TY - JOUR
AU - Ittner, Lars, M
AB - Abstract Hyperphosphorylation and deposition of tau in the brain characterizes frontotemporal dementia and Alzheimer’s disease. Disease-associated mutations in the tau-encoding MAPT gene have enabled the generation of transgenic mouse models that recapitulate aspects of human neurodegenerative diseases, including tau hyperphosphorylation and neurofibrillary tangle formation. Here, we characterized the effects of transgenic P301S mutant human tau expression on neuronal network function in the murine hippocampus. Onset of progressive spatial learning deficits in P301S tau transgenic TAU58/2 mice were paralleled by long-term potentiation deficits and neuronal network aberrations during electrophysiological and EEG recordings. Gene-expression profiling just prior to onset of apparent deficits in TAU58/2 mice revealed a signature of immediate early genes that is consistent with neuronal network hypersynchronicity. We found that the increased immediate early gene activity was confined to neurons harbouring tau pathology, providing a cellular link between aberrant tau and network dysfunction. Taken together, our data suggest that tau pathology drives neuronal network dysfunction through hyperexcitation of individual, pathology-harbouring neurons, thereby contributing to memory deficits. Alzheimer’s disease, frontotemporal dementia, memory, tau, neurodegeneration Introduction Alzheimer’s disease and frontotemporal dementia (FTD) are two of the most prevalent forms of dementia. Memory decline is the lead symptom of Alzheimer’s disease (Gotz et al., 2012) and is associated with behavioural, personality (including disinhibition and apathy) and/or language changes in FTD (Weder et al., 2007; Piguet et al., 2011). Both Alzheimer’s disease and FTD are histopathologically characterized by progressive neuronal loss and the deposition of hyperphosphorylated tau in the brain (Ballatore et al., 2007; Ittner and Gotz, 2011b). Tau belongs to the family of microtubule-associated proteins and is encoded by the MAPT gene (Gendron and Petrucelli, 2009). Under physiological conditions, tau binds to microtubules and regulates their dynamics and contributes to the maintenance of neuronal functions (Ballatore et al., 2007). In the CNS, tau is predominately enriched in neuronal axons, but also has physiological functions in other compartments of neurons, including the post-synapse (Ittner et al., 2010, 2016). The localization and physiological function of tau is determined by a large number of post-translational modifications (Mandelkow and Mandelkow, 2012). In disease, tau becomes increasingly phosphorylated at both physiological and pathological sites (hyperphosphorylation), compromising tau-microtubule interactions and eventually causing the detachment of tau from microtubules and its mislocalization from the axon to the somatodendritic compartment (Ittner et al., 2010). Eventually, hyperphosphorylated tau accumulates into insoluble neurofibrillary tangles, the hallmark lesions of Alzheimer’s disease and FTD. The identification of MAPT mutations in familial FTD established that tau dysfunction is sufficient to cause neurodegeneration and cognitive decline (Spillantini et al., 2000; Wszolek et al., 2005). Since then, transgenic mice carrying various MAPT mutations have been generated by us and others, recapitulating aspects of human neurodegenerative diseases, including tau hyperphosphorylation and neurofibrillary tangle formation (Lewis et al., 2000; Gotz et al., 2001; Allen et al., 2002; Santacruz et al., 2005; Yoshiyama et al., 2007; Ittner et al., 2008; van Eersel et al., 2015). Although behavioural dysfunctions have been studied in several tau transgenic mouse models of Alzheimer’s disease and FTD (Gotz and Ittner, 2008), underlying mechanisms for hippocampal-dependent learning and neuronal network dysfunctions, in particular in FTD, remain unclear. This may also be because of the assumption that memory dysfunctions are relatively spared or of late onset in FTD (Ahmed et al., 2017). However, recent studies revealed significant deficits in hippocampal-driven episodic memory function in FTD patients (Hornberger and Piguet, 2012) and further provide evidence that underlying memory impairments in FTD might even be comparable to those found in Alzheimer’s disease (Hornberger et al., 2010). In Alzheimer’s disease, tau pathology and synaptic failure correlate with cognitive decline (Serrano-Pozo et al., 2011). Similarly, post-mortem studies in FTD with tau pathology reported a reduction in key synaptic proteins positively correlating with cognitive dysfunction (Clare et al., 2010; Goetzl et al., 2016), further indicating that tau pathology may compromise neuronal network integrity in FTD. We and others have previously shown that tau mediates amyloid-β-induced neuronal network dysfunctions in Alzheimer’s disease mouse models (Roberson et al., 2007; Ittner et al., 2010). Although both in vitro and in vivo studies have demonstrated that tau contributes to neuronal network aberration in disease (Ittner et al., 2010; Fa et al., 2016; Busche et al., 2019), these mechanisms were described in conjunction with amyloid-β. Similarly, cortical recordings in tau transgenic mice suggests neuronal network dysfunction (Crimins et al., 2012; Garcia-Cabrero et al., 2013; Menkes-Caspi et al., 2015; Holth et al., 2017; Das et al., 2018). Whether aberrant tau per se is sufficient to drive neuronal network deficits needs to be elucidated further. Here, we show that memory deficits in the P301S mutant human tau transgenic TAU58/2 mouse strain are associated with progressive neuronal network dysfunction in the hippocampus. Interestingly, markers of neuronal hyperexcitation were confined to neurons harbouring tau pathology. Materials and methods Mice TAU58/2 mice express the human 0N4R tau isoform with the P301S mutation under the control of the mouse Thy1.2 promoter, as previously described (van Eersel et al., 2015). Mice were maintained heterozygous on a C57BL/6 background and non-transgenic littermates were used as controls. Mice were housed in filter top cages containing nesting material, a wooden stick and a transparent red dome and maintained on a 12-h light/dark cycle with food and water ad libidum. All animal experiments were conducted with male mice only due to earlier onset and faster progression of phenotypes in male TAU58/2 mice compared to females (van Eersel et al., 2015). Experimenters were blinded to the genotype of mice until data acquisition was completed. All animal experiments were approved by the Animal Ethics Committees of Macquarie University and the University of New South Wales. All procedures complied with the statement on animal experimentation issued by the National Health and Medical Research Council of Australia. Memory testing To assess spatial learning and memory, the Morris water maze was conducted on 2-, 4- and 6-month-old TAU58/2 transgenic mice (n = 9–13 males) and non-transgenic littermate controls (n = 9–12 males). The Morris water maze apparatus consisted of a 1.2-m diameter tank with a 40-cm high Perspex platform (diameter 10 cm), which was placed ∼20 cm from the edge of the wall. The tank was filled 0.5–1 cm above the surface of the platform and a non-toxic acrylic-based paint added to the water (19–21°C) to obscure the platform. Four signposts with different shapes were placed equidistant around the pool as visual cues. Mice were acclimatized to the room for 1 h prior to testing each day. Days 1–5 consisted of an acquisition phase, in which mice were placed in the quadrant opposite the platform (Q4) at one of four starting positions and given 60 s to locate the hidden platform (Q1). Mice that failed to find the hidden platform were guided to the platform and all mice remained on the platform for an additional 60 s before being removed from the maze. Mice had four consecutive trials per day, each starting from a different position, and the order of starting positions was altered each day. On the sixth day, the platform was removed, and the mice were given 30 s to explore the pool (probe trial). On the seventh day, the platform was placed back in the pool with a flag attached, and visual cues were removed from the outside of the pool, to ensure that all mice had normal vision. Average swim speed of all mice was determined to exclude motor deficits. Mice were tested in littermate groups, without knowledge of genotypes and video analysis was done before de-blinding experimenters. Videos were analysed using the ANY-maze® software (Stoelting Co., IL, USA). Morris water maze path analysis For all learning days, trace plots were obtained for each swim after video analysis using ANY-maze® (Stoelting Co., IL, USA). Swim traces were classified visually using a paradigm based on Garthe et al. (Garthe and Kempermann, 2013; Tan et al., 2018). Briefly, swim patterns were scored as follows: 1, thigmotaxis; 2, random swim; 3, scanning; 4, chaining; 5, directed search; 6, focal search; and 7, direct swim. Based on this scoring scheme, 1–3 reflect non-spatial hippocampal learning, whilst search strategies 4–7 are considered to reflect spatial hippocampal learning. Electrophysiology Horizontal brain slices (350 μm) were prepared from naïve 2- and 4-month-old TAU58/2 transgenic mice (n = 6–9) and non-transgenic littermates (n = 7) using a VT1200 vibratome (Leica) according to standard techniques. Briefly, mice were anaesthetized with isoflurane (5%), decapitated, and brains were removed and sectioned in oxygenated (95% O2, 5% CO2) ice-cold, modified artificial CSF containing the following (in mM): 125 NaCl, 3 KCl, 1.25 NaH2PO4, 25 NaHCO3, 6 MgCl2, 1 CaCl2, and 10 glucose. Slices were maintained at 31 ± 1°C thereafter in artificial CSF solution (in mM): 125 NaCl, 3 KCl, 1.25 NaH2PO4, 25 NaHCO3, 1 MgCl2, 2 CaCl2, and 25 glucose (95% O2, 5% CO2). After equilibration of at least 60 min, slices were transferred into a commercial brain slice recording chamber (Kerr Scientific Instruments Ltd), constantly superfused at 2–2.5 ml/min with oxygenated artificial CSF at 31 ± 1°C. Synaptic potentials were evoked using a bipolar stimulating electrode (Kerr Scientific Instruments Ltd.) and a constant voltage biphasic isolated stimulator (A-M Systems Model 2200). Field excitatory postsynaptic potentials (fEPSPs) were obtained using a recording electrode (Kerr Scientific Instruments Ltd.) placed in the stratum radiatum approximately 300–400 μm from