured (div 21) hippocampal neurons (perisomatic, proximal dendrites) (Supplementary Figure 3). In addition, to GABA_AR-γ2 immunoreactive puncta largely overlapping with gephyrin clusters and p35 immunoreactive puncta, we also observed overlapping immunoreactivities for pGeph and p35 at extrasynaptic sites in vivo (Fig. 3, inset, arrow) and in vitro (Supplementary Figure 3). Interestingly, results from PLA assays for studying the interaction between p35 or CDK5 and gephyrin in cultured hippocampal neurons are suggestive for a more extrasynaptic interaction of these proteins (Supplementary Figure 4).

Fig.3

Increased p35 colocalizes and correlates with elevated phospho-gephyrin and GABA_AR-γ2 in the hippocampus of young APP-PS1 mice. a) Triple-immunolabeling of perfusion-fixed cryosections for p35 (green), phospho-gephyrin (magenta), and GABA_A-γ2-R (red). Single channel acquisitions show the increased immunoreactivity of all three proteins in the pyramidal cell layer (PCL) and stratum radiatum (SR) of CA1 in APP-PS1 hippocampus in comparison to WT. Merged images demonstrate a higher number of overlapping p35/pGeph and p35/GABA_A-γ2-R immunoreactivities on the soma of pyramidal cells of CA1 in APP-PS1 hippocampus compared to WT, indicating the presence of perisomatic inhibitory synapses. Note: the white arrows in the fourth right inset pointing to extrasynaptic colocalization of p35 and pGeph. Confocal maximum intensity projections (8 optical sections, 3.5 μm thick Z-stack; inset: one optical section, 0.5 μm thick); Scale bars: 50 μm; insets: 5 μm. b) Quantification of mean IF intensities (see Methods) for p35, pGeph and GABA_AR-γ2 at overlapping sites measured in the pyramidal cell layer (PCL) of CA1. n = 5 animals/group; Means±SEM; *p < 0.05; Student’s t-test.

Increased p35/CDK5 colocalize and correlate with phospho-tau in the hippocampus of young APP-PS1 mice

Although the formation of neurofibrillary tangles is not detectable in the brain of APP-PS1 mice, tau hyperphosphorylation, thought to be the result of elevated Aβ levels, was shown in brain homogenates of these mice [54]. To test if the increased p35/CDK5 activity detected in the hippocampus of 3-month-old APP-PS1 mice affects phosphorylation of tau we performed immunostainings using antibodies that specifically recognize phospho-epitopes of tau and co-stained with anti-CDK5 antibodies (Fig. 4a, c). Interestingly, quantification of the resulting fluorescence intensities evidenced a significant increase in the phosphorylation of tau at Ser199 in the CA1 and DG (but not in the CA3) of APP-PS1 hippocampi compared to WT (Fig. 4d). The immunoreactivities of two other well-known CDK5 target sites on tau -Ser202 and Thr205- were not significantly increased in the afore mentioned subregions of APP-PS1 hippocampi (Fig. 4a, b), although all three epitopes are known to be phosphorylated by CDK5 and elevated levels could be detected in hippocampus homogenates of 3- and 5-month-old APP-PS1 mice [54, 67]. Thus, our results confirm that increased phosphorylation of tau occurs in the hippocampus in the APP-PS1 mice already in an early stage of amyloidogenesis, but underline that this process may occur in a subfield-, and age-specific manner and might be, at least partially the result of increased p35/CDK5 activity.

Fig.4

Increased phospho-CDK5 colocalizes and correlates with phospho-Tau in the hippocampus of young APP-PS1 mice. a, b) Representative images from perfusion-fixed cryosections immunolabeled for phospho-Tau (p-Tau) (Ser202/Th205) (red) (a) and double-labeled for pTau (Ser199) (green), and phospho-CDK5 (pCDK5) (Ser159) (red) (b). Representative single channel acquisitions show the increased expression of pTau and pCDK5 in the CA1 region of APP-PS1 hippocampus in comparison to WT. Merged images evidence that pTau (S199) and pCDK5 largely overlap in the PCL of CA1 subregion of the hippocampus. Confocal maximum intensity projections (8 optical sections, 3.5-μm thick Z-stack); Scale bar: 50 μm. c, d) Quantification of mean IF intensities (see Methods) for Tau phosphorylated at different epitopes (Ser202/Thr205 and Ser199) and CDK5 phosphorylated at Ser159 in CA1, CA3, and DG of hippocampus of WT and APP-PS1 mice. Expression of Tau phosphorylated at Ser199 is significantly increased in the CA1 subregion and DG of APP-PS1 hippocampus in comparison to WT. PCL, pyramidal cell layer; SR, stratum radiatum; GCL, granular cell layer; n = 3-4 animals/group; Means±SEM; *p < 0.05, **p < 0.01; Student’s t-test.

p35 knock-down diminishes mAb7a-positive-gephyrin- and γ2-GABA_AR-clusters in cultured hippocampal neurons

To get more insight into the functional role of p35/CDK5 in the phosphorylation of gephyrin clusters at GABAergic synapses, we established a knockdown for p35 in cultured hippocampal neurons using two different short hairpin RNAs (shRNAs) expressing lentiviruses (clone 47 and 48). Successful infection was detected by EGFP expression. Both p35 targeting shRNAs led to a 50–60% knockdown of p35 protein in transfected neurons compared to the control cultures as estimated by Western blot or fluorescence microscopy in the somatodendritic cell compartment (data not shown). Since the infection efficiency of clone 48 was very low (data not shown), we focused on clone 47 (p35kd1) for further experiments.

