Meta-Analysis in Transgenic Alzheimer’s Disease Mouse Models Reveals Opposite Brain Network Effects of Amyloid-β and Phosphorylated Tau Proteins

Abstract

Background:

Cognitive deficits observed in Alzheimer’s disease (AD) patients have been correlated with altered hippocampal activity. Although the mechanism remains under extensive study, neurofibrillary tangles and amyloid plaques have been proposed as responsible for brain activity alterations. Aiming to unveil the mechanism, researchers have developed several transgenic models of AD. Nevertheless, the variability in hippocampal oscillatory alterations found in different genetic backgrounds and ages remains unclear.

Objective:

To assess the oscillatory alterations in relation to animal developmental age and protein inclusion, amyloid-β (Aβ) load, and abnormally phosphorylated tau (pTau), we reviewed and analyzed the published data on peak power, frequency, and quantification of theta-gamma cross-frequency coupling (modulation index values).

Methods:

To ensure that the search was as current as possible, a systematic review was conducted to locate and abstract all studies published from January 2000 to February 2023 that involved in vivo hippocampal local field potential recording in transgenic mouse models of AD.

Results:

The presence of Aβ was associated with electrophysiological alterations that are mainly reflected in power increases, frequency decreases, and lower modulation index values. Concomitantly, pTau accumulation was associated with electrophysiological alterations that are mainly reflected in power decreases, frequency decreases, and no significant alterations in modulation index values.

Conclusions:

In this study, we showed that electrophysiological parameters are altered from prodromal stages to the late stages of pathology. Thus, we found that Aβ deposition is associated with brain network hyperexcitability, whereas pTau deposition mainly leads to brain network hypoexcitability in transgenic models

Keywords

Alzheimer’s disease amyloid-β cross-frequency coupling hippocampus network hyperexcitability network hypoexcitability oscillatory activity phospho-tau

INTRODUCTION

Alzheimer’s disease (AD) is related to cognitive decline and memory impairments, where aging is the main risk [1]. Although the etiology remains under extensive study, abnormal neural protein aggregation in several brain areas is proposed as the main cause of disease [2]. Additionally, protein aggregation strongly correlates with the severity of dementia [2]. In AD patients, the brain is affected by extracellular amyloid-β (Aβ) plaques and intracellular neurofibrillary tangles (NFTs), mainly composed of hyperphosphorylated tau protein (pTau) [2, 3]. The leading hypothesis for AD onset suggests that an increase in Aβ triggers the hyperphosphorylation of tau protein, resulting in synaptic dysfunction, dendritic spine loss, early network dysfunction, and neuronal death, with the hippocampus being one of the most affected brain regions [4, 5]. However, brain network impairments rather than cell loss are now suggested as the cause of cognitive decline [1 , 6]. Brain network activity is regulated by multiple neural oscillatory rhythms because of interactions between different brain regions [7]. It is proposed that oscillatory rhythms are fundamental for the temporal coding and decoding of active neural ensembles [8]. The largest synchronous signal is the hippocampal theta rhythm [9]. In humans, this activity is measured by the non-invasive technique of the electroencephalogram (EEG). Unsurprisingly, AD patients show a “slowing” of oscillatory brain activity in parameters like spectral power and functional connectivity associated with the severity of dementia and the progression of the pathology [10 –12]. The main changes are seen in lower rhythms like theta oscillation [12 –14]. Importantly, hippocampal theta oscillations are thought to drive cognitive processes by coding with gamma oscillations (faster oscillatory rhythms), a process called cross-frequency coupling (CFC) [7 , 15–19]. CFC has been described in mice, rats, monkeys, and humans in several brain regions like the hippocampus [15] and cortex [20, 21]. A strong CFC has been related to high cognitive performance such as learning, memory [15 , 22] and spatial resolution [16]. Accordingly, AD patients show a weak cortical CFC when performing memory tasks [23, 24].

Aiming to study brain network alterations, researchers have developed several transgenic mouse models characterized by Aβ and pTau deposition [25, 26]. While some studies have found evidence of brain network alterations, others have failed to replicate these findings. Moreover, the relationship between network alterations, cognition, and protein deposition is unclear. Although some review papers have provided an overview of AD transgenic (AD-tg) mouse models assessing protein deposition and changes in cognitive performance [27], they lack in-depth assessments of network alterations and overlook the methodological heterogeneity of the studies.

This meta-analysis review aims to determine the state of brain network alterations in AD-tg models during the progression of disease. In recent years, our laboratory has focused on evaluating network alterations in pre-symptomatic stages of AD-tg models [3 , 28–30]. Considering the prodromal AD stage, we conducted a meta-analysis review to locate and abstract data from all studies of local field potential (LFP) analysis in AD-tg models that provided frequency, peak power, and modulation index analysis. In this report, we focus on two types of AD mouse models: A) amyloid-only and B) tau-only (Table 1). Of the 84 studies identified, 15 proved useful. This approach has relevance in the field of AD diagnosis since symptoms can vary widely, but physiological parameters would allow for a better approach to disease progression.

Table 1

AD transgenic models included in this meta-analysis

Transgenic model	Mutations	Neuropathology	Cognitive impairment	Reference
APP/PS1	APPSwe(K670N/M671L); PSEN1 L166P	Hippocampal Aβ depositions by 3–4 months	7 months	Alzforum
APP/PS2	APPSwe(K670N/M671L); PSEN2 N141L	Hippocampal Aβ depositions by 6 months	8 months	Alzforum
TgF344	APPSwe(K670N/M671L); PSEN1:deltaE9	Hippocampal Aβ depositions by 6 months	6 months	Alzforum
Tg2576	APPSwe(K670N/M671L)	Aβ depositions by 11–13 months	6 months	Alzforum
J20	APPSwe(K670N/M671L); APP V717F	Hippocampal Aβ depositions by 5–7 months	4 months	Alzforum
5xFAD	APPSwe(K670N/M671L); APP I716 V; APP V717I; PSEN M146L (A > C); PSEN1 L286V	Hippocampal Aβ depositions by 2 months	3–6 months	Alzforum
P301S	MAPT P301S	Neurofibrillary tangles by 4 months	2.5 months	Alzforum
VLW	G272 V; P301L; R406W	Tau deposition in hippocampal pyramidal neurons 8 months	8 months	[34]
Tg4510	MAPT P301L	Pretangles by 2.5 months. Hippocampal tangles by 5.5 months	2.5–4 months	Alzforum