the stimulating electrode. Potentials were amplified (100–250×) with an KSI 2 amplifier (Kerr Scientific Instruments Ltd.) and digitized (20 kHz) using a multifunctional data acquisition card (National Instruments NI PCI-6221) connected to a PC running AxoGraph X (Axograph Scientific). Field EPSPs were evoked at 30-s intervals. After establishing a stable baseline, a stimulus response function was generated. The stimulus intensity that evoked a half-maximum fEPSP amplitude was used for the duration of the experiment. A theta-burst stimulation (TBS) was used to induce long-term potentiation (LTP). The TBS protocol for induction of LTP consisted of 10 trains of five pulses (0.1 ms pulse width) at 100 Hz with a 200 ms inter-train interval, repeated twice with a 20 s interval. The slope of the EPSP (10% and 90%) was used as an index of synaptic strength and TBS-induced synaptic potentiation was calculated and analysed relative to baseline. Unstable recordings and abnormal wave patterns were excluded from further analysis. EEG Hippocampal EEG recordings were conducted in freely moving 4- and 6-month-old TAU58/2 mice (n = 4–6) and littermate controls (n = 3–5) as previously described (Ittner et al., 2014). Briefly, after mice were anaesthetized with ketamine/xylazine, a scalp incision along the midline was performed and the head was fixed using a stereotaxic frame (Kopf instrument). Bregma was located and bone openings were drilled at previously described positions for hippocampus (x, y, z = 2.0, −2.0, −2 from bregma) using a bone micro-drill (Fine Science Tools, F.S.T). A wire EEG electrode was inserted at this position, whilst the reference electrode was placed above the cerebellum (x, y, z = 0, −6.0,0 with reference to bregma). Both electrodes were fixed in place using polyacrylate and the wound was closed with staples and rehydrated. Ten days after surgery, EEGs and activity were recorded using a DSI wireless receiver set-up (DSI) with amplifier matrices using the Dataquest A.R.T. recording software at 500 Hz sampling rate (Weiergraber et al., 2005). After EEG recordings were successfully conducted, animals were perfused with cold phosphate-buffered saline (PBS) and brains extracted and further processed for histological analysis. Correct electrode placement was confirmed by serial sections of paraffin-embedded brain tissue stained with haematoxylin and eosin. Only recordings from mice with correct placement were included in further analysis. EEG recordings were analysed using the NeuroScore software v3.0 (DSI) with integrated spike detection tool. Activity counts were analysed over 24 h. The number of spikes were thus detected automatically and statistical data on number and frequency of spikes obtained. All recordings were visually screened for movement artefacts and only artefact-free episodes were used for further analysis. Spectral analysis [i.e. analysis of signal power at individual frequencies expressed as square of the fast Fourier transform (FFT) magnitude] of interictal sequences was conducted using the integrated FFT spectral analysis function of NeuroScore. Frequency bands of theta and gamma wave forms were defined between 4–12 Hz (low theta 4–8 Hz, high theta 9–12 Hz) and 30–100 Hz (low gamma 30–60 Hz, high gamma 61–100 Hz), respectively. Gamma and theta spectral contributions were quantified by area under the curve (AUC) analysis across the defined frequency band in 8–10 artefact- and hypersynchronous spike-free sequences per recording (each 1 min in length). Cross-frequency coupling of theta phase and gamma amplitude was performed using MATLAB (Mathworks) as previously described (Tort et al., 2010). Briefly, for cross frequency coupling (CFC) analysis, raw local field potential (LFP) was noise filtered using a powerline noise filter (Neuroscore, DSI). Noise-filtered LFP was filtered at two frequency ranges of interest for gamma (fA) and theta (fp). The phase time series for theta [ Φfp(t)] and the amplitude envelope time series for gamma [AfA(t)] were obtained by Hilbert transformation of the filtered LFPs. The combined series [ Φfp(t) ⁠, AfA(t)] was then generated. After phase binning, the means A‾fA(j) of AfA for each bin j were calculated and normalized using the sum ∑j=1NA‾fA(j) of A‾fA(j)(1) over n bins to generate phase-amplitude distribution P(j). The modulation index, a measure of CFC (Tort et al., 2009; Tort et al., 2010), is based on calculating the Kullback-Leibler distance DKL between the non-uniform (i.e. coupled) phase-amplitude distribution P(j) over all phase bins and the uniform (i.e. uncoupled) distribution U(j). DKL(P,Q)=∑j=1NPjlog⁡[P(j)U(j)](2) The modulation index (MI) is defined as: MI=DKL(Pj,Uj)log⁡N(3) Phase amplitude distributions and modulation indices were determined from artefact- and hypersynchronous spike-free 8–10 sequences (each 1 min) per recording. Seizure induction Seizure induction with 50 mg/kg freshly prepared pentylenetetrazol (PTZ) intraperitoneally and subsequent response scoring was performed as previously described (Ittner et al., 2010; Bi et al., 2017). RNA sequencing To obtain an unbiased insight into the molecular changes associated with transgenic P301S human tau expression, we performed quantitative polyadenylated RNA sequencing (RNAseq) of hippocampal brain samples from 10-week-old TAU58/2 mice (n = 4) and their non-transgenic littermates (n = 4). The mRNA extraction and mRNA sequencing were performed by Macrogen (Korea). Briefly, Maxwell® 16 LEV simplyRNA tissue kits (Promega) was used for mRNA extraction, followed by library construction using TruSeq Stranded Total RNA with Ribo-Zero Gold sample prep kit and TruSeq rapid SBS kit. Sequencing was performed on the Illumina HiSeq 2500 platform according to the HiSeq 2500 System User Guide Document #15035786 v01 HCS 2.2.70 protocol, using HCS version 2.2 for sequencing control. To qualify filter reads and to remove adapter contamination, trimmomatic 0.35 was used. The first 10 and the last five bases were removed after visual inspection of the read quality distribution. After trimming, the final read length was 86 bp. For read mapping, Star aligner v2.5.2b was used on the Gencode M12 transcriptome (mm10) with settings ‘–outSAMmultNmax 20’. Reads were assigned to genes using feature Counts v1.5.2 with settings ‘-C -B -M -O’. To calculate log2 fold change estimates between groups, a negative binomial general linear model was fit using DESeq2 and only genes with a Benjamini-Hochberg corrected Wald test false discovery rate (FDR) of <30% were labelled significant (Supplementary Table 1). STRING annotation The list of differentially regulated genes obtained through RNAseq analysis was subjected to STRING v10.5 to identify possible gene clusters (Szklarczyk et al., 2017). Data source settings for nodal associations were set to default (‘on’ for text-mining, databases, experiments, co-expression, neighbourhood, gene-fusion, co-occurrence) with ‘medium confidence’ (≥0.400) interaction score minimums. Gene networks were visualized using Cytoscape v3.6.1 software (Shannon et al., 2003). Each node represents a protein, while each edge corresponds to a STRING-based physical and/or functional interaction. The node size reflects the fold change (upregulated genes appear larger than downregulated genes) and is proportional to the FPKM (fragments per kilobase million) fold-change in transgenic relative to non-transgenic samples. The node colour reflects whether the gene is up- (red) or downregulated (blue) or indicates whether the gene was found through RNAseq (red) or was proposed by string as being involved in similar pathways (grey). The thickness of each edge is proportional to the STRING-based physical and/or functional interaction between two genes (the thicker the edge, the stronger is their STRING-based association), whilst the edge colour indicates interactions between genes obtained through RNAseq (green) versus genes that were proposed by STRING (dotted black lines). DAVID gene ontology Protein ontology was analysed using DAVID gene ontology database (v6.8) using gene ontology annotations for the entire human genome. The top five functional annotated clusters of enriched protein sets according to DAVID Db (v6.8) were displayed. The input list consisted of all differentially regulated genes. Quantitative polymerase chain reaction Isolated hippocampal tissue of 10-week-old TAU58/2 mice (n = 7) and non-transgenic littermate controls (n = 5) was homogenized in TRIzol®, according to the manufacturer’s instructions. The aqueous phase was then processed further using a RNeasy® Mini Kit (Qiagen) following the manufacturer’s instructions. To remove contaminating genomic DNA, an on-column DNA-digest was performed with RNase-free DNaseI (Qiagen). Complementary DNA was synthesized from 1 µg total RNA using the SuperScript™ VILO™ cDNA synthesis kit (ThermoFisher Scientific). Messenger RNA levels were determined by quantitative PCR using Fast SYBR® Green (ThermoFisher Scientific) and gene-specific primer pairs listed in Supplementary Table 2, using an Applied Biosystems ViiA 7 Real-Time PCR System (ThemoFisher Scientific). CT values of genes determined by quantitative PCR were normalized to actin and values displayed as fold changes of non-transgenic. Staining Three, 6- and 12-month-old TAU58/2 transgenic males (n = 9–13) and non-transgenic male littermates (n = 6–8) were used for histological analysis. At the specific ages, mice were anaesthetized and transcardially perfused with PBS (pH 7.4) to remove blood. Brains were removed, hemispheres separated and fixed in 4% paraformaldehyde. Immersion-fixed brains were further processed using an Excelsior tissue processor (ThermoFisher), embedded in paraffin and coronally sectioned at 3 μm for immunohistochemistry. One to two sections per mouse were stained with antibodies against the immediate early gene (IEG) markers Egr1 (Cell Signaling), Egr2 (Abcam), cFos (Abcam), JunB (Santa Cruz), Npy (Sigma) and ARC (Santa Cruz) and antibodies against human tau (Tau13, Santa Cruz) and tau phosphorylated at Ser396/404 (PHF-1, kind gift by P. Davies) as described previously (van Eersel et al., 2015). All brain sections were scanned using Axio Scan.Z1. ARC- and Npy-positive neurons were counted throughout the hippocampus, while neurons with visually bright staining signal for ARC were counted as Archigh cells separately to Arclow cells with less but clear immunoreactivity. Once Npy and ARC immunopositive