Selective impairment of long-term depression in accumbal D1R+ MSNs involves calcium-permeable AMPARs in early Alzheimer's disease

Nicolas Riffo-Lepe; Juliana González-Sanmiguel; Isaías Meza; Paulina Saavedra-Sieyes; Lorena Armijo-Weingart; Helena Zambrano; Ailín Riquelme; Armando G Salinas; Loreto San Martín; Luis G Aguayo

doi:10.1016/j.nbd.2026.107568

. Author manuscript; available in PMC: 2026 Oct 6.

Published in final edited form as: Neurobiol Dis. 2026 Aug 7;228:107568. doi: 10.1016/j.nbd.2026.107568

Selective impairment of long-term depression in accumbal D1R+ MSNs involves calcium-permeable AMPARs in early Alzheimer's disease

Nicolas Riffo-Lepe ^a, Juliana González-Sanmiguel ^a, Isaías Meza ^a, Paulina Saavedra-Sieyes ^a, Lorena Armijo-Weingart ^a,^c, Helena Zambrano ^a, Ailín Riquelme ^a, Armando G Salinas ^e, Loreto San Martín ^a,^b,^d, Luis G Aguayo ^a,^b,^*

PMCID: PMC13635928 NIHMSID: NIHMS2209780 PMID: 42567367

Abstract

Alzheimer's disease (AD) is increasingly associated with early circuit dysfunction preceding cognitive decline, including neuronal hyperactivity and neuropsychiatric symptoms linked to mesolimbic pathways. The nucleus accumbens (NAc), a central regulator of reward and motivational processing, exhibits early alterations in excitation/inhibition balance in patients and experimental models, yet the synaptic mechanisms underlying its vulnerability remain unclear. Using a double transgenic APP/PS1 mice crossed with a Drd1a-tdTomato reporter line, we combined cell-type-specific electrophysiology, immunohistochemistry, ex vivo photometry, and behavioral assays. At a pre-plaque stage, intraneuronal Aβ accumulated in both dopamine D1 receptor-positive (D1R+) and D1R-negative medium spiny neurons (MSNs). Despite comparable Aβ levels, both high-frequency stimulation-induced long-term depression (LTD) and mGluR1/5-dependent LTD were selectively impaired in D1R+ MSNs. This vulnerability was accompanied by an increased contribution of calcium-permeable AMPA receptors (CP-AMPARs). Subsequent CP-AMPAR blockade reduced the residual evoked excitatory postsynaptic current that persisted after mGluR1/5 activation in APP/PS1 D1R+ MSNs. Because paired-pulse ratios remained unchanged, this residual response was consistent with a predominantly postsynaptic mechanism. These synaptic changes were accompanied by reduced evoked dopamine signaling, increased chocolate consumption, and altered baseline context preference, whereas standard pellet consumption, conditioned place preference, anxiety-like behavior, and social behavior were unchanged. These findings define a pre-plaque, cell-type-specific synaptic phenotype in male APP/PS1 mice in which impaired mGluR1/5-dependent plasticity and persistent CP-AMPAR signaling in D1R+ MSNs coincide with selective reward-related alterations.

Keywords: Alzheimer's disease, Nucleus accumbens, Synaptic plasticity, Long-term depression, Dopamine receptor, Calcium-permeable AMPARs, mGluRs, Amyloid-beta

1. Introduction

Alzheimer's disease (AD) has traditionally been conceptualized as a disorder of memory arising from cortical and hippocampal dysfunction (Selkoe and Hardy, 2016). However, neuropsychiatric symptoms, including alterations in motivation, reward processing, and impulse control, frequently emerge during early stages of disease progression and strongly predict clinical outcomes (Masters, 2015; Shah et al., 2025). Increasing clinical and experimental evidence indicates that these early manifestations are associated with network hyperactivity and disruption of excitation/inhibition (E/I) balance, with a substantial proportion of patients exhibiting epileptiform activity (Vossel, 2026). These observations implicate early alterations in synaptic function, yet the cellular mechanisms linking Aβ pathology to circuit hyperexcitability remain poorly defined.

A central feature of AD progression is the temporal dissociation between intracellular and extracellular Aβ accumulation. Intraneuronal Aβ can appear before extracellular plaque deposition and has been associated with disrupted calcium homeostasis, synaptic function, and neuronal excitability (Alcantara-Gonzalez et al., 2025; Iulita et al., 2014; LaFerla et al., 2007). In contrast, extracellular Aβ has been extensively linked to synaptic depression and receptor loss in cortical and hippocampal circuits during more advanced disease stages (Selkoe and Hardy, 2016). The synaptic consequences of endogenous intraneuronal Aβ accumulation during early, pre-plaque stages remain less well characterized, particularly in defined neuronal populations outside cortical and hippocampal regions.

Emerging evidence indicates that early pathological changes affect the mesolimbic system, a circuit critically involved in reward processing and motivational control (Cordella et al., 2018; Fernández-Pérez et al., 2020; Kelly et al., 2021; Wu et al., 2024). The NAc, a central integrative hub within this system, receives convergent glutamatergic and dopaminergic inputs and plays a key role in regulating hedonic drive and behavioral flexibility (Russo and Nestler, 2013). Both patients with AD and mouse models exhibit early alterations in NAc function, including changes in reward-related behaviors and neuronal excitability (Armijo-Weingart et al., 2024; Guo et al., 2022; Li et al., 2025; Nie et al., 2017), occurring prior to overt cognitive decline. Notably, intraneuronal Aβ accumulation has been reported in the NAc during pre-plaque stages (Fernández-Pérez et al., 2020), suggesting that early amyloid pathology may disrupt accumbal synaptic function. However, whether these alterations involve cell-type specific changes in synaptic plasticity and contribute to excitation/inhibition imbalance remains unknown.

MSNs, the predominant neuronal population in the NAc, are inhibitory GABAergic projection neurons that provide the main output of this structure and coordinate information flow via dopamine receptor type 1 (D1R) and dopamine receptor type 2 (D2R) expressing pathways (Russo and Nestler, 2013; Vieitas-Gaspar et al., 2025). Balanced excitation and inhibition across these pathways is critical for appropriate motivational salience and behavioral flexibility (Li et al., 2025; Scaduto et al., 2023). LTD represents a key mechanism of synaptic plasticity in the NAc for constraining excitatory drive (Kauer and Malenka, 2007; Thomas et al., 2001). Accumbal LTD critically depends on AMPA receptor (AMPAR) remodeling driven by group I metabotropic glutamate receptor (mGluR1/5) signaling, which promotes endocytosis of calcium-permeable AMPARs to maintain synaptic homeostasis (Luscher and Huber, 2010; Mango and Ledonne, 2023). Disruption of this mechanism favors CP-AMPAR accumulation shifting circuit output toward increased excitatory drive (Carr, 2020; McCutcheon et al., 2011; Wolf, 2016).

AMPARs are tetrameric ionotropic receptors composed of GluA1–4 subunits that mediate fast excitatory transmission (Henley and Wilkinson, 2016). AMPARs lacking edited GluA2, including receptors enriched in GluA1 subunits, are calcium-permeable, exhibit inward rectification, and contribute to synaptic plasticity processes involving dynamic receptor trafficking (Cull-Candy and Farrant, 2021). Increased CP-AMPAR expression or synaptic incorporation has been implicated in addiction, maladaptive feeding, and epilepsy (Guo and Ma, 2021; Italia et al., 2025; Wolf, 2016). However, whether CP-AMPAR remodeling contributes to early accumbal dysfunction in AD remains unclear.

Dopaminergic dysfunction may further modify this synaptic phenotype. Reduced dopamine availability, alterations in dopamine transporter expression, and degeneration or dysfunction of ventral tegmental area dopaminergic neurons have been reported in patients with AD and in mouse models, including at stages preceding marked cognitive impairment (Krashia et al., 2019; Pilotto et al., 2025; Storga et al., 1996). Because dopamine regulates MSN excitability and synaptic plasticity, early dopaminergic disruption may interact with glutamatergic abnormalities to destabilize accumbal function.

Here, we tested whether pre-plaque intraneuronal Aβ accumulation is associated with cell-type-specific alterations in excitatory transmission and LTD in accumbal MSNs. Using APP/PS1 mice crossed with Drd1a-tdTomato reporter mice, we combined cell-type-specific electrophysiology, synaptic plasticity assays, immunohistochemistry, and measurements of calcium and dopamine signaling. We identify a selective impairment of LTD in D1R+ MSNs associated with persistent CP-AMPAR contribution and reduced evoked dopaminergic signaling, providing a mechanistic framework for early reward-related alterations in this AD model.

2. Materials and methods

2.1. Animals

All experimental procedures were approved by the Institutional Animal Care and Use Committee of the University of Concepción (#CEBB 1194–2022) and were conducted in accordance with national and international guidelines for the care and use of laboratory animals. Male C57BL/6 J mice, double-transgenic APPswe/PS1dE9 mice (MMRRC:034832; B6.Cg-Tg(APPswe,PSEN1dE9)85Dbo/Mmjax), and Drd1a-tdTomato reporter mice (B6.Cg-Tg(Drd1a-tdTomato)6Calak/J; JAX stock #016204) were obtained from The Jackson Laboratory (Bar Harbor, ME, USA) and maintained at the Regional Center for Advanced Studies in Life Sciences (CREAV), University of Concepción. The trans genic line expresses the Swedish mutation (K594M/N595L) in amyloid precursor protein (APP) and the human presenilin-1 variant lacking exon 9 (PS1-dE9), leading to an increased Aβ production (Jankowsky et al., 2001). Drd1a-tdTomato mice express the fluorescent reporter tdTomato under the control of the dopamine D1 receptor (Drd1a) promoter, allowing selective visualization of D1R-expressing MSNs (Ade et al., 2011). To generate experimental cohorts enabling recordings from genetically fluorescent labeled D1R+ MSNs, D1RtdTomato mice were crossed with APP/PS1 mice to obtain APP/PS1;D1 + tdTom offspring (Fig. 1A); WT;D1 + tdTom littermates were used as controls for these experiments. Genotyping was performed according to the provider's instructions for each line. Mice were housed in groups of 3–6 under a 12 h light/dark cycle with ad libitum access to food and water. Animals were used between 3 and 12 months of age. Euthanasia was performed by decapitation following anesthesia with inhaled isoflurane.

Fig. 1. — Increased intraneuronal Aβ immunoreactivity and altered AMPAR-mediated excitatory synaptic transmission in accumbal D1R+ MSNs at an early pre-plaque stage in the NAc core. (A) Schematic representation of the experimental strategy. APP/PS1 mice were crossed with Drd1a-tdTomato reporter mice to identify D1R+ MSNs in the NAc core of 6-month-old WT;D1-tdTomato and APP/PS1;D1-tdTomato mice. The anterior commissure (ac), NAc core, and NAc shell are indicated as anatomical landmarks. (B, C) Representative confocal images of NAc core sections from 6-month-old WT;D1-tdTomato (B) and APP/PS1;D1-tdTomato (C) mice immunolabeled for Aβ using MOAB-2 (green), with tdTomato fluorescence shown in red. Filled arrowheads indicate D1R+ MSNs (tdTomato+), whereas open arrowheads indicate D1R− MSNs (tdTomato−). Right panels show representative magnifications of D1R+ and D1R− MSNs. Scale bars: 20 μm. (D) Quantification of Aβ immunofluorescence within D1R+ and D1R− MSN somata. APP/PS1 mice showed increased somatic Aβ immunoreactivity in both MSN subtypes compared with WT mice, with no significant difference between D1R+ and D1R− MSNs within each genotype (one-way ANOVA followed by Tukey's multiple-comparisons test: F(3,16) = 13.64, p < 0.001). Each point represents the mean value obtained from one animal; n = 5 mice per genotype. Total cells analyzed: WT, 50 D1R+ and 47 D1R− MSNs; APP/PS1, 40 D1R+ and 45 D1R− MSNs. (E) Schematic representation of whole-cell voltage-clamp recordings used to assess sEPSCs in NAc core MSNs. Recordings were performed using a CsCl-based internal solution containing QX-314 and TEA, in ACSF containing picrotoxin and D-APV. (F) Representative sEPSC traces recorded at −60 mV from D1R+ and D1R− MSNs in WT;D1-tdTomato (black) and APP/PS1;D1-tdTomato (red) mice. Scale bars: 6 pA, 5 s. (G) Quantification of AMPAR-mediated sEPSC amplitude. APP/PS1 D1R+ MSNs showed increased sEPSC amplitude compared with WT D1R+ MSNs, whereas no significant genotype-dependent difference was detected in D1R− MSNs (ordinary two-way ANOVA: genotype × MSN subtype interaction, F(1,29) = 5.311, p = 0.0285; genotype effect, F (1,29) = 4.272, p = 0.0478; MSN subtype effect, F(1,29) = 5.825, p = 0.0223; Šídák's multiple-comparisons test: WT D1R+ vs APP/PS1 D1R+, p = 0.0158; WT D1R− vs APP/PS1 D1R−, p > 0.9999). (H) Cumulative amplitude distributions illustrating a rightward shift in APP/PS1 D1R+ MSNs compared with WT D1R+ MSNs, with minimal separation between WT and APP/PS1 D1R− MSNs. Statistical comparisons based on mean sEPSC amplitude per neuron are reported in panel G. (I) Quantification of AMPAR-mediated sEPSC frequency. No significant genotype effect or genotype × MSN subtype interaction was detected, whereas a significant main effect of MSN subtype indicated a higher overall sEPSC frequency in D1R+ than in D1R− MSNs (ordinary two-way ANOVA: genotype × MSN subtype interaction, F (1,29) = 0.0457, p = 0.8323; genotype effect, F(1,29) = 0.2909, p = 0.5937; MSN subtype effect, F(1,29) = 4.699, p = 0.0385). No genotype-specific pairwise comparison was significant. For electrophysiological recordings, n = 12 WT D1R+, 7 WT D1R−, 7 APP/PS1 D1R+, and 7 APP/PS1 D1R− MSNs. sEPSC amplitude and frequency were quantified from the same recordings. The numbers of neurons/mice were: WT D1R+, n = 12/4; WT D1R−, n = 7/4; APP/PS1 D1R+, n = 7/4; and APP/PS1 D1R−, n = 7/4. Each point represents one neuron. Bars and error bars represent mean ± SD, and individual data points are shown. *p < 0.05, **p < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