METHODS

This study aimed to identify and summarize published literature that shows alterations in electrophysiological parameters in AD-tg models (Table 1). Our meta-analysis was conducted following with the preferred reporting items for systematic reviews and meta-analyses (PRISMA). To identify studies for inclusion in this meta-analysis, a PubMed search was conducted to locate all studies published in English from January 2000 to February 2023 using the terms [(amyloid OR Aβ) AND transgenic AND (local field potential OR oscillatory activity)] AND [(tau OR tauopathy) AND transgenic AND (local field potential OR oscillatory activity)]. A total of 84 articles were identified, 15 of which met the inclusion criteria for meta-analysis. Studies identified through database searches were initially screened based on their titles and abstracts. Each article was independently evaluated for full-text eligibility. Studies were included if they simultaneously fulfilled the following criteria: a) the study reported original results from analysis of AD-tg models; b) LFP was recorded in vivo on anesthetized or awake mice; c) basal activity was reported; and d) the report included peak power band, peak frequency, or modulation index for CFC. For this meta-analysis review, we gathered mean values of peak band power, frequency, and modulation index from reported LFP analyses, along with standard deviations, sample size, age, and transgenic model. In accordance with Cochran’s, at least two studies per parameter were required for the meta-analysis. Some of the papers provided insufficient information to conduct a proper meta-analysis, so they were only mentioned.

After subtracting all the data, the information was normalized on a scale of 1 to facilitate comparison. The data was divided into two groups, as follows: Group A: only amyloid mutation (9 studies) [31 –39] and Group B: only tau mutation (6 studies) [28 , 40–42]. Frequency bands were classified as 1.5–4 Hz delta, 4–10 Hz theta, 10–30 Hz beta, 30–80 Hz low-gamma, and 80–300 Hz high-gamma [43]. All analyses were based on published studies; therefore, no ethical approval is required. Two authors independently performed the data extraction, and any disagreements were resolved by joint discussions. Finally, data were arranged according to age to enhance comprehension of electrophysiological properties during disease progression.

Statistical methods

All data were analyzed using the software Review Manager (RevMan) [computer program] Version 5.4. The Cochrane Collaboration, 2020. The meta-analysis results were reported as a standard difference in mean (SDM) with 95% confidence intervals (CI) as the synthesized measure of effect size. Cochran’s Q-test statistic was performed to assess the heterogeneity of the studies, and the I-squared (I²) statistic was performed to indicate heterogeneity as a percentage.

RESULTS

A total of 84 scientific documents were screened, but only 64 resulted in potentially relevant articles. Fifteen were included based on selection criteria. Articles were excluded when mean values, standard error, and/or standard deviation were not reported or could not be obtained. All studies included electrophysiological assessments in a group of AD-tg models with a healthy age-matched control group. Considering that protein accumulation in the brain has been related to network alterations, we aimed to collect data from studies that compared the electrophysiological properties of AD-tg models and age-matched controls at different developmental ages. Two of the screened documents were reviews, one article was unavailable, eleven did not use AD-tg models, twelve did not evaluate LFP, six were not registered in-vivo, twelve did not record the hippocampus, and five did not report parameters of interest. A total of 15 articles contributed data to the final analysis. Importantly, all studies include values at various ages, or pathology stages (Table 1). All results are explained in terms of heterogeneity.

Oscillatory activity is altered in mice that only have the Aβ mutation (group A)

In the pooled analysis, Aβ load caused an increase in beta power (0.77, 95% CI 0.67 to 0.87, Fig. 1a) and an increase in low-gamma power levels (0.44, 95% CI 0.38 to 0.51, Fig. 1b) compared to the age-matched control group. Interestingly, Aβ accumulation did not lead to a significant change in high-gamma power (0.10, 95% CI –0.01 to 0.21, Fig. 1c). Although Kumari et al. [32] and Dávila-Bouziguet [34] reported alterations in theta power, there was no significant difference between Aβ-mouse models and the age-matched control group. Regarding frequency, the theta peak was reduced in Aβ-transgenic mice compared to age-matched non-transgenic mice (–0.12, 95% CI –0.13 to –0.11, Fig. 2). Importantly, there was insufficient information to make a proper meta-analysis of high-gamma peak frequency between Aβ-transgenics and age-matched controls.

Fig. 1

Standard mean difference and pooled estimate of each study included in the meta-analysis of beta (a), low-gamma (b) and high-gamma (c) peak power. All the analyses compared Aβ transgenic mouse models with age-matched related non-transgenic mice. Summary includes p = significance level; I²= percentage of heterogeneity; Q = Cochran’s Q. The horizontal lines represent the 95% confidence interval for each mean difference. Weights are from fixed effect analysis.

Fig. 2

Standard mean difference and pooled estimate of each study included in the meta-analysis of theta frequency. All the analyses compared Aβ transgenic mouse models with age-matched related non-transgenic mice. Summary includes p = significance level; I²= percentage of heterogeneity; Q = Cochran’s Q. The horizontal lines represent the 95% confidence interval for each mean difference. Weights are from fixed effect analysis.

Regarding the modulation index for theta/low-gamma and theta/high-gamma coupling in Aβ-transgenic mouse models, Aβ accumulation did not change the hippocampal modulation index values for theta/low-gamma coupling (–0.01, 95% CI –0.06 to 0.03, Fig. 3a), but it lowered those for theta/high-gamma coupling (–0.21, 95% CI –0.26 to –0.17, Fig. 3b).

Fig. 3

Standard mean difference and pooled estimate of each study included in the meta-analysis of hippocampal theta/low-gamma (a) and theta/high-gamma (b) cross frequency coupling (modulation index). All the analyses compared Aβ transgenic mouse models with age-matched related non-transgenic mice. Summary includes p = significance level; I²= percentage of heterogeneity; Q = Cochran’s Q. The horizontal lines represent the 95% confidence interval for each mean difference. Weights are from fixed effect analysis.