neurons were identified and marked, each neuron was checked for double-labelling with either human tau or tau phosphorylated at Ser396/404. To determine differences in numbers of stained cells, total and double-labelled neurons were counted and the total number of positive and double-labelled neurons calculated from two sections per mouse, between stereotaxic levels −1.70 mm and −2.06 mm posterior relative to bregma (Paxinos and Franklin, 2001). Zen software 2.6 was used to delineate and automatically calculate the hippocampal area. All cell counts were converted to a density value (cells/µm2). To determine differences in staining intensity, ImageJ software was used to determine and analyse the intensity of Npy expression in the CA3 area of the hippocampus, following a published protocol (Palop et al., 2005). Between three and four small non-overlapping and randomly selected regions per slide were chosen within the CA3 region using ring-shaped regions of interest and mean pixel intensity was measured by averaging all pixels within the regions of interest. For Egr2, the regions of interest comprised the entire hippocampus. Mean pixel intensity was then corrected for baseline fluorescence, which was determined using three mean pixel intensity measures. For all quantitation, repeated measures on different days gave an inter- and intra-rater variability of <5%. Western blotting Western blotting and quantification of blots was done as previously described (Ke et al., 2019). Primary antibodies were to human tau (Tau13, Santa Cruz), pan (human + mouse) tau (Dako), tau phosphorylated at Ser214 (pS214, Abcam), Ser396/S404 (PHF1, gift from P. Davies) and Ser422 (pS422, Abcam) and Gapdh (Sigma). Bound primary antibodies were detected with species-specific HRP-labelled secondary antibodies and visualized with the HRP substrate (Bio-Rad) on a Chemidoc imager (Bio-Rad). Quantification was done with ImageJ (NIH). Full membranes are shown. Statistical analysis All statistical analysis was done using the Graphpad Prism 6.0 software (GraphPad, La Jolla, CA, USA) using either Student’s t-tests for comparison of two datasets, ANOVA for comparison of more than two datasets, repeated measures ANOVA across days/trials/recordings or two-way ANOVA for comparison across time. Data met the assumptions for parametric statistical testing. P-values <0.05 were considered significant. All values are presented as mean ± standard error of the mean (SEM) with presentation of individual values or as box and whisker blots (median, minimum and maximum). Data availability The data that support the findings of this study are available from the corresponding author, upon reasonable request. Results Spatial memory deficits in TAU58/2 mice TAU58/2 mice present with a progressive tau neuropathology (van Eersel et al., 2015), with expression of P301S mutant human tau (including in the hippocampus) and progressive accumulation of hyperphosphorylated tau (Supplementary Fig. 1). Total tau levels were 1.9-, 2.0- and 2.4-fold increased in 3-, 6- and 12-month-old TAU58/2 mice as compared to littermate controls (Supplementary Fig. 1). We have previously shown progressive behavioural and moderate motor deficits in TAU58/2 mice, which were of earlier onset and faster progression in males than females (van Eersel et al., 2015; Przybyla et al., 2016), but did not assess their cognitive performance. To determine whether P301S mutant human tau expression causes spatial memory deficits in TAU58/2 mice, we tested them in the Morris water maze at 2, 4 and 6 months of age, focusing on male mice only throughout the study. At 2 months of age, TAU58/2 mice and non-transgenic littermate controls presented with similar learning during 5 days of acquisition trials, as suggested by comparable swim paths and progressive reduction in escape latency over subsequent test days in the Morris water maze (Fig. 1A–C). Swim paths were categorized into non-spatial (thigmotaxis, random swim or scanning) and spatial (chaining, directed search, focal search or direct swim) escape strategies (Garthe and Kempermann, 2013; Tan et al., 2018). In line with comparable learning performance, escape strategies changed similarly from non-spatial to spatial learning in TAU58/2 mice and non-transgenic controls over the learning days (Fig. 1D and E). In contrast, 4- and 6-month-old TAU58/2 mice showed delayed learning when compared to non-transgenic littermate controls, as suggested by swim path comparison and significantly increased escape latencies, as well as delayed and incomplete conversion from non-spatial to spatial escape strategies (Fig. 1F–O). For comparison, TAU58/2 mice of all ages spent similar time in the target quadrants (Q1) during probe trials compared to non-transgenic controls, indicative of intact memory consolidation and retrieval (Supplementary Fig. 2). Average swim speed was not reduced in TAU58/2 mice compared to non-transgenic controls at all ages tested, indicating motor competency (Supplementary Fig. 1). In fact, swim speed was marginally increased in TAU58/2 mice comparted to non-transgenic littermates, significantly at 2 and 6 months of age, and a trend at 4 months. This is in line with increased activity of TAU58/2 mice during 24-h home cage monitoring (Supplementary Fig. 3) (van Eersel et al., 2015; Przybyla et al., 2016). In summary, TAU58/2 mice presented with spatial learning deficits from 4 months of age onwards. Figure 1 Open in new tabDownload slide Progressive spatial learning deficits in TAU58/2 mice. (A) Representative swim traces of 2-month-old TAU58/2 (yellow) and non-transgenic littermates (non-tg; grey) on Day 3 of Morris water maze (MWM) testing. Black lines = quadrant borders; small broken circle = position of submerged escape platform. (B) Average escape latency over five Morris water maze test days [ns = not significant; n = 9 TAU58/2, n = 9 non-transgenic; two-way ANOVA (Sidak post hoc)]. (C) AUC analysis of escape latency shown in B (Student’s t-test). (D) Relative distribution of different search strategies used by non-transgenic (top) and TAU58/2 (bottom) over five Morris water maze test days (n = 9 TAU58/2, n = 9 non-transgenic). (E) Learning performance: mean search strategy scores [non-spatial (1–3) and spatial (4–7) swim pattern scores see the ‘Materials and methods’ section] on Day 3 of Morris water maze testing (n = 9 TAU58/2, n = 9 non-transgenic; Student’s t-test). (F) Representative swim traces of 4-month-old TAU58/2 (red) and non-transgenic (grey) on Day 3 of Morris water maze testing. (G) Average escape latency over five Morris water maze test days [*P < 0.05; ***P < 0.001; n = 12 TAU58/2, n = 12 non-transgenic; two-way ANOVA (Sidak post hoc)]. (H) AUC analysis of escape latency shown in G (**P < 0.01; Student’s t-test). (I) Relative distribution of different search strategies used by non-transgenic (top) and TAU58/2 (bottom) over five Morris water maze test days (n = 12 TAU58/2, n = 12 non-transgenic). (J) Learning performance: mean search strategy scores on Day 3 of Morris water maze testing (**P < 0.01; n = 12 TAU58/2, n = 12 non-transgenic; Student’s t-test). (K) Representative swim traces of 6-month-old TAU58/2 (blue) and non-transgenic (grey) on Day 3 of Morris water maze testing. (L) Average escape latency over five Morris water maze test days [*P < 0.05; ***P < 0.001; n = 8 TAU58/2, n = 13 non-transgenic; two-way ANOVA (Sidak post hoc)]. (M) AUC analysis of escape latency shown in L (**P < 0.01; Student’s t-test). (N) Relative distribution of different search strategies used by non-transgenic (top) and TAU58/2 (bottom) over five Morris water maze test days (n = 8 TAU58/2, n = 13 non-transgenic). (O) Learning performance: mean search strategy scores on Day 3 of Morris water maze testing (**P < 0.01; n = 8 TAU58/2, n = 13 non-transgenic; Student’s t-test). Error bars represent the standard error. mo = months. Figure 1 Open in new tabDownload slide Progressive spatial learning deficits in TAU58/2 mice. (A) Representative swim traces of 2-month-old TAU58/2 (yellow) and non-transgenic littermates (non-tg; grey) on Day 3 of Morris water maze (MWM) testing. Black lines = quadrant borders; small broken circle = position of submerged escape platform. (B) Average escape latency over five Morris water maze test days [ns = not significant; n = 9 TAU58/2, n = 9 non-transgenic; two-way ANOVA (Sidak post hoc)]. (C) AUC analysis of escape latency shown in B (Student’s t-test). (D) Relative distribution of different search strategies used by non-transgenic (top) and TAU58/2 (bottom) over five Morris water maze test days (n = 9 TAU58/2, n = 9 non-transgenic). (E) Learning performance: mean search strategy scores [non-spatial (1–3) and spatial (4–7) swim pattern scores see the ‘Materials and methods’ section] on Day 3 of Morris water maze testing (n = 9 TAU58/2, n = 9 non-transgenic; Student’s t-test). (F) Representative swim traces of 4-month-old TAU58/2 (red) and non-transgenic (grey) on Day 3 of Morris water maze testing. (G) Average escape latency over five Morris water maze test days [*P < 0.05; ***P < 0.001; n = 12 TAU58/2, n = 12 non-transgenic; two-way ANOVA (Sidak post hoc)]. (H) AUC analysis of escape latency shown in G (**P < 0.01; Student’s t-test). (I) Relative distribution of different search strategies used by non-transgenic (top) and TAU58/2 (bottom) over five Morris water maze test days (n = 12 TAU58/2, n = 12 non-transgenic). (J) Learning performance: mean search strategy scores on Day 3 of Morris water maze testing (**P < 0.01; n = 12 TAU58/2, n = 12 non-transgenic; Student’s t-test). (K) Representative swim traces of 6-month-old TAU58/2 (blue) and non-transgenic (grey) on Day 3 of Morris water maze testing. (L) Average escape latency over five Morris water maze test days [*P < 0.05; ***P < 0.001; n = 8 TAU58/2, n = 13 non-transgenic; two-way ANOVA (Sidak post hoc)]. (M) AUC analysis of escape latency shown in L (**P < 0.01; Student’s t-test). (N) Relative distribution of different search strategies used by non-transgenic (top) and TAU58/2 (bottom) over five Morris water maze test days (n = 8 TAU58/2, n = 13 non-transgenic). (O) Learning performance: mean search strategy scores on Day 3 of Morris water maze testing (**P < 0.01; n = 8 TAU58/2, n = 13 non-transgenic; Student’s t-test). Error bars represent the standard error. mo = months. Reduced synaptic plasticity in TAU58/2 mice Impaired synaptic plasticity has been linked to memory deficits in the Morris water