2.2. Immunohistochemistry

Immunofluorescence experiments were done as previously reported (Armijo-Weingart et al., 2024). Briefly, mice were anesthetized with Ketamine (100 mg/kg, i.p.) and Xylazine (10 mg/kg, i.p.) and transcardially perfused with pre-warmed saline (0.9% NaCl, 35 °C), followed by freshly prepared ice-cold 4% paraformaldehyde (PFA). Brains were then dissected, post-fixed for 24 h at 4 °C, and cryoprotected in 30% sucrose for 3–5 days at 4 °C. Samples were embedded in NEG50, cooled at −20 °C for 2–4 h, and stored at −80 °C for at least 24 h before sectioning with a cryostat. Free-floating coronal sections (30 μm) were rinsed in Tris-phosphate buffer, permeabilized in Trisphosphate containing 1% bovine serum albumin (BSA) and 0.2% Triton X-100, and incubated for 24 h at 4 °C with primary antibodies: MOAB-2 (1:200, mouse, Novus Biologicals, USA), a pan-specific monoclonal antibody directed against the N-terminal residues 1–4 of Aβ that recognizes unaggregated, oligomeric, and fibrillar Aβ42, as well as unaggregated Aβ40, without detecting full-length APP or APP C-terminal fragments (Youmans et al., 2012), microtubule-associated protein 2 (MAP2) (1:200, guinea pig, Synaptic Systems, Germany), and ionized calcium-binding adaptor molecule 1 (Iba1) (1:1000, rabbit, Alomone Labs, Germany). Sections were then incubated for 2 h with secondary antibodies (Alexa Fluor 488, Alexa Fluor 594, Alexa Fluor 647). Stained samples were mounted with DAKO fluorescent medium on glass slides and imaged using confocal microscopy (LSM 780 NLO Zeiss) at the Advanced Microscopy Center (CMA, Biobío). For each animal, at least two coronal sections of the NAc were analyzed, with three distinct regions of interest (ROIs) per section containing the NAc core. Each ROI was consistently acquired as a Z-stack (~20 μm) for subsequent processing and quantification using FIJI and Zen software. Fluorescence intensity of MOAB-2 was quantified by generating neuronal soma masks based on MAP2 and tdTomato signals, allowing selective measurement within identified MSN somata.

2.3. Thioflavin-S staining

Thioflavin-S (Th-S; Sigma, T1892), which binds β-sheet–rich structures present in compact amyloid aggregates, was used to assess fibrillar amyloid plaque deposition as previously described (Fernández-Pérez et al., 2020). Coronal brain sections (30 μm) containing the NAc core and cortex were mounted on glass slides and processed at room temperature (~22 °C). Sections were dehydrated through a graded ethanol series (50%, 70%, 80%, 90%, 95%, and 100%; 5 min each), incubated in xylene (Winkler, XI-1670) for 10 min, and subsequently rehydrated through descending ethanol concentrations (100%, 95%, 90%, 80%, and 70%; 5 min each). Freshly prepared Th–S solution (0.05% in 50% ethanol) was filtered before use, and sections were incubated for 10 min protected from light. Sections were then washed in 70% ethanol for 3 min, rinsed in distilled water for 2 min, cover slipped, and stored protected from light until imaging. Images were acquired using confocal microscopy (LSM 780 NLO Zeiss) at the Advanced Microscopy Center (CMA Biobío). Th–S signal was acquired using excitation at ~405 nm and emission collection at 430–480 nm. The tdTomato fluorescence signal was also acquired and used as an anatomical reference to identify the NAc core and cortical regions within the same coronal sections, together with anatomical landmarks including the anterior commissure and corpus callosum. Acquisition parameters, including laser power, gain, offset, pixel resolution, and z-step, were kept constant across all samples. Th–S positive deposits were quantified in WT;D1-tdTomato and APP/PS1;D1-tdTomato mice at 6 months of age, and in APP/PS1; D1-tdTomato mice at 12 months of age. For each animal, five 30-μm coronal sections were analyzed, sampled every 300 μm, and five animals per experimental group were included. Regions of interest corresponding to the NAc core and cortex were manually delineated using the tdTomato signal and anatomical landmarks. Background fluorescence in the Th–S channel was estimated from negative control sections processed without Th–S and subtracted from all images. Compact Th-S-positive deposits were defined as discrete fluorescent puncta larger than 5 μm, identified using uniform thresholding criteria across all samples. Plaque density was expressed as the number of Th-S-positive deposits per mm². Absence of detectable Th-S-positive deposits under these conditions was interpreted as absence of detectable compact fibrillar plaque deposition, but not as absence of soluble or non-fibrillar Aβ species. All analyses were performed blinded to genotype and age.

2.4. Electrophysiological recordings in coronal brain slices

Patch clamp recordings were done as previously reported (Fernández-Pérez et al., 2020). Acute coronal brain slices containing the NAc were prepared from male mice anesthetized with isoflurane and euthanized by decapitation. Brains were rapidly removed and transferred to an ice-cold, oxygenated cutting solution containing (in mM): 194 sucrose, 30 NaCl, 4.5 KCl, 1.2 NaH₂PO₄ x H₂O, 1 MgCl₂ × 6H₂O, 26 NaHCO₃, and 10 Glucose (pH 7.4, equilibrated with 95% O₂/5% CO₂). Coronal slices (300 μm) were prepared using a vibratome (VT1200, Leica, Germany) and allowed to recover for 1 h at 30 °C in artificial cerebrospinal fluid (aCSF) containing (in mM): 124 NaCl, 26 NaHCO₃, 10 Glucose, 4.5 KCl, 2 CaCl₂ × 2H₂O, 1 MgCl₂ × 6H₂O, and 1.2 NaH₂PO₄ x H₂O, continuously bubbled with 95% O₂/5% CO₂.

Whole-cell patch-clamp recordings were performed in the NAc core, identified according to its anatomical position surrounding the anterior commissure, and recordings were restricted to this region and its MSNs. Synaptic currents were recorded using an Axopatch 200B amplifier coupled to a 1322A Digidata and pClamp 10 software (Axon Instruments). Recording pipettes (4–5 MΩ) were pulled from borosilicate glass capillaries (WPI) using a horizontal puller (P-1000, Sutter Instruments). During recordings, slices were continuously perfused with oxygenated aCSF at 30 °C. Signals were low-pass filtered at 2 kHz and digitized at 20 kHz. Before data acquisition, recordings were excluded if the holding current required to maintain the cell at −60 mV was more positive than −50 pA or more negative than −100 pA. During the recording, cells were additionally discarded if holding current changed by ≥30% relative to baseline or showed progressive drift. Recordings were also discarded if access resistance changed by ≥25%. Series resistance was compensated at 70%. Holding current and access resistance stability across LTD recordings, including uncompensated Ra measurements obtained before and after recordings, are shown in Supplementary Fig. S6. To avoid pseudo-replication, only a single neuron was recorded from each brain slice. In experiments using Drd1a-tdTomato mice, D1R+ neurons were identified by tdTomato fluorescence, while D1R− neurons were defined by the absence of fluorescence.

For voltage-clamp experiments, the internal pipette solution contained (in mM): 120 CsCl, 10 HEPES, 4 MgCl₂·6H₂O, 2 Mg-ATP, 0.5 Na₂-GTP, and 10 BAPTA (tetra-Cs) (pH 7.4, adjusted with CsOH; 295 mOsm), together with QX-314 (5 mM) and TEA-Cl (5 mM). For rectification index experiments, Spermine (100 μM) was also included. For DHPG-induced LTD experiments, EGTA (1 mM) was used instead of BAPTA. Bath solutions were continuously perfused at a rate of 1 mL/min. For spontaneous and evoked EPSC recordings, neurons were voltage-clamped at −60 mV. Although NAc core SPNs have been reported to display a physiological resting membrane potential close to −80 mV (Pennartz et al., 1992), −60 mV was selected as a standardized holding potential for voltage-clamp recordings to obtain stable baseline conditions and reliable inward AMPAR-mediated EPSCs. This holding potential was used consistently across genotypes and experimental groups.

2.5. Spontaneous synaptic currents

Spontaneous excitatory postsynaptic currents (sEPSCs) were recorded in the NAc core at a holding potential of −60 mV. AMPAR-mediated events were isolated by bath application of picrotoxin (PTX, 100 μM) to block GABAA and glycine receptors, and D-APV (50 μM) to block NMDA receptors. After break-in, cells were allowed to stabilize for at least 5 min before recording. sEPSCs were recorded under baseline conditions and subsequently in the presence of PTX and D-APV for a minimum of 15 min. Events were detected using a template-matching algorithm in Clampfit v11, with manual verification to exclude false positives, and at least 300 events per cell were analyzed. Parameters quantified included event frequency, amplitude, rise time, and decay time.

2.6. Synaptic stimulation and evoked responses

Evoked excitatory postsynaptic currents (eEPSCs) were elicited using a tungsten bipolar stimulating electrode (World Precision Instruments) positioned approximately 100 μm from the recorded neuron and connected to an isolated pulse stimulator (A-M Systems). Square current pulses (100 μs, 0.05–1 mA) were delivered to evoke stable responses with amplitudes ≤200 pA. Stimulation intensity was adjusted to evoke stable responses and kept constant throughout each recording. Baseline recordings were initiated only after achieving stable eEPSCs that varied by no more than 30% over a period exceeding 1 min (corresponding to at least three consecutive sweeps, delivered every 20 s).

2.7. AMPA/NMDA ratio

AMPAR- and NMDAR-mediated components were measured from eEPSCs recorded in the presence of PTX (100 μM). AMPAR responses were obtained at −60 mV. The holding potential was then shifted to +40 mV to record mixed AMPA+NMDA responses (30 sweeps, one every 20 s). D-APV (50 μM) was subsequently applied to isolate the AMPAR component at +40 mV. The NMDA component was calculated by subtracting the averaged AMPAR trace from the mixed response, and its amplitude was measured 20 ms after the peak of the AMPAR current. The AMPA/NMDA ratio was calculated by dividing the peak amplitude of the pharmacologically isolated AMPAR-mediated current recorded at +40 mV by the NMDAR-mediated current amplitude obtained by digital subtraction at the same holding potential. These recordings were obtained from visually identified NAc core MSNs without Drd1a-tdTomato-based subtype identification.

2.8. Paired-pulse ratio (PPR)

Paired-pulse ratio (PPR) was assessed at −60 mV using two consecutive stimuli delivered with a 70 ms inter-stimulus interval producing a double eEPSC. PPR was calculated as the ratio between the second and first eEPSC amplitudes (R2/R1), using responses ≤200 pA.

2.9. Rectification index (RI)

Rectification properties of AMPAR-mediated currents were assessed using CsCl-based internal solution containing spermine (100 μM), and in the presence of PTX (100 μM) and D-APV (50 μM) in the ACSF. eEPSCs were recorded at holding potentials ranging from −60 to +40 mV in 20 mV increments. For each potential, 30 sweeps were collected and averaged. The rectification index was calculated as the ratio of the absolute current amplitude at +40 mV to that at −60 mV.

2.10. NASPM sensitivity

To assess the contribution of calcium-permeable AMPARs, MSNs were voltage-clamped at −60 mV. After establishing a stable baseline for at least 10 min (one eEPSC every 20 s), 1-naphthylacetyl spermine (NASPM) (150 μM) was bath-applied for 15 min. Inhibition was expressed as the percentage reduction in mean eEPSC amplitude, comparing baseline responses with those recorded during the last 5 min of NASPM application. The rectification-index and NASPM-sensitivity experiments shown in Fig. 4E–L were performed in visually identified NAc core MSNs without Drd1a-tdTomato-based subtype identification.