In summary, mutations that lead to Aβ accumulation generate electrophysiological alterations that are reflected in: a) power increases in low- and high-frequency bands; b) frequency decreases in low-frequency bands; and c) lower modulation index values in theta/high-gamma coupling (Figs. 1–3 present a summary of hippocampal electrophysiological parameters for group A).

Oscillatory activity is altered in mice that only have the pTau mutation (group B)

In the pooled analysis, pTau presence caused a decrease in delta peak power (–0.12, 95% CI –0.24 to –0.00, Fig. 4a) and theta peak power (–0.11, 95% CI –0.14 to –0.09, Fig. 4b) compared to the age-matched control group. Concomitantly, high-gamma power was increased in pTau-transgenic mouse models (0.28, 95% CI 0.17 to 0.39, Fig. 4c). Accordingly, we recently found an increase in low-gamma peak power in one-month-old rTg4510 mice [28]. When evaluating peak frequency values, pTau-transgenic mice showed a reduction in theta (–0.14, 95% CI –0.18 to –0.10, Fig. 5a) and high-gamma peak frequency (–0.09, 95% CI –0.11 to –0.07, Fig. 5b).

Fig. 4

Standard mean difference and pooled estimate of each study included in the meta-analysis of hippocampal delta (a), theta (b) and high-gamma (c) peak power. All the analyses compared Tau transgenic mouse models with age-matched related non-transgenic mouse. Summary includes p = significance level; I²= percentage of heterogeneity; Q = Cochran’s Q. The horizontal lines represent the 95% confidence interval for each mean difference. Weights are from fixed effect analysis.

Fig. 5

Standard mean difference and pooled estimate of each study included in the meta-analysis of hippocampal theta (a) and high-gamma (b) peak frequency. All the analyses compared Tau transgenic mouse models with age-matched related non-transgenic mice. Summary includes p = significance level; I²= percentage of heterogeneity; Q = Cochran’s Q. The horizontal lines represent the 95% confidence interval for each mean difference. Weights are from fixed effect analysis.

Regarding the modulation index in pTau-transgenic mouse models, pTau accumulation did not alter the hippocampal modulation index for theta/gamma coupling (0.06, 95% CI –0.05 to 0.17, Fig. 6).

Fig. 6

Standard mean difference and pooled estimate of each study included in the meta-analysis of hippocampal theta/low-gamma cross-frequency coupling (modulation index). All the analyses compared Tau transgenic mouse models with age-matched related non-transgenic mice. Summary includes p = significance level; I²= percentage of heterogeneity; Q = Cochran’s Q. The horizontal lines represent the 95% confidence interval for each mean difference. Weights are from fixed effect analysis.

In summary, mutations that lead to pTau accumulation generate electrophysiological alterations that are reflected in: a) a power and frequency decrease in low-frequency bands; b) an increase in gamma peak power; c) a significant decrease in gamma peak frequency; and d) no alterations in the modulation index for theta/gamma coupling (Figs. 4 to 6 present a summary of hippocampal electrophysiological parameters for group B).

DISCUSSION

Aβ impact on brain network activity

The hippocampus, essential for learning and memory, is one of the most affected structures by Aβ and pTau protein accumulation during AD development [3]. In pathological conditions, molecular alterations such as protein accumulation affect neural activities, resulting in circuit disruption and, ultimately, neurological function failure [1, 4]. Transgenic models have made it possible for researchers to learn more about the underlying mechanisms of AD and to identify biomarkers that can be used for early detection and diagnosis [25, 26]. Since electrophysiological parameters have been suggested as a potential biomarker [4, 28], in this study we analyzed the effect of mutations that lead to protein accumulation in the hippocampus of different AD-tg mouse models (Table 1). Specifically, we focused on brain network alterations caused by Aβ (group A) and pTau (group B) deposition through disease development (Table 1). To our knowledge, this is the first meta-analysis to collect spontaneous hippocampal oscillatory activity in AD transgenic mouse models. First, we aimed to identify power changes induced by Aβ accumulation from prodromal to advanced stages of the disease. Despite the limited number of studies with similar experimental designs, we found that Aβ accumulation caused hippocampal power increases in low- and high-frequency bands (Fig. 1). Interestingly, the power increases were evident from the prodromal stages of disease and persisted through the advanced stages (Fig. 1). Altogether, the data suggested that the Aβ transgenic mice exhibit hippocampal network hyperexcitability from the early stages of the disease. In this regard, it has been shown that several Aβ transgenic lines exhibit network hyperexcitability accompanied by spontaneous epileptiform activity [44 –48]. Further supporting these findings, studies by Palop and colleagues demonstrated that Aβ transgenic mice are characterized by spontaneous epileptiform discharges at the parietal cortex, indicating network hypersynchrony, hyperexcitability, and behavioral abnormalities [48].

The coordinated firing of ensembles of neurons gives rise to oscillatory activity that plays a key role in information processing in neuronal networks [18 , 50]. Interneurons play a significant role in coordinating network activity [49 –51]. In vitro experiments demonstrated that interneurons expressing parvalbumin (PV) and somatostatin (SOM) participated in intrinsic rhythm generation and PC coordination in distal CA1 and subiculum [52 –54]. While SOM interneurons strongly modulated temporoammonic inputs, PV interneurons controlled PC network and rhythm generation at low-frequency bands [50 , 56].

Slow gamma rhythms are driven by inputs from CA3 that recruit CA1 fast-spiking PV cells which in turn drive slow gamma activity in other interneurons [50]. Fast gamma rhythms are driven by inputs from the media entorhinal cortex and appear to drive an unidentified group of interneurons, that in turn, entrain the local network to fast gamma activity [50]. In vitro studies in CA3 have confirmed that PV cells were phase-locked to gamma oscillations [55]. Subsequent studies indicated that fast-spiking perisomatic targeting interneurons were essential for generating gamma oscillations [57, 58].