maze paradigm in mice (Barnhart et al., 2015). Given the onset of spatial memory deficits in TAU58/2 mice, and normal memory formation at 2 months of age (Fig. 1), we next performed electrophysiological recordings in the hippocampal CA1 area of acute brain slices from 2- and 4-month-old TAU58/2 mice and non-transgenic littermates to probe plasticity of fEPSPs. LTP of synaptic responses in the CA1 area were recorded following excitation of Schäffer collateral/commissural fibres with three trains of θ burst stimulus. Input-output (I/O) functions of stimulus intensity versus EPSP magnitude recorded prior to stimulation were not significantly different in TAU58/2 and non-transgenic slices at 2 and 4 months of age (Supplementary Fig. 4), and hence, comparable stimulus intensities were used for LTP induction. In slices of 2-month-old TAU58/2 mice LTP formation was slightly reduced as compared to non-transgenic control slices, but only linear regression analysis of fEPSP was significantly changed (Fig. 2A–C). At 4 months of age, however, LTP formation was consistently and significantly reduced in TAU58/2 mice compared to non-transgenic littermate controls (Fig. 2D–F). Taken together, these data suggest concomitant onset of reduced synaptic plasticity, memory and spatial learning deficits in TAU58/2 mice. Figure 2 Open in new tabDownload slide Impaired synaptic plasticity in TAU58/2 mice. (A) LTP formation in 2-month-old TAU58/2 (yellow) and non-transgenic littermates (non-tg; grey) acute brain slices (n = 9 TAU58/2, n = 7 non-transgenic). Average sample traces at baseline (dashed lines) and 60 min after stimulation (solid lines) are shown for TAU58/2 (yellow box) and non-transgenic slices (grey box). (B) AUC analysis of fEPSP slopes after stimulation shown in A (ns = not significant; Student’s t-test). (C) Linear regression analysis of fEPSP slopes shown in A (**P < 0.01; n = 9 TAU58/2, n = 7 non-transgenic; Student’s t-test). (D) LTP formation in 4-month-old TAU58/2 (red) and non-transgenic littermates (grey) acute brain slices (n = 6 TAU58/2, n = 7 non-transgenic). Average sample traces at baseline (dashed lines) and 60 min after stimulation (solid lines) are shown for TAU58/2 (red box) and non-transgenic slices (grey box). (E) AUC analysis of fEPSP slopes after stimulation shown in D (**P < 0.01; Student’s t-test). (F) Linear regression analysis of fEPSP slopes shown in D (****P < 0.0001; n = 6 TAU58/2, n = 7 non-transgenic; Student’s t-test). Error bars represent the standard error. mo = months; TBS = theta-burst stimulation. Figure 2 Open in new tabDownload slide Impaired synaptic plasticity in TAU58/2 mice. (A) LTP formation in 2-month-old TAU58/2 (yellow) and non-transgenic littermates (non-tg; grey) acute brain slices (n = 9 TAU58/2, n = 7 non-transgenic). Average sample traces at baseline (dashed lines) and 60 min after stimulation (solid lines) are shown for TAU58/2 (yellow box) and non-transgenic slices (grey box). (B) AUC analysis of fEPSP slopes after stimulation shown in A (ns = not significant; Student’s t-test). (C) Linear regression analysis of fEPSP slopes shown in A (**P < 0.01; n = 9 TAU58/2, n = 7 non-transgenic; Student’s t-test). (D) LTP formation in 4-month-old TAU58/2 (red) and non-transgenic littermates (grey) acute brain slices (n = 6 TAU58/2, n = 7 non-transgenic). Average sample traces at baseline (dashed lines) and 60 min after stimulation (solid lines) are shown for TAU58/2 (red box) and non-transgenic slices (grey box). (E) AUC analysis of fEPSP slopes after stimulation shown in D (**P < 0.01; Student’s t-test). (F) Linear regression analysis of fEPSP slopes shown in D (****P < 0.0001; n = 6 TAU58/2, n = 7 non-transgenic; Student’s t-test). Error bars represent the standard error. mo = months; TBS = theta-burst stimulation. Aberrant neuronal network activity in TAU58/2 mice To determine neuronal network activity in TAU58/2 mice in vivo, we implanted telemetric EEG transmitters into the hippocampus of 4- and 6-month-old mice to record hippocampal EEGs. Implantation of electrodes in younger mice is limited by their body size. EEG recordings were analysed for the presence of spontaneous hypersynchronicity and epileptiform discharges (hyperactivity), spectral power across wave frequencies and CFC, using our established algorithms (Ittner et al., 2014, 2016). Average numbers of spikes per hour in 24-h recordings were increased in 4-month-old TAU58/2 mice compared to non-transgenic controls (Fig. 3A and B). At this age, spectral power of θ waves were significantly reduced at 9–12 Hz in TAU58/2 mice as compared to non-transgenic littermate controls, while low frequency θ power (4–8 Hz) was similar during no-spike episodes of EEG recordings (Fig. 3C). In contrast, spectral power of low and high frequency γ-waves were significantly increased (Fig. 3D). Spike numbers were further increased at 6 months of age in TAU58/2 mice (Fig. 3E and F). In parallel, both low and high frequency θ power was reduced in TAU58/2 compared to non-transgenic controls at 6 months of age (Fig. 3G), while high frequency γ power remained increased (Fig. 3H). Next, we determined CFC of θ phase modulation of γ power during no-spike episodes of EEG recordings, a modality linked to memory formation including in humans (Canolty et al., 2006; Goutagny et al., 2009; Tort et al., 2009; Buzsaki and Moser, 2013). Four-month-old non-transgenic mice showed strong CFC in contrast to TAU58/2 mice with disrupted CFC (Fig. 3I). While θ phase amplitudes were similar in TAU58/2 mice and non-transgenic controls (Fig. 3J), the modulation index was significantly reduced in TAU58/2 mice (Fig. 3K). Six-month-old TAU58/2 mice also showed disrupted CFC of θ phase and γ amplitude (Fig. 3L). Furthermore, θ phase amplitude was reduced (Fig. 3M) and the modulation index was further reduced in TAU58/2 mice when compared to non-transgenic controls (Fig. 3N). Consistent with increased network activity, TAU58/2 mice challenged with PTZ presented with a reduced latency to develop seizures and higher mean seizure severity (Supplementary Fig. 5). Taken together, TAU58/2 mice showed hippocampal neuronal network aberrations already at 4 months of age, with further progression of changes as mice aged. Figure 3 Open in new tabDownload slide Neuronal network aberrations in TAU58/2 mice. (A) Examples of 3-min hippocampal EEG traces from 4-month-old non-transgenic (non-tg; top) and TAU58/2 (bottom) mice. Insets: Magnification of traces to demonstrate spike activity in TAU58/2 mice. (B) Number of spikes per hour in TAU58/2 (red) and non-transgenic (grey) mice during 24 h of EEG recording (**P < 0.01; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (C) Spectral power at low and high theta (θ) frequencies in TAU58/2 and non-transgenic mice (left). AUC analysis of low θ (4–8 Hz) (middle) and high θ (9–12 Hz) (right) (****P < 0.0001; ns = not significant; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (D) Spectral power at low and high gamma (γ) frequencies in TAU58/2 and non-transgenic mice (left). AUC analysis of low γ (30–60 Hz) (middle) and high γ (61–100 Hz) (right) (***P < 0.001; ****P < 0.0001; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (E) Examples of 3-min hippocampal EEG traces from 6-month-old non-transgenic (top) and TAU58/2 (bottom) mice. Insets: Magnification of traces to demonstrate spike activity in TAU58/2 mice. (F) Number of spikes per hour in TAU58/2 (blue) and non-transgenic (grey) mice during 24 h of EEG recording (**P < 0.01; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). (G) Spectral power at low and high θ in TAU58/2 and non-transgenic mice (left). AUC analysis of low θ (4–8 Hz) (middle) and high θ (9–12 Hz) (right) (***P < 0.001; ****P < 0.0001; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). (H) Spectral power at low and high γ in TAU58/2 and non-transgenic mice (left). AUC analysis of low γ (30–60 Hz) (middle) and high γ (61–100 Hz) (right) (**P < 0.01; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). (I) Representative phase-amplitude co-modulograms of interictal hippocampal EEG recordings showed CFC at ∼8 Hz in 4-month-old non-transgenic (left) but not in TAU58/2 (right) mice. (J) Phase-amplitude plot computed for interictal hippocampal EEG recordings (n = 4 TAU58/2, n = 4 non-transgenic). (K) Reduced modulation index in TAU58/2 mice compared with non-transgenic controls (**P < 0.01; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (L) Representative phase-amplitude co-modulograms of interictal hippocampal EEG recordings showed CFC at ∼7 Hz in 6-month-old non-transgenic (left) but not in TAU58/2 (right) mice. (M) Phase-amplitude plot computed for interictal hippocampal EEG recordings (n = 7 TAU58/2, n = 7 non-transgenic). (N) Reduced modulation index in TAU58/2 mice compared with non-transgenic controls (****P < 0.0001; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). Minimum–maximum whisker blots are shown. mo = months. Figure 3 Open in new tabDownload slide Neuronal network aberrations in TAU58/2 mice. (A) Examples of 3-min hippocampal EEG traces from 4-month-old non-transgenic (non-tg; top) and TAU58/2 (bottom) mice. Insets: Magnification of traces to demonstrate spike activity in TAU58/2 mice. (B) Number of spikes per hour in TAU58/2 (red) and non-transgenic (grey) mice during 24 h of EEG recording (**P < 0.01; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (C) Spectral power at low and high theta (θ) frequencies in TAU58/2 and non-transgenic mice (left). AUC analysis of low θ (4–8 Hz) (middle) and high θ (9–12 Hz) (right) (****P < 0.0001; ns = not significant; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (D) Spectral power at low and high gamma (γ) frequencies in TAU58/2 and non-transgenic mice (left). AUC analysis of low γ (30–60 Hz) (middle) and high γ (61–100 Hz) (right) (***P < 0.001; ****P < 0.0001; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (E) Examples of 3-min hippocampal EEG traces from 6-month-old non-transgenic (top) and TAU58/2 (bottom) mice. Insets: Magnification of traces to demonstrate spike activity in TAU58/2 mice. (F) Number of spikes per hour in TAU58/2 (blue) and non-transgenic (grey) mice during 24 h of EEG recording (**P < 0.01; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). (G) Spectral power at low and high θ in TAU58/2 and non-transgenic mice (left). AUC analysis of low θ (4–8 Hz) (middle) and high θ (9–12 Hz) (right) (***P < 0.001; ****P < 