Abstract Alzheimer’s disease (AD) is a progressive neurodegenerative disorder marked by β-amyloid (βA) accumulation, neuroinflammation, excessive synaptic pruning, and cognitive decline. Despite extensive research, effective treatments remain elusive. Here, we identify potassium channel-interacting protein 3 (KChIP3) as a key driver of AD pathology using the 5XFAD mouse model. KChIP3 levels were significantly elevated in the hippocampus of 5XFAD mice, correlating with βA burden and neuroinflammation. This upregulation was triggered by inflammatory signaling via the NLRP3 inflammasome and Caspase-1 activation. Notably, genetic deletion of KChIP3 (5XFAD/KChIP3^−/−) markedly reduced βA plaque deposition, pro-inflammatory cytokines, reactive gliosis, and expression of inflammation-related proteins (APO, CLU, MDK). Transcriptomic and proteomic analyses revealed restored synaptic markers (CD47, CD200, CACNB4, GDA) and a shift of the disease-associated microglial (DAM-1) phenotype. Mechanistically, we propose that KChIP3 amplifies AD pathology through two key mechanisms: (1) sustaining neuroinflammation by upregulating pro-inflammatory genes and (2) impairing synaptic integrity by repressing genes critical for neuronal function. Consistently, KChIP3 deletion enhanced dendritic complexity, synaptic plasticity, and cognitive performance in 5XFAD mice. These findings position KChIP3 as a potential therapeutic target for mitigating neuroinflammation and synaptic dysfunction in AD and highlight its potential as a biomarker for disease progression. Graphical abstract [52]graphic file with name 12974_2025_3426_Figa_HTML.jpg Supplementary Information The online version contains supplementary material available at 10.1186/s12974-025-03426-2. Keywords: Alzheimer’s disease, Neuroinflammation, Synaptic plasticity, Learning, Memory, LTP, Microglia, 5XFAD mice, KChIP3, Aging Highlights * Genetic deletion of potassium channel interacting protein 3 (KChIP3) in 5XFAD mice, significantly reduced βA plaque formation and neuroinflammation in the hippocampus. * KChIP3 deletion restored synaptic markers and reduced microglial DAM1 phenotype in 5XFAD mice. * KChIP3 deficiency improved dendritic tree complexity and synaptic plasticity in 5XFAD mice. * The absence of KChIP3 in 5XFAD mice confers neuroprotection, resulting in enhanced learning and memory capacities. Supplementary Information The online version contains supplementary material available at 10.1186/s12974-025-03426-2. Introduction The nervous system, comprising a complex network of neuronal and glial cells, plays a pivotal role in receiving, processing, and transmitting information to facilitate high-order cognitive functions. Recent research has uncovered that both central and peripheral inflammation, induced by various factors such as genetic mutations, infections, ischemia (cardiac or cerebral), obesity, lifestyle choices, and aging, represent significant risk factors contributing to the dysregulation of brain function. Chronic neurodegenerative diseases share common pathological features, including the accumulation of misfolded proteins, failure in long-term potentiation (LTP) within neurons, reduced synaptic markers, increased glial immunoreactivity, toxicity, and chronic inflammation. Remarkable examples of such diseases encompass Parkinson’s disease, Huntington’s disease, vascular dementia, frontotemporal dementia, and Alzheimer’s disease (AD) [[53]1, [54]2]. AD is a multifactorial disorder with serious social, health, and economic implications, affecting over 30 million individuals worldwide. Epidemiological predictions suggest that this number may triple by 2050 [[55]3]. Data from the United Nations Program on Aging and the US Centers for Disease Control and Prevention Project predict a substantial increase in the global elderly population (aged 65 and above), from 420 million in 2000 to nearly one billion by 2030, representing a rise from 7 to 12%. This surge in the elderly population is anticipated to be more pronounced in developed nations compared to emerging economies. AD manifests in two main forms, familial and sporadic. The familial type is linked to mutations in the amyloid precursor protein (App), presenilin 1 (Psen1), and presenilin 2 (Psen2) genes. These mutations follow an autosomal dominant inheritance pattern and account for less than 5% of AD cases. The remaining 95% of AD cases correspond to the sporadic form which is primarily associated with aging, environmental factors, and single nucleotide polymorphisms (SNPs) in the apolipoprotein E (Apoe) gene. AD is characterized by the presence of extracellular senile plaques, intracellular neurofibrillary tangles, microglia activation, synapse loss, inflammation leading to neuronal degeneration and cognitive decline [[56]4]. These distinctive hallmarks are commonly observed in various brain regions, with the hippocampus and cerebral cortex exhibiting the most severe manifestations. The aggregation of extracellular senile plaques results from the accumulation of beta-amyloid (βA) peptides, typically 40–42 amino acids (aa) in length, derived from the processing of APP. βA peptides have the propensity to oligomerize and initiate signaling cascades that lead to increased tau protein phosphorylation, extracellular senile plaque formation, synaptic dysfunction, neuronal death, long-term potentiation (LTP) inhibition, and cognitive impairment, as observed in both humans and mouse models of AD [[57]5–[58]8]. The potassium channel interacting protein 3 (KChIP3) protein, encoded by the Kcnip3 gene, serves as a multifunctional cellular entity. Also known as Casenilin or downstream regulatory element antagonist modulator (DREAM), this protein possesses four EF-HANDS domains, also referred to as Ca^2+ binding motifs. At the cell membrane, KChIP3 binds Kv4 potassium channels and modulates potassium conductance [[59]9–[60]11]. In contrast with this, within the endoplasmic reticulum and Golgi apparatus, KChIP3 interacts with PSEN1 and PSEN2, crucial components of the γ-secretase enzyme complex responsible for cleaving the APP in the amyloidogenic pathway [[61]12–[62]14]. KChIP3 also serves as a transcriptional factor in the nucleus. It binds to specific DNA regions, referred to as DRE sites (GTCAG) [[63]15, [64]16] and acts as a Ca^2+-regulated transcriptional repressor [[65]17, [66]18]. Notably, increased levels of KChIP3 have been reported in the cortex and hippocampus of postmortem AD brains, as well as in the TG2576 mouse model of AD [[67]19]. In vitro studies have shown that the overexpression of KChIP3 in a neuroblastoma cell line correlates with increased levels of βA42 peptides; and the administration of βA peptides in the neuronal cell line SK-N-BE2 induces the accumulation of KChIP3 protein, which correlates with increased cell death [[68]20]. Conversely, mice lacking KChIP3 (calsenilin knock out mouse model) exhibit reduced levels of βA40 and βA42 peptides in the cortex and cerebellum [[69]21]. Collectively, these findings suggest that KChIP3 modulates the activity of the γ-secretase complex through its interaction with presenilins. Previous studies have demonstrated that the accumulation of βA 40–42 peptides leads to the establishment of a neuroinflammatory process that contributes to neuronal damage and loss of cognitive capacity [[70]22, [71]23]. Here, we propose that KChIP3 promotes a pro-inflammatory environment in AD, that ultimately impacts cognitive function. To test this hypothesis, we generated a KChIP3 knockout model (KChIP3^−/−) in the background of the 5XFAD mouse model of AD (5XFAD/KChIP3^−/−). Our results indicate that 5XFAD/KChIP3^−/− mice exhibit decreased βA peptide levels and reduced accumulation of senile plaques, particularly in the hippocampus. Interestingly, the pro-inflammatory environment associated with plaque accumulation was also reduced. Using both transcriptomic and quantitative proteomic approaches, we identified enrichment of molecules related to synaptic transmission, LTP, learning and memory. Accordingly, we found that the reduced dendritic tree complexity observed in 5XFAD mice’ hippocampal neurons was rescued in the 5XFAD/KChIP3^−/− mice, even to the same extent as observed in wild type (WT) mice. Importantly, this finding correlates with the restoration of the LTP of hippocampal synapses and with the recovery in the learning and memory capacity in the 5XFAD mice lacking KChIP3. Since AD is recognized for increased synaptic pruning [[72]24], our data propose that in the absence of KChIP3 the levels of the neuroimmune regulatory proteins CD200 (type 1 glycoprotein belonging to the IgSF family) and CD47 (cluster of differentiation 47) are increased lowering the levels of complement components such as C1q promoting “don’t eat me signals” which results in healthy synapses and cognitive improvement in the 5XFAD mouse. Our findings collectively support the notion that KChIP3 protein levels accumulate as AD pathology progresses, contributing to altered gene expression profiles that promote AD development. Specifically, this dysregulation negatively affects the expression of genes encoding proteins involved in synaptic transmission and maintenance, while simultaneously increasing the expression of genes associated with inflammation. Importantly, some of these changes in gene expression were mirrored by changes in protein levels. Therefore, we propose that monitoring KChIP3 protein levels could serve as a valuable indicator of AD progression. Materials and methods Animals 5XFAD mice in the C57BL/6 J genetic background were obtained from The Jackson Laboratory. These mice were maintained as heterozygotes through mating them with C57BL/6 J mice. As a control group, we used littermates without the 5XFAD transgene, referred to as WT mice. The KChIP3^−/− mouse was generated by the insertion of a neomycin cassette into exons 6–8 of the Kcnip3 gene [[73]25]; and it was kindly provided by Dr. W. Frank An from The Broad Institute of Harvard and MIT, Boston, Massachusetts, USA. KChIP3^−/− mouse colony was maintained in C57BL/6 J genetic background. To knock out KChIP3 into the 5XFAD AD model, 5XFAD mice and KChIP3^−/− mice (5XFAD/KChIP3^−/−) were crossbred. Only male littermates of 5XFAD, KChIP3^−/−, and WT mice (C57BL/6 genetic background) were used. 5XFAD/Caspase 1 (Casp1)^−/− mice were obtained as previously described [[74]22]. Only male littermates of 5XFAD, Casp1^−/−, 5XFAD/Casp1^−/− and WT mice (C57BL/SJL genetic background) were used. Genotype (5XFAD, KChIP3^−/− and 5XFAD/KChIP3^−/− mice) was confirmed by PCR and Western-blot assays (Supplementary information, Fig. S1). All experiments were carried out using 9–12 months-old mice since in the C57BL/6 genetic background the phenotype of the 5XFAD mouse model is established at the age of 8 months ([75]https://www.alzforum.org/research-models/5xfad-c57bl6). Details about specific age and number of animals used in each experimental condition in supplementary STable 1. Animals were housed with ad libitum access to food and water, under standard conditions with a 12-h light/12-h dark cycle. All animal procedures were approved by the Bioethics Committee of the Institute of Biotechnology at the National Autonomous University of Mexico. Protein extracts and immunoblotting Hippocampal tissue homogenates were prepared in TLB lysis buffer (20 mM Tris pH 7.4, 137 mM NaCl, 2 mM PPiNa, 2 mM EDTA, 1% Triton X- 100, 10% glycerol, 0.5 mM DTT, 25 mM β-glycerol phosphate, 200 mM Na[3]VO[4], 1 mM PMSF, and complete protease inhibitor) by sonication. After a 10-min incubation at 4 °C (on ice), protein extracts were spun at 14,000 rpm for 10 min. The aqueous phase was separated and stored at − 70 °C. Protein levels were quantified using the BCA assay and 30–40 µg of protein were loaded into polyacrylamide gels (10–17%) for SDS-PAGE. Subsequently, proteins were transferred to a nitrocellulose membrane (Hybond-ECL- GE Healthcare Life Sciences). Membranes were incubated in TBS-T 0.1% with 5% milk for 1 h. Then, they were incubated with primary antibodies at 4 °C overnight. Then, the secondary antibodies were incubated at room temperature for 1 h. Proteins were detected using the chemiluminescence and the Western Lightning Plus-ECL kit. Images were acquired with a C-Digit Blot Scanner (LICOR-model 3600), and analyses were conducted using the Image Studio Lite Version 5.2.5 software. For Immunoblotting the following antibodies were used: anti-KChIP3 K + channel antibody (K66/38, NeuroMab; dilution 1:1000); purified anti-β-Amyloid, 1–16 antibody- 6E10 (previously Covance catalog# SIG- 39320; dilution 1:3000); anti-GFAP (GA5) mouse mAb (Cell signaling technology #3670; dilution 1:6000); anti-β-actin mouse-mAb (Cell Signaling technology 8H10D10, dilution 1:10,000). The used secondary antibodies were: KPL peroxidase, labeled affinity purified antibody to mouse in IgG (H + L) produced in goat; dilutions: 1:3000 (KChIP3), 1:6000 (β amyloid peptides), 1:12,000 (GFAP), and 1:15,000 (actin) with their corresponding isotopes. Brain tissue preparation Mice were anesthetized by intraperitoneal injection with a ketamine/xylazine mixture (1:1) and then transcardially perfused using an isotonic saline solution, followed by 4% paraformaldehyde (PFA) in 0.1 M sodium phosphate buffer (pH 7.4). Brains were fixed with 4% PFA overnight and cryoprotected in a 30% sucrose solution at 4ºC. For immunostaining, sagittal cryostat Sects. (30 μm-thick) were prepared. For each condition, we analyzed at least three tissue sections from each animal, and three or four animals were included per genotype. Thioflavin S-staining and quantification of β-Amyloid plaques Sagittal brain sections were stained with 1% thioflavin-S for 5 min, washed sequentially in 80% and 70% ethanol for 5 min each, and finally in distilled water for 1 min. Slides were dried and mounted on glass slides, and coverslipped using Fluoroshield (F6182, Sigma Aldrich) or, when indicated, immunofluorescence assays were followed. Brain sections were analyzed using a Zeiss Axioskop microscope using an emission spectrum of 550 nm. For each mouse, sagittal sections corresponding to the prefrontal cortex and dentate gyrus (lateral coordinates 0.84–2.28 mm [[76]26]) were collected. The number of β-Amyloid plaques per mm^2 was calculated using Image J software (National Institutes of Health, Bethesda, MD, USA) in 3 slices per mouse. The average of each slice was determined, each slice was taken as an independent n. Three mice were analyzed per genotype. GFAP and Iba1 immunofluorescence Brain tissue was prepared as described above. Free-floating sagittal sections were washed with 0.1% Triton TBS (TBS-T) and permeabilized with 1% Triton TBS for 15 min at room temperature. Sections underwent antigen retrieval by steaming in sodium citrate buffer (10 mM, pH 6) at 95 °C for 5 min. Subsequently, sections were blocked with 2% fetal bovine serum TBS-T (blocking solution) for 1 h at room temperature. After washing three times with TBS-T, sections were incubated overnight at 4 °C with the primary antibody (GFAP, dilution 1:400 (#3670, Cell Signaling Technology) or Iba1, dilution 1:300 (#NB100 - 1028, Novus) in blocking buffer. Sections were washed three times in TBS-T and incubated with a goat anti-mouse secondary antibody conjugated to Alexa Fluor 488 (A11011, Thermo Scientific) at a dilution of 1:400 in blocking buffer at room temperature for 1 h for GFAP, and with a donkey anti-goat antibody conjugated to Alexa Fluor 555 (A21432, Thermo Scientific) at a dilution of 1:300 in blocking buffer at room temperature for 2 h for Iba1. After three additional washes in TBS-T, sections were mounted on glass slides and coverslipped using Fluoroshield (F6182, Sigma Aldrich). Thirty µm image stacks were acquired sequentially using a laser-scanning confocal microscope (Olympus FV1000 Inverted Confocal Multiphoton Microscope) with 20 × and 60 × objective lenses as indicated. The GFAP immunoreactivity area was calculated using a thresholding protocol and the ImageJ software. For each mouse, sagittal sections corresponding to the hippocampus and subiculum (lateral coordinates 0.84–2.28 mm; [[77]26]) were collected. The sections were grouped into four areas based on lateral coordinates: 0.84–1.08 mm, 1.20–1.44 mm, 1.56–1.80 mm, and 1.92–2.04 mm. One section per area was processed for GFAP immunofluorescence, and three mice were analyzed per genotype. Percentage of Iba1 positive cells per βA plaque was determined for each sagittal section corresponding to the prefrontal cortex (CTX) and dentate gyrus (DG), under 20 × magnification. Prior to Iba1 immunolocalization, thioflavin S-staining was performed as described above in this section. Iba1 signal around each plaque (100 plaques total) was quantified and normalized by area using ImageJ software. 150X magnification shows the overlapping and orthogonal view of 39 optical slices acquired sequentially with 60 × objective lens and 2.5 digital zoom. For each mouse, sagittal sections corresponding to the prefrontal cortex and dentate gyrus (lateral coordinates 0.84–1.32 mm [[78]26]) were collected. Two slices were processed for Iba1, two mice were analyzed per genotype. Inflammatory cytokines-antibodies array Hippocampal protein extracts were prepared as described above. To determine the inflammatory profile, protein extracts (250 μg) were exposed to the mouse cytokine antibody array C1 RayBio® C-Series (#AAM-INF- 1–4) following the manufacturer’s directions. Antibody-antigen interactions were detected using chemiluminescence and a C-Digit Blot Scanner (LICOR-model 3600). Densitometric analyses were conducted using the Image Studio version 5.2.5 software, Experimental data’ fold changes were normalized to the control (WT). Enzyme-linked immunosorbent assay (ELISA) Concentrations of IL-1β, TNF, IL-6 and IL-10 were determined in 100 µg of hippocampal protein extracts using the IL-1β (432,604), TNF (430,904), IL-6 (431,304), and IL-10 (431,414) ELISA MAX ™ Deluxe Sets (BioLegend) kits following manufacturer’s instructions. Transcriptomic analysis RNA extraction and library preparation Total RNA was extracted from freshly dissected hippocampus using Trizol, following manufacturer’s instructions (these samples are the opposite hemisphere of the ones used for the proteomics analyses). Triplicates of the following genotypes were used: WT, KChIP3^−/−, 5XFAD, 5XFAD/KChIP3^−/−. RNA quality was verified using a bioanalyzer low-sensitivity chip. Illumina RNA-seq libraries were prepared from 1 μg of total RNA (RNA integrity number of at least 8) using the TruSeq mRNA Library prep kit. Libraries were sequenced in an Illumnina NextSeq 500 System using a NextSeq 500/550 High Output Kit v2.5 (150 Cycles) in a 2X75 cycles configuration. RNA-seq analyses Paired-end RNA-seq reads were QC’d using fastQC v. 0.11.9. All of them had good quality and were used in the analysis. Reads were aligned against the mouse transcriptome (Ensembl annotation v99, based on mm10) with HISAT 2.1.0 using the following non-standard parameters: –no-unal –no-mixed –no-discordant. Alignments with QS < 10 or falling within Encode blacklisted regions were eliminated. Read count tables were obtained with featureCounts from the Rsubred package (2.10.5, R version 4.2.1) using the following parameters: isGTFAnnotationFile = T, useMetaFeatures = T, minMQS = 10, largestOverlap = T, isPairedEnd = T, requireBothEndsMapped = T, with the Ensembl annotation v99 as reference. The read count table was analyzed with edgeR v3.38.4. Lowly expressed genes (cpm < 3) were eliminated. After calculating normalization factors and estimating dispersion, differentially expressed genes between the four groups were calculated using glm modeling. Only genes with an absolute log2 fold change > 0.5 and an FDR < 1e- 2 were considered as DEG. Heatmaps of DEG were generated using heatmap.2 from the gplots package v3.1.3. Gene ontology analysis was performed using DAVID. Proteomic analysis The proteomic study design is illustrated in Supplementary information, Fig. S1 A, consisting of four experimental groups: WT, KChIP3^−/−, 5XFAD, and 5XFAD/KChIP3^−/−. The hippocampal region of three animals (2 experimental samples per mouse) per genotype (WT, KChIP3^−/−, 5XFAD, and 5XFAD/KChIP3^−/−) were analyzed. All samples were labeled using iTRAQ reagents, and then, subjected to a prefractionation by RP-HPLC. Subsequently, the generated peptides were injected into a LC–MS/MS Q-exactive quadrupole Orbitrap, leading to the identification of 4474 proteins. After applying a filter to only include proteins with 2 or more razor + unique peptides, 3860 proteins were identified based on their abundance profile in six replicates per condition. Data normalization and standardization were performed to correct for differences between replicates. Protein extraction Tissue samples were washed with cold PBS and homogenized in lysis buffer (containing 7 M urea, 2 M thiourea, 4% CHAPS, 0.1% SDS, 50 mM DTT, 0.2 mM PMSF, and 40 mM Tris; pH 7.4). For each 15 mg of tissue, 100 μL of lysis buffer were added. After homogenization, samples were centrifuged at 12,000 rpm at 4 °C for 15 min. The resulting pellet was discarded, and the protein content of the supernatant was estimated using the 2D Quant kit (GE Healthcare). Cysteine residues were alkylated with 100 mM IAA in the dark and at room temperature for 30 min. Tris was added to a final concentration of 100 mM (pH 8.6). Samples were stored at − 70 °C until use. Protein precipitation 30 μg of protein extracts from the hippocampus were precipitated with nine volumes of cold ethanol at − 20 °C overnight. The resulting pellet was washed three times with a 90% ethanol solution. The precipitate was solubilized with a solution containing 0.1% SDS, 0.5% SDC, and 0.1 M TEAB (Triethylammonium bicarbonate buffer), pH 8.0. Trypsin was added to achieve a ratio of 1:50 (enzyme:substrate), and the sample was incubated at 37 °C for 16 h [[79]27]. After digestion, the generated peptides were chemically modified with iTRAQ 8plex reagents (AB Sciex Pte. Ltd.), following the manufacturer’s instructions. These reactions were pooled, and detergent was eliminated. Peptide fractionation Peptides were further fractionated using an Ultra High Performance Liquid Chromatography (RP-HPLC) (Dionex Ultimate 3000 UHPLC, Thermo Fisher) with a Xterra MS C18, 5 um 3.9 × 150 mm, column (Waters). The mobile phases consisted of A (0.1% formic acid (FA), pH 10) and B (60% acetonitrile, 40% water, 0.1% FA, pH 10). Peptides were eluted at a flow rate of 0.8 mL/min following the gradient: 2% of mobile phase B for 20 min, 2% to 20% of mobile phase B for 70 min, 20% to 45% mobile phase B for 65 min, 45% mobile phase B for 23 min, 100% of mobile phase B for 5 min, 100% to 2% of mobile phase B for 5 min, and 2% of mobile phase B for 5 min. Following a carousel arrangement, fifteen fractions were collected every 2 min during chromatography separation, following a carousel arrangement. Mass spectrometry analysis All fractions were desalted using Sep-Pak C18 cartridges (Waters). Peptides were resuspended in the initial chromatographic conditions and separated on a Dionex Ultimate 3000, RSLCnano UPLC system, coupled to a Q-Exactive Plus high-resolution mass spectrometer (Thermo Fischer Scientific). Samples were trapped on a precolumn (C18 PepMap 100, 5 µm, 100 Å, 300 µm inner diameter × 5 mm) and then separated over a 250-min elution gradient using a capillary column (EASY Spray Column, PepMap RSLC, C18, 3 µm, 100 Å, 75 µm × 15 cm). The mobile phases consisted of A (0.1% formic acid in water) and B (90:10 (v/v) acetonitrile: water, 0.1% formic acid). The mass spectrometer operated in positive data-dependent acquisition mode, with a full MS range from 300 to 2,000 m/z. The ten most intense ions were isolated in the quadrupole and fragmented under high-energy collisional dissociation with a normalized collision energy of 27%. Precursor ions were measured at a resolution of 70,000 (at 200 m/z), and the fragments were measured at 17,500. Only ions with charge states of 2 and higher were fragmented with an isolation window of 2 m/z. Data analysis Protein identification and quantification were performed using the MaxQuant v1.5.3.30 software. Trypsin/P was selected as the digestion enzyme, and carbamidomethyl-cysteine was set as a fixed modification. N-terminal protein acetylation and oxidation (M) were included as variable modifications. The database used for protein identification was the Mus musculus reviewed reference proteome Taxon ID 10090 from the UniProt repository, downloaded on September 6 th, 2022. To consider proteins correctly identified, we accepted those with at least two razor-unique peptides and one unique peptide. Proteins were reported with an FDR of 1% based on the target-decoy strategy to ensure the quality of the identifications. For relative protein quantification, we incorporated additional filters: all peptides were filtered by the distribution of the isotopic precursor (PIF 0.75), to avoid reporting quantitative values for peptides with clear isotopic contamination. According to above criteria, quantitative data were reported only for proteins with at least two identified/quantified unique peptides according to the above criteria [[80]27]. Dendritic tree evaluation Golgi-Cox staining and sample preparation Brains were dissected and rinsed with cold Milli-Q water to wash away the blood and immediately transferred to a Golgi–Cox staining solution. The staining procedure followed the manufacturer recommendations of the FD Rapid GolgiStain kit (FD Neurotechnologies, Columbia, MD, US). Each brain was immersed in 4 ml of the impregnation solution prepared with an equal volume mixture of solutions A and B (mercuric chloride, potassium chromate, and potassium dichromate). After 24 h, the solution was replaced with a new mix of A and B solutions. Brains were stored in darkness at 25 °C for two weeks. Before tissue sectioning, brains were immersed in the cryoprotectant solution C and stored under the same conditions for a week. Afterward, the whole brains were frozen, and 80 µm-thick slices (coronal sections) were cut in a cryostat. The slices were mounted in gelatin-coated microscope slides (2.5 g gelatin, 0.2 g Cr[2](SO[4])3/500 ml). For the developing step, slices were treated with a mixture of the solutions D and E solutions. Next, the slides were rinsed with Milli-Q water, air dried naturally, and finally protected with Cytoseal resin (Thermo Fisher, Waltham, MA, US) and cover slipped. Image acquisition 40 well-stained pyramidal neurons from the CA1 region of the dorsal hippocampus (bregma –2.03 to –2.27 mm), [[81]26] of each experimental group (ten neurons per mouse) were identified and captured in an Olympus BX51-WI-DSU (Disk Scanning Unit) microscope (Olympus, Tokyo, Japan; Objective Microscope: 20x) coupled to the Stereoinvestigator 9 R software (MBF Bioscience, Williston, VT, US). Series of z-stack images were taken (distance between images, 2 µm) for each neuron. A well-stained neuron shows a complete dendrite filling with no breaks in dendritic branches and is isolated from surrounding neurons. Sholl analysis. A semi-automated Sholl analysis was performed following the method reported by Langhammer et al., 2010 [[82]28]. Z-stack images of hippocampal neurons were imported to the ImageJ software (NIH, MD, US) and merged into single 8-bit images. Using the NeuronJ plugin, the dendritic morphology of each neuron was digitized to generate the trace files (.ndf). Then, traces were converted into the SWC format using the Bonfire program, which consists of a series of personalized scripts in MATLAB (Mathworks, MA, US). The connectivity pattern between the dendritic segments was defined by using the NeuronStudio software (CNIC, NY, US). The Center-Out Sholl analysis was used to evaluate dendrites’ number, length, and branching pattern. The evaluation considers primary dendrites, those which extend from the soma; secondary dendrites, those emanating from the primary dendrites; tertiary and higher order (<) dendrites which are those derived from the secondary ones. Fluorescence analysis of single synapse Long-Term Potentiation (FASS-LTP) Isolation of synaptosomes The hippocampus from both hemispheres of each mouse was dissected and kept on ice at 4ºC. Before isolating tissues, tubes, pestles, and buffers were pre-chilled on ice. Each hippocampal tissue was gently homogenized in 1.5 mL of sucrose solution (320 mM sucrose, 10 mM HEPES) containing protease inhibitors (5 μg/mL pepstatin, 5 μg/mL leupeptin, 5 μg/mL antipain, 5 μg/mL aprotinin, 1 mM PMSF) and a phosphatase inhibitor (1 μM Na[3]VO[4]). The homogenate was then centrifuged at 1200 xg at 4ºC for 10 min, and the resulting supernatant was transferred to two new tubes. This supernatant was centrifuged at 12,000 xg at 4ºC for 20 min. The supernatant was discarded, and the pellet, referred to as the P2 crude synaptosomal fraction, was resuspended in 1.5 mL of either basal solution (120 mM NaCl, 3 mM KCl, 2 mM CaCl[2], 2 mM MgCl[2], 15 mM glucose, 15 mM HEPES, pH 7.4) or cLTP solution (125 mM NaCl, 2 mM CaCl[2], 5 mM KCl, 10 mM HEPES, 30 mM glucose, pH 7.4). Synaptosomes were brought to a final volume of 2 mL and for recovery, they were gently shaken at room temperature for 15 min. Stimulation of synaptosomes After recovery, 180 μL of synaptosomes (50–200 μg) were maintained in either basal or cLTP conditions. These synaptosomes were then transferred to cytometer tubes (preheated for 5 min at 37 ºC). Synaptosomes in basal solution were used to determine basal levels of GluA1^+/Nrx1^+ levels, while synaptosomes in cLTP solution were stimulated with 20 μL of 5 mM glycine supplemented with 0.01 mM strychnine and 0.2 mM bicuculline at 37ºC for 15 min. To induce synaptosome depolarization after priming NMDA receptors with glycine, synaptosomes were destined for evaluating LTP response at 45 or 75 min. Briefly, the P2 crude synaptosomal fraction was stimulated for either 30 or 60 min with 100 μL of high KCl solution (50 mM NaCl, 2 mM CaCl[2], 100 mM KCl, 10 mM HEPES, 30 mM glucose, 0.5 mM glycine, 0.001 strychnine, and 0.02 mM bicuculline, pH 7.4) at 37ºC to induce synaptosome depolarization. The reaction was stopped by adding 0.5 mL of 0.1 mM EDTA-PBS (cold) and 4 mL of blocking solution (FACS solution: PBS with 5% FBS). Subsequently, tubes were centrifuged at 2,500 xg at 4ºC for 6 min. Supernatants were decanted and pellets were gently resuspended [[83]29]. Flow cytometry analysis Anti-GluA1 and anti-Nrx1β antibodies were added at a 1:400 dilution in 400 μL of FACS solution and incubated at 4ºC for 30 min in agitation. Four milliliters of FACS solution were added and then, samples were centrifuged at 2,500 xg for 6 min. Supernatants were decanted, and pellets were resuspended. Synaptosomes were incubated with secondary antibodies (rabbit-Alexa 488 and mouse-Alexa 647) at a concentration of 2.5 μg/mL in 400 μL of FACS solution in the dark for 30 min. Synaptosomes were then resuspended in 400 μL of FACS solution and acquired immediately. During FASS-LTP experiments, we used an aliquot of the P2 crude synaptosomal fraction to determinate the membrane integrity. Briefly, the P2 crude synaptosomal fraction of each experimental group that were treated under the same conditions for FASS-LTP experiments were stained with 100 nM of calcein-AM. We used as negative control a P2 crude synaptosomal fraction incubated at 37ºC with 90% EtOH for 1 min. Data were acquired using a FACSCanto II flow cytometer (BD Bioscience) with FACS Diva software. Briefly, we set up a gate based on calibrated beads with sizes ranging from 1 to 3 μm to select size-gated synaptosomes [[84]29]. The fluorescence signal was collected using log amplification and detected using BP filter 530/30 (for Alexa Fluor- 488) and BP filter 660/20 (for Alexa Fluor- 647). We recorded ten thousand events, and the analysis was conducted on doublets exclusion using FSC-A vs FSC-H bivariate plot. Data were analyzed with FlowJo 10.8.1 software. Cognitive tests Novel object recognition (NOR) For the Novel Object Recognition (NOR) test, two black wooden boxes, each measuring 33 × 33 cm and 30 cm in height were utilized. Walls of the boxes had distinctive markings to help mice’ orientation, boxes’ floor was covered with sawdust. We employed two different objects for this test: one was a falcon culture flask (9.5 cm high) filled with sand, while the other one was a multicolored tower made of Lego bricks (8 cm high). The NOR test was conducted in two sessions. The first session was the training or familiarization phase, during which mice were presented to two identical objects and allowed to explore them for 5 min. Following the training phase, short-term memory (STM) was assessed using intervals of 1 h and 6 h, while long-term memory (LTM) was evaluated at 24 and 48 h. To prevent odor-based preferences, a thorough cleaning procedure was