Considering that oscillatory activity is mainly generated by inhibitory interneurons, Palop and colleagues demonstrated that network dysfunction arises from PV cells impaired at the voltage-gated sodium channel (Nav1.1) [59]. Aiming to establish that this mechanism was consistent at the hippocampus level, we recently recorded PV cell activity from Aβ transgenic mice as young as one month old [29]. Our data showed that the spike frequency rate is significantly lower in Aβ transgenic mice when compared to age-matched related non-transgenic mice [29], therefore contributing to the increased activity in PC from the hippocampus and the power increase [48].

Further supporting our findings, Chung and colleagues demonstrated the dysfunction of PV and SOM interneurons’ input to CA1 PC as the synaptic mechanism underlying Aβ-induced impairments of hippocampal network oscillations [60].

The exact mechanism underlying Aβ-induced impairments of hippocampal network oscillations remains under extensive study. However, it was shown that perineuronal nets, consisting of chondroitin sulfate proteoglycans and hyaluronic acid, protect against Aβ neurotoxicity [61]. Thus, it was reported that perineuronal nets are affected during neurodegeneration [62]. The functions of perineuronal nets include synaptic stabilization and plasticity [61, 63]. Interestingly, perineuronal nets are also associated with GABAergic interneurons [64]. Thus, it was reported that perineuronal nets protect against lipid peroxidation and reactive oxygen damage [61]. Taken together, these results suggest that perineuronal nets might be involved in the modulation of network excitability.

The second result from our meta-analysis showed that frequency decreased in low oscillatory frequency bands (Fig. 2). Knowing that epilepsy is a disorder associated with increased network excitability accompanied by rewiring in the brain circuits [65], most likely these frequency changes may contribute to the altered hippocampal network activity. The third result from our meta-analysis further supported this hypothesis by revealing that Aβ accumulation causes a significant decrease in modulation index values (Fig. 3). In agreement with the power findings, modulation index alterations started from prodromal stages of disease and persisted throughout advanced stages of disease (Fig. 3). Supporting previous findings, we recently reported that modulation indexes, theta/low-gamma and theta/high-gamma, were significantly reduced in one-month-old Aβ transgenic mice when compared to age-matched related non-transgenic mice [29]. Importantly, the reported modulation index disturbances started at the prodromal stages of the disease [29].

In neuroscience, the “communication through coherence” hypothesis is widely accepted. This hypothesis proposes that efficient anatomic communications occur when synchronization increases [66]. Indeed, it has been shown that phase coupling reflects several cognitive processes in humans [21, 67]. Thus, it was suggested that phase coupling between different brain regions could serve as an inter-area communication vehicle [21, 67]. Similarly, the phase coupling between theta and gamma oscillations reflects the interrelations between local microscale and system-level macroscale neuronal networks [19 , 69]. Not surprisingly, disturbances in rhythm coherence are highly related to brain pathology development [10 –12]. Therefore, a decrease in frequency and modulation index values may contribute to the observed cognitive deficit in the AD-tg models.

In summary, lower modulation index values between theta and gamma, theta power increases, network hyperexcitability, and reduced theta frequency in the Aβ transgenic mouse models indicate that brain network function and communication are affected at several scales. Interestingly, the pooled analysis of the studies showed that these alterations, mainly caused by Aβ accumulation and deposition, were detected at the prodromal stages of disease progression, suggesting that they could be preclinical biomarkers based on electrophysiological changes.

pTau impact on brain network activity

Regarding tau mutations altering brain network activity in transgenic mouse models, the pooled analysis reveals that pTau deposition was reflected in power and frequency decreases in low-frequency bands (Figs. 4 and 5). Thus, the meta-analysis also revealed an increase in gamma peak power but a significant decrease in frequency (Figs. 4 and 5). Overall, these findings suggest that pTau mainly causes a significant decrease in the excitability levels of the hippocampal circuits. The findings of this study are consistent with reports that demonstrated that pTau reduces the activity of single neocortical PC and the neocortical network in young pTau transgenic mice [70]. Aiming to demonstrate that these alterations in brain network activity were consistent at the hippocampal level, we recorded brain network activity from pTau transgenic mice as young as one month old [28]. Our data showed that young pTau transgenic mice exhibit significant reductions in the power and frequency of lower frequency bands such as theta and delta [28]. Thus, we reported that the electrophysiological alterations were correlated with pTau increases in hippocampal PC and PV cells [28]. As previously discussed, PC and PV cells interactions drive network and oscillatory synchrony, and the synchronized inhibitory synaptic input onto excitatory PC produces accurate rhythmic patterns [6 , 60]. PC firing is synchronized with low-frequency oscillations like theta, whereas interneurons, particularly PV cells, are synchronized with low- and high-frequency oscillations like theta and gamma [6 , 54].

Notably, our previous research demonstrated that increases in pTau levels directly regulate the activity of the N-methyl-D-aspartate (NMDA)-sensitive glutamate receptor [71], leading to increase in phosphorylation of postsynaptic tau, therefore regulating the interaction of tau, Fyn, and the postsynaptic scaffolding protein PSD-95 [71]. In other words, increases in postsynaptic pTau lead to the breakdown of the PSD-95 complex, which reduces NMDA receptor activity [3 , 71–73]. Overall, the increase in pTau through disease development contributes to the decrease in power and frequency and the brain network hypoexcitability. The hypoexcitable state caused by pTau indicates that brain network function and communication at several scales are affected. Nevertheless, pooled analysis of the studies showed that theta/gamma modulation index values were not altered (Fig. 6). Indeed, our data showed that pTau accumulation was associated with enhanced cognitive performance [28]. Further supporting our data, Boekhoorn and colleagues [74] reported that young pTau transgenic mice showed increased long-term potentiation (LTP), which is a correlate for learning and memory [75]. According to our findings, Boekhoorn and colleagues also reported that the increase in LTP correlated with improved cognitive performance [74].

Despite the positive role of pTau, the pooled analysis of the studies showed that power and frequency alterations were detected at the prodromal stages of disease progression, suggesting that they could be preclinical biomarkers.