0.0001; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). (H) Spectral power at low and high γ in TAU58/2 and non-transgenic mice (left). AUC analysis of low γ (30–60 Hz) (middle) and high γ (61–100 Hz) (right) (**P < 0.01; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). (I) Representative phase-amplitude co-modulograms of interictal hippocampal EEG recordings showed CFC at ∼8 Hz in 4-month-old non-transgenic (left) but not in TAU58/2 (right) mice. (J) Phase-amplitude plot computed for interictal hippocampal EEG recordings (n = 4 TAU58/2, n = 4 non-transgenic). (K) Reduced modulation index in TAU58/2 mice compared with non-transgenic controls (**P < 0.01; n = 4 TAU58/2, n = 4 non-transgenic; Student’s t-test). (L) Representative phase-amplitude co-modulograms of interictal hippocampal EEG recordings showed CFC at ∼7 Hz in 6-month-old non-transgenic (left) but not in TAU58/2 (right) mice. (M) Phase-amplitude plot computed for interictal hippocampal EEG recordings (n = 7 TAU58/2, n = 7 non-transgenic). (N) Reduced modulation index in TAU58/2 mice compared with non-transgenic controls (****P < 0.0001; n = 7 TAU58/2, n = 7 non-transgenic; Student’s t-test). Minimum–maximum whisker blots are shown. mo = months. Immediate early gene response in TAU58/2 mice To determine early molecular changes in hippocampal neurons at an age before or at onset of functional deficits in TAU58/2 mice, we next performed gene expression profiling by RNA sequencing of hippocampal extracts at 10 weeks of age. Comparable RNA quality between samples was confirmed by total RNA integrity numbers of eight or higher. To account for the heterogeneity between samples and to increase sensitivity, differential expression analysis was carried out with a relaxed cut-off of >30% FDR revealing 44 differentially regulated genes (31 down- and 13 upregulated genes) (Fig. 4A and Supplementary Table 1). Note the marked upregulation of Mapt and Thy1 mRNA reflective of the transgenic construct used to generate TAU58/2 mice (van Eersel et al., 2015). Importantly, differentially regulated genes from TAU58/2 mice and non-transgenic controls clustered according to the genotypes, suggesting changes to gene expression are due to transgenic tau expression and/or tau pathology (Fig. 4B). When subjecting the list of differentially regulated genes to STRING analysis, we found only two significant clusters; the major cluster, IEGs, included 7 of the 13 upregulated genes, while none of the downregulated genes mapped to this cluster, indicating increased overall activity (Fig. 4C). Additional genes, not part of the significantly differentially regulated genes detected by RNA sequencing, were predicted to be associated with this cluster, including Arc and Fos. Interestingly, both were upregulated in RNA sequencing, but did not reach significance after multiple hypothesis testing. One of the remaining upregulated genes, Nr1d1, mapped to an associated cluster of circadian rhythm genes. Notably, none of the downregulated genes were annotated to gene clusters. DAVID analysis identified behaviour and synaptic plasticity as highest ranked significant GO terms within the synaptic signalling and function annotation cluster, in line with the phenotypical presentation of TAU58/2 mice (Supplementary Fig. 6). Other enriched annotation clusters include SH3 domain-cell junction and plasma membrane, suggesting altered signalling processes. Quantitative RT-PCR validated differential expression of selected IEGs Npy and Egr2 in TAU58/2 mice at 3 months of age (Fig. 4D). Furthermore, Arc mRNA was significantly increased and both FosB and cFos showed a trend to increased mRNA levels in TAU58/2 mice compared to non-transgenic controls. Quantification of staining of brain sections from TAU58/2 mice and non-transgenic controls at 3, 6 and 12 months of age for differentially regulated genes Egr1, Egr2, Npy and JunB, and cFos showed significant increases in staining intensity and/or numbers of stained cells in hippocampal regions at distinct ages (Fig. 4E and Supplementary Fig. 7). Number of Egr1-positive cells significantly and progressively increased with age of TAU58/2 mice. Quantification furthermore showed that intensity of widespread hippocampal Egr2 staining was significantly increased in 6-month-old TAU58/2 mice and showed trends towards increased levels at 3 and 12 months. Numbers of Npy-positive neurons were significantly increased in 3- and 6-month-old TAU58/2 mice compared to non-transgenic controls, while numbers were comparable at 12 months of age between TAU58/2 mice and littermate controls. Similarly, Npy staining intensity of the hippocampal CA3 region was increased in TAU58/2 mice at all ages, although these changes were only significant in the 6 month age group. Frequency of JunB-positive cells only significantly increased in 12-month-old TAU58/2 mice. Finally, the density of cFos expressing cells was increased the dentate gyrus of TAU58/2 mice at all ages, but this reached significance only for the 12 month cohort. Taken together, TAU58/2 mice presented with IEG signature consistent with increased neuronal activity. Figure 4 Open in new tabDownload slide IEG signature in TAU58/2 mice. (A) RNA sequencing (RNAseq): volcano blot of differentially up- (red) and downregulated (blue) mRNAs in the hippocampus of 10-week-old TAU58/2 compared to non-transgenic (non-tg) littermates (n = 4). (B) Heat map shows clustering of differentially up- (red) and downregulated (blue) gene expression in non-transgenic (n = 4; grey) and TAU58/2 (n = 4; orange) samples. (C) STRING analysis reveals two clusters for immediate early genes (green) and circadian rhythm genes (yellow). Genes not annotated to clusters are presented individually (left). Circle size reflects level of differential regulation; red = upregulation; blue = downregulation; grey = genes annotated to cluster by STRING; lines = gene interaction; green lines = interaction between genes obtained by RNAseq analysis; black dashed lines = interaction between genes annotated to cluster by STRING. (D) Hippocampal gene expression at 3 months of age in TAU58/2 (coloured bars; red = genes found by RNAseq analysis; grey = genes annotated by STRING) relative to non-transgenic (white bars) mRNA levels as determined by quantitative PCR (*P = 0.05; **P < 0.01; triplicates of n = 7 TAU58/2, n = 5 non-transgenic; Student’s t-test). (E) Quantification of immunofluorescence staining of 3-, 6- and 12-month-old TAU58/2 and non-transgenic hippocampus; numbers of Egr1-, Npy- and JunB-positive neurons in the hippocampal CA1 region and of cFos in the hippocampal dentate gyrus, as well as Egr2 (entire hippocampus) and Npy (CA3 region) staining intensity in the hippocampus of TAU58/2 and non-transgenic mice at indicated ages (*P < 0.05; **P < 0.01; ****P < 0.0001; ns = not significant; n = 13–19 TAU58/2, n = 6–12 non-transgenic; Student’s t-test). Error bars represent standard error. Figure 4 Open in new tabDownload slide IEG signature in TAU58/2 mice. (A) RNA sequencing (RNAseq): volcano blot of differentially up- (red) and downregulated (blue) mRNAs in the hippocampus of 10-week-old TAU58/2 compared to non-transgenic (non-tg) littermates (n = 4). (B) Heat map shows clustering of differentially up- (red) and downregulated (blue) gene expression in non-transgenic (n = 4; grey) and TAU58/2 (n = 4; orange) samples. (C) STRING analysis reveals two clusters for immediate early genes (green) and circadian rhythm genes (yellow). Genes not annotated to clusters are presented individually (left). Circle size reflects level of differential regulation; red = upregulation; blue = downregulation; grey = genes annotated to cluster by STRING; lines = gene interaction; green lines = interaction between genes obtained by RNAseq analysis; black dashed lines = interaction between genes annotated to cluster by STRING. (D) Hippocampal gene expression at 3 months of age in TAU58/2 (coloured bars; red = genes found by RNAseq analysis; grey = genes annotated by STRING) relative to non-transgenic (white bars) mRNA levels as determined by quantitative PCR (*P = 0.05; **P < 0.01; triplicates of n = 7 TAU58/2, n = 5 non-transgenic; Student’s t-test). (E) Quantification of immunofluorescence staining of 3-, 6- and 12-month-old TAU58/2 and non-transgenic hippocampus; numbers of Egr1-, Npy- and JunB-positive neurons in the hippocampal CA1 region and of cFos in the hippocampal dentate gyrus, as well as Egr2 (entire hippocampus) and Npy (CA3 region) staining intensity in the hippocampus of TAU58/2 and non-transgenic mice at indicated ages (*P < 0.05; **P < 0.01; ****P < 0.0001; ns = not significant; n = 13–19 TAU58/2, n = 6–12 non-transgenic; Student’s t-test). Error bars represent standard error. Immediate early gene induction locates to tau pathology harbouring neurons in TAU58/2 mice Next, we determined whether transcriptional activation was a generalized neuronal event in TAU58/2 mice or if it was directly linked to tau expression and/or pathology. Therefore, we stained brain sections from 3-, 6- and 12-month-old TAU58/2 mice and non-transgenic littermates for ARC in combination with tau antibodies. ARC has been shown to stain hyperexcited neurons, for example during seizures or targeted network activation (Palop et al., 2005). Low but significantly increased numbers of ARC-positive neurons were detected in the hippocampus of TAU58/2 mice at 3 months of age compared to non-transgenic littermate controls (Fig. 5A and B), consistent with differentially regulated Arc mRNA at this age (Fig. 4D). At 3 months of age, cells staining intensively for ARC (Archigh) were only detected in TAU58/2 mice (Fig. 5B). Over 80% of Archigh cells co-labelled with the staining for transgenic human tau (Tau13) and for tau phosphorylated at the late-stage sites S396 and S404 (PHF-1), while cells that labelled less intensively for ARC (Arclow) only showed co-labelling with Tau13 and PHF-1 in ∼40% of cells (Fig. 5A and C). Arclow cells were also occasionally found in the hippocampus of non-transgenic littermate controls (Fig. 5B). Numbers of both Arclow and Archigh cells further increased in 6-month-old TAU58/2 mice compared to non-transgenic controls (Fig. 5D and E), with high co-labelling of ARC with Tau13 and PHF-1 (Fig. 5F). Similarly, 12-month-old TAU58/2 mice presented with higher numbers of ARC-positive cells as compared to aged non-transgenic controls (Fig. 5G and H). Almost all Archigh cells co-labelled with Tau13 and >90% with