Fig. 4. — Increased AMPAR-dependent calcium activity and functional contribution of calcium-permeable AMPARs in the NAc core of APP/PS1 mice. (A) Schematic representation of the experimental design used for calcium imaging. AAV-Syn-GCaMP6s was injected into the NAc of 5-month-old APP/PS1 mice and WT littermates and allowed to express for 2–3 weeks before acute brain slice preparation at 6 months-old. Electrically evoked GCaMP6s fluorescence was recorded in the NAc core. A representative image shows neuronal GCaMP6s expression, with the anterior commissure (ac) indicated as an anatomical landmark. Scale bar: 60 μm. (B) Representative electrically evoked calcium transients recorded in NAc core slices from WT (black) and APP/PS1 (red) mice before and after CNQX application. Scale bars: 25% ΔF/F, 2 s. (C) Time course of normalized electrically evoked calcium activity before and during CNQX application. APP/PS1 slices showed a greater reduction in calcium activity following AMPAR blockade than WT slices. Symbols and shaded regions represent mean ± SD. (D) Quantification of CNQX-induced inhibition of electrically evoked calcium activity, expressed relative to baseline. CNQX produced greater inhibition in APP/PS1 than in WT slices (unpaired two-tailed t-test: t(15) = 3.725, p = 0.002). Numbers of recordings/mice: WT, n = 9/4; APP/PS1, n = 8/4. Each point represents one calcium-imaging recording obtained from a separate acute brain slice. (E) Schematic representation of whole-cell voltage-clamp recordings used to assess AMPAR-mediated eEPSCs in visually identified NAc core MSNs from 6-month-old APP/PS1 mice and WT littermates. These recordings were performed without Drd1a-tdTomato-based subtype identification. Recordings were performed using a CsCl-based internal solution containing QX-314, TEA, and spermine, in ACSF containing PTX and D-APV. (F) Representative AMPAR-mediated eEPSCs recorded at holding potentials ranging from −60 to +40 mV in WT (black) and APP/PS1 (red) MSNs. Selected holding potentials are indicated. Scale bars: 50 pA, 15 ms. (G) Normalized current–voltage relationships showing greater inward rectification in APP/PS1 MSNs than in WT MSNs. (H) Quantification of the rectification index showing a lower RI in APP/PS1 MSNs than in WT MSNs, consistent with greater inward rectification and an increased functional contribution of CP-AMPARs (unpaired two-tailed t-test: t(14) = 2.989, p = 0.0098). Numbers of neurons/mice: WT, n = 10/4; APP/PS1, n = 6/4. (I) Schematic representation of the experimental protocol used to assess NASPM-sensitive AMPAR-mediated transmission. Following a 10-min baseline period, NASPM (150 μM) was applied for 15 min in the continuous presence of PTX and D-APV. (J) Representative eEPSCs recorded before NASPM application (1, baseline) and during NASPM exposure (2) in WT and APP/PS1 MSNs. Scale bars: 30 pA, 25 ms. (K) Time course of normalized eEPSC amplitude before and during NASPM application. APP/PS1 MSNs showed greater NASPM-induced inhibition than WT MSNs. Symbols and shaded regions represent mean ± SD. (L) Quantification of NASPM-induced inhibition of eEPSC amplitude, expressed relative to baseline and calculated from the final 5 min of NASPM application (20–25 min). NASPM produced greater inhibition in APP/PS1 MSNs than in WT MSNs (unpaired two-tailed t-test: t(16) = 3.244, p = 0.005). Numbers of neurons/mice: WT, n = 9/4; APP/PS1, n = 9/4. Each point in panels H and L represents one neuron recorded from a separate acute brain slice, with only one neuron recorded per slice. Bars and error bars represent mean ± SD, and individual data points are shown. **p < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

2.11. Long-term depression

All LTD experiments were performed in the presence of picrotoxin (PTX, 100 μM) and D-APV (50 μM). For HFS-induced LTD, recordings were obtained using the CsCl-based internal solution described above, which contained QX-314 (5 mM) to suppress voltage-gated Na⁺channel-dependent action potentials in the recorded postsynaptic MSN. After recording a stable baseline for at least 10 min at a holding potential of −60 mV, LTD was induced by extracellular tetanic stimulation of afferent fibers in the NAc core while maintaining the recorded neuron under voltage clamp at −60 mV. The HFS protocol consisted of four trains delivered at 100 Hz for 1 s, with 10 s between train onsets. Each individual stimulus pulse had a duration of 0.5 ms. This protocol differs from classical suprathreshold intracellular/current-clamp striatal plasticity paradigms (Calabresi et al., 1992), because HFS here refers to tetanic extracellular stimulation of afferent inputs under voltage-clamp conditions rather than action potential generation in the recorded postsynaptic neuron. Similar voltage-clamp approaches using intracellular QX-314 have been used to study LTD in accumbal and striatal MSNs (Mathur et al., 2011).

Evoked EPSCs were recorded for at least 40 min after HFS induction. Responses were normalized to the baseline period before quantification, and LTD magnitude was calculated as the percentage change in normalized eEPSC amplitude during the last 5 min of recording relative to baseline.

For mGluR1/5-dependent LTD, recordings were performed using an internal solution containing EGTA (1 mM) instead of BAPTA. After recording a stable baseline for at least 10 min, (RS)-3,5-dihydroxyphenylglycine (DHPG, 50 μM) was bath-applied for 5 min, and eEPSCs were monitored for at least 25 min thereafter. LTD magnitude was calculated as the percentage reduction in mean eEPSC amplitude during the final 5 min relative to baseline.

2.12. Stereotaxic injections

Five-month-old male WT and APP/PS1 mice were used. Stereotaxic surgery was performed to deliver adeno-associated viruses (AAVs) expressing the genetically encoded calcium indicator GCaMP6s under the synapsin promoter, as previously described (Armijo-Weingart et al., 2024). A total of 200 nL of AAV1-Syn-GCaMP6s.WPRE.SV40 (1.76 × 10¹³ GC/ml; Addgene #100843-AAV1) or 400 nL of pAAV-CAG-dLight1.1 (7 × 10¹² vg/ml; Addgene #111067-AAV5) was injected bilaterally into the NAc core using a stereotaxic alignment system (Kopf Instruments) at 50 nL per minute. Injection coordinates relative to bregma were: AP +0.13 mm, ML ±0.11 mm, and DV −0.4 mm (Allen Brain Atlas). Animals received post-operative analgesia (Meloxicam 2 mg/kg). Mice were anesthetized with 4% isoflurane/oxygen and positioned in a stereotaxic frame; anesthesia was maintained with 2–3% isoflurane/oxygen throughout the procedure. After leveling the skull, a small craniotomy was made at the target site. A 1 μL Neuros Hamilton syringe was lowered slowly to the desired depth, and viral solution was delivered. The syringe was left in place for 3 min post-infusion before withdrawal, and incisions were closed with Leukosan adhesive.

2.13. Calcium and dopamine photometry

Brain slice photometry were evaluated as previously reported (Salinas et al., 2023). Two to three weeks after AAV injection, mice were 6 months old at the time of experiments. Acute coronal slices (300 μm) containing the NAc core were prepared for calcium or dopamine imaging. Slices were transferred to an upright microscope and continuously perfused with oxygenated aCSF (1 mL/min). The recording region of interest (medial to the anterior commissure, corresponding to the NAc core) was visualized under fluorescence to confirm GCaMP6s or dLight1.1 expression. A bipolar stimulating electrode (DS3 Isolated Current Stimulator, Digitimer, UK) was placed on the slice surface near the area of interest. Stimulation consisted of single electrical pulses (400–800 μA, 1 ms duration) delivered every 2 min, generating one evoked fluorescence transient per stimulus. Transients were measured by slice photometry using a Horiba PTI D-104 Microscope Photometer equipped with a 710 nm photomultiplier tube mounted on an Olympus BX51 microscope, with a 120 LED Boost High-Power illumination system and appropriate excitation/emission filters (488 nm for GCaMP6s and 405 nm for dLight1.1).

Fluorescence signals were acquired using PatchMaster software and expressed as ΔF/F₀, where F₀ was defined as the mean baseline fluorescence prior to stimulation. A stable baseline was recorded for 12 min (six responses), followed by bath application of the AMPAR antagonist CNQX and continued stimulation for an additional 12 min until responses reached a plateau. For analysis, ΔF/F₀ values were normalized to baseline and expressed as percentage change. Drug effects were quantified by comparing the mean response during the last 6 min in the presence of CNQX with baseline for each slice. At least two recordings per mouse were obtained from independent NAc slices; each slice was treated as an individual observation, while the number of animals contributing to each dataset is reported in the corresponding figure legends.

2.14. Western blot

Protein expression analysis were done as previously described (Fernández-Pérez et al., 2020). Total NAc (shell and core included) was dissected from coronal brain slices obtained from 6-month-old WT and APP/PS1 male mice. Each biological replicate corresponded to the entire NAc isolated from a single animal. Tissue samples were homogenized in ice-cold RIPA buffer supplemented with protease and phosphatase inhibitor cocktails. Lysates were centrifuged at 14,000 ×g for 15 min at 4 °C, and supernatants were collected for protein quantification. Equal amounts of protein (50 μg per lane) were denatured in buffer, separated by SDS–PAGE at constant 80 Volt for 3 h, and transferred onto PVDF membranes at 250 mA for 3 h. Membranes were blocked in 5% non-fat dry milk prepared in TBS-T (Tris-buffered saline containing 0.1% Tween-20) for 1 h at room temperature and then incubated overnight at 4 °C with primary antibodies against SV2 (DSHB #AB2315387), PSD95 (Synaptic Systems #124011), GluA1 (Synaptic Systems #182011), GluA2 (Synaptic Systems #182103). Protein levels were normalized to Gβ (Santa Cruz, sc-166,123), which was used as an internal loading control based on previous Western blot analyses of nucleus accumbens lysates from our laboratory and related studies (Fernández-Pérez et al., 2020; San Martin et al., 2020). Gβ was used only for normalization of protein loading and not as a readout of G-protein signaling. GluA1 and GluA2 were selected a priori because the primary hypothesis concerned altered contribution of GluA2-lacking CP-AMPARs. GluA3 was not included in the original experimental design and was not excluded on the basis of preliminary results. Consequently, the present analysis does not provide a complete assessment of AMPAR subunit composition.

2.15. qRT-PCR

The total NAc (core and shell included) was dissected from 300 μm coronal brain slices. Total RNA was extracted using TRIzol reagent according to the manufacturer's instructions and treated with DNase to eliminate potential genomic DNA contamination. Complementary DNA (cDNA) was synthesized from 2 μg of total RNA using reverse transcriptase and oligo(dT) primers. Quantitative real-time PCR was performed for 40 cycles using SYBR Green Universal Master Mix (Agilent Technologies) and gene-specific primers targeting NMDA receptor subunits and AMPA receptor subunits. The following primer pairs were used: Gria1 (GluA1), forward 5′-ACCCTCCATGTGATCGAAATG-3′ and reverse 5′-GGTTCTATTCTGGACGCTTGAG-3′; Gria2 (GluA2), forward 5′-AAAGAATACCCTGGAGCACAC-3′ and reverse 5′-CCAAACAATCTCCTGCATTTCC-3′; Grin1 (NMDA receptor subunit 1), forward 5′-AAATGTGTCCCTGTCCATACTC-3′ and reverse 5′-CCTGCCATGTTCTCAAAAGTG-3′; Grin2b (NMDA receptor subunit 2B), forward 5′-GAACGAGACTGACCCAAAGAG-3′ and reverse 5′-CAGAAGCTTGCTGTTCAATGG-3′. Cyclophilin A was used as the housekeeping gene, with forward primer 5′-ATAATGGCACTGGTGGCAAGTC-3′ and reverse primer 5′-ATTCCTGGACCCAAAACGCTCC-3′. Relative expression levels were normalized to Cyclophilin A and calculated using the 2^−ΔΔCt method.

2.16. Palatable food-conditioned place preference

A modified palatable food-conditioned place preference (CPP) paradigm was used to assess reward-related contextual preference. Six-month-old WT and APP/PS1 mice were tested in a two-compartment apparatus (33 × 27 × 20 cm) located inside individual ventilated, light- and sound-attenuating enclosures. The two compartments were distinguished by tactile floor cues consisting of either a wide-grid or a small-grid floor. A removable transparent divider was used to separate the compartments during conditioning sessions, whereas the divider was removed during pre- and post-conditioning preference tests to allow free exploration of both contexts.

On day 1, mice were placed in the apparatus and allowed to freely explore both grid-associated contexts for 15 min. Time spent on each grid was recorded to determine baseline grid preference. For each mouse, the chocolate-paired context (CS+) was assigned to the grid that was less preferred during the pre-conditioning session, whereas the standard pellet-paired context (CS−) was assigned to the initially preferred grid. This biased assignment was used to minimize the contribution of innate tactile-context preference and to favor the detection of conditioning driven by the palatable food stimulus. From days 2 to 17, mice underwent 16 conditioning sessions, alternating between the chocolate-paired context (CS+) and the standard pellet-paired context (CS−). During each conditioning session, mice were confined to the corresponding compartment for 30 min. Chocolate and standard pellet consumption were measured after each session and normalized to body weight (g/kg/30 min). Twenty-four hours after the final conditioning session, mice were tested in a 30-min post-conditioning CPP session in the absence of food. During this test, the divider was removed and mice were allowed to freely explore both contexts. Time spent in each context was recorded by video tracking and analyzed using ANY-maze software (Stoelting Co.). CPP was quantified as the time spent in the chocolate-paired context (CS+) compared with the pellet-paired context (CS−).

2.17. Elevated plus maze (EPM)

Anxiety-related behavior was evaluated using the EPM test in 6-month-old WT and APP/PS1 mice during the light phase of light cycle. Before testing, mice were habituated to the experimental room for 1 h. The apparatus consisted of two open arms and two closed arms arranged in a plus-shaped configuration. At the beginning of the test, each mouse was placed in the central platform facing an open arm and was allowed to freely explore the maze for 10 min. Behavior was recorded by video tracking and analyzed using ANY-maze software (Stoelting Co.). The variables quantified were time spent in the open arms, time spent in the closed arms, number of entries into open and closed arms, and distance traveled in open and closed arms. The apparatus was cleaned between animals to minimize olfactory cues.