Conclusion and perspective

The pooled analysis revealed brain network alterations from the prodromal stage to the advanced stage of disease progression. Accordingly, electrophysiological changes detected in transgenic models could be valuable biomarkers of disease. However, the pooled analysis also revealed that, even though both proteins generate changes in the brain network activity, they also trigger an opposite effect in brain network activity: Aβ causes hyperexcitability, and pTau mainly causes hypoexcitability.

Another point to consider is that accumulating evidence suggests that both pathologies have synergistic effects [76]. Aβ causes neuronal hyperexcitability leading to impaired network oscillations [76]. However, Aβ-related hyperexcitability depends on endogenous tau levels [3 , 76]. Further supporting this relationship, it was reported that the default mode network (DMN), that is a set of widely distributed brain regions in the parietal, temporal, and frontal cortex [77], is hyperexcited with increasing levels of Aβ, which drives hyperexcitability within the medial temporal lobe and this directed hyperexcitation of the medial temporal lobe by the DMN predicts the rate of pTau accumulation within the entorhinal cortex [78].

Interestingly, triple AD-tg mice, marked by the accumulation of Aβ and pTau, are mainly characterized by hypoexcitability at the initial stages of the disease [3]. In agreement, it was demonstrated that Aβ and tau co-expression also suppressed neuronal activity as the tau phenotype seemed to dominate [76]. Overall, evidence suggests that tau, by reducing synaptic overexcitation, suppresses neuronal activity [3 , 76]. Taken together, data suggests that co-pathogenic interaction between Aβ and tau drives disease progression [3 , 76].

Although the exact mechanism remains under extensive study, we know that both proteins are directly connected through NMDA receptor activity. In other words, Aβ communicates with tau to participate in synapse strengthening or weakening depending on the stage of disease development [71 , 79–82]. Therefore, before attempting to use AD transgenic models for biomarker development, the physiological relationship between Aβ and tau at the synapse needs to be fully elucidated. Additionally, more research into the role of Aβ and tau in PC and PV cell modulation is needed to understand brain network alterations. Not only could new biomarkers emerge, but also new therapeutic strategies based on brain circuits remodeling.

AUTHOR CONTRIBUTIONS

Siddhartha Mondragon-Rodriguez (Conceptualization; Data curation; Formal analysis; Funding acquisition; Investigation; Supervision; Project administration; Writing – original draft; Writing – Review & Editing); Carlos A. Garcia-Carlos (Data curation; Formal analysis; Investigation; Methodology; Software; Writing – original draft); Gustavo Basurto-Islas (Data curation; Funding acquisition; Supervision); George Perry (Funding acquisition; Resources).

Footnotes

ACKNOWLEDGMENTS

We thank Jessica Gonzalez Norris for proofreading. We thank the following facilities from INB-UNAM: Proteogenomic Facility Unit, Behavioral Analysis Core Facility and Vivarium Facility. Carlos Antonio García-Carlos is a doctoral student from Programa de Doctorado en Ciencias Biológicas, Universidad Nacional Autónoma de México (UNAM).

FUNDING

This research was supported by Consejo Nacional de Ciencia y Tecnología CONAHCYT (grant number 269021, 319863 and A1-S-29906). Carlos Antonio García-Carlos was awarded by CONAHCYT (fellowship number 1082520), México. Dr. Mondragón-Rodríguez was awarded a Cátedra position by CONAHCYT, México.

CONFLICT OF INTEREST STATEMENT

The authors declare no competing financial interests. Dr. Mondragón-Rodríguez is an Editorial Board Member of this journal but was not involved in the peer-review process nor had access to any information regarding its peer-review. Dr. Perry is the Editor-in-Chief of JAD but was not involved in the peer-review process nor had access to any information regarding its peer-review.

GP serves on Nervgen (Scientific Advisory Board) and Synaptogenix (Equity and Scientific Advisory Board).

DATA AVAILABILITY

Data is available on request due to privacy/ethical restrictions.

References

Palop

, Chin

, Mucke

(2006) A network dysfunction perspective on neurodegenerative disease. Nature 443, 768–773.

Braak

, Braak

,(1991) . Neuropathological stageing of Alzheimer-related changes. Acta Neuropathol 82, 239–259.

Mondragón-Rodríguez

, Salas-Gallardo

, González-Pereyra

, Macías

, Ordaz

, Peña-Ortega

, Aguilar-Vázquez

, Orta-Salazar

, Díaz-Cintra

, Perry

, Williams

(2018) Phosphorylation of Tau protein correlates with changes in hippocampal theta oscillations and reduces hippocampal excitability in Alzheimer’s model. J Biol Chem 293, 8462–8472.

Mondragón-Rodríguez

, Salgado-Burgos

, Peña-Ortega

(2020) Circuitry and synaptic dysfunction in Alzheimer’s disease: a new tau hypothesis. Neural Plast 2020, 2960343.

Ittner

, Ke

, Delerue

, Bi

, Gladbach

, van Eersel

, Wölfing

, Chieng

, Christie

, Napier

, Eckert

, Staufenbiel

, Hardeman

, Götz

(2010) Dendritic function of tau mediates amyloid-β toxicity in Alzheimer’s disease mouse models. Cell 142, 387–397.

Palop

, Mucke

(2016) Network abnormalities and interneuron dysfunction in Alzheimer disease. Nat Rev Neurosci 17, 777–792.

Buzsáki

, Watson

(2012) Brain rhythms and neural syntax: Implications for efficient coding of cognitive content and neuropsychiatric disease. Dialogues Clin, Neurosci 14, 345–367.

Buzsáki

, (2002) . Theta oscillations in the hippocampus. Neuron 33, 325–340.

Goutagny

, Jackson

, Williams

(2009) Self-generated theta oscillations in the hippocampus. Nat Neurosci 12, 1491–1493.

10.

Moretti

D V.

, Babiloni

, Binetti

, Cassetta

, Dal Forno

, Ferreric

, Ferri

, Lanuzza

, Miniussi

, Nobili

, Rodriguez

, Salinari

, Rossini

(2004) Individual analysis of EEG frequency and band power in mild Alzheimer’s disease. Clin Neurophysiol 115, 299–308.

11.