PHF-1 at this age (Fig. 5I). Taken together, progressively increased labelling of neurons with the IEG maker protein ARC was largely confined to hippocampal neurons that harboured tau pathology. Figure 5 Open in new tabDownload slide Progressive ARC accumulation in transgenic tau-harbouring neurons in TAU58/2 mice. (A) Representative immunofluorescence co-staining of ARC (green) and Tau13 (red; top) or PHF-1 (red; bottom) in hippocampal sections of TAU58/2 and non-transgenic (non-tg) mice at 3 months of age. Scale bar = 200 µm for all images. Insets: Magnification of areas indicated by broken boxes. Filled white arrowheads indicate cells with high ARC staining intensity (=Archigh) and ARC/Tau13 or ARC/PHF-1 co-labelling. Filled orange arrowheads indicate cells with low ARC staining (=Arclow) and ARC/Tau13 or ARC/PHF-1 co-labelling. (B) Numbers of cells with low ARC staining (Arclow, filled bars) and of those with intensive ARC staining (Archigh; open bars) in non-transgenic (grey) and TAU58/2 (yellow) hippocampi at 3 months of age (*P < 0.05; **P < 0.01; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). (C) Percentage of low ARC-positive (Arclow, filled bars) and Archigh (open bars) cells that co-label with Tau13 (top) and PHF-1 (bottom) in TAU58/2 hippocampi at 3 months of age (*P < 0.05; **P < 0.01; ****P < 0.0001 versus non-transgenic; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). (D) Representative immunofluorescence co-staining of ARC (green) and Tau13 (red; top) or PHF-1 (red; bottom) in hippocampal sections of TAU58/2 and non-transgenic mice at 6 months of age. Insets: Magnification of areas indicated by broken boxes. Filled white arrowheads indicate cells with ARC/Tau13 or ARC/PHF-1 co-labelling. Filled white arrowheads indicate ARC-positive cells that do not co-label with Tau13 or PHF-1. (E) Numbers of cells with low ARC staining (Arclow, filled bars) and of those with intensive ARC staining (Archigh; open bars) in non-transgenic (grey) and TAU58/2 (blue) hippocampi at 6 months of age (**P < 0.01; n = 11 TAU58/2, n = 8 non-transgenic; Student’s t-test). (F) Percentage of low ARC-positive (Arclow, filled bars) and Archigh (open bars) cells that co-label with Tau13 (top) and PHF-1 (bottom) in TAU58/2 hippocampi at 6 months of age (***P < 0.001; ****P < 0.0001 versus non-transgenic; n = 11 TAU58/2, n = 8 non-transgenic; Student’s t-test). (G) Representative immunofluorescence co-staining of ARC (green) and Tau13 (red; top) or PHF-1 (red; bottom) in hippocampal sections of TAU58/2 and non-transgenic mice at 12 months of age. Insets: Magnification of areas indicated by broken boxes. Filled white arrowheads indicate cells with ARC/Tau13 or ARC/PHF-1 co-labelling. Filled white arrowheads indicate ARC-positive cells that do not co-label with Tau13 or PHF-1. (H) Numbers of cells with low ARC staining (Arclow, filled bars) and of those with intensive ARC staining (Archigh; open bars) in non-transgenic (grey) and TAU58/2 (green) hippocampi (**P < 0.01; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). (I) Percentage of low ARC-positive (Arclow, filled bars) and Archigh (open bars) cells that co-label with Tau13 (top) and PHF-1 (bottom) in TAU58/2 hippocampi (**P < 0.01; ****P < 0.0001 versus non-transgenic; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). Error bars represent the standard error. Scale bar = 200 μm. Figure 5 Open in new tabDownload slide Progressive ARC accumulation in transgenic tau-harbouring neurons in TAU58/2 mice. (A) Representative immunofluorescence co-staining of ARC (green) and Tau13 (red; top) or PHF-1 (red; bottom) in hippocampal sections of TAU58/2 and non-transgenic (non-tg) mice at 3 months of age. Scale bar = 200 µm for all images. Insets: Magnification of areas indicated by broken boxes. Filled white arrowheads indicate cells with high ARC staining intensity (=Archigh) and ARC/Tau13 or ARC/PHF-1 co-labelling. Filled orange arrowheads indicate cells with low ARC staining (=Arclow) and ARC/Tau13 or ARC/PHF-1 co-labelling. (B) Numbers of cells with low ARC staining (Arclow, filled bars) and of those with intensive ARC staining (Archigh; open bars) in non-transgenic (grey) and TAU58/2 (yellow) hippocampi at 3 months of age (*P < 0.05; **P < 0.01; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). (C) Percentage of low ARC-positive (Arclow, filled bars) and Archigh (open bars) cells that co-label with Tau13 (top) and PHF-1 (bottom) in TAU58/2 hippocampi at 3 months of age (*P < 0.05; **P < 0.01; ****P < 0.0001 versus non-transgenic; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). (D) Representative immunofluorescence co-staining of ARC (green) and Tau13 (red; top) or PHF-1 (red; bottom) in hippocampal sections of TAU58/2 and non-transgenic mice at 6 months of age. Insets: Magnification of areas indicated by broken boxes. Filled white arrowheads indicate cells with ARC/Tau13 or ARC/PHF-1 co-labelling. Filled white arrowheads indicate ARC-positive cells that do not co-label with Tau13 or PHF-1. (E) Numbers of cells with low ARC staining (Arclow, filled bars) and of those with intensive ARC staining (Archigh; open bars) in non-transgenic (grey) and TAU58/2 (blue) hippocampi at 6 months of age (**P < 0.01; n = 11 TAU58/2, n = 8 non-transgenic; Student’s t-test). (F) Percentage of low ARC-positive (Arclow, filled bars) and Archigh (open bars) cells that co-label with Tau13 (top) and PHF-1 (bottom) in TAU58/2 hippocampi at 6 months of age (***P < 0.001; ****P < 0.0001 versus non-transgenic; n = 11 TAU58/2, n = 8 non-transgenic; Student’s t-test). (G) Representative immunofluorescence co-staining of ARC (green) and Tau13 (red; top) or PHF-1 (red; bottom) in hippocampal sections of TAU58/2 and non-transgenic mice at 12 months of age. Insets: Magnification of areas indicated by broken boxes. Filled white arrowheads indicate cells with ARC/Tau13 or ARC/PHF-1 co-labelling. Filled white arrowheads indicate ARC-positive cells that do not co-label with Tau13 or PHF-1. (H) Numbers of cells with low ARC staining (Arclow, filled bars) and of those with intensive ARC staining (Archigh; open bars) in non-transgenic (grey) and TAU58/2 (green) hippocampi (**P < 0.01; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). (I) Percentage of low ARC-positive (Arclow, filled bars) and Archigh (open bars) cells that co-label with Tau13 (top) and PHF-1 (bottom) in TAU58/2 hippocampi (**P < 0.01; ****P < 0.0001 versus non-transgenic; n = 9 TAU58/2, n = 6 non-transgenic; Student’s t-test). Error bars represent the standard error. Scale bar = 200 μm. Discussion In the present study, we showed that TAU58/2 mice develop progressive spatial learning deficits that are associated with impaired LTP, neuronal network aberrations and an IEG expression signature that is indicative of neuronal hyperexcitation. Importantly, staining with excitation marker protein ARC was largely confined to neurons harbouring hyperphosphorylated tau, suggesting that tau drives neuronal hyperexcitation via cell intrinsic processes. Previous studies in alternate mutant tau transgenic models showed altered cortical neuronal networks during electrophysiological and EEG recordings (Crimins et al., 2012; Garcia-Cabrero et al., 2013; Menkes-Caspi et al., 2015; Holth et al., 2017; Das et al., 2018; Van Erum et al., 2020). For comparison, we show hyperexcitation, spectral power changes and compromised CFC of hippocampal neuronal networks in TAU58/2 mice. In line with these findings, TAU58/2 were more susceptible to induced excitotoxic seizures, a phenotype previously reported for other tau transgenic lines (Garcia-Cabrero et al., 2013; Liu et al., 2017; Van Erum et al., 2020). We have previously reported an array of behavioural and motor deficits in TAU58/2 mice (van Eersel et al., 2015; Przybyla et al., 2016); TAU58/2 mice present with early onset disinhibition and moderate motor deficits. These behavioural changes align with phenotypes reported for another widely used P301S mutant tau transgenic line, PS19 (Yoshiyama et al., 2007; Takeuchi et al., 2011). New data shown here demonstrate progressive learning deficits in TAU58/2 mice, extending their value for studying molecular and cellular processes driven by tau pathology in cognitive decline. To which extent motor impairments and/or hyperactivity may have contributed to learning deficits of TAU58/2 mice remains to be fully examined. However, motor deficits had no impact on disinhibition testing in previous studies (Przybyla et al., 2016; Van der Jeugd et al., 2016). Similarly, swim speed during Morris water maze testing was increased, rather than decreased in TAU58/2 mice compared to non-transgenic littermates, suggesting that motor deficits were not contributing to the progressively reduced performance as mice aged. PS19 mice similarly present with reduced LTP formation and learning deficits during Morris water maze testing but, in contrast to TAU58/2 mice, fail to retain memory (Takeuchi et al., 2011; Lasagna-Reeves et al., 2016). This difference may be explained by the pronounced neuronal loss in PS19, which is not observed in TAU58/2 brains (Yoshiyama et al., 2007; van Eersel et al., 2015). Furthermore, PS19 and TAU58/2 mice differ in promotors used to drive transgene expression (mouse prion promotor versus mouse Thy1.2 promoter) and human tau isoforms expressed (1N4R versus 0N4R), as well as levels of transgenic tau (5-fold versus 2-fold overexpression). Hence, different tau isoforms and/or expression levels/patterns may impact differentially on pathways that modulate memory formation and retention. Furthermore, it is possible that deficits in both behavioural and memory tasks reflect similar underlying disease mechanisms affecting different neuronal networks. Gene expression profiling in TAU58/2 mice at/prior to the onset of LTP, memory and neuronal network deficits revealed only one significant cluster of differentially regulated genes, IEGs. IEGs are expressed in response to neuronal hyperexcitation, including in mouse models of acute brain damage (e.g. epilepsy or stroke) (Bi et al., 2017). The prominent IEG cluster indicates a coordinated hyperactivation of neuronal circuits in TAU58/2, while pathways that lead to downregulation of genes were diverse and mechanistically unrelated. Together, this suggests that neuronal dysfunction due to hyperexcitation is a major contributor to the onset of functional deficits in TAU58/2. However, a direct link between LTP deficits and network