2.18. Social behavior experiments

Social behavior was assessed in 6-month-old WT and APP/PS1 mice using the three-chamber social interaction test. The apparatus consisted of three interconnected chambers, with identical wire cups located in the two lateral chambers. Mice were habituated to the experimental room for 1 h and then allowed to explore the empty apparatus for 10 min. For social preference test, an unfamiliar stimulus mouse (S1) was enclosed in one wire cup, while the opposite cup remained empty. The test mouse was allowed to explore the apparatus for 10 min. Social preference was assessed by comparing sniffing time and time spent in the chamber containing S1 with the corresponding measures for the empty cup and chamber. For the social novelty phase, S1 remained in place and a second unfamiliar mouse (S2) was introduced into the opposite wire cup. The test mouse explored the apparatus for an additional 10 min. Social novelty was assessed by comparing sniffing time and chamber occupancy for S1 and S2. Behavior was video-recorded and analyzed using ANY-maze software (Stoelting Co.). The apparatus and wire cups were cleaned between trials to minimize olfactory cues. Stimulus mice were unfamiliar, age- and sex-matched conspecifics.

2.19. Statistical analysis

Data acquisition and analysis were performed blind to genotype. In Drd1a-tdTomato experiments, fluorescence was used only to identify D1R+ and D1R− MSNs. The experimental unit was defined according to the assay: mouse for histological, molecular, and behavioral analyses; neuron for electrophysiology; and slice recording for photometry. When multiple measurements were obtained from the same animal, values were averaged when required to generate one biological replicate. Sample sizes are reported in the figure legends.

Electrophysiological data were analyzed using Clampfit v11, and statistical analyses were performed in GraphPad Prism v10. Normality was assessed using the Shapiro–Wilk test. Two-group comparisons were performed using two-tailed unpaired Student's t-tests, Welch's t-tests, or Mann–Whitney U tests, as appropriate. Comparisons involving multiple groups were analyzed using one- or two-way ANOVA or Kruskal–Wallis tests. Repeated-measures data were analyzed using repeated-measures ANOVA or mixed-effects models fitted by restricted maximum likelihood when observations were missing or unbalanced. Geisser–Greenhouse correction and Sidak's, Tukey's, Bonferroni's, or Dunn's multiple-comparison tests were applied when appropriate. RT-qPCR data were analyzed after log₂ transformation. Statistical significance was defined as p < 0.05. Data are presented as mean ± SD unless otherwise indicated, and exact test statistics, degrees of freedom, p values, and sample sizes are reported in the corresponding figure legends.

3. Results

3.1. Early intraneuronal Aβ accumulation precedes detectable compact fibrillar plaque deposition and is associated with selective enhancement of excitatory synaptic transmission in D1R+ MSNs

To establish the stage of amyloid pathology in the NAc, we first evaluated neuronal Aβ immunoreactivity and compact fibrillar plaque deposition across ages. At 3 months, MOAB-2 immunoreactivity was increased within MAP2-positive neuronal somata of APP/PS1 mice compared with WT controls (Fig. S1A, B). Thioflavin-S staining revealed no detectable compact fibrillar deposits in the NAc core of either genotype at 6 months, whereas sparse deposits were detected in 12-month-old APP/PS1 mice (Fig. S1C, D). In the cortex, Thioflavin-S-positive deposits were already detectable in APP/PS1 mice at 6 months and remained evident at 12 months (Fig. S1D, E). At 6 months, neither Iba1 fluorescence intensity nor Iba1-positive cell density differed between genotypes (Fig. S1F–H). Thus, the 6-month-old APP/PS1 NAc core was characterized by increased neuronal Aβ immunoreactivity without detectable compact fibrillar plaque deposition or changes in the Iba1-associated measures examined.

We next determined whether somatic Aβ accumulation differed between MSN subtypes. In 6-month-old APP/PS1;D1-tdTomato mice, Aβ immunoreactivity was increased in both D1R+ and D1R− MSN somata compared with their corresponding WT groups, with no significant difference between MSN subtypes within either genotype (Fig. 1A–D). Therefore, increased somatic Aβ immunoreactivity was not restricted to D1R+ MSNs. Despite this comparable Aβ burden, AMPA receptor-mediated synaptic transmission was differentially altered between MSN subtypes. Whole-cell recordings showed increased mean spontaneous excitatory postsynaptic current (sEPSC) amplitude in APP/PS1 D1R+ MSNs, whereas no genotype-dependent difference was detected in D1R− MSNs (Fig. 1E–G). Pooled event-level amplitude distributions, shown descriptively, displayed a rightward shift in APP/PS1 D1R+ MSNs but minimal separation between genotypes in D1R− MSNs (Fig. 1H). Inferential analyses were based on mean values obtained from individual neurons. sEPSC frequency showed no genotype effect or genotype × MSN subtype interaction, although an overall main effect of MSN subtype indicated higher frequency in D1R+ than in D1R− MSNs (Fig. 1I). These alterations were not detected at 3 months, when sEPSC amplitude, frequency, rise time, and decay time were comparable between WT and APP/PS1 MSNs (Fig. S2A–I). In non-subtype-resolved recordings at 6 months, APP/PS1 MSNs showed increased sEPSC amplitude without changes in frequency or kinetics (Fig. S2J–R), supporting the emergence of altered excitatory synaptic transmission between 3 and 6 months.

Together, these findings show that increased somatic Aβ immunoreactivity occurs in both MSN subtypes during an early stage without detectable compact fibrillar plaques in the NAc core. However, enhanced AMPA receptor-mediated synaptic strength was selective to D1R+ MSNs, indicating differential functional vulnerability despite comparable somatic Aβ accumulation.

3.2. Selective impairment of long-term depression in the nucleus accumbens emerges at pre-plaque stages and selectively affects D1R+ MSNs

To determine whether early intraneuronal Aβ accumulation is associated with altered synaptic plasticity, we evaluated long-term depression (LTD) in NAc core MSNs from WT and APP/PS1 mice. At 3 months, HFS-induced LTD did not differ between genotypes. In contrast, at 6 months, LTD magnitude was markedly reduced in APP/PS1 MSNs compared with WT controls (Fig. S3A–G). Similarly, LTD induced by the group I metabotropic glutamate receptor agonist DHPG was attenuated in APP/PS1 MSNs (Fig. S3H–K). Basal paired-pulse ratio (PPR) did not differ between genotypes (Fig. S3L–N), arguing against a major alteration in presynaptic release probability under these conditions. These recordings were obtained from visually identified MSNs without D1R reporter-based identification and therefore provide population-level evidence only. Because visual patch-clamp sampling does not ensure equal representation of D1R+ and D1R− MSNs, the cellular composition of this dataset cannot be reconstructed retrospectively.

We next assessed whether the HFS-LTD deficit was restricted to a specific MSN subtype. Recordings from D1R+ and D1R− MSNs of 6-month-old WT;D1-tdTomato and APP/PS1;D1-tdTomato mice revealed a significant genotype × MSN subtype interaction. HFS-induced LTD was reduced in APP/PS1 D1R+ MSNs compared with WT D1R+ MSNs, whereas no genotype-dependent difference was detected in D1R− MSNs. Within APP/PS1 mice, LTD magnitude was greater in D1R− than in D1R+ MSNs (Fig. 2A–D). The post-LTD/pre-LTD PPR ratio did not differ among groups (Fig. 2E), providing no evidence for a major presynaptic contribution to the cell-type-specific LTD deficit. Together, these results show that impaired LTD is detectable in the NAc core at 6 months but not at 3 months and that this deficit is selectively expressed in D1R+ MSNs during the pre-plaque stage.

Fig. 2. — HFS-induced LTD is selectively impaired in D1R+ MSNs but preserved in D1R− MSNs in the NAc core of 6-month-old APP/PS1 mice. (A) Schematic representation of the experimental configuration used for whole-cell voltage-clamp recordings of AMPAR-mediated eEPSCs in NAc core MSNs. Recordings were performed using a CsCl-based internal solution containing QX-314 and TEA, in ACSF containing picrotoxin and D-APV. After a 10-min baseline period, LTD was induced using four high-frequency stimulation trains (HFS; 100 Hz for 1 s), with 10 s between train onsets, and synaptic responses were monitored for 50 min after induction. (B) Representative paired-pulse eEPSC traces recorded from D1R+ and D1R− MSNs in WT;D1-tdTomato (black) and APP/PS1;D1-tdTomato (red) mice before HFS induction (1 and 1′, baseline) and during established LTD (2 and 2′). Unprimed and primed labels indicate the first and second responses of each paired-pulse recording, respectively. Overlaid traces illustrate changes in eEPSC amplitude following HFS. Scale bars: 50 pA, 35 ms. (C) Time course of normalized eEPSC amplitude in D1R+ and D1R− MSNs from WT;D1-tdTomato and APP/PS1;D1-tdTomato mice. Symbols and shaded regions represent mean ± SEM. (D) Quantification of HFS-induced LTD magnitude, calculated as 100 minus the mean normalized eEPSC amplitude during the final 10 min of recording (40–50 min). A significant genotype × MSN subtype interaction was detected (ordinary two-way ANOVA: interaction, F(1,20) = 16.30, p < 0.001; MSN subtype effect, F(1,20) = 5.635, p = 0.030; genotype effect, F(1,20) = 0.6676, p = 0.420). Šídák's multiple-comparisons test showed reduced LTD in APP/PS1 D1R+ MSNs compared with WT D1R+ MSNs (p = 0.010), whereas no genotype-dependent difference was detected in D1R− MSNs (p = 0.190). Within APP/PS1 mice, LTD magnitude was greater in D1R− than in D1R+ MSNs (p = 0.002). (E) Paired-pulse ratio expressed as the ratio between post-LTD and pre-LTD values. No significant differences were detected among groups (Kruskal–Wallis test: H(3) = 4.149, p = 0.246; Dunn's multiple-comparisons test, all adjusted p > 0.05). Each point represents one neuron recorded from a single acute brain slice, with only one neuron recorded per slice. For panels B–D, the numbers of neurons/mice were: WT D1R+, n = 6/5; WT D1R−, n = 7/5; APP/PS1 D1R+, n = 6/5; and APP/PS1 D1R−, n = 5/5. For panel E, the numbers of neurons/mice were: WT D1R+, n = 6/5; WT D1R−, n = 8/5; APP/PS1 D1R+, n = 6/5; and APP/PS1 D1R−, n = 5/5. Bars and error bars represent mean ± SD, and individual data points are shown. *p < 0.05, **p < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.3. Alterations in AMPAR-mediated synaptic transmission in the nucleus accumbens are not explained by transcriptional changes

To further characterize glutamatergic synaptic alterations in the NAc core, we measured AMPAR- and NMDAR-mediated eEPSCs in MSNs from 6-month-old WT and APP/PS1 mice. The AMPA/NMDA ratio was reduced in APP/PS1 MSNs compared with WT controls (Fig. 3A–C). Because the AMPA/NMDA ratio depends on both receptor-mediated components, its reduction does not determine whether AMPAR-mediated current, NMDAR-mediated current, or both were altered. NMDAR subunit protein levels were not examined, and unchanged Grin1 and Grin2b transcript levels do not exclude changes in NMDAR protein abundance, subunit composition, synaptic localization, or function. Moreover, the AMPA/NMDA measurements were obtained from non-subtype-resolved NAc core MSNs. Therefore, this result was interpreted as evidence of an altered relative contribution of AMPAR- and NMDAR-mediated currents, rather than as evidence of increased CP-AMPAR function. We next quantified glutamatergic synaptic proteins in total NAc lysates. These measurements were performed in total RIPA-soluble lysates prepared from the whole NAc, including core and shell, and therefore reflect regional total protein abundance. They do not distinguish synaptic from intracellular or extrasynaptic protein pools and cannot be assigned specifically to D1R+ or D1R− MSNs. Expression of the presynaptic marker SV2 did not differ between genotypes. In contrast, the postsynaptic scaffold protein PSD95 and the AMPAR subunits GluA1 and GluA2 were increased in APP/PS1 mice compared with WT controls (Fig. 3D, E).