Gaubert

, Raimondo

, Houot

, Corsi

, Naccache

, Sitt

, Hermann

, Oudiette

, Gagliardi

, Habert

, Dubois

, De Vico Fallani

, Bakardjian

, Epelbaum

(2019) EEG evidence of compensatory mechanisms in preclinical Alzheimer’s disease. Brain 142, 2096–2112.

12.

Musaeus

, Engedal

, Høgh

, Jelic

, Mørup

, Naik

, Oeksengaard

, Snaedal

, Wahlund

, Waldemar

, Andersen

(2018) EEG theta power is an early marker of cognitive decline in dementia due to Alzheimer’s disease. J Alzheimers Dis 64, 1359–1371.

13.

Jelic

, Johansson

, Almkvist

, Shigeta

, Julin

, Nordberg

, Winblad

, Wahlund

(2000) Quantitative electroencephalography in mild cognitive impairment: Longitudinal changes and possible prediction of Alzheimer’s disease. Neurobiol Aging 21, 533–540.

14.

Huang

, Wahlund

, Dierks

, Julin

, Winblad

, Jelic

(2000) Discrimination of Alzheimer’s disease and mild cognitive impairment by equivalent EEG sources: A cross-sectional and longitudinal study. Clin Neurophysiol 111, 1961–1967.

15.

Tort

ABL

, Komorowski

, Manns

, Kopell

, Eichenbaum

(2009) Theta-gamma coupling increases during the learning of item-context associations. Proc Natl Acad Sci U S A 106, 20942–20947.

16.

Fernández-Ruiz

, Oliva

, Nagy

, Maurer

, Berényi

, Buzsáki

(2017) Entorhinal-CA3 dual-input control of spike timing in the hippocampus by theta-gamma coupling. Neuron 93, 1213–1226.

17.

Lisman

, Jensen

(2013) The theta-gamma neural code. Neuron 77, 1002–1016.

18.

Buzsáki

, Moser

(2013) Memory, navigation and theta rhythm in the hippocampal-entorhinal system. Nat Neurosci 16, 130–138.

19.

Belluscio

, Mizuseki

, Schmidt

, Kempter

, Buzsáki

(2012) Cross-frequency phase-phase coupling between theta and gamma oscillations in the hippocampus. J Neurosci 32, 423–435.

20.

, Bai

, Liu

, Yi

, Tian

(2012) Increases of theta-low gamma coupling in rat medial prefrontal cortex during working memory task. Brain Res Bull 89, 115–123.

21.

Canolty

, Edwards

, Dalal

, Soltani

, Nagarajan

, Kirsch

, Berger

, Barbare

, Knight

(2006) High gamma power is phase-locked to theta oscillations in human neocortex. Science 313, 1626–1628.

22.

Nakazono

, Takahashi

, Sakurai

(2019) Enhanced theta and high-gamma coupling during late stage of rule switching task in rat hippocampus. Neuroscience 412, 216–232.

23.

Schack

, Vath

, Petsche

, Geissler

, Möller

(2002) Phase-coupling of theta-gamma EEG rhythms during short-term memory processing. Int J Psychophysiol 44, 143–163.

24.

Goodman

, Kumar

, Zomorrodi

, Ghazala

, Cheam

ASM

, Barr

, Daskalakis

, Blumberger

, Fischer

, Flint

, Mah

, Herrmann

, Bowie

, Mulsant

, Rajji

, Pollock

, Lourenco

, Butters

, Gallagher

, Golas

, Graff

, Kennedy

, Ovaysikia

, Rapoport

, Thorpe

, Verhoeff

NPLG

, Voineskos

(2018) Theta-Gamma coupling and working memory in Alzheimer’s dementia and mild cognitive impairment. Front Aging Neurosci 10, 101.

25.

Sanchez-Varo

, Mejias-Ortega

, Fernandez-Valenzuela

, Nuñez-Diaz

, Caceres-Palomo

, Vegas-Gomez

, Sanchez-Mejias

, Trujillo-Estrada

, Garcia-Leon

, Moreno-Gonzalez

, Vizuete

, Vitorica

, Baglietto-Vargas

, Gutierrez

(2022) Transgenic mouse models of Alzheimer’s disease: an integrative analysis. Int J Mol Sci 23, 5404.

26.

Ashe

(2009) Alzheimer’s disease: transgenic mouse models. Encyclopedia Neurosci 77, 283–287.

27.

Foley

, Ammar

, Lee

, Mitchell

(2015) Systematic review of the relationship between amyloid-β levels and measures of transgenic mouse cognitive deficit in Alzheimer’s disease. J Alzheimers Dis 44, 787–795.

28.

Xolalpa-Cueva

, García-Carlos

, Villaseñor-Zepeda

, Orta-Salazar

, Díaz-Cintra

, Peña-Ortega

, Perry

, Mondragón-Rodríguez

(2022) Hyperphosphorylated tau relates to improved cognitive performance and reduced hippocampal excitability in the young rTg4510 mouse model of tauopathy. J Alzheimers Dis 87, 529–543.

29.

Mondragón-Rodríguez

, Gu

, Manseau

, Williams

(2018) Alzheimer’s transgenic model is characterized by very early brain network alterations and β-CTF fragment accumulation: Reversal by β-secretase inhibition. Front Cell Neurosci 12, 121.

30.

Mondragón-Rodríguez

, Gu

, Fasano

, Peña-Ortega

, Williams

(2019) Functional connectivity between hippocampus and lateral septum is affected in very young Alzheimer’s transgenic mouse model. Neuroscience 401, 96–105.

31.

Fontana

, Agostini

, Murana

, Mahmud

, Scremin

, Rubega

, Sparacino

, Vassanelli

, Fasolato

(2017) Early hippocampal hyperexcitability in PS2APP mice: Role of mutant PS2 and APP. Neurobiol Aging 50, 64–76.

32.

Kumari

, Li

, Yang

, Zhang

(2020) Tacrine accelerates spatial long-term memory via improving impaired neural oscillations and modulating GAD isomers including neuro-receptors in the hippocampus of APP/PS1 AD mice. Brain Res Bull 161, 166–176.

33.