hyperactivity remains to be established. Furthermore, it remains to be shown whether the second associated cluster of circadian rhythm genes is a result of neuronal network hyperactivity or independently contributes to functional deficits in TAU58/2 mice. Here, TAU58/2 mice show increased activity, including during normal resting times. Furthermore, we formally cannot rule out that other processes that contribute to the onset of functional deficits are under-represented in the gene profiling because of linear signalling pathways (i.e. pathway activity is reflected by differential regulation of a single gene rather than clusters). To our knowledge, previous transcriptomic studies did not determine differential mRNA expression in the hippocampus of tau transgenic mice at or prior to the onset of pathology and phenotypic changes. However, mRNA profiling of young, 2-month-old rTg4510 mice, that express early on very high P301L mutant tau (0N4R) levels and develop rapid pathology (Santacruz et al., 2005), neurodegeneration and functional deficits, identified differentially regulated gene clusters associated with neuroinflammation and inflammatory response (Sharma et al., 2018). Notably, IEGs that we found upregulated in TAU58/2 were downregulated in rTg4510 mice at 2 months, possibly reflecting different disease courses and stages (Sharma et al., 2018). For comparison, transcriptomic analysis of young, 3-month-old TRP50 mice, which express P301S mutant tau (2N4R) driven by the prion promoter and present more comparable disease progression to the TAU58/2 line but show moderate neuronal loss (Onishi et al., 2014), identified a synaptic gene module in addition to an inflammatory response gene cluster (Swarup et al., 2019). Hence, differences in promoters, tau isoforms and phenotype onset/progression between models impact on gene profiles. Our data may suggest that the IEG signature of TAU58/2 reflects disease pathways involved in or prior to onset and slower progression of pathology and deficits without overt neuronal loss. The IEG marker protein ARC co-localized with hyperphosphorylated tau in hippocampal neurons, suggesting that tau pathology renders neurons prone to responding with hyperexcitation to network activity, in turn compromising the function of entire networks and resulting in functional deficits, including progressive memory decline. Interestingly, a single nucleotide polymorphism in ARC has been associated with a reduced risk for Alzheimer’s disease (Landgren et al., 2012). Furthermore, ARC contributes to LTP (Bramham et al., 2010; Carmichael and Henley, 2018). Hence, a progressive increase in ARC in tau pathology-harbouring neurons may therefore not only be a marker of hyperexcitation (Palop et al., 2005), but may possibly be directly involved in pathways that contribute to neuronal dysfunction. A previous study has reported reduced hippocampal ARC activity in P301L mutant tau transgenic mice (Fox et al., 2011); however, ARC activation was only assessed in the context of targeted stimulation, different from our study that found increased numbers of ARC positive neurons in non-stimulated TAU58/2 mice. Accordingly, it has been shown that network-wide activity-driven ARC expression was not affected by tau pathology (Rudinskiy et al., 2014). To this end, it remains to be shown why only a subset of tau pathology-harbouring neurons increase ARC levels, and whether their aberrant activity is enough to confer neuronal network dysfunction. Interestingly, neuronal network hyperactivity in APP transgenic Alzheimer’s disease mice was similarly associated with ARC staining of only a subset of hippocampal neurons (Palop et al., 2005). This may suggest a similar contribution by selected neurons to neuronal network changes in APP and tau transgenic models, given the connection between tau and amyloid-β in network dysfunction in Alzheimer’s disease (Ittner and Ittner, 2018). Finally, the temporal contribution of Archigh neurons to the onset and progression of neuronal network aberrations and memory deficits requires further investigation. We have previously shown that tau is a mediator of neuronal network hypersynchronicity in amyloid-β-based mouse models of Alzheimer’s disease (Ittner et al., 2010, 2016; Ittner and Gotz, 2011a; Ittner and Ittner, 2018). In line with work by others (Hoover et al., 2010), the present study suggests that pathological tau on its own is possibly at the root of dysfunctional plasticity within neuronal networks driving neurodegenerative diseases. Therefore, the idea of targeting neuronal network hypersynchronicity in Alzheimer’s disease may also be translatable to tau-only neurodegenerative conditions such as FTD. Acknowledgements The authors would like to thank the staff of the animal facilities at Macquarie University and UNSW for their help with animal husbandry. Funding The authors received funding from the National Health and Medical Research Council (1081916, 1123564, 1132524, 1136241, 1143848, 1143978, 1102012, 1141789, 1095215), the Australian Research Council (DP150104321, DP170100781, DP170100843) and CurePSP (609-2015-07). Y.K. is a NHMRC Career Development Fellow. M.K. and L.I. are NHMRC Principal Research Fellows. Competing interests The authors report no competing interests. References Ahmed RM , Irish M, van Eersel J, Ittner A, Ke YD, Volkerling A, et al. Mouse models of frontotemporal dementia: a comparison of phenotypes with clinical symptomatology . Neurosci Biobehav Rev 2017 ; 74 : 126 – 38 . Google Scholar Crossref Search ADS PubMed WorldCat Allen B , Ingram E, Takao M, Smith MJ, Jakes R, Virdee K, et al. Abundant tau filaments and nonapoptotic neurodegeneration in transgenic mice expressing human P301S tau protein . J Neurosci 2002 ; 22 : 9340 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat Ballatore C , Lee VM, Trojanowski JQ. Tau-mediated neurodegeneration in Alzheimer’s disease and related disorders . Nat Rev Neurosci 2007 ; 8 : 663 – 72 . Google Scholar Crossref Search ADS PubMed WorldCat Barnhart CD , Yang D, Lein PJ. Using the Morris water maze to assess spatial learning and memory in weanling mice . PLoS One 2015 ; 10 : e0124521 . Google Scholar Crossref Search ADS PubMed WorldCat Bi M , Gladbach A, van Eersel J, Ittner A, Przybyla M, van Hummel A, et al. Tau exacerbates excitotoxic brain damage in an animal model of stroke . Nat Commun 2017 ; 8 : 473 . Google Scholar Crossref Search ADS PubMed WorldCat Bramham CR , Alme MN, Bittins M, Kuipers SD, Nair RR, Pai B, et al. The Arc of synaptic memory . Exp Brain Res 2010 ; 200 : 125 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat Busche MA , Wegmann S, Dujardin S, Commins C, Schiantarelli J, Klickstein N, et al. Tau impairs neural circuits, dominating amyloid-beta effects, in Alzheimer models in vivo . Nat Neurosci 2019 ; 22 : 57 – 64 . Google Scholar Crossref Search ADS PubMed WorldCat Buzsaki G , Moser EI. Memory, navigation and theta rhythm in the hippocampal-entorhinal system . Nat Neurosci 2013 ; 16 : 130 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Canolty RT , Edwards E, Dalal SS, Soltani M, Nagarajan SS, Kirsch HE, et al. High gamma power is phase-locked to theta oscillations in human neocortex . Science 2006 ; 313 : 1626 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Carmichael RE , Henley JM. Transcriptional and post-translational regulation of Arc in synaptic plasticity . Semin Cell Dev Biol 2018 ; 77 : 3 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Clare R , King VG, Wirenfeldt M, Vinters HV. Synapse loss in dementias . J Neurosci Res 2010 ; 88 : 2083 – 90 . Google Scholar Crossref Search ADS PubMed WorldCat Crimins JL , Rocher AB, Luebke JI. Electrophysiological changes precede morphological changes to frontal cortical pyramidal neurons in the rTg4510 mouse model of progressive tauopathy . Acta Neuropathol 2012 ; 124 : 777 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat Das M , Maeda S, Hu B, Yu GQ, Guo W, Lopez I, et al. Neuronal levels and sequence of tau modulate the power of brain rhythms . Neurobiol Dis 2018 ; 117 : 181 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Fa M , Puzzo D, Piacentini R, Staniszewski A, Zhang H, Baltrons MA, et al. Extracellular tau oligomers produce an immediate impairment of LTP and memory. Sci Rep 2016 ; 6 : 19393 . Google Scholar Crossref Search ADS PubMed WorldCat Fox LM , William CM, Adamowicz DH, Pitstick R, Carlson GA, Spires-Jones TL, et al. Soluble tau species, not neurofibrillary aggregates, disrupt neural system integration in a tau transgenic model . J Neuropathol Exp Neurol 2011 ; 70 : 588 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat Garcia-Cabrero AM , Guerrero-Lopez R, Giraldez BG, Llorens-Martin M, Avila J, Serratosa JM, et al. Hyperexcitability and epileptic seizures in a model of frontotemporal dementia . Neurobiol Dis 2013 ; 58 : 200 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Garthe A , Kempermann G. An old test for new neurons: refining the Morris water maze to study the functional relevance of adult hippocampal neurogenesis . Front Neurosci 2013 ; 7 : 63 . Google Scholar Crossref Search ADS PubMed WorldCat Gendron TF , Petrucelli L. The role of tau in neurodegeneration . Mol Neurodegener 2009 ; 4 : 13 . Google Scholar Crossref Search ADS PubMed WorldCat Goetzl EJ , Kapogiannis D, Schwartz JB, Lobach IV, Goetzl L, Abner EL, et al. Decreased synaptic proteins in neuronal exosomes of frontotemporal dementia and Alzheimer’s disease . FASEB J 2016 ; 30 : 4141 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Gotz J , Chen F, Barmettler R, Nitsch RM. Tau filament formation in transgenic mice expressing P301L tau . J Biol Chem 2001 ; 276 : 529 – 34 . Google Scholar Crossref Search ADS PubMed WorldCat Gotz J , Ittner A, Ittner LM. Tau-targeted treatment strategies in Alzheimer’s disease . Br J Pharmacol 2012 ; 165 : 1246 – 59 . Google Scholar Crossref Search ADS PubMed WorldCat Gotz J , Ittner LM. Animal models of Alzheimer’s disease and frontotemporal dementia . Nat Rev Neurosci 2008 ; 9 : 532 – 44 . Google Scholar Crossref Search ADS PubMed WorldCat Goutagny R , Jackson J, Williams S. Self-generated theta oscillations in the hippocampus . Nat Neurosci 2009 ; 12 : 1491 – 3 . Google Scholar Crossref Search ADS PubMed WorldCat Holth JK , Mahan TE, Robinson GO, Rocha A, Holtzman DM. Altered sleep and EEG power in the P301S Tau transgenic mouse model . Ann Clin Transl Neurol 2017 ; 4 : 180 – 90 . Google Scholar