Fig. 3. — Reduced AMPA/NMDA ratio in NAc core MSNs and altered glutamatergic protein expression in the NAc of APP/PS1 mice. (A) Schematic representation of the experimental design used to assess AMPAR- and NMDAR-mediated eEPSCs in NAc core MSNs from 6-month-old APP/PS1 mice and WT littermates. Whole-cell voltage-clamp recordings were performed at +40 mV in the presence of picrotoxin. Mixed AMPAR- and NMDAR-mediated eEPSCs were initially recorded, followed by pharmacological isolation of the AMPAR-mediated component after application of D-APV. The NMDAR-mediated component was obtained by digital subtraction of the AMPAR-mediated current from the mixed response. (B) Representative AMPAR- and NMDAR-mediated eEPSCs recorded at +40 mV from WT (black) and APP/PS1 (red) MSNs. Scale bars: 50 pA, 100 ms. (C) Quantification of the AMPA/NMDA ratio showing a significant reduction in APP/PS1 MSNs compared with WT MSNs (unpaired two-tailed t-test: t(23) = 2.363, p = 0.0270). Each point represents one neuron recorded from a single acute brain slice, with only one neuron recorded per slice. Numbers of neurons/mice: WT, n = 12/4; APP/PS1, n = 13/4. (D) Representative Western blots of SV2, PSD95, GluA1, and GluA2 in NAc lysates from WT and APP/PS1 mice. Gβ was used as the internal loading control for each protein. (E) Quantification of total protein levels relative to Gβ. SV2 levels did not differ between WT and APP/PS1 mice (unpaired two-tailed t-test: t(15) = 0.1462, p = 0.886; WT, n = 8; APP/PS1, n = 9). PSD95 levels were increased in APP/PS1 mice (Welch's unpaired two-tailed t-test: t(16.19) = 2.762, p = 0.014; WT, n = 12; APP/PS1, n = 12). GluA1 values were higher in APP/PS1 mice (two-tailed Mann–Whitney test: U = 80.50, p = 0.016; WT, n = 19; APP/PS1, n = 16). GluA2 levels were also increased in APP/PS1 mice (unpaired two-tailed t-test: t(18) = 3.314, p = 0.0039; WT, n = 9; APP/PS1, n = 11). For Western blot analyses, each point represents one animal. Bars and error bars represent mean ± SD, and individual data points are shown. *p < 0.05, **p < 0.01. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

To determine whether these protein-level changes were accompanied by altered transcription, we quantified Gria1, Gria2, Grin1, and Grin2b mRNA in the NAc of 6- and 9-month-old mice. No main effect of genotype or age × genotype interaction was detected for any transcript. However, all four transcripts showed significant main effects of age (Fig. S4A–D). The absence of corresponding genotype-dependent transcript changes indicates that the differences in total protein abundance were not accompanied by detectable changes in the mRNAs examined. However, the present experiments do not determine whether this dissociation reflects post-transcriptional regulation, protein stability, trafficking, or differences among cellular and subregional protein pools. Moreover, Western blot analyses were performed in total whole-NAc lysates and therefore cannot establish synaptic localization, AMPAR stoichiometry, or cell-type specificity. Increased total GluA1 and GluA2 abundance cannot be assigned to D1R+ MSNs or interpreted as direct evidence of CP-AMPAR composition. GluA3 was not assessed, further limiting conclusions regarding AMPAR subunit composition. Together, these findings show that APP/PS1 mice exhibit a reduced AMPA/NMDA ratio in NAc core MSNs and increased abundance of selected postsynaptic proteins in total whole-NAc lysates, without genotype-dependent changes in the glutamate receptor transcripts examined.

3.4. Increased AMPAR-dependent calcium signaling and functional contribution of calcium-permeable AMPARs in the nucleus accumbens of APP/PS1 mice

To determine whether altered AMPAR-mediated transmission in APP/PS1 mice was accompanied by changes in AMPAR functional properties, we first assessed electrically evoked calcium signals in NAc core slices expressing GCaMP6s. AMPAR blockade with CNQX produced greater inhibition of evoked calcium responses in APP/PS1 slices than in WT slices (Fig. 4A–D), indicating an increased AMPAR-dependent component of the evoked calcium signal. We next examined the rectification properties of AMPAR-mediated eEPSCs in NAc core MSNs. Current–voltage relationships showed greater inward rectification in APP/PS1 MSNs, reflected by a lower rectification index compared with WT MSNs (Fig. 4E–H). This electrophysiological profile is consistent with an increased functional contribution of GluA2-lacking, calcium-permeable AMPA receptors (CP-AMPARs). To further assess CP-AMPAR contribution, we measured the sensitivity of AMPAR-mediated eEPSCs to NASPM. NASPM produced greater inhibition of eEPSC amplitude in APP/PS1 MSNs than in WT MSNs (Fig. 4I–L), demonstrating an increased NASPM-sensitive component of excitatory synaptic transmission. The GCaMP6s measurements were obtained at the slice level and were not cell-type specific. Likewise, the rectification-index and NASPM-sensitivity recordings were obtained from visually identified NAc core MSNs without Drd1a-tdTomato-based subtype identification. Therefore, these datasets provide slice- or population-level evidence and do not establish whether the increased CP-AMPAR contribution is preferentially expressed in D1R+ or D1R− MSNs. Together, the enhanced AMPAR-dependent calcium signal, greater inward rectification, and increased NASPM sensitivity provide convergent evidence for an increased functional contribution of CP-AMPARs to excitatory synaptic transmission in the NAc core of 6-month-old APP/PS1 mice.

3.5. Impaired mGluR1/5-dependent LTD in APP/PS1 D1R+ MSNs is accompanied by persistent CP-AMPAR contribution

Because APP/PS1 mice exhibited both impaired LTD and increased CP-AMPAR contribution in the NAc core, we next examined whether these alterations converged in specific MSN subtypes. Whole-cell recordings were performed in D1R+ and D1R− MSNs from 6-month-old WT;D1-tdTomato and APP/PS1;D1-tdTomato mice using a sequential protocol comprising baseline recording, induction of mGluR1/5-dependent LTD with DHPG, and subsequent CP-AMPAR blockade with NASPM (Fig. 5A, B). DHPG induced significant LTD in WT D1R+, WT D1R−, and APP/PS1 D1R− MSNs. In contrast, no significant depression was detected in APP/PS1 D1R+ MSNs (Fig. 5C, D). During the LTD period, normalized eEPSC amplitude was higher in APP/PS1 D1R+ MSNs than in WT D1R+ and APP/PS1 D1R− MSNs, whereas LTD did not differ between WT and APP/PS1 D1R− MSNs. These results identify a selective impairment of mGluR1/5-dependent LTD in APP/PS1 D1R+ MSNs.

Fig. 5. — mGluR-dependent LTD is selectively impaired in D1R+ MSNs, and residual synaptic transmission is selectively NASPM-sensitive in the NAc core of 6-month-old APP/PS1 mice. (A) Schematic representation of whole-cell voltage-clamp recordings of AMPAR-mediated eEPSCs in NAc core MSNs from 6-month-old WT;D1-tdTomato and APP/PS1;D1-tdTomato mice. Recordings were performed using a CsCl-based internal solution containing QX-314 and TEA, in ACSF containing picrotoxin and D-APV. After a 10-min baseline period, mGluR-dependent LTD was induced by bath application of DHPG for 5 min. NASPM was subsequently applied from 35 to 50 min. Period 1 corresponds to baseline (0–10 min), period 2 to the final 5 min before NASPM application (30–35 min), used to quantify DHPG-induced LTD, and period 3 to the final 5 min of NASPM application (45–50 min). (B) Representative paired-pulse AMPAR-mediated eEPSCs recorded from D1R+ and D1R− MSNs in WT;D1-tdTomato (black) and APP/PS1;D1-tdTomato (red) mice during baseline (1), after DHPG-induced LTD (2), and during NASPM application (3). Scale bars: 50 pA, 40 ms. (C) Time course of normalized eEPSC amplitude in D1R+ and D1R− MSNs. DHPG-induced LTD was impaired in APP/PS1 D1R+ MSNs, whereas D1R− MSNs showed comparable synaptic depression between genotypes. Subsequent NASPM application produced an additional reduction in residual eEPSC amplitude selectively in APP/PS1 D1R+ MSNs. Horizontal bars indicate the periods of DHPG and NASPM application. Symbols and shaded regions represent mean ± SD. (D) Quantification of normalized eEPSC amplitude during baseline (1; 0–10 min), after DHPG-induced LTD (2; 30–35 min), and during the final 5 min of NASPM application (3; 45–50 min). The mixed-effects model showed significant effects of experimental stage and group, as well as a significant stage × group interaction (stage, F(1.841,38.67) = 137.8, p < 0.001; experimental group, F(3,21) = 4.459, p = 0.014; stage × experimental group interaction, F(5.524,38.67) = 5.484, p < 0.001). Tukey's multiple-comparisons test showed significant LTD relative to baseline in WT D1R+ (p < 0.001), WT D1R− (p < 0.001), and APP/PS1 D1R− MSNs (p < 0.001), whereas no significant depression was detected in APP/PS1 D1R+ MSNs after DHPG application (p = 0.293). During the LTD period, normalized eEPSC amplitude was higher in APP/PS1 D1R+ MSNs than in WT D1R+ MSNs (p = 0.031) and APP/PS1 D1R− MSNs (p = 0.014), whereas WT and APP/PS1 D1R− MSNs did not differ (p = 0.995). NASPM significantly reduced residual eEPSC amplitude in APP/PS1 D1R+ MSNs relative to the LTD period (p = 0.020) but produced no additional significant inhibition in WT D1R+ (p = 0.939), WT D1R− (p = 0.788), or APP/PS1 D1R− MSNs (p = 0.921). No significant differences among groups remained during the final NASPM period. (E) Paired-pulse ratio expressed as the ratio between post-LTD and pre-LTD values. No significant differences were detected among groups (Kruskal–Wallis test: H(3) = 2.02, p = 0.570). Each point represents one neuron recorded from a separate acute brain slice, with only one neuron recorded per slice. For panels C–D, the numbers of neurons/mice were: WT D1R+, n = 7/4; WT D1R−, n = 5/4; APP/PS1 D1R+, n = 7/5; and APP/PS1 D1R−, n = 6/5. For panel E, the numbers of neurons/mice were: WT D1R+, n = 6/4; WT D1R−, n = 5/4; APP/PS1 D1R+, n = 7/5; and APP/PS1 D1R−, n = 6/5. Bars and error bars represent mean ± SD, and individual data points are shown. *p < 0.05. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Subsequent application of NASPM further reduced residual eEPSC amplitude specifically in APP/PS1 D1R+ MSNs. NASPM produced no additional significant inhibition in WT D1R+, WT D1R−, or APP/PS1 D1R− MSNs, and no differences among groups remained during the final NASPM period (Fig. 5C, D). Thus, CP-AMPAR-sensitive transmission persisted selectively in APP/PS1 D1R+ MSNs after the failure of DHPG-induced LTD. The post-LTD/pre-LTD paired-pulse ratio did not differ among groups (Fig. 5E), providing no evidence for group-dependent differences in the presynaptic component associated with LTD expression. Together, these findings show that mGluR1/5-dependent LTD is selectively impaired in D1R+ MSNs of APP/PS1 mice and that this deficit is accompanied by a persistent NASPM-sensitive CP-AMPAR component.

3.6. Reduced evoked dopamine signaling co-occurs with selective alterations in reward-related behavior in pre-plaque APP/PS1 mice

To determine whether synaptic alterations in the NAc core were accompanied by changes in dopaminergic signaling, we measured electrically evoked dopamine responses using the genetically encoded fluorescent sensor dLight1.1 (Fig. 6A). Electrical stimulation elicited dLight1.1 transients in slices from both genotypes; however, response amplitudes were lower in APP/PS1 slices than in WT slices at intermediate and maximal stimulation intensities (Fig. 6B, C). These results indicate reduced evoked dopaminergic signaling in the NAc core of 6-month-old APP/PS1 mice.

We next assessed reward-related behavior using a chocolate-conditioned place preference paradigm (Fig. 6D). Before conditioning, APP/PS1 mice spent more time on the small-grid floor than on the wide-grid floor, whereas WT mice showed no significant baseline grid preference (Fig. 6E). Following conditioning, both genotypes spent more time in the chocolate-paired context (CS+) than in the pellet-paired context (CS−), indicating acquisition of conditioned contextual preference in both groups (Fig. 6F). APP/PS1 mice consumed more chocolate across conditioning sessions than WT mice, whereas standard pellet consumption did not differ between genotypes (Fig. 6G, H). No genotype-dependent differences were detected in the elevated-plus-maze or three-chamber social interaction measures examined (Fig. S5A–M). Thus, the altered baseline contextual preference and increased chocolate consumption were not accompanied by detectable differences in anxiety-related or social behavior under these experimental conditions.

Together, these findings show that reduced evoked dopaminergic signaling in the NAc core co-occurs with selective alterations in reward-related behavior in pre-plaque APP/PS1 mice. The data establish temporal coexistence, but not a causal relationship, between the dopaminergic and behavioral phenotypes.

4. Discussion

Our study identifies a cell-type-specific disruption of synaptic plasticity in the NAc during early stages of AD pathology. We show that long-term depression is selectively impaired in accumbal D1R+ MSNs despite comparable somatic Aβ immunoreactivity across MSN subtypes. This deficit was accompanied by an increased functional contribution of calcium-permeable AMPARs, whereas separate slice-photometry experiments revealed reduced evoked dopaminergic signaling in the NAc core. These alterations occur before detectable compact fibrillar plaque deposition in the NAc core and co-occur with selective changes in reward-related behavior. Together, these findings identify an early, cell-type-specific vulnerability of D1R+ MSNs and support an association between increased intraneuronal Aβ immunoreactivity and synaptic dysfunction in the NAc, as summarized in the proposed working model (Fig. 7).

Fig. 7. — Proposed model of cell-type-specific synaptic dysfunction in the nucleus accumbens core during early pre-plaque Alzheimer's disease pathology. The coronal brain section identifies the NAc core, and the enlarged central inset depicts the local population of MSNs. The synaptic schematics compare D1R+ MSNs (left) and D1R− MSNs (right) in APP/PS1 mice at an early stage characterized by increased intraneuronal Aβ immunoreactivity without detectable compact fibrillar plaques in the NAc core. In D1R+ MSNs, mGluR1/5-dependent LTD is selectively impaired, and residual glutamatergic transmission remains sensitive to NASPM, consistent with persistent functional contribution of calcium-permeable AMPARs and altered AMPAR regulation. In contrast, D1R− MSNs retain mGluR1/5-dependent LTD, and NASPM produces no additional significant inhibition after LTD induction, consistent with comparatively preserved AMPAR regulation under these experimental conditions. Electrically evoked dopamine signals were reduced in the NAc core of APP/PS1 mice. Separately, APP/PS1 mice exhibited increased palatable food consumption, whereas conditioned contextual preference was detected in both genotypes. The synaptic, dopaminergic, and behavioral alterations are presented as co-occurring findings, and causal relationships among these outcomes were not established. Dashed pathways represent hypothetical intracellular signaling relationships that were not directly tested in this study. Arrowheads and blunt-ended lines indicate activation and inhibition, respectively. CP-AMPAR, calcium-permeable AMPA receptor; CI-AMPAR, calcium-impermeable AMPA receptor; D1R, dopamine D1 receptor; mGluR1/5, group I metabotropic glutamate receptors; PKA, protein kinase A; PKC, protein kinase C; PSD95, postsynaptic density protein 95; SV2, synaptic vesicle glycoprotein 2.