Wang

, Li

, Zhao

, Zhang

, Yang

, Zhang

(2020) Early-stage dysfunction of hippocampal theta and gamma oscillations and its modulation of neural network in a transgenic 5xFAD mouse model. Neurobiol Aging 94, 121–129.

34.

Dávila-Bouziguet

, Casòliba-Melich

, Targa-Fabra

, Galera-López

, Ozaita

, Maldonado

, Ávila

, Delgado-García

, Gruart

, Soriano

, Pascual

(2022) Functional protection in J20/VLW mice: a model of non-demented with Alzheimer’s disease neuropathology. Brain 145, 729–743.

35.

Hollnagel

, Elzoheiry

, Gorgas

, Kins

, Beretta

, Kirsch

, Kuhse

, Kann

, Kiss

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

36.

Cayzac

, Mons

, Ginguay

, Allinquant

, Jeantet

, Cho

(2015) Altered hippocampal information coding and network synchrony in APP-PS1 mice. Neurobiol Aging 36, 3200–3213.

37.

Jin

, Lipponen

, Koivisto

, Gurevicius

, Tanila

(2018) Increased cortical beta power and spike-wave discharges in middle-aged APP/PS1 mice. Neurobiol Aging 71, 127–141.

38.

Stoiljkovic

, Kelley

, Horvath

, Hajós

(2018) Neurophysiological signals as predictive translational biomarkers for Alzheimer’s disease treatment: effects of donepezil on neuronal network oscillations in TgF344-AD rats. Alzheimers Res Ther 10, 105.

39.

Stoiljkovic

, Gutierrez

, Kelley

, Horvath

, Hajós

(2022) TREM2 deficiency disrupts network oscillations leading to epileptic activity and aggravates amyloid-β-related hippocampal pathophysiology in mice. J Alzheimers Dis 88, 837–847.

40.

Ahnaou

, Rodriguez-Manrique

, Biermans

, Embrechts

, Manyakov

N V.

, Drinkenburg

(2020) Functional alterations in the olfactory neuronal circuit occur before hippocampal plasticity deficits in the P301S mouse model of tauopathy: implications for early diagnosis and translational research in Alzheimer’s disease. Int J Mol Sci 21, 5431.

41.

Ciupek

, Cheng

, Ali

, Lu

, Ji

(2015) Progressive functional impairments of hippocampal neurons in a tauopathy mouse model. J Neurosci 35, 8118–8131.

42.

Witton

, Staniaszek

, Bartsch

, Randall

, Jones

, Brown

(2016) Disrupted hippocampal sharp-wave ripple-associated spike dynamics in a transgenic mouse model of dementia. J Physiol 594, 4615–4630.

43.

Buzsáki

, Draguhn

(2004) Neuronal oscillations in cortical networks. Science 304, 1926–1929.

44.

Palop

, Chin

, Roberson

, Wang

, Thwin

, Bien-Ly

, Yoo

, Ho

, Yu

, Kreitzer

, Finkbeiner

, Noebels

, Mucke

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

45.

Busche

, Eichhoff

, Adelsberger

, Abramowski

, Wiederhold

, Haass

, Staufenbiel

, Konnerth

, Garaschuk

(2008) Clusters of hyperactive neurons near amyloid plaques in a mouse model of Alzheimer’s disease. Science 321, 1686–1689.

46.

Minkeviciene

, Rheims

, Dobszay

, Zilberter

, Hartikainen

, Fülöp

, Penke

, Zilberter

, Harkany

, Pitkänen

, Tanila

(2009) Amyloid β-induced neuronal hyperexcitability triggers progressive epilepsy. J Neurosci 29, 3453–3462.

47.

Roberson

, Halabisky

, Yoo

, Yao

, Chin

, Yan

, Wu

, Hamto

, Devidze

, Yu

, Palop

, Noebels

, Mucke

(2011) Amyloid-β/fyn-induced synaptic, network, and cognitive impairments depend on tau levels in multiple mouse models of Alzheimer’s disease. J Neurosci 31, 700–711.

48.

Harris

, Devidze

, Verret

, Ho

, Halabisky

, Thwin

, Kim

, Hamto

, Lo

, Yu

, Palop

, Masliah

, Mucke

(2010) Transsynaptic progression of amyloid-β-induced neuronal dysfunction within the entorhinal-hippocampal network. Neuron 68, 428–441.

49.

Tzilivaki

, Tukker

, Maier

, Poirazi

, Sammons

, Schmitz

(2023) Hippocampal GABAergic interneurons and memory. Neuron 111, 3154–3175.

50.

Pelkey

, Chittajallu

, Craig

, Tricoire

, Wester

, McBain

(2017) Hippocampal GABAergic inhibitory interneurons. Physiol Rev 97, 1619–1747.

51.

Pancotti

, Topolnik

(2022) Cholinergic modulation of dendritic signaling in hippocampal GABAergic inhibitory interneurons. Neuroscience 489, 44–56.

52.

Amilhon

, Huh

CYL

, Manseau

, Ducharme

, Nichol

, Adamantidis

, Williams

(2015) Parvalbumin interneurons of hippocampus tune population activity at theta frequency. Neuron 86, 1277–1289.

53.

Huh

CYL

, Amilhon

, Ferguson

, Manseau

, Torres-Platas

, Peach

, Scodras

, Mechawar

, Skinner

, Williams

(2016) Excitatory inputs determine phase-locking strength and spike-timing of CA1 stratum oriens/alveus parvalbumin and somatostatin interneurons during intrinsically generated hippocampal theta rhythm. J Neurosci 36, 6605–6622.

54.

Park

, Lee

, Jang

, Richards

, Kohl

, Kwag

(2020) Optogenetic activation of parvalbumin and somatostatin interneurons selectively restores theta-nested gamma oscillations and oscillation-induced spike timing-dependent long-term potentiation impaired by amyloid β oligomers. BMC Biol 18, 1–20.

55.

Hájos

, Pálhalini

, Mann

, Nèmeth

, Paulsen

, Freund

(2004) Spike timing of distinct types of GABAergic interneuron during hippocampal gamma oscillations. J Neurosci 24, 9127–9137.

56.

Fisahn

, Pike

, Buhl

, Paulsen

(1998) Cholinergic induction of network oscillations at 40Hz in the hippocampus. Nature 394, 186–189.

57.

Oren

, Hájos

, Paulsen

(2010) Identification of the current generator underlying cholinergically induced gamma frequency field potential oscillations in the hippocampal CA3 region. J Physiol 588, 785–797.

58.

Mann

, Suckling

, Hajos

, Greenfield

, Paulsen

(2005) Perisomatic feedback inhibition underlies cholinergically induced fast network oscillations in the rat hippocampus. Neuron 45, 105–117.

59.

Martinez-Losa

, Tracy

, MA

, Verret

, Clemente-Perez

, Khan

, Cobos

, Ho

, Gan

, Mucke

, Alvarez-Dolado

, Palop

(2018) Nav1.1-overexpressing interneuron transplants restore brain rhythms and cognition in a mouse model of Alzheimer’s disease. Neuron 98, 75–89.

60.

Chung

, Park

, Jang

, Kohl

, Kwag

(2020) Dissociation of somatostatin and parvalbumin interneurons circuit dysfunctions underlying hippocampal theta and gamma oscillations impaired by amyloid β oligomers. Brain Struct Funct 225, 935–954.

61.

Miyata

, Nishimura

, Nakashima

(2007) Perineuronal nets protect against amyloid β-protein neurotoxicity in cultured cortical neurons. Brain Res 1150, 200–206.

62.

Reichelt

(2020) Is loss of perineuronal nets a critical pathological event in Alzheimer’s disease? . EBioMedicine 59, 102946.

63.

Hockfield

, Kalb

, Zaremba

, Fryer

(1990) Expression of neural proteoglycans correlates with the acquisition of mature neuronal properties in the mammalian brain. Cold Spring Harb Symp Quant Biol 55, 505–514.

64.

Kosaka

, Heizmann

(1989) Selective staining of a population of parvalbumin-containing GABAergic neurons in the rat cerebral cortex by lectins with specific affinity for terminal N-acetylgalactosamine. Brain Res 483, 158–163.

65.

Morimoto

, Fahnestock

, Racine

(2004) Kindling and status epilepticus models of epilepsy: Rewiring the brain. Prog Neurobiol 73, 1–60.

66.

Fries

(2005) A mechanism for cognitive dynamics: Neuronal communication through neuronal coherence. Trends Cogn Sci 9, 474–480.

67.

Axmacher

, Henseler

, Jensen

, Weinreich

, Elger

, Fell

(2010) Cross-frequency coupling supports multi-item working memory in the human hippocampus. Proc Natl Acad Sci U S A 107, 3228–3233.

68.

Colgin

, Denninger

, Fyhn

, Hafting

, Bonnevie

, Jensen

, Moser

(2009) Frequency of gamma oscillations routes flow of information in the hippocampus. Nature 462, 353–357.

69.

Canolty

, Ganguly

, Kennerley

, Cadieu

, Koepsell

, Wallis

, Carmena

(2010) Oscillatory phase coupling coordinates anatomically dispersed functional cell assemblies. Proc Natl Acad Sci U S A 107, 17356–17361.

70.

Menkes-Caspi

, Yamin

, Kellner

, Spires-Jones

, Cohen

, Stern

(2015) Pathological tau disrupts ongoing network activity. Neuron 85, 959–966.

71.

Mondragón-Rodríguez

, Trillaud-Doppia

, Dudilot

, Bourgeois

, Lauzon

, Leclerc

, Boehm

(2012) Interaction of endogenous tau protein with synaptic proteins is regulated by N-methyl-D-aspartate receptor-dependent tau phosphorylation. J Biol Chem 287, 32040–32053.

72.

Mondragón-Rodríguez

, Perry

, Peña-Ortega

, Williams

(2017) Tau, amyloid beta and deep brain stimulation: aiming to restore cognitive deficit in Alzheimer’s disease. Curr Alzheimer Res 14, 40–46.

73.

Mondragón-Rodríguez

, Perry

, Zhu

, Moreira

, Acevedo-Aquino

, Williams

(2013) Phosphorylation of tau protein as the link between oxidative stress, mitochondrial dysfunction, and connectivity failure: Implications for Alzheimer’s disease. Oxid Med Cell Longev 2013, 940603.

74.

Boekhoorn

, Terwel

, Biemans

, Borghgraef

, Wiegert

, Ramakers

GJA

, De Vos

, Krugers

, Tomiyama

, Mori

, Joels

, Van Leuven

, Lucassen

(2006) Improved long-term potentiation and memory in young tau-P301L transgenic mice before onset of hyperphosphorylation and tauopathy. J Neurosci 26, 3514–3523.

75.

Collingridge

, Peineau

, Howland

, Wang

(2010) Long-term depression in the CNS. Nat Rev Neurosci 11, 459–473.

76.

Busche

, Hyman

(2020) Synergy between amyloid-β and tau in Alzheimer’s disease. Nat Neurosci 23, 1183–1193.

77.

Smallwood

, Bernhardt

, Leech

, Bzdok

, Jefferies

, Margulies

(2021) The default mode network in cognition: a topographical perspective. Nat Rev Neurosci 22, 503–513.

78.

Giorgio

, Adams

, Maass

, Jagust

, Breakspear

(2024) Amyloid induced hyperexcitability in default mode network drives medial temporal hyperactivity and early tau accumulation. Neuron 112, 676–686.

79.

Mondragón-Rodríguez

, Perry

, Zhu

, Moreira

, Williams

(2012) Glycogen synthase kinase 3: A point of integration in alzheimer’s disease and a therapeutic target? . Int J Alzheimers Dis 2012, 276803.

80.

Spires-Jones

, Hyman

(2014) The intersection of amyloid beta and tau at synapses in Alzheimer’s disease. Neuron 82, 756–771.

81.

Zempel

, Thies

, Mandelkow

(2010) Aβ oligomers cause localized Ca²⁺ elevation, missorting of endogenous tau into dendrites, tau phosphorylation, and destruction of microtubules and spines. J Neurosci 30, 11938–11950.

82.

Selkoe

(2002) Alzheimer’s disease is a synaptic failure. Science 298, 789–791.