Crossref Search ADS PubMed WorldCat Hoover BR , Reed MN, Su J, Penrod RD, Kotilinek LA, Grant MK, et al. Tau mislocalization to dendritic spines mediates synaptic dysfunction independently of neurodegeneration . Neuron 2010 ; 68 : 1067 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat Hornberger M , Piguet O. Episodic memory in frontotemporal dementia: a critical review . Brain 2012 ; 135 : 678 – 92 . Google Scholar Crossref Search ADS PubMed WorldCat Hornberger M , Piguet O, Graham AJ, Nestor PJ, Hodges JR. How preserved is episodic memory in behavioral variant frontotemporal dementia? Neurology 2010 ; 74 : 472 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Ittner A , Chua SW, Bertz J, Volkerling A, van der Hoven J, Gladbach A, et al. Site-specific phosphorylation of tau inhibits amyloid-beta toxicity in Alzheimer’s mice . Science 2016 ; 354 : 904 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Ittner LM , Fath T, Ke YD, Bi M, van Eersel J, Li KM, et al. Parkinsonism and impaired axonal transport in a mouse model of frontotemporal dementia . Proc Natl Acad Sci U S A 2008 ; 105 : 15597 – 6002 . Google Scholar Crossref Search ADS WorldCat Ittner AA , Gladbach A, Bertz J, Suh LS, Ittner LM. p38 MAP kinase-mediated NMDA receptor-dependent suppression of hippocampal hypersynchronicity in a mouse model of Alzheimer inverted question marks disease . Acta Neuropathol Commun 2014 ; 2 : 149 . Google Scholar Crossref Search ADS PubMed WorldCat Ittner LM , Gotz J. Amyloid-beta and tau - a toxic pas de deux in Alzheimer’s disease . Nat Rev Neurosci 2011 a; 12 : 67 – 72 . Google Scholar Crossref Search ADS WorldCat Ittner LM , Gotz J. Amyloid-beta and tau–a toxic pas de deux in Alzheimer’s disease . Nat Rev Neurosci 2011 b; 12 : 65 – 72 . Google Scholar Crossref Search ADS PubMed WorldCat Ittner A , Ittner LM. Dendritic Tau in Alzheimer’s Disease . Neuron 2018 ; 99 : 13 – 27 . Google Scholar Crossref Search ADS PubMed WorldCat Ittner LM , Ke YD, Delerue F, Bi M, Gladbach A, van Eersel J, et al. Dendritic function of tau mediates amyloid-beta toxicity in Alzheimer’s disease mouse models . Cell 2010 ; 142 : 387 – 97 . Google Scholar Crossref Search ADS PubMed WorldCat Ke YD , Chan G, Stefanoska K, Au C, Bi M, Muller J, et al. CNS cell type-specific gene profiling of P301S tau transgenic mice identifies genes dysregulated by progressive tau accumulation . J Biol Chem 2019 ; 294 : 14149 – 62 . Google Scholar Crossref Search ADS PubMed WorldCat Landgren S , von Otter M, Palmer MS, Zetterstrom C, Nilsson S, Skoog I, et al. A novel ARC gene polymorphism is associated with reduced risk of Alzheimer’s disease . J Neural Transm 2012 ; 119 : 833 – 42 . Google Scholar Crossref Search ADS PubMed WorldCat Lasagna-Reeves CA , de Haro M, Hao S, Park J, Rousseaux MW, Al-Ramahi I, et al. Reduction of Nuak1 decreases tau and reverses phenotypes in a tauopathy mouse model . Neuron 2016 ; 92 : 407 – 18 . Google Scholar Crossref Search ADS PubMed WorldCat Lewis J , McGowan E, Rockwood J, Melrose H, Nacharaju P, Van Slegtenhorst M, et al. Neurofibrillary tangles, amyotrophy and progressive motor disturbance in mice expressing mutant (P301L) tau protein . Nat Genet 2000 ; 25 : 402 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat Liu S , Shen Y, Shultz SR, Nguyen A, Hovens C, Adlard PA, et al. Accelerated kindling epileptogenesis in Tg4510 tau transgenic mice, but not in tau knockout mice . Epilepsia 2017 ; 58 : e136 – e41 . Google Scholar Crossref Search ADS PubMed WorldCat Mandelkow EM , Mandelkow E. Biochemistry and cell biology of tau protein in neurofibrillary degeneration . Cold Spring Harb Perspect Med 2012 ; 2 : a006247 . Google Scholar Crossref Search ADS PubMed WorldCat Menkes-Caspi N , Yamin HG, Kellner V, Spires-Jones TL, Cohen D, Stern EA. Pathological tau disrupts ongoing network activity . Neuron 2015 ; 85 : 959 – 66 . Google Scholar Crossref Search ADS PubMed WorldCat Onishi T , Matsumoto Y, Hattori M, Obayashi Y, Nakamura K, Yano T, et al. Early-onset cognitive deficits and axonal transport dysfunction in P301S mutant tau transgenic mice . Neurosci Res 2014 ; 80 : 76 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat Palop JJ , Chin J, Bien-Ly N, Massaro C, Yeung BZ, Yu GQ, et al. Vulnerability of dentate granule cells to disruption of arc expression in human amyloid precursor protein transgenic mice . J Neurosci 2005 ; 25 : 9686 – 93 . Google Scholar Crossref Search ADS PubMed WorldCat Paxinos G , Franklin K. The mouse brain in stereotaxic coordinates . 2 edn. San Diego : Academic Press ; 2001 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Piguet O , Hornberger M, Mioshi E, Hodges JR. Behavioural-variant frontotemporal dementia: diagnosis, clinical staging, and management . Lancet Neurol 2011 ; 10 : 162 – 72 . Google Scholar Crossref Search ADS PubMed WorldCat Przybyla M , Stevens CH, van der Hoven J, Harasta A, Bi M, Ittner A, et al. Disinhibition-like behavior in a P301S mutant tau transgenic mouse model of frontotemporal dementia . Neurosci Lett 2016 ; 631 : 24 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Roberson ED , Scearce-Levie K, Palop JJ, Yan F, Cheng IH, Wu T, et al. Reducing endogenous tau ameliorates amyloid beta-induced deficits in an Alzheimer’s disease mouse model . Science 2007 ; 316 : 750 – 4 . Google Scholar Crossref Search ADS PubMed WorldCat Rudinskiy N , Hawkes JM, Wegmann S, Kuchibhotla KV, Muzikansky A, Betensky RA, et al. Tau pathology does not affect experience-driven single-neuron and network-wide Arc/Arg3.1 responses . Acta Neuropathol Commun 2014 ; 2 : 63 . Google Scholar Crossref Search ADS PubMed WorldCat Santacruz K , Lewis J, Spires T, Paulson J, Kotilinek L, Ingelsson M, et al. Tau suppression in a neurodegenerative mouse model improves memory function . Science 2005 ; 309 : 476 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat Serrano-Pozo A , Frosch MP, Masliah E, Hyman BT. Neuropathological alterations in Alzheimer disease . Cold Spring Harb Perspect Med 2011 ; 1 : a006189 . Google Scholar Crossref Search ADS PubMed WorldCat Shannon P , Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks . Genome Res 2003 ; 13 : 2498 – 504 . Google Scholar Crossref Search ADS PubMed WorldCat Sharma S , Khadimallah I, Corya AW, Ali YO, Rao X, Liu Y, et al. Presymptomatic change in microRNAs modulates Tau pathology . Sci Rep 2018 ; 8 : 9251 . Google Scholar Crossref Search ADS PubMed WorldCat Spillantini MG , Van Swieten JC, Goedert M. Tau gene mutations in frontotemporal dementia and parkinsonism linked to chromosome 17 (FTDP-17) . Neurogenetics 2000 ; 2 : 193 – 205 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Swarup V , Hinz FI, Rexach JE, Noguchi KI, Toyoshiba H, Oda A, et al. Identification of evolutionarily conserved gene networks mediating neurodegenerative dementia . Nat Med 2019 ; 25 : 152 – 64 . Google Scholar Crossref Search ADS PubMed WorldCat Szklarczyk D , Morris JH, Cook H, Kuhn M, Wyder S, Simonovic M, et al. The STRING database in 2017: quality-controlled protein-protein association networks, made broadly accessible . Nucleic Acids Res 2017 ; 45 : D362 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Takeuchi H , Iba M, Inoue H, Higuchi M, Takao K, Tsukita K, et al. P301S mutant human tau transgenic mice manifest early symptoms of human tauopathies with dementia and altered sensorimotor gating . PLoS One 2011 ; 6 : e21050 . Google Scholar Crossref Search ADS PubMed WorldCat Tan DCS , Yao S, Ittner A, Bertz J, Ke YD, Ittner LM, et al. Generation of a new tau knockout (tauDeltaex1) line using CRISPR/Cas9 genome editing in mice . J Alzheimer's Dis 2018 ; 62 : 571 – 8 . Tort AB , Komorowski R, Eichenbaum H, Kopell N, Measuring phase-amplitude coupling between neuronal oscillations of different frequencies . J Neurophysiol 2010 ; 104 : 1195 – 210 . Google Scholar Crossref Search ADS PubMed WorldCat Tort AB , Komorowski RW, Manns JR, Kopell NJ, Eichenbaum H, Theta-gamma coupling increases during the learning of item-context associations . Proc Natl Acad Sci U S A 2009 ; 106 : 20942 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat Van der Jeugd A , Vermaercke B, Halliday GM, Staufenbiel M, Impulsivity GJ, decreased social exploration, and executive dysfunction in a mouse model of frontotemporal dementia . Neurobiol Learn Mem 2016 ; 130 : 34 – 43 . Google Scholar Crossref Search ADS PubMed WorldCat van Eersel J , Stevens CH, Przybyla M, Gladbach A, Stefanoska K, Chan CK, et al. Early-onset axonal pathology in a novel P301S-Tau transgenic mouse model of frontotemporal lobar degeneration . Neuropathol Appl Neurobiol 2015 ; 41 : 906 – 25 . Google Scholar Crossref Search ADS PubMed WorldCat Van Erum J , Valkenburg F, Van Dam D, De Deyn PP, Pentylenetetrazole-induced seizure susceptibility in the Tau58/4 transgenic mouse model of tauopathy . Neuroscience 2020 ; 425 : 112 – 22 . Google Scholar Crossref Search ADS PubMed WorldCat Weder ND , Aziz R, Wilkins K, Tampi RR, Frontotemporal dementias: a review . Ann Gen Psychiatry 2007 ; 6 : 15 . Google Scholar Crossref Search ADS PubMed WorldCat Weiergraber M , Henry M, Hescheler J, Smyth N, Schneider T, Electrocorticographic and deep intracerebral EEG recording in mice using a telemetry system . Brain Res Brain Res Protoc 2005 ; 14 : 154 – 64 . Google Scholar Crossref Search ADS PubMed WorldCat Wszolek ZK , Slowinski J, Golan M, Dickson DW, Frontotemporal dementia and parkinsonism linked to chromosome 17 . Folia Neuropathol 2005 ; 43 : 258 – 70 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Yoshiyama Y , Higuchi M, Zhang B, Huang SM, Iwata N, Saido TC, et al. Synapse loss and microglial activation precede tangles in a P301S tauopathy mouse model . Neuron 2007 ; 53 : 337 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat AUC = area under the curve CFC = cross frequency coupling fEPSP = field excitatory postsynaptic potential FTD = frontotemporal dementia IEG = immediate early gene LTP = long-term potentiation Author notes Magdalena Przybyla and Janet van Eersel contributed equally to this work. © The Author(s) (2020). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Onset of hippocampal network aberration and memory deficits in P301S tau mice are associated with an early gene signature
JF - Brain
DO - 10.1093/brain/awaa133
DA - 2020-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/onset-of-hippocampal-network-aberration-and-memory-deficits-in-p301s-oOF07RLk29
SP - 1889
VL - 143
IS - 6
DP - DeepDyve
ER -