First, we analyzed the effects of p35 knock-down (p35kd) on the postsynaptic immunoreactivity of pGeph using the phosphospecific antibody (mAb7a) [45]. In order to differentiate between synaptic and extrasynaptic clusters, we performed double labeling with the presynaptic marker of inhibitory synapses VGAT. The reduction of p35 levels in developing hippocampal neurons led to reduction of both pGeph+ clusters density and fluorescence intensity/cluster. As seen in Fig. 5 (a–d), 12 days after infection (div 14) with a shRNA targeting p35, hippocampal pyramidal neurons displayed 2–3 times fewer pGeph positive clusters, both around the soma and at the proximal dendrite (soma: 6.5±2.2 in p35kd in comparison to 15.8±2.3 in scrambled, p < 0.05; dendrite: 6.0±1.5 in p35 knock-down in comparison to 21.0±2.9 in scrambled, p < 0.01). These effects were seen for both synaptic (pGeph+/VGAT+) and extrasynaptic (pGeph+ only) clusters (Fig. 5b). The remaining pGeph clusters in p35 knock-down neurons also had a moderate reduction in pGeph fluorescence intensities (Fig. 5c, d). However, as also those neurons expressing a scrambled shRNA construct showed a reduction in cluster fluorescence for the somatic synaptic compartment, parts of these results could be explained by unspecific effects on general siRNA pathways. Previously, we have shown that CDK5 phosphorylates gephyrin at Ser 270 in vitro and that CDK5 knock-down reduced the phosphorylation of gephyrin detected by the mAb7a antibody, while being ineffective on total gephyrin protein expression and clustering [38, 45]. Thus, we next tested the effect of p35 knock-down on the clustering of total gephyrin in cultured hippocampal neurons. No significant differences were detected either in the density or in the cluster fluorescence intensity for total gephyrin at both synaptic and extrasynaptic sites in p35 knock-down neurons when compared to GFP or scrambled shRNA expressing neurons, however a tendency to decreased density of synaptic sites could be observed (Fig. 5e–h). In addition, in p35 knock-down neurons a probably compensatory increase in the favor of extrasynaptic Geph+ clusters could be observed. It is known that GABA_A receptors harboring GABA_AR-γ2 subunits, depend on gephyrin for synaptic localization [63] and we have previously shown that they are reduced upon pharmacological inhibition or knock-down of CDK5 [38]. Thus, next we investigated whether p35 reduction can also influence the clustering of synaptic GABA_A receptors in cultured hippocampal neurons. As shown in Fig. 6 (a–d) similarly to pGeph, the density of GABA_AR-γ2 clusters were also reduced at the soma and proximal dendrites upon p35 knock-down compared to EGFP-scrambled infected controls, an effect that was seen for both synaptic and extrasynaptic clusters (GABA_AR-γ2+/VGAT+) (soma:5.9±1.4 in p35 knock-down in comparison to 30.6±5.2 in scrambled p < 0.001; dendrite: 7.5±1.5 in p35 knock-down in comparison to 19.7±3.2 in scrambled, p < 0.01) (Fig. 6). Thus, our results show that reducing the CDK5 activator p35 in hippocampal neurons leads to a decrease in phosphorylation of predominantly synaptic gephyrin and in consequence to a reduced number of γ2-GABA_A receptors at inhibitory synapses. To prove whether the CDK5/p35 dependent phosphorylation of gephyrin in cultured neurons is direct rather than indirect, we performed a proximity ligation assay using antibody combinations for gephyrin and CDK5 as well as gephyrin and p35 in cultured hippocampal neurons at div15 and div 21 stages. The use of two different CDK5 specific antibodies together with a gephyrin antibody resulted in a robust number of PLA- signals indicating a close proximity of gephyrin and CDK5 predominantly in the cell soma and to some extent within dendrites (Supplementary Figure 4a and not shown). Similar results were obtained using the combination of anti-p35 and anti-gephyrin antibodies (Supplementary Figure 4b) supporting the hypothesis that the phosphorylation of gephyrin at Ser270 by CDK5/p35 is direct. Comparing the subcellular localization of the “proximity signals” with the immunoreactivity of the presynaptic vesicle protein VGAT revealed that only few of these signals overlap with this presynaptic marker protein suggesting that most of the gephyrin phosphorylation occurs in the somata of neurons in agreement with former results [38].

Fig.5

p35 knock-down (p35kd) results in reduction of both density and fluorescent intensity of phospho-gephyrin (pGeph) but not total gephyrin (Geph) clusters. Hippocampal neurons (div14) were immunolabeled for GFP (white) to detect infected neurons, together with pGeph (a-d) or Geph (e-h) (green) and VGAT (red). a) Representative IF images indicate a strong reduction in density and fluorescent intensities of both somatic and dendritic pGeph (mAb7a+) clusters. Scale bars: 10 μm and 2 μm. b-d) Quantification of synaptic and extrasynaptic pGeph cluster density (b) and total fluorescence intensity (c, d), both at the soma and proximal dendrites (see Methods). Note: 1) a significant reduction of the density for synaptic and non-synaptic pGeph clusters in p35kd neurons as compared to controls in both the somatic and the dendritic compartment and 2) scrambled shRNA construct showed a reduction in cluster fluorescence intensities for the somatic synaptic compartment, indicating potential unspecific effects on general siRNA pathways. e) Representative IF images evidencing similar density and fluorescent intensities of gephyrin (3B11) clusters both at the soma and proximal dendrites. Scale bars: 10 μm and 2 μm. f-h) Quantification of synaptic and extrasynaptic gephyrin cluster density (f) and total fluorescence intensity (g, h) both at the soma and proximal dendrites (see Methods). *p< 0.05, **p < 0.01, ***p < 0.001 (one-way ANOVA with Tukeys multiple comparison test). Three independent experiments; n = 24 cells.

Fig.6

p35 knock-down (p35kd) results in reduction of the density of GABA_AR-γ2 clusters. Hippocampal neurons (div14) were immunolabeled for GFP (white), the GABA_AR-γ2 subunit (green) and VGAT (red). a) The representative IF images evidence a reduction in the density of both somatic and dendritic GABA_AR-γ2 clusters. Scale bars: 10 μm and 2 μm. b-d) Quantification of synaptic and extrasynaptic GABA_AR-γ2 cluster density (b) and cluster fluorescence intensity (c, d) at the soma and proximal dendrite. GABA_AR-γ2 clusters density, but not cluster fluorescence intensity was significantly reduced at both somatic and dendritic compartments in p35 knock-down neurons as compared to controls. *p < 0.05, **p < 0.01, ***p < 0.001 (one-way ANOVA with Tukeys multiple comparison test). n = 24 cells from four independent experiments.

DISCUSSION

In the present study, we demonstrated that 1) hAPP and Aβ fosters p35/CDK5 signaling in cultured hippocampal neurons and hippocampal subregions of 3-month-old APP-PS1 mice in comparison to WT littermates and 2) a decrease in p35/CDK5 signaling in cultured hippocampal neurons leads to a reduction of postsynaptic gephyrin phosphorylation and presumably thereby to a reduced density of GABA_A receptors at inhibitory synapses. Our findings support the regulatory role of p35/CDK5 signaling in inhibitory synapse organization and plasticity early in the pathogenesis of cerebral amyloidosis.

In accordance with other studies [2, 68], and our earlier findings showing that the levels of various inhibitory synaptic proteins are upregulated, and the synaptic output of PV+interneurons onto CA1 and CA3 pyramidal cells is augmented in young APP-PS1 mice underscore the view that enhanced GABAergic neurotransmission involving remodeling of inhibitory synapses could be early hallmarks in the pathogenesis of AD [7, 8]. Increasing evidence indicates that gephyrin, a principal organizer of the ligand gated ion channels at the inhibitory postsynaptic density might regulate the plasticity of inhibitory synapses even differentially depending on its phosphorylation status [42 –44]. Previously, we demonstrated that gephyrin can be phosphorylated by the proline-directed serine/threonine kinase CDK5 at position S270 in vitro [38, 45]. Additionally, using antibodies for both total and phosphorylated (mAb7a) gephyrin we showed a steady increase in gephyrin phosphorylation at S270 in hippocampal subregions of 1–3-month-old APP-PS1 mice [7, 8].