Early dysfunction of the NAc has been reported during early stages of Alzheimer's disease, including alterations in neuronal activity and network hyperexcitability, yet the synaptic mechanisms underlying these changes remain poorly defined. Here, we provide evidence for a postsynaptic contribution characterized by an increased functional contribution of CP-AMPARs and selective impairment of mGluR1/5-dependent LTD in D1R+ MSNs. As this form of plasticity normally promotes AMPAR endocytosis and constrains excitatory synaptic strength, its disruption is consistent with the persistence of CP-AMPAR–mediated signaling and a shift toward increased excitatory synaptic strength in D1R+ MSNs. These findings may be relevant to the broader context of early neuronal hyperexcitability in AD (Vossel, 2026), although network hyperactivity and epileptiform activity were not assessed in the present study. Recent computational and neuroimaging approaches have identified a progressive disruption of excitation–inhibition balance across AD stages, with limbic regions showing early vulnerability. Notably, increased excitability within the NAc has been reported at the stage of mild cognitive impairment, preceding overt dementia (Li et al., 2025). These observations raise the possibility that cell-type-specific synaptic alterations in the NAc may contribute to broader circuit dysfunction.

AD is increasingly recognized as a disorder that perturbs distributed neural circuits before the onset of memory impairment, with early non-cognitive symptoms linked to dysfunction in limbic and reward-related networks (Frank, 2025; Masters, 2015). In this context, the NAc functions as a central integrative hub where glutamatergic and dopaminergic inputs converge to regulate motivational behavior (Bayassi-Jakowicka et al., 2022). Notably, structural and functional alterations of the NAc have been reported in patients, including reduced volume and disrupted connectivity with prefrontal regions involved in decision-making and inhibitory control (Contreras et al., 2020; Nie et al., 2017). By focusing on pre-plaque stages, our findings provide a framework for understanding the association between early intraneuronal Aβ immunoreactivity, altered synaptic plasticity, and reduced evoked dopaminergic signaling in this circuit. These results extend current models of AD by identifying the NAc as an additional site of early synaptic and neuromodulatory alteration.

A critical factor in interpreting Alzheimer's disease mechanisms is the disease stage at which pathology is examined. While most experimental studies have focused on advanced phases characterized by extracellular amyloid plaque deposition and overt cognitive impairment, our findings emphasize an earlier stage in which synaptic alterations are present before detectable compact fibrillar plaque deposition in the NAc core. In this context, the selective impairment of LTD together with increased CP-AMPAR contribution and reduced dopaminergic signaling supports the notion that early pathology involves a shift in synaptic balance that precedes classical neurodegenerative features. Rather than reflecting late-stage synaptic loss, these findings are consistent with an early synaptic phenotype characterized by enhanced AMPAR-mediated transmission in D1R+ MSNs, impaired LTD, and increased CP-AMPAR contribution. This synaptic phenotype is consistent with evidence that intraneuronal Aβ-associated alterations can precede prominent extracellular plaque pathology (Chang et al., 2025). Accordingly, the population-level LTD impairment observed in non-subtype-resolved recordings should not be interpreted as evidence that D1R+ and D1R− MSNs were uniformly affected.

The selective vulnerability of D1R+ MSNs observed in this study parallels findings from addiction and food restriction murine models, in which synaptic plasticity and AMPAR remodeling are preferentially disrupted in D1R+ MSNs within the NAc, linked with mGluR-dependent synaptic plasticity dysfunction (Hwang et al., 2025; Italia et al., 2025; Kawa et al., 2022). Notably, our experiments showed that NASPM further reduced residual synaptic transmission selectively in APP/PS1 D1R+ MSNs after DHPG failed to induce LTD, indicating persistent CP-AMPAR-sensitive transmission under these conditions. These findings align with forms of synaptic plasticity dysfunction described in other pathological contexts. Together, these results indicate that mGluR1/5-dependent plasticity deficits in this APP/PS1 model depend on neuronal subtype and disease stage.

The selective vulnerability of D1R+ MSNs may be influenced by the reduction in evoked dopaminergic signaling observed in the NAc core. D1 receptors exhibit lower affinity for dopamine than D2 receptors, rendering D1R+ MSNs particularly sensitive to reductions in dopamine availability (Gerlach et al., 2003; Richfield et al., 1989). Consistent with this framework, we observed reduced electrically evoked dLight1.1 responses in the NAc core at 6 months of age. These findings align with previous reports showing that dopamine release in the NAc is already diminished during pre-plaque stages in APPswe mice, together with compensatory changes in dopamine transporter expression (Nobili et al., 2017). Although the present data do not allow discrimination between impaired presynaptic release, dopaminergic terminal dysfunction, or early alterations in the integrity of dopaminergic VTA neurons, reduced evoked dopamine signaling could disproportionately affect D1R-mediated responses (Beaulieu and Gainetdinov, 2011), although this possibility was not directly tested.

Reduced evoked dopaminergic signaling may coexist with Aβ-associated postsynaptic alterations in the NAc, although the present experiments do not establish functional convergence between these processes. In support of this idea, Whitcomb et al. demonstrated that intracellular perfusion of Aβ oligomers into hippocampal neurons rapidly increases surface GluA1 expression and promotes CP-AMPAR insertion through a PKA-dependent mechanism (Whitcomb et al., 2015). In this line, one possible hypothesis would be that intracellular Aβ may alter PKA-dependent GluA1 trafficking. However, PKA activity, GluA1 phosphorylation, and surface receptor insertion were not measured in the present study. In addition, since D1R− neurons were defined by the absence of tdTomato fluorescence rather than by direct D2R labeling, the relative preservation of this population cannot be attributed specifically to D2R signaling. Together, the present data are consistent with impaired mGluR1/5-dependent depression and persistent functional CP-AMPAR contribution in D1R+ MSNs, without directly resolving the underlying receptor-trafficking mechanism.

In agreement with the present findings, Aguado et al. reported increased CP-AMPAR expression in the hippocampus of cognitive vulnerable APPswe mice at advanced disease stages (Aguado et al., 2024). Notably, mGluR5 expression was selectively reduced in vulnerable APPswe mice but preserved in cognitively vulnerable APPswe mice, paralleling the normalization of CP-AMPAR levels (Aguado et al., 2024). These observations suggest that coordinated regulation of mGluR1/5 signaling and CP-AMPAR composition may modulate synaptic vulnerability across disease stages and brain regions. Complementing this view, Guo et al. demonstrated that acute exposure to exogenous oligomeric Aβ in the NAc of young WT mice induces synaptic incorporation of CP-AMPARs, leading to spine loss, synaptic weakening, and motivational deficits (Guo et al., 2022). Importantly, this model involves acute exposure to exogenous extracellular Aβ oligomers, in contrast to our observation of persistent CP-AMPAR-sensitive transmission in APP/PS1 D1R+ MSNs following DHPG application during endogenously produced Aβ aggregates. Together, these studies suggest that CP-AMPAR dysregulation may represent a convergent feature across models, whereas disease stage, cellular context, and Aβ localization may influence the resulting synaptic phenotype.

The NAc functions as a key inhibitory hub within the mesolimbic circuit, regulating reward-related signal gain through the integration of glutamatergic inputs and dopaminergic modulation (Russo and Nestler, 2013). The selective loss of LTD in D1R+ MSNs may alter the output of this MSN population, although downstream circuit activity was not directly assessed and the exact relationship between reduced evoked dopaminergic signaling and increased chocolate consumption remains unresolved. Consistent with this framework, both genotypes developed a preference for the chocolate-paired context after conditioning, indicating that reward-context conditioning remained detectable under this protocol despite altered baseline grid preference in APP/PS1 mice. However, APP/PS1 mice displayed a selective increase in consumption of palatable solid food, while intake of standard chow remained unchanged. Together, these findings indicate that early synaptic dysfunction within the NAc is associated with altered palatable-food consumption and baseline contextual preference, without detectable genotype differences in the anxiety-related and social measures examined.

Some limitations should be considered when interpreting these findings. First, D1R− MSNs were identified by the absence of tdTomato fluorescence and were not directly confirmed as D2R-expressing neurons; therefore, conclusions regarding D2R-specific mechanisms cannot be directly drawn. In addition, all experiments were conducted exclusively in male mice, precluding assessment of sex-specific mechanisms. This is relevant given evidence that estradiol signaling modulates synaptic plasticity in the NAc through mGluR5- and endocannabinoid-dependent pathways and influences reward circuit function (Peterson et al., 2016). Moreover, synaptic plasticity in this region can rely on distinct molecular mechanisms in males and females despite similar functional outcomes, as shown by differences in the signaling pathways underlying long-term potentiation (Copenhaver and LeGates, 2024). Accordingly, the present conclusions are restricted to male mice, and whether the same synaptic and behavioral phenotypes occur in females remains to be determined. Second, although our data reveal a strong association between intraneuronal Aβ accumulation and synaptic alterations, the APP/PS1 model does not allow definitive attribution of these effects exclusively to Aβ and the absence of detectable Thioflavin-S-positive deposits does not exclude soluble, oligomeric, or non-fibrillar Aβ species in the NAc core. Nevertheless, previous studies are consistent with the possibility that intraneuronal Aβ can modulate excitatory synaptic function (Fernandez-Perez et al., 2021; Saavedra-Sieyes et al., 2025). However, these findings do not establish that intraneuronal Aβ is the causal driver of the synaptic phenotype observed in the present APP/PS1 model. The electrophysiological and pharmacological data support an increased functional contribution of CP-AMPARs but do not directly demonstrate GluA1/GluA2 surface trafficking or synaptic insertion. Moreover, Western blot analyses were performed in total whole-NAc lysates and did not distinguish synaptic from intracellular or extrasynaptic protein pools, receptor stoichiometry, or MSN subtype. GluA3 was not assessed, further limiting conclusions regarding AMPAR subunit composition.

While the present data do not establish direct causality, they indicate that intraneuronal Aβ immunoreactivity co-occurs with increased functional CP-AMPAR contribution and impaired synaptic plasticity in the NAc. Future studies incorporating sex as a biological variable and experimental strategies enabling selective manipulation of intraneuronal Aβ will be required to determine causal relationships.

In conclusion, this study identifies an early synaptic phenotype in the NAc core of male APP/PS1 mice, characterized by altered AMPAR functional properties, increased functional CP-AMPAR contribution, and selective impairment of LTD in D1R+ MSNs before detectable compact fibrillar plaque deposition. These synaptic alterations co-occurred with reduced evoked dopaminergic signaling and selective changes in reward-related behavior, although causal relationships among these outcomes remain unresolved. Together, these findings identify the NAc as an early site of vulnerability and neuronal subtype as a factor associated with differential synaptic susceptibility in this model. They further provide a framework for investigating the relationships among intraneuronal Aβ-associated pathology, CP-AMPAR signaling, mGluR1/5-dependent plasticity, and early accumbal dysfunction.

Supplementary Material

NIHMS2209780-supplement-1.pdf^{(24.7MB, pdf)}

Acknowledgements

We thank Laurie Aguayo, Ixia Cid and Gerson Ramos for technical assistance. We also acknowledge Carolina Benítez, Jocelyn González, and Claudia Ramírez for veterinary assistance. We thank Lauren Aguayo and Mauricio Avendaño Valenzuela (Universidad de Concepción, Chile) for assistance with language editing.

Funding

This work was supported by ANID Fondecyt Regular grant 1221080 (L.G.A.), NIH grant R01AA025718 (L.G.A.), ANID PhD fellowship 21202521 (N.R.L.), ANID Fondecyt Iniciación 11251074 (L.A.W.), and ANID Fondecyt Iniciación 11250551 (L.S.M.).

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.nbd.2026.107568.

Footnotes

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

CRediT authorship contribution statement

Nicolas Riffo-Lepe: Writing – review & editing, Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Juliana González-Sanmiguel: Writing – review & editing, Investigation, Formal analysis, Conceptualization. Isaías Meza: Writing – review & editing, Investigation, Formal analysis, Conceptualization. Paulina Saavedra-Sieyes: Writing – review & editing, Investigation, Formal analysis, Conceptualization. Lorena Armijo-Weingart: Writing – review & editing, Resources, Investigation, Formal analysis. Helena Zambrano: Writing – review & editing, Validation, Methodology, Investigation, Formal analysis. Ailín Riquelme: Writing – review & editing, Visualization, Methodology, Investigation, Formal analysis. Armando G. Salinas: Writing – review & editing, Validation, Supervision, Conceptualization. Loreto San Martín: Writing – review & editing, Supervision, Resources, Conceptualization. Luis G. Aguayo: Writing – review & editing, Supervision, Resources, Project administration, Funding acquisition, Conceptualization.

Data availability

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

References

Ade KK, et al. , 2011. An improved BAC transgenic fluorescent reporter line for sensitive and specific identification of striatonigral medium spiny neurons. Front. Syst. Neurosci. 5, 32. [DOI] [PMC free article] [PubMed] [Google Scholar]
Aguado C, et al. , 2024. Resilience to structural and molecular changes in excitatory synapses in the hippocampus contributes to cognitive function recovery in Tg2576 mice. Neural Regen. Res. 19, 2068–2074. [DOI] [PMC free article] [PubMed] [Google Scholar]
Alcantara-Gonzalez D, et al. , 2025. Increased excitability of dentate gyrus mossy cells occurs early in life in the Tg2576 model of Alzheimer’s disease. Alzheimer’s Res. Ther. 17, 105. [DOI] [PMC free article] [PubMed] [Google Scholar]
Armijo-Weingart L, et al. , 2024. Loss of glycine receptors in the nucleus accumbens and ethanol reward in an Alzheimer s Disease mouse model. Prog. Neurobiol. 237, 102616. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bayassi-Jakowicka M, et al. , 2022. More than addiction-the nucleus accumbens contribution to development of mental disorders and neurodegenerative diseases. Int. J. Mol. Sci. 23. [DOI] [PMC free article] [PubMed] [Google Scholar]
Beaulieu JM, Gainetdinov RR, 2011. The physiology, signaling, and pharmacology of dopamine receptors. Pharmacol. Rev. 63, 182–217. [DOI] [PubMed] [Google Scholar]
Calabresi P, et al. , 1992. Long-term potentiation in the striatum is unmasked by removing the voltage-dependent magnesium block of NMDA receptor channels. Eur. J. Neurosci. 4, 929–935. [DOI] [PubMed] [Google Scholar]
Carr KD, 2020. Homeostatic regulation of reward via synaptic insertion of calcium-permeable AMPA receptors in nucleus accumbens. Physiol. Behav. 219, 112850. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chang M, et al. , 2025. Intercellular communication in the brain through a dendritic nanotubular network. Science 390, eadr7403. [DOI] [PubMed] [Google Scholar]
Contreras JA, et al. , 2020. Functional connectivity among brain regions affected in Alzheimer’s disease is associated with CSF TNF-alpha in APOE4 carriers. Neurobiol. Aging 86, 112–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
Copenhaver AE, LeGates TA, 2024. Sex-specific mechanisms underlie long-term potentiation at hippocampus–>medium spiny neuron synapses in the medial shell of the nucleus accumbens. J. Neurosci. 44. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cordella A, et al. , 2018. Dopamine loss alters the hippocampus-nucleus accumbens synaptic transmission in the Tg2576 mouse model of Alzheimer’s disease. Neurobiol. Dis. 116, 142–154. [DOI] [PubMed] [Google Scholar]
Cull-Candy SG, Farrant M, 2021. Ca(2+) -permeable AMPA receptors and their auxiliary subunits in synaptic plasticity and disease. J. Physiol. 599, 2655–2671. [DOI] [PMC free article] [PubMed] [Google Scholar]
Fernández-Pérez EJ, et al. , 2020. Changes in neuronal excitability and synaptic transmission in nucleus accumbens in a transgenic Alzheimer’s disease mouse model. Sci. Rep. 10, 19606. [DOI] [PMC free article] [PubMed] [Google Scholar]
Fernandez-Perez EJ, et al. , 2021. Synaptic dysregulation and hyperexcitability induced by intracellular amyloid beta oligomers. Aging Cell 20, e13455. [DOI] [PMC free article] [PubMed] [Google Scholar]
Frank P, et al. , 2025. Specific midlife depressive symptoms and long-term dementia risk: a 23-year UK prospective cohort study. Lancet Psychiatr. 13 (2), 100–111. 10.1016/S2215-0366(25)00331-1. [DOI] [PubMed] [Google Scholar]
Gerlach M, et al. , 2003. Dopamine receptor agonists in current clinical use: comparative dopamine receptor binding profiles defined in the human striatum. J. Neural Transm. (Vienna) 110, 1119–1127. [DOI] [PubMed] [Google Scholar]
Guo C, Ma YY, 2021. Calcium permeable-AMPA receptors and excitotoxicity in neurological disorders. Front. Neural Circuits 15, 711564. [DOI] [PMC free article] [PubMed] [Google Scholar]
Guo C, et al. , 2022. Amyloid-beta oligomers in the nucleus accumbens decrease motivation via insertion of calcium-permeable AMPA receptors. Mol. Psychiatry 27, 2146–2157. [DOI] [PMC free article] [PubMed] [Google Scholar]
Henley JM, Wilkinson KA, 2016. Synaptic AMPA receptor composition in development, plasticity and disease. Nat. Rev. Neurosci. 17, 337–350. 10.1038/nrn.2016.37. [DOI] [PubMed] [Google Scholar]
Hwang E-K, et al. , 2025. Retinoic acid-mediated homeostatic plasticity drives cell type-specific CP-AMPAR accumulation in nucleus accumbens core and incubation of cocaine craving. Mol. Psychiatry 30, 3175–3187. [DOI] [PMC free article] [PubMed] [Google Scholar]
Italia M, et al. , 2025. What’s in your plate? AMPA receptors modulation at the crossroads of homeostatic and hedonic eating. Pharmacol. Res. 217, 107821. [DOI] [PubMed] [Google Scholar]
Iulita MF, et al. , 2014. Intracellular Abeta pathology and early cognitive impairments in a transgenic rat overexpressing human amyloid precursor protein: a multidimensional study. Acta Neuropathol. Commun. 2, 61. [DOI] [PMC free article] [PubMed] [Google Scholar]
Jankowsky JL, et al. , 2001. Co-expression of multiple transgenes in mouse CNS: a comparison of strategies. Biomol. Eng. 17, 157–165. [DOI] [PubMed] [Google Scholar]
Kauer JA, Malenka RC, 2007. Synaptic plasticity and addiction. Nat. Rev. Neurosci. 8, 844–858. [DOI] [PubMed] [Google Scholar]
Kawa AB, et al. , 2022. Positive allosteric modulation of mGlu(1) reverses cocaine-induced behavioral and synaptic plasticity through the integrated stress response and oligophrenin-1. Biol. Psychiatry 92, 871–879. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kelly L, et al. , 2021. Identification of intraneuronal amyloid beta oligomers in locus coeruleus neurons of Alzheimer’s patients and their potential impact on inhibitory neurotransmitter receptors and neuronal excitability. Neuropathol. Appl. Neurobiol. 47, 488–505. [DOI] [PubMed] [Google Scholar]
Krashia P, et al. , 2019. Unifying hypothesis of dopamine neuron loss in neurodegenerative diseases: focusing on Alzheimer’s disease. Front. Mol. Neurosci. 12, 123. [DOI] [PMC free article] [PubMed] [Google Scholar]
LaFerla FM, et al. , 2007. Intracellular amyloid-beta in Alzheimer’s disease. Nat. Rev. Neurosci. 8, 499–509. [DOI] [PubMed] [Google Scholar]
Li G, et al. , 2025. Revealing excitation-inhibition imbalance in Alzheimer’s disease using multiscale neural model inversion of resting-state functional MRI. Commun. Med. (Lond.) 5, 17. [DOI] [PMC free article] [PubMed] [Google Scholar]
Luscher C, Huber KM, 2010. Group 1 mGluR-dependent synaptic long-term depression: mechanisms and implications for circuitry and disease. Neuron 65, 445–459. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mango D, Ledonne A, 2023. Updates on the physiopathology of group I metabotropic glutamate receptors (mGluRI)-dependent long-term depression. Cells 12. [DOI] [PMC free article] [PubMed] [Google Scholar]
Masters M. Roe, 2015. “Noncognitive” symptoms of early Alzheimer disease: a longitudinal analysis. Neurology 84, 617–622. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mathur BN, et al. , 2011. Serotonin induces long-term depression at corticostriatal synapses. J. Neurosci. 31, 7402–7411. [DOI] [PMC free article] [PubMed] [Google Scholar]
McCutcheon JE, et al. , 2011. Group I mGluR activation reverses cocaine-induced accumulation of calcium-permeable AMPA receptors in nucleus accumbens synapses via a protein kinase C-dependent mechanism. J. Neurosci. 31, 14536–14541. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nie X, et al. , 2017. Subregional structural alterations in hippocampus and nucleus accumbens correlate with the clinical impairment in patients with Alzheimer’s disease clinical spectrum: parallel combining volume and vertex-based approach. Front. Neurol. 8, 399. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nobili A, et al. , 2017. Dopamine neuronal loss contributes to memory and reward dysfunction in a model of Alzheimer’s disease. Nat. Commun. 8, 14727. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pennartz CM, et al. , 1992. Differential membrane properties and dopamine effects in the shell and core of the rat nucleus accumbens studied in vitro. Neurosci. Lett. 136, 109–112. [DOI] [PubMed] [Google Scholar]
Peterson BM, et al. , 2016. Estradiol impacts the endocannabinoid system in female rats to influence behavioral and structural responses to cocaine. Neuropharmacology 110, 118–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pilotto A, et al. , 2025. Dopaminergic deficits along the spectrum of Alzheimer’s disease. Mol. Psychiatry 30, 3069–3076. 10.1038/s41380-025-02913-5. [DOI] [PubMed] [Google Scholar]
Richfield EK, et al. , 1989. Anatomical and affinity state comparisons between dopamine D1 and D2 receptors in the rat central nervous system. Neuroscience 30, 767–777. [DOI] [PubMed] [Google Scholar]
Russo SJ, Nestler EJ, 2013. The brain reward circuitry in mood disorders. Nat. Rev. Neurosci. 14, 609–625. [DOI] [PMC free article] [PubMed] [Google Scholar]
Saavedra-Sieyes P, et al. , 2025. Alterations in neuronal excitability and AMPA-mediated neurotransmission in an APP/PS1 neuronal culture model derived from nucleus accumbens. J. Alzheimer’s Dis. 108 (2), 531–538. 10.1177/13872877251378468. [DOI] [PubMed] [Google Scholar]
Salinas AG, et al. , 2023. Distinct sub-second dopamine signaling in dorsolateral striatum measured by a genetically-encoded fluorescent sensor. Nat. Commun. 14, 5915. [DOI] [PMC free article] [PubMed] [Google Scholar]
San Martin L, et al. , 2020. Ethanol consumption and sedation are altered in mice lacking the glycine receptor alpha2 subunit. Br. J. Pharmacol. 177, 3941–3956. [DOI] [PMC free article] [PubMed] [Google Scholar]
Scaduto P, et al. , 2023. Functional excitatory to inhibitory synaptic imbalance and loss of cognitive performance in people with Alzheimer's disease neuropathologic change. Acta Neuropathol. 145, 303–324. [DOI] [PMC free article] [PubMed] [Google Scholar]
Selkoe DJ, Hardy J, 2016. The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol. Med. 8, 595–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
Shah J, et al. , 2025. Predicting cognitive decline from neuropsychiatric symptoms and Alzheimer’s disease biomarkers: a machine learning approach to a population-based data. J. Alzheimer’s Dis 103, 833–843. [DOI] [PubMed] [Google Scholar]
Storga D, et al. , 1996. Monoaminergic neurotransmitters, their precursors and metabolites in brains of Alzheimer patients. Neurosci. Lett. 203, 29–32. [DOI] [PubMed] [Google Scholar]
Thomas MJ, et al. , 2001. Long-term depression in the nucleus accumbens: a neural correlate of behavioral sensitization to cocaine. Nat. Neurosci. 4, 1217–1223. [DOI] [PubMed] [Google Scholar]
Vieitas-Gaspar N, et al. , 2025. From valence encoding to motivated behavior: a focus on the nucleus accumbens circuitry. Neurosci. Biobehav. Rev. 172, 106125. [DOI] [PubMed] [Google Scholar]
Vossel K, et al. , 2026. Epileptic activity in Alzheimer’s disease: emerging insights and therapeutic implications. Lancet Neurol. 25 (3), 294–307. 10.1016/S1474-4422(25)00425-9. [DOI] [PubMed] [Google Scholar]
Whitcomb DJ, et al. , 2015. Intracellular oligomeric amyloid-beta rapidly regulates GluA1 subunit of AMPA receptor in the hippocampus. Sci. Rep. 5, 10934. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wolf ME, 2016. Synaptic mechanisms underlying persistent cocaine craving. Nat. Rev. Neurosci. 17, 351–365. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wu XR, et al. , 2024. Amygdala neuronal dyshomeostasis via 5-HT receptors mediates mood and cognitive defects in Alzheimer’s disease. Aging Cell 23, e14187. [DOI] [PMC free article] [PubMed] [Google Scholar]
Youmans KL, et al. , 2012. Intraneuronal Abeta detection in 5xFAD mice by a new Abeta-specific antibody. Mol. Neurodegener. 7, 8. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS2209780-supplement-1.pdf^{(24.7MB, pdf)}

Data Availability Statement

The data supporting the findings of this study are available from the corresponding author upon reasonable request.