It is largely accepted that CDK5 signaling plays a pivotal role in the regulation of synaptic plasticity, although the exact mechanism behind it is not clearly understood [69]. Many studies addressed the role of CDK5 signaling on NMDA receptor-dependent plasticity and showed that phosphorylation of various pre- and postsynaptic components by CDK5 represents an important molecular mechanism of its regulatory functions and has a major impact on plasticity at excitatory synapses. CDK5-dependent phosphorylation of the NR2A subunit of the NMDA receptor has been demonstrated to alter NMDA receptor channel conductance and affect the induction of long-term potentiation (LTP) in CA1 hippocampal neurons [25]. In addition to regulation of the NR2B subunit degradation by calpain [19] phosphorylation of NR2B by CDK5 at Ser1116 has been shown to control NMDA-receptor trafficking preventing cell surface expression of NR2B-containing NMDA receptors and thus attenuating synaptic transmission [20]. PSD-95, the main scaffold protein that anchor NMDA receptors at the postsynaptic density as well as its interacting protein SPAR (Spine Associated RapGAP) can also be phosphorylated by CDK5 resulting in decreased recruitment of glutamate-gated ion-channels to postsynaptic membrane specializations [13 , 23]. GKAP (guanylate kinase-associated protein), another scaffold protein crucial for the organization of the PSD has also been identified as a direct substrate of CDK5. Here, the Aβ-induced degradation of GKAP in rat cortical neurons involved CDK5-dependent phosphorylation of the protein and resulted in disassembly of the synaptic actin cytoskeleton [24] sustaining together with earlier findings an important role of CDK5 in mediating the effects of Aβ on PSD-95 and NMDARs at excitatory synapses [70].

Considering the involvement of CDK5 signaling in the regulation of excitatory synaptic plasticity as well as in the pathogenesis of AD [29 –35], we wanted to gain insight whether the CDK5 signaling might also be linked to changes of inhibitory synapses in this early stage of amyloidosis in the APP-PS1 mice. These mice coexpress the Swedish mutated APP along with mutated human PS1, which additionally impairs amyloid protein processing leading to early elevated Aβ levels in these mice [49]. To mimic the in vivo situation of APP-PS1 mice, we expressed hAPPswe in cultured hippocampal neurons and studied the expression pattern of p35 protein, known to correlate with CDK5 activity [71]. In homogenates of hAPPswe expressing cultured hippocampal neurons, the protein level of p35 was significantly increased along with gephyrin phosphorylation. Importantly, by immunoblot analyses of hippocampus homogenates we disclosed a significantly increased level of p35 protein also in 3-month-old APP-PS1 mice in comparison to WT, thus confirming earlier findings [54]. Interestingly, we found that the level of CDK5 protein, which is actually not a decisive parameter for its activity, is also significantly increased in APP-PS1 hippocampus homogenates. This is in concordance with the results of Sadleir et al. in 2-month-old 5xFAD mice [60], although several other studies reported unchanged levels of CDK5 by increased p35 [54 , 73], indicating that various context-dependent factors might influence CDK5 levels. Evaluation of brain coronal sections after immunostaining for subregional distribution of these proteins confirmed the immunoblot results, yielding significant higher mean values for fluorescence intensities and documenting an overall increase of p35 and CDK5 proteins in CA1, CA3, and DG of APP-PS1 hippocampus. Enhancement of CDK5 activity/level has been consistently linked to AD mainly in later stages of the disease [29 –31]. The underlying mechanism has been attributed to elevation of intracellular calcium caused by cellular stress, thus activating calpain to cleave the regulatory p35 subunit of CDK5 into p25 and p10. p25 is more resistant to ubiquitin-mediated proteolysis because the eliminated p10 peptide contains the signal for degradation and therefore allows prolonged CDK5 activation, promoting cell death and neurodegeneration [10 , 57]. However, data concerning p25 levels in AD brains are contradictory [61]. We did not observe any apparent increase in p25 levels in APP-PS1 hippocampus homogenates nor p25 production in hAPPswe expressing cultured hippocampal neurons corresponding to the findings of Chen et al. who reported no obvious p25 production in Aβ-treated neurons and in 6-month-old APP-PS1 mouse hippocampus [73]. Different factors were identified to control p35/CDK5 levels and function by regulating their synthesis and degradation [69] and some of them might be active in AD. Alvarez et al. showed that fibrillary Aβ induces CDK5 activation implying the formation of a stable p35-CDK5 complex rather than generation of p25 [74]. Furthermore, the presence of high levels of p35 can block the ubiquitination of CDK5 and attenuate its degradation, as the ubiquitination site of CDK5 localizes in the p35 binding area [75]. These findings might provide one explanation for the high levels of p35 and CDK5 detected in the young pre-symptomatic APP-PS1 hippocampus. Glutamate, acting through the ionotropic glutamate receptors, is also considered an important driver of the p35/CDK5 pathway mainly through modulation of p35 stability and with seemingly different outcome depending on physiological and pathological Ca^2 + concentrations, specifically leading to p35 degradation and CDK5 inactivation [76] or cleavage of p35 to p25 and CDK5 activation, respectively [77]. Thus, a direct Aβ driven increased synaptic activity might be involved in the regulation of p35/CDK5 signaling in hAPP expressing cells and APP-PS1 mice as suggested also by our observations (unpublished data) that inhibition of synaptic activity by TTX, that inhibits neurotransmission by blocking voltage-gated sodium channels, resulted in decreased p35 protein levels.

In addition, both p35 and CDK5 can be regulated at transcriptional and poststranscriptional levels [66]. Transcription of p35 has been shown to be increased by neurotrophic factors such as brain-derived neurotrophic factor (BDNF) and nerve growth factor (NGF) as well as inflammatory mediators such as interferon-gamma (IFN-γ) and tumor necrosis factor-alpha (TNFα), which all mainly activate the Erk1/2–EGR1 (extracellular regulated MAP kinases 1 and 2 - early growth response protein 1) pathway [69]. The expression of CDK5 was also upregulated through the Erk1/2 pathway but mediated by the transcription factors Fos and CREB (cAMP-responsive element binding protein) [78]. MicroRNAs (miRNAs) were involved in post-transcriptionally modulating p35 and CDK5 expression and activity [69], some of them (e.g., miR-15/107 family) with possible implications for the pathogenesis of AD as reported recently by Moncini et al. [79].

p35/CDK5 are predominantly cytoplasmic and membrane-associated [57, 59]. The mislocalization, especially the nuclear translocation of the p25/CDK5 complex, induced by loss of the p35 membrane binding sites (localized on p10) is considered a crucial factor that leads to CDK5-induced neurotoxicity [80]. In our immunostaining experiments, both p35 and CDK5 showed predominantly somato-cytoplasmic localization and largely overlapping intracellular distribution, without apparent increase in nuclear localization in APP-PS1 hippocampus sections, which might correspond to putative more p35/CDK5 complexes.

Both p35 and CDK5 have been shown to be present in synaptosomes and co-localize with pre- and postsynaptic proteins at the excitatory synapse such as bassoon and PSD95 [81, 82]. Analogous data concerning inhibitory synapses are limited. In lysates of adult mouse visual cortex, CDK5 has been found to form a complex with GAD67, the major enzyme converting glutamate into GABA as demonstrated by co-immunoprecipitation [37]. To investigate whether p35/CDK5 are also localized at inhibitory synapses, we analyzed different subregions of the hippocampus as well as cultured hippocampal neurons double or triple labeled with antibodies against p35, pGeph, and γ2-GABA_AR. We found that p35 immunoreactivity often overlaps with pGeph and GABA_ARγ2, or VGAT immunoreactivities on the soma and proximal dendrites of pyramidal cells, suggesting colocalization of these proteins at the inhibitory synapse. However, the proximity ligation assay (PLA) which permits the detection of protein-protein interactions with a high sensitivity and specificity (recognizing two proteins which are closer than ≤30 nm to each other) revealed that the phosphorylation of gephyrin by CDK5 may occur predominantly extrasynaptically in the cell soma, which is consistent with our earlier in vitro findings showing that the inhibition of phosphatases P1/P2 resulted in a strong increase of immunofluorescence intensity for pGeph (mAb7a+) in the neuronal cell soma and dendrites [38]. Thus, these results led to the conclusion that p35/CDK5 and phosphorylated gephyrin may be located both at the inhibitory postsynapse; however, the phosphorylation of gephyrin probably occurs predominantly at extrasynaptic sites.

To analyze whether p35 is essential for CDK5-dependent phosphorylation of gephyrin and clustering of γ2-subunits containing GABA_A receptors at the inhibitory synapse, we generated p35-knockdown hippocampal neurons using shRNA-expressing lentiviruses. By confocal immunofluorescence microscopy, we examined the phosphorylation of synaptic and extrasynaptic gephyrin and the clustering of γ2-GABA_AR at the soma and proximal dendrites. In neurons with reduced expression of p35, both number and fluorescence intensity of synaptic pGeph clusters were robustly reduced as was the number of synaptic γ2-GABA_AR clusters. These results agree with our earlier findings in CDK5-deficient hippocampal neurons [45] and suggest that CDK5-dependent phosphorylation of gephyrin may positively regulate the clustering of GABA_ARs at the synapse in the hippocampus of young APP-PS1 mice. A positive role of CDK5 in promoting synaptic plasticity was identified in both p35- or CDK5-deficient mice which showed impaired synaptic plasticity and hippocampus-dependent spatial learning [28 , 55] and in vivo models of acute CDK5 gain of function which presented a dramatic increase in dendritic spine and synaptophysin positive synapse number in CA1 hippocampal neurons correlating with enhanced learning ability [28]. In contrast, other studies reported negative effects of p35/CDK5 activity on NMDA-receptor-mediated synaptic plasticity and hippocampal learning [19, 83]. Interestingly, in a recent study on p35 cKO mice in addition to an increase in firing rates of CA1 pyramidal cells a decrease in theta oscillations, known to be dependent upon perisomatic targeting inhibitory neurons, has been recorded, indicating a possible cell type-specific role of p35/CDK5 [84]. In our study, the surprisingly similar effects observed in all hippocampal subregions of APP-PS1 mice as well as in cultured hippocampal neurons suggest a general and highly effective probably Aβ related mechanism fostering the p35/CDK5 pathway which might promote the increase of GABAergic synapses in early cerebral amyloidosis.

Our results imply the involvement of gephyrin phosphorylation at S270 by CDK5 in the regulation of postsynaptic plasticity of inhibitory synapses, specifically increasing GABA_AR clustering at the synapse. This is in line with recent findings of Niwa et al. indicating that gephyrin de-phosphorylation at S270 weakens the GABA_AR-gephyrin interaction and leads to selective increase of GABA_ARs diffusion and their loss from inhibitory synapses in spinal cord neurons [85]. Similarly, a higher fraction of phosphorylation of gephyrin at S270 might contribute to the maintenance of synaptic GABA_AR levels in response to long-term application of diazepam, despite a substantial reduction in total gephyrin content [86]. The phosphorylation sites of gephyrin, predominantly mapped to the linker region of the protein, are targets of various kinases. In addition to CDK5 also GSK3β has been shown to phosphorylate gephyrin at S270 [45, 87] thereby decreasing the number of gephyrin clusters [82]. More recently, site mutant analyses identified a complementary effect of S270 along with PKA- and CaMKIIα- dependent phosphorylation at Ser303 and 305 as being essential for GABA_AR diffusion at synapses [44]. In organotypic slice cultures CaMKII-dependent phosphorylation of Ser305 increased gephyrin cluster size and promoted the formation of new gephyrin clusters mediating NMDA receptor–dependent compensatory adaptations at the GABAergic postsynaptic site [43]. A possible explanation for these partial contradictory findings could be that gephyrin is dynamically regulated by different kinases and the specific combinatorial patterns of phosphorylation contribute to input-specific adaptations at postsynaptic sites, sometimes with divergent functional consequences [42, 88].

The increased p35/CDK5 activity in 3-month-old APP-PS1 mice is likely not restricted only on gephyrin phosphorylation. It cannot be excluded that in addition to gephyrin the phosphorylation of other protein/s necessary for GABA_AR clustering at the synapse, e.g., different GABA_AR subunits [89], might depend on CDK5. Moreover, CDK5 can be active also in astrocytes affecting their reactivity and functions, e.g. increasing their neuroprotective capacities by transcriptional induction of antioxidant genes as recently reported [90, 91]. Additionally, emerging evidence denote the role of astrocytes in regulating GABAergic neurotransmission [92]. As astrocytes are activated by Aβ already early in amyloidogenesis a CDK5-mediated modulation of astrocyte functions might also contribute to the observed changes of GABAergic synapses in APP-PS1 mice.

Aβ-induced p35/CDK5 activation has been associated with changes in phosphorylation patterns and intraneuronal distribution of tau protein in vitro and in vivo [35, 74]. In hippocampal homogenates of 3-month-old APP-PS1 mice elevated levels of phosphorylated tau at various epitopes have been previously reported [54]. In addition, CDK5 (but not GSK3β) has been shown to be involved in GABA_AR-induced phosphorylation of tau. Tau as a microtubule-binding protein could participate in regulation of clustering and/or trafficking of gephyrin and GABA_ARs at the inhibitory synapse [93], and consequently modulate GABAergic synaptic signaling. Thus, we analyzed putative correlations of increased p35/CDK5 expression to tau phosphorylation in different hippocampal subregions of APP-PS1 mice by double labeling for CDK5 and different tau phosphoepitopes. We detected colocalization and a significant increase of pTau especially in the CA1 subregion and DG of APP-PS1 hippocampus at this early stage of amyloidosis, which in the absence of an obvious increase in GSK-3β expression [54], the other major tau kinase, indicate that CDK5 may be involved in the early phosphorylation cascade of tau eventually in a subregion-specific manner in the 3-month-old APP-PS1 mice. In this line, Li et al. observed elevated levels of CDK5 seven days after injecting soluble oligomeric Aβ₄₂ peptide into the hippocampus of mice, correlating with increased phosphorylation of tau in the CA1 and subgranular zone of DG [35].

In summary, our data demonstrate that upregulated p35/CDK5 dependent phosphorylation of gephyrin can represent an early molecular control mechanism at hippocampal inhibitory synapses probably to maintain synaptic homeostasis at the onset of cerebral amyloidosis in APP-PS1 mice. In addition to the findings of Li et al. showing reduced GAD-expression and mIPSC (miniature inhibitory postsynaptic currents) after CDK5-inhibition [37], our study underscores the possible role of the p35/CDK5 pathway in regulating the organization and plasticity of the inhibitory synapse.

Footnotes

ACKNOWLEDGMENTS

This work was supported by a grant of Romanian Ministry of Research and Innovation CNCS- UEFISCDI (PN-III-P4-ID-PCE-2016-0052, within PNCDI III) to E.K. and a grant of the Alzheimer Forschung Initiative e.V. (AFI) (Project number: 17024) to J.K., E.K. and S.K.

We thank Dagmar Groß (Department of Human Biology and Human Genetics, University of Kaiserslautern, Germany) and Rita Rosner (Department of Anatomy and Cell Biology, University of Heidelberg, Germany) for excellent technical assistance, Dr. Simone Eggert (Department of Human Biology and Human Genetics, University of Kaiserslautern, Germany) for providing the APPswe expression construct and Dr. Carlo Antonio Beretta (CellNetworks Math-Clinic Core Facility at Heidelberg, University of Heidelberg, Germany) for developing ImageJ/Fiji macro.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Mufson

, Mahady

, Waters

, Counts

, Perez

, DeKosky

, Ginsberg

, Ikonomovic

, Scheff

, Binder

(2015) Hippocampal plasticity during the progression of Alzheimer’s disease. Neuroscience 309, 51–67.

Palop

, Chin

, Roberson

, Wang

, Thwin

, Bien-Ly

, Yoo

, Ho

, Yu

, Kreitzer

, Finkbeiner

, Noebels

, Mucke

(2007) Aberrant excitatory neuronal activity and compensatory remodeling of inhibitory hippocampal circuits in mouse models of Alzheimer’s disease. Neuron 55, 697–711.

, Sun

, Chen

, Xu

, Bu

, Zheng

(2016) Implications of GABAergic neurotransmission in Alzheimer’s disease. Front Aging Neurosci 8, 31.

Vico Varela

, Etter

, Williams

(2019) Excitatory-inhibitory imbalance in Alzheimer’s disease and therapeutic significance. Neurobiol Dis 127, 605–615.

Fuhrer

, Palpagama

, Waldvogel

, Synek

BJL

, Turner

, Faull

, Kwakowsky

(2017) Impaired expression of GABA transporters in the human Alzheimer’s disease hippocampus, subiculum, entorhinal cortex and superior temporal gyrus. Neuroscience 351, 108–118.

Kwakowsky

, Calvo-Flores Guzmán

, Govindpani

, Waldvogel

, Faull

(2018) Gamma-aminobutyric acid A receptors in Alzheimer’s disease: Highly localized remodeling of a complex and diverse signaling pathway. Neural Regen Res 13, 1362–1363.

Kiss

, Gorgas

, Schlicksupp

, Groß

, Kins

, Kirsch

, Kuhse

(2016) Biphasic alteration of the inhibitory synapse scaffold protein gephyrin in early and late stages of an Alzheimer disease model. Am J Pathol 186, 2279–2291.

Hollnagel

, Elzoheiry

, Gorgas

, Kins

, Beretta

, Kirsch

, Kuhse

, Kann

, Kiss

(2019) Early alterations in hippocampal perisomatic GABAergic synapses and network oscillations in a mouse model of Alzheimer’s disease amyloidosis. PLoS One 14, e0209228.

Woolfrey

, Dell’Acqua

(2015) Coordination of protein phosphorylation and dephosphorylation in synaptic plasticity. J Biol Chem 290, 28604–28612.

10.

Ohshima

, Ward

, Huh

, Longenecker

, Veeranna , Pant

, Brady

, Martin

, Kulkarni

(1996) Targeted disruption of the cyclin-dependent kinase 5 gene results in abnormal corticogenesis, neuronal pathology and perinatal death. Proc Natl Acad Sci U S A 93, 11173–11178.

11.

, Tsai

(2011) Cyclin-dependent kinases in brain development and disease. Annu Rev Cell Dev Biol 27, 465–491.

12.

Lai

, Ip

(2009) Recent advances in understanding the roles of Cdk5 in synaptic plasticity. Biochim Biophys Acta 1792, 741–745.

13.

Seeburg

, Feliu-Mojer

, Gaiottino

, Pak

, Sheng

(2008) Critical role of CDK5 and Polo-like kinase 2 in homeostatic synaptic plasticity during elevated activity. Neuron 58, 571–583.

14.

Sheng

, Zhang

, Su

, Tsai

, Julius Zhu

(2016) Cdk5 is a new rapid synaptic homeostasis regulator capable of initiating the early Alzheimer-like pathology. Cereb Cortex 26, 2937–2951.

15.

Dixit

, Banerjee

, Tripathi

, Sarkar

, Chandra

(2017) Synaptic roles of cyclin-dependent kinase 5 & its implications in epilepsy. Indian J Med Res 145, 179–188.

16.

Kim

SH1

, Ryan

(2010) CDK5 serves as a major control point in neurotransmitter release. Neuron 67, 797–809.

17.

Mishiba

, Tanaka

, Mita

, He

, Sasamoto

, Itohara

, Ohshima

(2014) Cdk5/p35 functions as a crucial regulator of spatial learning and memory. Mol Brain 7, 82.

18.

Mita

, He

, Sasamoto

, Mishiba

, Ohshima

(2016) Cyclin-dependent kinase 5 regulates dendritic spine formation and maintenance of cortical neuron in the mouse brain. Cereb Cortex 26, 967–976.

19.

Hawasli

, Benavides

, Nguyen

, Kansy

, Hayashi

, Chambon

, Greengard

, Powell

, Cooper

, Bibb

(2007) Cyclin-dependent kinase 5 governs learning and synaptic plasticity via control of NMDAR degradation. Nat Neurosci 10, 880–886.

20.

Plattner

, Hernández

, Kistler

, Pozo

, Zhong

, Yuen

, Tan

, Hawasli

, Cooke

, Nishi

, Guo

, Wiederhold

, Yan

, Bibb

(2014) Memory enhancement by targeting Cdk5 regulation of NR2B. Neuron 81, 1070–1083.

21.

, Fu

, Cheung

, Tsim

, Ip

, Wang

, Ip

(2001) Cdk5 is involved in neuregulin-induced AChR expression at the neuromuscular junction. Nat Neurosci 4, 374–381.

22.

Morabito

, Sheng

, Tsai

(2004) Cyclin-dependent kinase 5 phosphorylates the N-terminal domain of the postsynaptic density protein PSD-95 in neurons. J Neurosci 24, 865–876.

23.

Zhang

, Edelmann

, Liu

, Crandall

, Morabito

(2008) CDK5 regulates the phosphorylation of tyrosine 1472 NR2B and the surface expression of NMDA receptors. J Neurosci 28, 415–424.

24.

Roselli

, Livrea

, Almeida

(2011) CDK5 is essential for soluble amyloid β-induced degradation of GKAP and remodeling of the synaptic actin cytoskeleton. PLoS One 6, e23097.

25.

, Sun

, Zhang

, Takahashi

, Ma

, Vinade

, Kulkarni

, Brady

, Pant

(2001) Regulation of NMDA receptors by cyclin-dependent kinase-5. Proc Natl Acad Sci U S A 98, 12742–12747.

26.

Tsai

, Delalle

, Caviness

Jr , Chae

, Harlow

(1994) p35 is a neural-specific regulatory subunit of cyclin-dependent kinase 5. Nature 371, 419–423.

27.

Tang

, Yeung

, Lee

, Matsushita

, Matsui

, Tomizawa

, Hatase

, Wang

(1995) An isoform of the neuronal cyclin-dependent kinase 5 (Cdk5) activator. J Biol Chem 270, 26897–26903.

28.

Fischer

, Sananbenesi

, Pang

, Lu

, Tsai

(2005) Opposing roles of transient and prolonged expression of p25 in synaptic plasticity and hippocampus-dependent memory. Neuron 48, 825–838.

29.

Shukla

, Skuntz

, Pant

(2012) Deregulated Cdk5 activity is involved in inducing Alzheimer’s disease. Arch Med Res 43, 655–662.

30.

Liu

, Wang

, Jiang

, Tan

, Xing

, Yu

(2016) The role of Cdk5 in Alzheimer’s disease. Mol Neurobiol 53, 4328–4342.

31.

Cortés

, Guzmán-Martínez

, Andrade

, González

, Maccioni

(2019) CDK5: A unique CDK and its multiple roles in the nervous system. J Alzheimers Dis 68, 843–855.

32.

, Nakamura

, Cao

, Holland

, McKercher

, Lipton

(2011) S-Nitrosylation activates Cdk5 and contributes to synaptic spine loss induced by beta-amyloid peptide. Proc Natl Acad Sci U S A 108, 14330–14335.

33.

Takahashi

, Iseki

, Kosaka

(2000) Cdk5 and munc-18/p67 co-localization in early stage neurofibrillary tangles-bearing neurons in Alzheimer type dementia brains. J Neurol Sci 172, 63–69.

34.

Lopes

, Oliveira

, Agostinho

(2010) Neurodegeneration in an Abeta-induced model of Alzheimer’s disease: The role of Cdk5. Aging Cell 9, 64–77.

35.

, Tian

, Sun

, Liang

, Cheng

, Wang

, Ma

, Liu

, Li

, Wang

(2015) Association of Cdk5 and soluble oligomeric species of β-amyloid induced tau hyperphosphorylation. J Neurol Neurobiol 1, doi: 10.16966/2379-7150.108

36.

Shupp

, Casimiro

, Pestell

(2017) Biological functions of CDK5 and potential CDK5 targeted clinical treatments. Oncotarget 8, 17373–17382.

37.

, Wang

, Zhang

, Huang

, Li

, Wang

, Chen

, Jiang

, Yang

(2018) Inhibition of Cdk5 rejuvenates inhibitory circuits and restores experience-dependent plasticity in adult visual cortex. Neuropharmacology 128, 207–220.

38.

Kalbouneh

, Schlicksupp

, Kirsch

, Kuhse

(2014) Cyclin-dependent kinase 5 is involved in the phosphorylation of gephyrin and clustering of GABAA receptors at inhibitory synapses of hippocampal neurons. PLoS One 9,, e104256.

39.

Groeneweg

, Trattnig

, Kuhse

, Nawrotzki

, Kirsch

(2018)Gephyrin: A key regulatoryrotein of inhibitory synapses and beyond. Histochem Cell Biol 150, 489–508 .

40.

Pizzarelli

, Griguoli

, Zacchi

, Petrini

, Barberis

, Cattaneo

, Cherubini

(2019) Tuning GABAergic inhibition: Gephyrin molecular organization and functions. Neuroscience, doi: 10.1016/j.neuroscience.2019.07.036

41.

Ghosh

, Auguadri

, Battaglia

, Simone Thirouin

, Zemoura

, Messner

, Acuña

, Wildner

, Yévenes

, Dieter

, Kawasaki

, O Hottiger

, Zeilhofer

, Fritschy

, Tyagarajan

(2016) Several posttranslational modifications act in concert to regulate gephyrin scaffolding and GABAergic transmission. Nat Commun 7, 13365–13381.

42.

Kasaragod

, Schindelin

(2018) Structure-function relationships of glycine and GABAA receptors and their interplay with the scaffolding protein gephyrin. Front Mol Neurosci 11, 317–338.

43.

Flores

, Nikonenko

, Mendez

, Fritschy

, Tyagarajan

, Muller

(2015) Activity-dependent inhibitory synapse remodeling through gephyrin phosphorylation. Proc Natl Acad Sci U S A 112, E65–72.

44.

Battaglia

, Renner

, Russeau

, Côme

, Tyagarajan

, Lévi

(2018) Activity-dependent inhibitory synapse scaling is determined by gephyrin phosphorylation and subsequent regulation of GABAA receptor diffusion. eNeuro 5,, ENEURO.0203-17.2017.

45.

Kuhse

, Kalbouneh

, Schlicksupp

, Mükusch

, Nawrotzki

, Kirsch

(2012) Phosphorylation of gephyrin in hippocampal neurons by cyclin-dependent kinase CDK5 at Ser-270 is dependent on collybistin. J Biol Chem 287, 30952–30966.

46.

Tienari

, De Strooper

, Ikonen

, Simons

, Weidemann

, Czech

, Hartmann

, Ida

, Multhaup

, Masters

, Van Leuven

, Beyreuther

, Dotti

(1996) The beta-amyloid domain is essential for axonal sorting of amyloid precursor protein. EMBO J 15,, 5218–5229.

47.

Lois

, Hong

, Pease

, Brown

, Baltimore

(2002) Germline transmission and tissue-specific expression of transgenes delivered by lentiviral vectors. Science 295, 868–872.

48.

Dresbach

, Hempelmann

, Spilker

, tom Dieck

, Altrock

, Zuschratter

, Garner

, Gundelfinger

(2003) Functional regions of the presynaptic cytomatrix protein bassoon: Significance for synaptic targeting and cytomatrix anchoring. Mol Cell Neurosci 23, 279–291.

49.

Radde

, Bolmont

, Kaeser

, Coomaraswamy

, Lindau

, Stoltze

, Calhoun

, Jäggi

, Wolburg

, Gengler

, Haass

, Ghetti

, Czech

, Hölscher

, Mathews

, Jucker

(2006) Abeta42-driven cerebral amyloidosis in transgenic mice reveals early and robust pathology. EMBO Rep 7, 940–946.

50.

Serneels

, Van Biervliet

, Craessaerts

, Dejaegere

, Horré

, Van Houtvin

, Esselmann

, Paul

, Schäfer

, Berezovska

, Hyman

, Sprangers

, Sciot

, Moons

, Jucker

, Yang

, May

, Karran

, Wiltfang

, D’Hooge

, De Strooper

(2009) gamma-Secretase heterogeneity in the Aph1 subunit: Relevance for Alzheimer’s disease. Science 324, 639–642.

51.

Gengler

, Hamilton

, Hölscher

(2010) Synaptic plasticity in the hippocampus of a APP/PS1 mouse model of Alzheimer’s disease is impaired in old but not young mice. PLoS One 5, e9764.

52.

Söderberg

, Gullberg

, Jarvius

, Ridderstråle

, Leuchowius

, Jarvius

, Wester

, Hydbring

, Bahram

, Larsson

, Landegren

(2006) Direct observation of individual endogenous protein complexesby proximity ligation. Nat Methods 3, 995–1000.

53.

Johnston

, Cowburn

, Norgren

, Wiehager

, Venizelos

, Winblad

, Vigo-Pelfrey

, Schenk

, Lannfelt

, O’Neill

(1994) Increased beta-amyloid release and levels of amyloid precursor protein (APP) in fibroblast cell lines from family members with the Swedish Alzheimer’s disease APP670/671 mutation. FEBS Lett 354, 274–278.

54.

Pedrós

, Petrov

, Allgaier

, Sureda

, Barroso

, Beas-Zarate

, Auladell

, Pallàs

, Vázquez-Carrera

, Casadesús

, Folch

, Camins

(2014) Early alterations in energy metabolism in the hippocampus of APPswe/PS1dE9 mouse model of Alzheimer’s disease. Biochim Biophys Acta 1842, 1556–1566.

55.

Guan

, Su

, Gao

, Joseph

, Xie

, Zhou

, Durak

, Zhang

, Zhu

, Clauser

, Carr

, Tsai

(2011) Cdk5 is required for memory function and hippocampal plasticity via the cAMP signaling pathway. PLoS One 6, e25735.

56.

Lagace

, Benavides

, Kansy

, Mapelli

, Greengard

, Bibb

, Eisch

(2008) Cdk5 is essential for adult hippocampal neurogenesis. Proc Natl Acad Sci U S A 105, 18567–18571.

57.

Patrick

, Zukerberg

, Nikolic

, de la Monte

, Dikkes

, Tsai

(1999) Conversion of p35 to p25 deregulates Cdk5 activity and promotes neurodegeneration. Nature 402, 615–622.

58.

, Choi

, Qu

, Yu

, Cheung

, Qi

(2006) Identification of nuclear import mechanisms for the neuronal Cdk5 activator. J Biol Chem 281, 39014–39021.

59.

Asada

, Yamamoto

, Gohda

, Saito

, Hayashi

, Hisanaga

(2008) Myristoylation of p39 and p35 is a determinant of cytoplasmic or nuclear localization of active cyclin-dependent kinase 5 complexes. J Neurochem 106, 1325–1336.

60.

Sadleir

, Vassar

(2012) Cdk5 protein inhibition and Aβ42 increase BACE1 protein level in primary neurons by a post-transcriptional mechanism: Implications of CDK5 as a therapeutic target for Alzheimer disease. J Biol Chem 287, 7224–7235.

61.

Giese

(2014) Generation of the Cdk5 activator p25 is a memory mechanism that is affected in early Alzheimer’s disease. Front Mol Neurosci 7, 36.

62.

O’Hare

, Kushwaha

, Zhang

, Aleyasin

, Callaghan

, Slack

, Albert

, Vincent

, Park

(2005) Differential roles of nuclear and cytoplasmic cyclin-dependent kinase 5 in apoptotic and excitotoxic neuronal death. J Neurosci 25, 8954–8966.

63.

Essrich

, Lorez

, Benson

, Fritschy

, Lüscher

(1998) Postsynaptic clustering of major GABAA receptor subtypes requires the gamma 2 subunit and gephyrin. Nat Neurosci 1, 563–571.

64.

Heller

, Zhang

, Selimi

, Earnheart

, Ślimak

, Santos-Torres

, Ibañez-Tallon

, Aoki

, Chait

, Heintz

(2012) The biochemical anatomy of cortical inhibitory synapses. PLoS One 7, e39572.

65.

Yamasaki

, Hoyos-Ramirez

, Martenson

, Morimoto-Tomita

, Tomita

(2017) GARLH family proteins stabilize GABAA receptors at synapses. Neuron 93, 1138–1152.

66.

Davenport

, Pendolino

, Kontou

, McGee

, Sheehan

, López-Doménech

, Farrant

, Kittler

(2017) An essential role for the tetraspanin LHFPL4 in the cell-type-specific targeting and clustering of synaptic GABAA receptors. Cell Rep 21, 70–83.

67.

Styr

, Slutsky

(2018) Imbalance between firing homeostasis and synaptic plasticity drives early-phase Alzheimer’s disease. Nat Neurosci 21, 463–473.

68.

Bell

, de Kort

, Steggerda

, Shigemoto

, Ribeiro-da-Silva

, Cuello

(2003) Structural involvement of the glutamatergic presynaptic boutons in a transgenic mouse model expressing early onset amyloid pathology. Neurosci Lett 353, 143–147.

69.

Shah

, Lahiri

(2014) Cdk5 activity in the brain - multiple paths of regulation. J Cell Sci 127, 2391–2400.

70.

Roselli

, Tirard

, Lu

, Hutzler

, Lamberti

, Livrea

, Morabito

, Almeida

(2005) Soluble beta-amyloid1-40 induces NMDA-dependent degradation of postsynaptic density-95 at glutamatergic synapses. J Neurosci 25, 11061–11070.

71.

, Humbert

, Bronson

, Takahashi

, Kulkarni

, Li

, Tsai

(2001) p35 and p39 are essential for cyclin-dependent kinase 5 function during neurodevelopment. J Neurosci 21, 6758–6771.

72.

Ding

, Qiao

, Wang

, Goodwin

, Lee

, Block

, Allsbrook

, McDonald

, Fan

(2008) Retinoic acid attenuates beta-amyloid deposition and rescues memory deficits in an Alzheimer’s disease transgenic mouse model. J Neurosci 28, 11622–11634.

73.

Chen

, Li

, Sun

, Li

(2015) Anti-diabetes drug pioglitazone ameliorates synaptic defects in AD transgenic mice by inhibiting cyclin-dependent kinase5 activity. PLoS One 10, e0123864.

74.

Alvarez

, Muñoz

, Maccioni

(2001) A Cdk5-p35 stable complex is involved in the beta-amyloid-induced deregulation of Cdk5 activity in hippocampal neurons. Exp Cell Res 264, 266–274.

75.

Zhang

, Li

, Zhou

, Herrup

(2012) Cdk5 levels oscillate during the neuronal cell cycle: Cdh1 ubiquitination triggers proteosome-dependent degradation during S-phase. J Biol Chem 287, 25985–25994.

76.

Wei

, Tomizawa

, Ohshima

, Asada

, Saito

, Nguyen

, Bibb

, Ishiguro

, Kulkarni

, Pant

, Mikoshiba

, Matsui

, Hisanaga

(2005) Control of cyclin-dependent kinase 5 (Cdk5) activity by glutamatergic regulation of p35 stability. J Neurochem 93, 502–512.

77.

Kerokoski

, Suuronen

, Salminen

, Soininen

, Pirttilä

(2004) Both N-methyl-D-aspartate (NMDA) and non-NMDA receptors mediate glutamate-induced cleavage of the cyclin-dependent kinase 5 (cdk5) activator p35 in cultured rat hippocampal neurons. Neurosci Lett 368, 181–185.

78.

Lee J

, Kim

(2004) Induction of cyclin-dependent kinase 5 and its activator p35 through the extracellular-signal-regulated kinase and protein kinase A pathways during retinoic-acid mediated neuronal differentiation in human neuroblastoma SK-N-BE(2)C cells. J. Neurochem 91, 634–647.

79.

Moncini

, Lunghi

, Valmadre

, Grasso

, Del Vescovo

, Riva

, Denti

, Venturin

(2017) The miR-15/107 family of microRNA genes regulates CDK5R1/p35 with implications for Alzheimer’s disease pathogenesis. Mol Neurobiol 54, 4329–4334.

80.

Zhang

, Li

, Herrup

(2010) Cdk5 nuclear localization is p27-dependent in nerve cells: Implications for cell cycle suppression and caspase-3 activation. J Biol Chem 285, 14052–14061.

81.

Tomizawa

, Ohta

, Matsushita

, Moriwaki

, Li

, Takei

, Matsui

(2002) Cdk5/p35 regulates neurotransmitter release through phosphorylation and downregulation of P/Q-type voltage-dependent calcium channel activity. J Neurosci 22, 2590–2597.

82.

Diaz

, Jeanneret

, Merino

, McCann

, Yepes

(2019) Tissue-type plasminogen activator regulates p35-mediated Cdk5 activation in the postsynaptic terminal. J Cell Sci 132, jcs224196.

83.

Sato

, Xu

, Okuyama

, Martinez

, Walsh

, Jacobsen

, Swan

, Schlautman

, Ciborowski

, Ikezu

(2008) Spatial learning impairment, enhanced CDK5/p35 activity, and downregulation of NMDA receptor expression in transgenic mice expressing tau-tubulin kinase 1. J Neurosci 28, 14511–14521.

84.

Kamiki

, Boehringer

, Polygalov

, Ohshima

, McHugh

(2018) Inducible knockout of the cyclin-dependent kinase 5 activator p35 alters hippocampal spatial coding and neuronal excitability. Front Cell Neurosci 12, 138.

85.

Niwa

, Patrizio

, Triller

, Specht

(2019) cAMP-EPAC-dependent regulation of gephyrin phosphorylation and GABAAR trapping at inhibitory synapses. iScience 22, 453–465.

86.

Lorenz-Guertin

, Bambino

, Das

, Weintraub

, Jacob

(2019) Diazepam accelerates GABAAR synaptic exchange and alters intracellular trafficking. Front Cell Neurosci 13, 163.

87.

Tyagarajan

, Ghosh

, Yévenes

, Nikonenko

, Ebeling

, Schwerdel

, Sidler

, Zeilhofer

, Gerrits

, Muller

, Fritschy

(2010) Regulation of GABAergic synapse formation and plasticity by GSK3beta-dependent phosphorylation of gephyrin. Proc Natl Acad Sci U S A 108, 379–384.

88.

Villa

, Berry

KP1

, Subramanian

, Cha

, Oh

, Kwon

, Kubota

, So

, Nedivi

(2016) Inhibitory synapses are repeatedly assembled and removed at persistent sites in vivo. Neuron 89, 756–769.

89.

Saliba

, Kretschmannova

, Moss

(2012) Activity-dependent phosphorylation of GABAA receptors regulates receptor insertion and tonic current. EMBO J 31, 2937–2951.

90.

Becerra-Calixto

, Cardona-Gómez

(2017) The role of astrocytes in neuroprotection after brain stroke: Potential in cell therapy. Front Mol Neurosci 10, 88.

91.

Jimenez-Blasco

, Santofimia-Castaño

, Gonzalez

, Almeida

, Bolaños

(2015) Astrocyte NMDA receptors’ activity sustains neuronal survival through a Cdk5-Nrf2 pathway. Cell Death Differ 22, 1877–1889.

92.

Ishibashi

, Egawa

, Fukuda

(2019) Diverse actions of astrocytes in GABAergic signaling. Int J Mol Sci 20, E2964.

93.

Nykänen

NP1

, Kysenius

, Sakha

, Tammela

, Huttunen

(2012) γ-Aminobutyric acid type A (GABAA) receptor activation modulates tau phosphorylation. J Biol Chem 287, 6743–6752.

Amyloid-β Fosters p35/CDK5 Signaling Contributing to Changes of Inhibitory Synapses in Early Stages of Cerebral Amyloidosis

Abstract

Keywords

INTRODUCTION

MATERIALS AND METHODS

Hippocampal cell cultures and lentiviral infection

Construction of expression plasmids

Construction of p35-knock-down vectors (Supplementary Figure 1)

Lentivirus preparation

Cell culture

Animals and tissue preparation

Protein extracts and immunoblot analysis

Immunolabeling

For brain sections

For cultured hippocampal neurons

Confocal laser scanning microscopy and quantitative immunofluorescence analysis

For brain sections

For cultured hippocampal neurons

Proximity ligation assay (PLA) of hippocampal neurons

Statistical analysis

RESULTS

hAPP-swe expression leads to both increased phospho-gephyrin and p35 protein levels in cultured hippocampal neurons

p35 and CDK5 are increased in the hippocampus of young APP-PS1 mice

Increased p35/CDK5 colocalize and correlate with elevated phospho-gephyrin and GABAAR-γ2 in the hippocampus of young APP-PS1 mice

Increased p35/CDK5 colocalize and correlate with phospho-tau in the hippocampus of young APP-PS1 mice

p35 knock-down diminishes mAb7a-positive-gephyrin- and γ2-GABAAR-clusters in cultured hippocampal neurons

DISCUSSION

Footnotes

ACKNOWLEDGMENTS

References

Increased p35/CDK5 colocalize and correlate with elevated phospho-gephyrin and GABA_AR-γ2 in the hippocampus of young APP-PS1 mice

p35 knock-down diminishes mAb7a-positive-gephyrin- and γ2-GABA_AR-clusters in cultured hippocampal neurons