[R1] Ade KK, et al. , 2011. An improved BAC transgenic fluorescent reporter line for sensitive and specific identification of striatonigral medium spiny neurons. Front. Syst. Neurosci. 5, 32. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R2] Aguado C, et al. , 2024. Resilience to structural and molecular changes in excitatory synapses in the hippocampus contributes to cognitive function recovery in Tg2576 mice. Neural Regen. Res. 19, 2068–2074. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R3] Alcantara-Gonzalez D, et al. , 2025. Increased excitability of dentate gyrus mossy cells occurs early in life in the Tg2576 model of Alzheimer’s disease. Alzheimer’s Res. Ther. 17, 105. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] Armijo-Weingart L, et al. , 2024. Loss of glycine receptors in the nucleus accumbens and ethanol reward in an Alzheimer s Disease mouse model. Prog. Neurobiol. 237, 102616. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] Bayassi-Jakowicka M, et al. , 2022. More than addiction-the nucleus accumbens contribution to development of mental disorders and neurodegenerative diseases. Int. J. Mol. Sci. 23. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] Beaulieu JM, Gainetdinov RR, 2011. The physiology, signaling, and pharmacology of dopamine receptors. Pharmacol. Rev. 63, 182–217. [DOI] [PubMed] [Google Scholar]

[R7] Calabresi P, et al. , 1992. Long-term potentiation in the striatum is unmasked by removing the voltage-dependent magnesium block of NMDA receptor channels. Eur. J. Neurosci. 4, 929–935. [DOI] [PubMed] [Google Scholar]

[R8] Carr KD, 2020. Homeostatic regulation of reward via synaptic insertion of calcium-permeable AMPA receptors in nucleus accumbens. Physiol. Behav. 219, 112850. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] Chang M, et al. , 2025. Intercellular communication in the brain through a dendritic nanotubular network. Science 390, eadr7403. [DOI] [PubMed] [Google Scholar]

[R10] Contreras JA, et al. , 2020. Functional connectivity among brain regions affected in Alzheimer’s disease is associated with CSF TNF-alpha in APOE4 carriers. Neurobiol. Aging 86, 112–122. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R11] Copenhaver AE, LeGates TA, 2024. Sex-specific mechanisms underlie long-term potentiation at hippocampus–>medium spiny neuron synapses in the medial shell of the nucleus accumbens. J. Neurosci. 44. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R12] Cordella A, et al. , 2018. Dopamine loss alters the hippocampus-nucleus accumbens synaptic transmission in the Tg2576 mouse model of Alzheimer’s disease. Neurobiol. Dis. 116, 142–154. [DOI] [PubMed] [Google Scholar]

[R13] Cull-Candy SG, Farrant M, 2021. Ca(2+) -permeable AMPA receptors and their auxiliary subunits in synaptic plasticity and disease. J. Physiol. 599, 2655–2671. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] Fernández-Pérez EJ, et al. , 2020. Changes in neuronal excitability and synaptic transmission in nucleus accumbens in a transgenic Alzheimer’s disease mouse model. Sci. Rep. 10, 19606. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] Fernandez-Perez EJ, et al. , 2021. Synaptic dysregulation and hyperexcitability induced by intracellular amyloid beta oligomers. Aging Cell 20, e13455. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] Frank P, et al. , 2025. Specific midlife depressive symptoms and long-term dementia risk: a 23-year UK prospective cohort study. Lancet Psychiatr. 13 (2), 100–111. 10.1016/S2215-0366(25)00331-1. [DOI] [PubMed] [Google Scholar]

[R17] Gerlach M, et al. , 2003. Dopamine receptor agonists in current clinical use: comparative dopamine receptor binding profiles defined in the human striatum. J. Neural Transm. (Vienna) 110, 1119–1127. [DOI] [PubMed] [Google Scholar]

[R18] Guo C, Ma YY, 2021. Calcium permeable-AMPA receptors and excitotoxicity in neurological disorders. Front. Neural Circuits 15, 711564. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] Guo C, et al. , 2022. Amyloid-beta oligomers in the nucleus accumbens decrease motivation via insertion of calcium-permeable AMPA receptors. Mol. Psychiatry 27, 2146–2157. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] Henley JM, Wilkinson KA, 2016. Synaptic AMPA receptor composition in development, plasticity and disease. Nat. Rev. Neurosci. 17, 337–350. 10.1038/nrn.2016.37. [DOI] [PubMed] [Google Scholar]

[R21] Hwang E-K, et al. , 2025. Retinoic acid-mediated homeostatic plasticity drives cell type-specific CP-AMPAR accumulation in nucleus accumbens core and incubation of cocaine craving. Mol. Psychiatry 30, 3175–3187. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R22] Italia M, et al. , 2025. What’s in your plate? AMPA receptors modulation at the crossroads of homeostatic and hedonic eating. Pharmacol. Res. 217, 107821. [DOI] [PubMed] [Google Scholar]

[R23] Iulita MF, et al. , 2014. Intracellular Abeta pathology and early cognitive impairments in a transgenic rat overexpressing human amyloid precursor protein: a multidimensional study. Acta Neuropathol. Commun. 2, 61. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] Jankowsky JL, et al. , 2001. Co-expression of multiple transgenes in mouse CNS: a comparison of strategies. Biomol. Eng. 17, 157–165. [DOI] [PubMed] [Google Scholar]

[R25] Kauer JA, Malenka RC, 2007. Synaptic plasticity and addiction. Nat. Rev. Neurosci. 8, 844–858. [DOI] [PubMed] [Google Scholar]

[R26] Kawa AB, et al. , 2022. Positive allosteric modulation of mGlu(1) reverses cocaine-induced behavioral and synaptic plasticity through the integrated stress response and oligophrenin-1. Biol. Psychiatry 92, 871–879. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] Kelly L, et al. , 2021. Identification of intraneuronal amyloid beta oligomers in locus coeruleus neurons of Alzheimer’s patients and their potential impact on inhibitory neurotransmitter receptors and neuronal excitability. Neuropathol. Appl. Neurobiol. 47, 488–505. [DOI] [PubMed] [Google Scholar]

[R28] Krashia P, et al. , 2019. Unifying hypothesis of dopamine neuron loss in neurodegenerative diseases: focusing on Alzheimer’s disease. Front. Mol. Neurosci. 12, 123. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] LaFerla FM, et al. , 2007. Intracellular amyloid-beta in Alzheimer’s disease. Nat. Rev. Neurosci. 8, 499–509. [DOI] [PubMed] [Google Scholar]

[R30] Li G, et al. , 2025. Revealing excitation-inhibition imbalance in Alzheimer’s disease using multiscale neural model inversion of resting-state functional MRI. Commun. Med. (Lond.) 5, 17. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R31] Luscher C, Huber KM, 2010. Group 1 mGluR-dependent synaptic long-term depression: mechanisms and implications for circuitry and disease. Neuron 65, 445–459. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] Mango D, Ledonne A, 2023. Updates on the physiopathology of group I metabotropic glutamate receptors (mGluRI)-dependent long-term depression. Cells 12. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] Masters M. Roe, 2015. “Noncognitive” symptoms of early Alzheimer disease: a longitudinal analysis. Neurology 84, 617–622. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] Mathur BN, et al. , 2011. Serotonin induces long-term depression at corticostriatal synapses. J. Neurosci. 31, 7402–7411. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] McCutcheon JE, et al. , 2011. Group I mGluR activation reverses cocaine-induced accumulation of calcium-permeable AMPA receptors in nucleus accumbens synapses via a protein kinase C-dependent mechanism. J. Neurosci. 31, 14536–14541. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] Nie X, et al. , 2017. Subregional structural alterations in hippocampus and nucleus accumbens correlate with the clinical impairment in patients with Alzheimer’s disease clinical spectrum: parallel combining volume and vertex-based approach. Front. Neurol. 8, 399. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] Nobili A, et al. , 2017. Dopamine neuronal loss contributes to memory and reward dysfunction in a model of Alzheimer’s disease. Nat. Commun. 8, 14727. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] Pennartz CM, et al. , 1992. Differential membrane properties and dopamine effects in the shell and core of the rat nucleus accumbens studied in vitro. Neurosci. Lett. 136, 109–112. [DOI] [PubMed] [Google Scholar]

[R39] Peterson BM, et al. , 2016. Estradiol impacts the endocannabinoid system in female rats to influence behavioral and structural responses to cocaine. Neuropharmacology 110, 118–124. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R40] Pilotto A, et al. , 2025. Dopaminergic deficits along the spectrum of Alzheimer’s disease. Mol. Psychiatry 30, 3069–3076. 10.1038/s41380-025-02913-5. [DOI] [PubMed] [Google Scholar]

[R41] Richfield EK, et al. , 1989. Anatomical and affinity state comparisons between dopamine D1 and D2 receptors in the rat central nervous system. Neuroscience 30, 767–777. [DOI] [PubMed] [Google Scholar]

[R42] Russo SJ, Nestler EJ, 2013. The brain reward circuitry in mood disorders. Nat. Rev. Neurosci. 14, 609–625. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] Saavedra-Sieyes P, et al. , 2025. Alterations in neuronal excitability and AMPA-mediated neurotransmission in an APP/PS1 neuronal culture model derived from nucleus accumbens. J. Alzheimer’s Dis. 108 (2), 531–538. 10.1177/13872877251378468. [DOI] [PubMed] [Google Scholar]

[R44] Salinas AG, et al. , 2023. Distinct sub-second dopamine signaling in dorsolateral striatum measured by a genetically-encoded fluorescent sensor. Nat. Commun. 14, 5915. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] San Martin L, et al. , 2020. Ethanol consumption and sedation are altered in mice lacking the glycine receptor alpha2 subunit. Br. J. Pharmacol. 177, 3941–3956. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R46] Scaduto P, et al. , 2023. Functional excitatory to inhibitory synaptic imbalance and loss of cognitive performance in people with Alzheimer's disease neuropathologic change. Acta Neuropathol. 145, 303–324. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] Selkoe DJ, Hardy J, 2016. The amyloid hypothesis of Alzheimer’s disease at 25 years. EMBO Mol. Med. 8, 595–608. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] Shah J, et al. , 2025. Predicting cognitive decline from neuropsychiatric symptoms and Alzheimer’s disease biomarkers: a machine learning approach to a population-based data. J. Alzheimer’s Dis 103, 833–843. [DOI] [PubMed] [Google Scholar]

[R49] Storga D, et al. , 1996. Monoaminergic neurotransmitters, their precursors and metabolites in brains of Alzheimer patients. Neurosci. Lett. 203, 29–32. [DOI] [PubMed] [Google Scholar]

[R50] Thomas MJ, et al. , 2001. Long-term depression in the nucleus accumbens: a neural correlate of behavioral sensitization to cocaine. Nat. Neurosci. 4, 1217–1223. [DOI] [PubMed] [Google Scholar]

[R51] Vieitas-Gaspar N, et al. , 2025. From valence encoding to motivated behavior: a focus on the nucleus accumbens circuitry. Neurosci. Biobehav. Rev. 172, 106125. [DOI] [PubMed] [Google Scholar]

[R52] Vossel K, et al. , 2026. Epileptic activity in Alzheimer’s disease: emerging insights and therapeutic implications. Lancet Neurol. 25 (3), 294–307. 10.1016/S1474-4422(25)00425-9. [DOI] [PubMed] [Google Scholar]

[R53] Whitcomb DJ, et al. , 2015. Intracellular oligomeric amyloid-beta rapidly regulates GluA1 subunit of AMPA receptor in the hippocampus. Sci. Rep. 5, 10934. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R54] Wolf ME, 2016. Synaptic mechanisms underlying persistent cocaine craving. Nat. Rev. Neurosci. 17, 351–365. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] Wu XR, et al. , 2024. Amygdala neuronal dyshomeostasis via 5-HT receptors mediates mood and cognitive defects in Alzheimer’s disease. Aging Cell 23, e14187. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R56] Youmans KL, et al. , 2012. Intraneuronal Abeta detection in 5xFAD mice by a new Abeta-specific antibody. Mol. Neurodegener. 7, 8. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Selective impairment of long-term depression in accumbal D1R+ MSNs involves calcium-permeable AMPARs in early Alzheimer's disease

Nicolas Riffo-Lepe

Juliana González-Sanmiguel

Isaías Meza

Paulina Saavedra-Sieyes

Lorena Armijo-Weingart

Helena Zambrano

Ailín Riquelme

Armando G Salinas

Loreto San Martín

Luis G Aguayo

Abstract

1. Introduction

2. Materials and methods

2.1. Animals

Fig. 1.

2.2. Immunohistochemistry

2.3. Thioflavin-S staining

2.4. Electrophysiological recordings in coronal brain slices

2.5. Spontaneous synaptic currents

2.6. Synaptic stimulation and evoked responses

2.7. AMPA/NMDA ratio

2.8. Paired-pulse ratio (PPR)

2.9. Rectification index (RI)

2.10. NASPM sensitivity

Fig. 4.

2.11. Long-term depression

2.12. Stereotaxic injections

2.13. Calcium and dopamine photometry

2.14. Western blot

2.15. qRT-PCR

2.16. Palatable food-conditioned place preference

2.17. Elevated plus maze (EPM)

2.18. Social behavior experiments

2.19. Statistical analysis

3. Results

3.1. Early intraneuronal Aβ accumulation precedes detectable compact fibrillar plaque deposition and is associated with selective enhancement of excitatory synaptic transmission in D1R+ MSNs

3.2. Selective impairment of long-term depression in the nucleus accumbens emerges at pre-plaque stages and selectively affects D1R+ MSNs

Fig. 2.

3.3. Alterations in AMPAR-mediated synaptic transmission in the nucleus accumbens are not explained by transcriptional changes

Fig. 3.

3.4. Increased AMPAR-dependent calcium signaling and functional contribution of calcium-permeable AMPARs in the nucleus accumbens of APP/PS1 mice

3.5. Impaired mGluR1/5-dependent LTD in APP/PS1 D1R+ MSNs is accompanied by persistent CP-AMPAR contribution

Fig. 5.

3.6. Reduced evoked dopamine signaling co-occurs with selective alterations in reward-related behavior in pre-plaque APP/PS1 mice

Fig. 6.

4. Discussion

Fig. 7.

Supplementary Material

Acknowledgements

Funding

Appendix A. Supplementary data

Footnotes

Data availability

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases