Neurobiology of Disease 173 (2022) 105841 Contents lists available at ScienceDirect Neurobiology of Disease journal homepage: www.elsevier.com/locate/ynbdi Postnatal environmental enrichment enhances memory through distinct neural mechanisms in healthy and trisomic female mice Maria Alemany-González a, Marta Vilademunt a, Thomas Gener a, b, c, M. Victoria Puig a, b, c, * a Integrative Pharmacology and Systems Neuroscience, Hospital del Mar Medical Research Institute, Barcelona Biomedical Research Park, 08003 Barcelona, Spain b Catalan Institute of Nanoscience and Nanotechnology (ICN2), the Barcelona Institute of Science and Technology (BIST), Campus UAB, Bellaterra, 08193 Barcelona, Spain c Institut de Neurociències, Universitat Autònoma de Barcelona, 08193 Bellaterra, Barcelona, Spain A R T I C L E I N F O A B S T R A C T Keywords: Stimulating lifestyles have powerful effects on cognitive abilities, especially when they are experienced early in Intellectual disability life. Cognitive therapies are widely used to improve cognitive impairment due to intellectual disability, aging, Down syndrome and neurodegeneration, however the underlying neural mechanisms are poorly understood. We investigated the Prefrontal cortex neural correlates of memory amelioration produced by postnatal environmental enrichment (EE) in diploid mice Hippocampus Functional connectivity and the Ts65Dn mouse model of Down syndrome (trisomy 21). We recorded neural activities in brain structures Cognitive stimulation key for memory processing, the hippocampus and the prefrontal cortex, during rest, sleep and memory perfor­ Neural synchrony mance in mice reared in non-enriched or enriched environments. Enriched wild-type animals exhibited enhanced Ts65Dn Genetic mouse model neural synchrony in the hippocampus across different brain states (increased gamma oscillations, theta-gamma Memory impairment coupling, sleep ripples). Trisomic females showed increased theta and gamma rhythms in the hippocampus and REM sleep prefrontal cortex across different brain states along with enlarged ripples and disrupted circuit gamma signals that were associated with memory deficits. These pathological activities were attenuated in their trisomic EE- reared peers. Our results suggest distinct neural mechanisms for the generation and rescue of healthy and pathological brain synchrony, respectively, by EE and put forward hippocampal-prefrontal hypersynchrony and miscommunication as major targets underlying the beneficial effects of EE in intellectual disability. 1. Introduction dependent brain reorganization have not been elucidated. Environmental enrichment (EE) mimics a stimulating lifestyle in A rich social, physical, and cognitive lifestyle has powerful effects on laboratory animals and has demonstrated beneficial effects on various cognitive abilities and brain function. Cognitive training programs are aspects of brain structure and function, including cognition. It promotes the main nonpharmaceutical interventions used to ameliorate cognitive dendritic branching and synaptic plasticity in the hippocampus (HPC) decline due to aging or neurodegenerative diseases (Livingston et al., and the prefrontal cortex (PFC; Hirase and Shinohara, 2014; van Praag 2017) and are the only therapy currently available for people with Down et al., 2000), two brain regions crucial for learning and memory, and syndrome (DS, trisomy 21). In DS subjects, cognitive therapy improves improves cognitive performance in several tasks such as learning, executive functions and adaptive functionality (de la Torre et al., 2016; working memory, and recognition memory (Brenes et al., 2016; Leger Karaaslan and Mahoney, 2013; Martínez Cué and Dierssen, 2020) via a et al., 2015; Mesa-Gresa et al., 2013; Wang et al., 2020). Studies con­ neuroplastic reorganization of the brain (Anagnostopoulou et al., 2021). ducted in anesthetized rodents report that animals reared in enriched DS brains display increased brain synchrony as assessed by EEG, MEG environments present enhanced gamma rhythms within hippocampal and fMRI (Anderson et al., 2013; Figueroa-Jimenez et al., 2021; Pujol microcircuits (Hirase and Shinohara, 2014; Shinohara et al., 2013; et al., 2015; Ramírez-Toraño et al., 2021; Velikova et al., 2011), how­ Tanaka et al., 2017; van Praag et al., 2000). However, no studies have ever the exact neural mechanisms involved in the cognitive training- yet investigated how experience modulates long-range connectivity Abbreviations: Down syndrome, DS; environmental enrichment, EE; prefrontal cortex, PFC; hippocampus, HPC; trisomy, TS; discrimination index, DI; rapid eye movement sleep, REM sleep. * Corresponding author at: Catalan Institute of Nanoscience and Nanotechnology (ICN2), Barcelona, Spain. E-mail address: victoria.puig@icn2.cat (M.V. Puig). https://doi.org/10.1016/j.nbd.2022.105841 Received 9 June 2022; Received in revised form 12 August 2022; Accepted 15 August 2022 Available online 18 August 2022 0969-9961/© 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 during wakefulness and natural sleep. individually from the day of the surgery until the end of the experiment EE has also been found to attenuate molecular, cellular, and to prevent the implants from being damaged by other cage mates. Mice behavioral deficits in animal models of numerous neurological and in the enriched group had toys in their cages that were replaced every psychiatric disorders (Nithianantharajah and Hannan, 2006) and it ex­ three days until the end of the experiment. erts profound changes in the brains and behaviors of DS animal models. The Ts65Dn partial trisomic DS model, with triplication of ~90 genes 2.3. Stereotaxic surgeries ortholog for genes located in human chromosome 21, exhibits behav­ ioral, cellular, and molecular phenotypes relevant for DS. Over­ Mice were anesthetized with a mixture of ketamine/xylazine (keta­ expression of genes disrupts neurogenesis and synaptic plasticity in mine: Imalgene 1000, Distrivet SA; xylazine: X1251-1G, Sigma-Aldrich) Ts65Dn mice, simplifies dendritic architecture, and imbalances excita­ and placed in a stereotaxic apparatus. Anesthesia was maintained with tion–inhibition that leads to a general overinhibition (Belichenko et al., continuous 0.5–4% isoflurane (Zoetis Spain, S.L). Small craniotomies 2007; Cramer and Galdzicki, 2012; Kleschevnikov et al., 2004; Ruiz- were drilled above the medial PFC and the HPC. Five micro-screws were Mejias, 2019). EE decreases inhibition and ameliorates dendritic screwed into the skull to stabilize the implant, and the one on top of the complexity, synaptic plasticity and cognitive performance in DS mice cerebellum was used as a general ground. Three tungsten electrodes, one both during development and adulthood (Begenisic et al., 2015, 2011; stereotrode and one single electrode, were implanted in the PFC and Martínez-Cué et al., 2005, 2002; Pons-espinal et al., 2013). three more were implanted in the HPC. The electrodes were positioned Here we aimed to unravel the neural substrates of memory amelio­ stereotaxically in the prelimbic PFC (AP: 1.5, 2.1 mm; ML: ± 0.6, 0.25 ration produced by EE experienced early in life in diploid healthy mice mm; DV: − 1.7 mm from bregma) and in the CA1 area of the HPC (AP: and trisomic DS mice. We used Ts65Dn females because they are more − 1.8, − 2.5 mm; ML: − 1.3, − 2.3 mm; DV: − 1.15, − 1.25 mm). In addi­ sensitive to EE than Ts65Dn males (Martínez-Cué et al., 2002). In tion, three reference electrodes were implanted in corpus callosum and consonance with the increased brain synchrony reported in clinical lateral ventricles (AP: 1, 0.2, − 1; ML: 1, 0.8, 1.7; DV: − 1.25, − 1.4, − 1.5, studies, we have recently demonstrated hypersynchronous neural ac­ respectively). The electrodes were made by twisting a strand of tungsten tivities in hippocampal-prefrontal pathways of Ts65Dn males that are wire 25 μm wide (Advent, UK), had impedances that ranged from 100 to partially rescued by green tea extracts containing epigallocatechin-3- 400 kΩ at the time of implantation and were implanted unilaterally with gallate (EGCG) (Alemany-González et al., 2020; Puig et al., 2020). dental cement. Electrode wires were pinned to an adaptor to facilitate This study allowed us to assess abnormal neural activities in male and their connection to the recording system. After surgery, animals were female trisomic subjects and compare the actions of EGCG and EE on the allowed to recover during at least one week in which they were exten­ same circuits. sively monitored and received both analgesia and anti-inflammatory treatments. Additionally, animals were handled and familiarized with 2. Materials and methods the implant connected to the recording cable. After the experiments ended, a mild electrical current (100 Hz, 0.1 mA, 2 s) was applied to the 2.1. Animals electrodes to mark the placement of electrode tips, which were later confirmed histologically by staining the brain slices with Cresyl violet Ts65Dn female mice (n = 13) and their wild-type littermates (n = 12) (Sigma-Aldrich). Electrodes with tips outside the targeted areas (CA1 were obtained by breeding B6EiC4Sn.BLiA-Ts(1716)65Dn/DnJ females and prelimbic mPFC) were discarded for data analyses. with C57BL/6 × 6JOlaHsd (B6C3F1/OlaHsd) hybrid males. The parental generation was obtained from the Jackson Laboratory (Bar 2.4. Behavioral and neurophysiological characterization Harbor, ME) and a colony was generated and maintained at the Barce­ lona Biomedical Research Park (PRBB) Animal Facility. Mice were We recorded single-unit activity (SUA) and local field potentials genotyped and ~25% of the offspring showed trisomy. Aged-matched (LFPs) in freely moving mice with the multi-channel Open Ephys system diploid littermates served as controls. Animals were 2 to 3 months old at a sampling rate of 30 kHz with Intan RHD2132 amplifiers equipped at the start of all experiments and weighed between 20 and 30 g. All with an accelerometer. We used the accelerometer’s signals in the X, Y procedures were conducted in compliance with EU directive 2010/63/ and Z axis to monitor general mobility of mice. We found that the EU and Spanish guidelines (Laws 32/2007, 6/2013 and Real Decreto variance of the instantaneous acceleration module (Acc), that quantifies 53/2013) and were authorized by the Barcelona Biomedical Research the variation of movement across the three spatial dimensions, was Park (PRBB) Animal Research Ethics Committee and the local maximal during exploration and decreased as the animals were in quiet government. alertness and natural sleep (Alemany-González et al., 2020; Gener et al., 2019). More specifically, raw X, Y and Z signals were first downsampled 2.2. Environmental enrichment protocol to 1 kHz, then we calculated the instantaneous module (root square [X2, Y2, Z2]) from which we measured the variance. In addition, recorded Female mice were housed in non-enriched or enriched conditions for signals from each electrode were filtered offline to extract SUA and LFPs. 7 weeks after weaning. The non-enriched environment consisted in SUA was estimated by first subtracting the raw signal from each elec­ standard Plexiglas cages (37 × 16.5 × 13.5 cm) where mice were housed trode with the signal from a nearby referencing electrode to remove in groups of 2 to 3 individuals. Mice reared in enriched conditions were artifacts related to the animal’s movement. Next, referenced signals housed in groups of 5 to 7 individuals in large two-level cages (55 × 80 were filtered between 450 and 6000 Hz and the spikes from individual × 50 cm) with plastic toys that were replaced every three days. This neurons were sorted using principal component analysis with Offline enriched environment promoted sociability, physical exercise, and Sorter v4 (Plexon Inc). To obtain LFPs, signals were downsampled to 1 cognitive development. Diploid and trisomic mice were housed together kHz, detrended and notch-filtered to remove noise line artifacts (50 Hz in both rearing conditions. Typically, one trisomic mouse was housed and its harmonics) with custom-written scripts in Python. Signals were with one or two diploid mice in the non-enriched environment. In the then imported into MATLAB. The frequency bands considered for the enriched environment, two trisomic mice were housed with 2 to 4 band-specific analyses included theta (8–12 Hz), low gamma (30–50 diploid mice. All the trisomic mice and the same number of diploid Hz), high gamma (50–80 Hz), and higher frequencies (100 to 150 Hz) cagemates underwent surgery for chronic electrode implantation during when appropriate. the first week after completing the seven-week rearing period. The rest of the diploid animals were euthanized. The surgery was followed by 2.4.1. Quiet wakefulness, NREM and REM sleep one week of post-surgical recovery (Fig. 1A). All the mice were housed Recordings during quiet wakefulness were performed in an empty 2 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 (caption on next page) 3 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 Fig. 1. EE enhances pyramidal activity and gamma synchrony in the HPC of healthy females during rest. A: Experimental timeline. After weaning, mice were reared in enriched or non-enriched environments for 7 weeks. Typically, one trisomic mouse was housed with one or two diploid mice in the non-enriched environment while two trisomic mice were housed with 2 to 4 diploid mice in the enriched environment. All the trisomic mice and the same number of diploid cage mates underwent surgery for chronic electrode implantation in the HPC and the mPFC during the first week after completing the seven-week rearing period. The surgery was followed by one week of post-surgical recovery after which we recorded neural activities during resting periods, sleep, and memory performance. The results for the diploid groups are shown in Figs. 1 to 3 and those of the trisomic groups are shown in Figs. 4 to 6. Also shown are representative examples of histological validations in CA1 and prelimbic PFC. Note the small lesions caused by low-intensity electrical stimulations, which were used to mark the tips of the electrodes after the last recording session. B: Neurons were classified based on three electrophysiological properties: peak-to-valley ratio and half-valley width of the action po­ tentials, and the firing rate (Kim et al., 2016; Alemany-González et al., 2020). Purple dots denote putative pyramidal neurons (wide spiking), and red dots represent unclassified putative non-pyramidal neurons. Shown is a representative 3D plot including all the neurons recorded in the HPC in the wild-type NE group (n = 6 mice). Also shown are the mean firing rates per animal of presumed pyramidal and non-pyramidal neurons in HPC and PFC. Number of pyramidal neurons isolated in each brain region in wild-type NE and EE mice: [HPC] n = 23, 23 neurons; [PFC] n = 18, 26 neurons. Non-pyramidal neurons: [HPC] n = 13, 11 neurons; [PFC] n = 12, 11 neurons. C: Power spectra of signals in HPC and PFC during rest. D: Co-modulation maps of cross-frequency coupling in the HPC. Color scale indicates modulation index (MI) in all figs. E: Quantification of theta-high gamma MI in the HPC. Data are represented as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.) cage for 10 min. Briefly, low mobility was determined by a defined of familiar object exploration]/total exploration time). DIs varied be­ threshold in the output of the accelerometer signals and normal move­ tween +1 and − 1, where a positive score indicated more time spent ment was defined as above that threshold. The animals typically rested with the novel object, a negative score indicated more time spent with for brief periods between 2 and 10 consecutive seconds. Recordings the familiar object, and a zero score indicated a null preference (Leger during natural sleep were implemented following the familiarization et al., 2013). LFP measures associated with memory acquisition and phase of the NOR task to better capture neural signals related to memory retrieval were obtained by averaging one-second non-overlapping win­ consolidation (Fig. 2A). Accelerometer measures and LFP signals from dows triggered by the button presses on the joystick. Epochs that con­ the PFC and the HPC were used to score epochs of REM and NREM sleep. tained electrical artifacts were discarded from the analyses. Each one- NREM sleep was behaviorally defined as presenting immobility (low second window was only considered if the mouse explored the object variations of the accelerometer) and large amplitude slow oscillations for at least 600 ms. This allowed us to include many visits that were (1–4 Hz) in the PFC and HPC. Only periods of deep NREM sleep prior to particularly short while also granting a good estimation of power at low REM episodes were considered for analysis. REM epochs were defined as frequencies. Memory retrieval was investigated in the 24 h memory test immobility and prominent hippocampal theta rhythms. Analyses of LFPs during the visits to familiar objects. Memory acquisition and novelty signals during quiet wakefulness and natural sleep were performed seeking, which cannot be disentangled in this task, were investigated in averaging neural signals over discrete epochs of different duration (one the 24 h memory test during the visits to novel objects. As in our pre­ continuous epoch per experiment; quiet wakefulness: 3 min; NREM vious study (Alemany-González et al., 2020), we found that these sleep: 1 min; REM sleep: 10 s) that were chosen based on the stability of criteria provided the most robust results. the brain state (McShane et al., 2012). 2.4.2. Novel object recognition test 2.5. Data analyses We tested recognition memory using a well-established task that relies on mice’ innate instinct to explore novel objects in the environ­ 2.5.1. Power spectral analysis ment (McShane et al., 2012). We used a custom-designed T-maze made We used the multitaper fast Fourier transform method (time fre­ of aluminum with wider and higher arms than the standard mazes (8 cm quency bandwidth; TW = 5 and K = 9 tapers; 1–100 Hz range; non- wide × 30 cm long × 20 high). The maze was shielded and grounded for overlapping sliding window of 5 s for power spectra, 2 s for spectro­ electrophysiological recordings and was placed on an aluminum plat­ grams, 1 s for object explorations) with the Chronux toolbox in MATLAB form. The T shape of the maze also favored more visits to the objects by (Bokil et al., 2010). the mice. The novel-familiar object pairs were previously validated as in Gulinello et al., 2018. The arm of the maze where the novel object was 2.5.2. Phase-amplitude modulation index placed was randomly chosen across experiments. The test was imple­ To quantify the intensity of phase-amplitude coupling we used the mented in three phases of 10 min each: habituation and familiarization modulation index (MI) as described in (Tort et al., 2010). First, low during the first day and long-term memory test 24 h after familiariza­ frequency (delta and theta) phases were divided into 18◦ bins and tion. Mice were first habituated to an empty maze. Five minutes later, gamma (low, high gamma and high frequency oscillations-HFO) mice were placed again in the maze where they could explore two amplitude was calculated for each phase bin. MI measures the diver­ identical objects located at the end of the two lateral arms. Twenty-four gence of the phase-amplitude distribution and is higher as it is further hours later, in the test phase, mice were presented with one familiar and away from the uniform distribution. To choose specific frequency bands one novel object. Each session was videotaped via a video camera pairs for the MI quantification we represented overall MI in a two- located on top of the maze. Any investigative behavior of objects, dimensional pseudocolor comodulation map. A warmer color indicates including head orientation towards the objects or sniffing at a distance coupling between the phase of the low frequency band (x axis) and the below or equal to 2 cm or when the mice touched the objects with the amplitude of the high frequency band (y axis) while blue depicts absence nose, was considered object exploration. Exploratory events were of coupling. identified online by looking at the video using a custom-designed joystick with a right and left button that were pressed continuously 2.5.3. Phase slope index (PSI) during the time of explorations. Button presses were automatically To estimate the flow of information between neural signals of the aligned to the electrophysiological file by sending TTL pulses to the PFC and HPC, we used the Phase Slope Index as in Nolte et al., 2008 with acquisition system so that two more event channels were added to the the data2psi.m function. PSI is a robust measure based on the conceptual recording files. Object recognition memory was defined by the temporal argument supporting that the driver is earlier than the recip­ discrimination index (DI) for the novel object using the difference in ient and contains information about the future of the recipient. It exploration time for the familiar object divided by the total amount of quantifies the consistency of the direction of the change in the phase exploration of both objects (DI = [time of novel object exploration-time difference across frequencies. Positive slope reflects an HPC-to-PFC (HPC→PFC) flow of information in a specific frequency range while a 4 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 Fig. 2. EE entrains hippocampal gamma oscil­ A Habituation Familiarization Memory test lations and hippocampal-prefrontal gamma connectivity in healthy mice during memory retrieval. A: The novel object recognition task. B: Discrimination indices (DIs) and number of visits to familiar and novel objects were similar Sleep Recording between rearing groups. Visits to novel objects were longer than to familiar objects in both 5 min 24h groups. C: Power spectra of signals in HPC and 24h Memory test PFC. Quantifications of relevant bands are also shown. During the visits to familiar objects EE mice exhibited enhanced low/high gamma Memory retrieval Memory aquisition power in the HPC (mixed ANOVA, region x (Visits to familar objects) (Visits to novel objects) rearing interaction: F1,9 = 9.105, 14.689; P = B 0.015, 0.004; HPC: P = 0.029, 0.007; PFC: P = Number of visits Duration of visits 0.99, 0.65). During the visits to novel objects, 0,7 35 1,2 ** low/high gamma power also tended to increase 0,6 30 1,0 * in the HPC of EE mice, but less consistently than during the visits to familiar objects (F1,9 = 3.47, 25 WT NE Counts 0,5 5.11; P = 0.09, 0.05). D: Co-modulation maps of Time (s) 0,8 20 DI 0,4 WT EE cross-frequency coupling in the HPC and corre­ 15 0,6 sponding quantification of theta-high gamma 0,3 10 0,4 MI. E: HPC-to-PFC directionality (PSI) at low 0,2 5 gamma frequencies and corresponding correla­ 0,1 0 0,2 tions with DIs. FAM NOV FAM NOV FAM NOV FAM NOV Familiar object Novel object C HPC 60 x10 2 HPC x10 2 * * x10 3 x10 3 x10 2 50 ** x10 3 x10 2 70 16 30 40 15 x10 3 30 20 20 50 Power (mV/Hz) Power (mV/Hz) 15 25 30 25 12 20 15 30 * 10 20 0 10 20 10 5 10 10 LG HG 15 15 5 LG HG 8 Theta 10 5 Theta 10 4 5 5 0 0 0 0 0 5 10 15 20 20 40 60 80 100 0 5 10 15 20 20 40 60 80 100 x10 3 PFC x10 2 x10 2 x10 3 PFC x10 2 x10 2 14 12 3 14 9 3 16 30 x10 11 16 10 x10 12 12 8 Power (mV/Hz) Power (mV/Hz) 10 12 12 10 20 9 8 10 5 7 8 8 10 8 4 8 6 7 HG LG 4 6 6 6 0 5 LG HG Theta Theta 4 5 4 4 4 2 3 2 3 0 2 0 2 0 5 10 15 20 20 40 60 80 100 0 5 10 15 20 20 40 60 80 100 Frequency (Hz) Frequency (Hz) Frequency (Hz) Frequency (Hz) D -3 HPC WT NE HPC WT EE 10 HPC WT NE HPC WT EE 10-3 200 5 200 5 180 4,5 180 4,5 160 4 160 4 Amp Freq (Hz) Amp Freq (Hz) 140 Theta-HG 3,5 140 Theta-HG 3,5 120 -3 3 -2 3 7 x10 ** 2,5 120 60 x10 2,5 100 6 100 80 2 50 2 5 80 1,5 40 1,5 60 4 60 3 1 30 1 40 40 2 0,5 20 0,5 20 20 14 15 16 17 18 19 20 21 14 15 16 17 18 19 20 21 14 15 16 17 18 19 20 21 14 15 16 17 18 19 20 21 Phase freq (Hz) Phase freq (Hz) Phase freq (Hz) Phase freq (Hz) E Directionality Directionality 0,7 HPC-to-PFC 6 0,7 R = 0,69 2 0,6 P = 0,026 0,6 R = 0,08 PFC-to-HPC HPC-to-PFC P = 0,98 4 0,5 0 0,5 2 0,4 -2 0,4 PSI PSI DI DI PFC-to-HPC 0,3 -4 0,3 0 0,2 0,2 -6 -2 0,1 0,1 -8 0 0 -4 -2 0 2 4 6 8 -10 -8 -6 -4 -2 0 2 4 PSI (Directionality) PSI (Directionality) 5 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 negative slope reflects the opposite directionality of signals. 2.61 spikes/s, respectively), whereas spiking activity in the PFC was similar between groups (4.13 and 4.66, respectively). The firing rate of 2.5.4. Analysis of ripples during NREM sleep presumed non-pyramidal neurons was similar between rearing groups in Raw signals were downsampled to 1.25 kHz and bandpass filtered both brain regions (Fig. 1B). between 100 and 600 Hz. Sharp-wave ripple events were detected by Concomitantly, low gamma power (30–50 Hz) was increased in the thresholding (>3 SDs) the filtered signals using ripple detection scripts HPC of the enriched group (ANOVA, F1,10 = 9.739, P = 0.011; post hoc from M. Valero (Valero et al., 2017). Time-frequency analysis of ripples comparisons, HPC: P = 0.009, PFC: P = 0.11; Fig. 1C). We further was calculated using multitaper spectral estimation and frequency res­ quantified the coupling between the phase of slow rhythms (5–20 Hz) olution of 10 Hz in the 100–600 Hz range. and the amplitude of faster oscillations (20–200 Hz) in the HPC (Del­ gado-Sallent et al., 2021; Jensen and Colgin, 2007; Scheffer-Teixeira and 2.5.5. Analyses of entropy Tort, 2017). Coupling was strengthened between phases from 5 to 9 To quantify the level of frequency dispersion (and consequently the cycles/s (theta) and amplitudes from 60 to 80 cycles/s (high gamma) in level of synchronization in the neuronal activity), we performed analysis enriched females (Mann-Whitney test, P = 0.008; Fig. 1D,E). Together, of the spectral entropy (based on Shannon entropy) of the LFP (30–150 these results identified relevant roles for hippocampal gamma syn­ Hz), where a narrower power distribution results in lower entropy chrony in the cellular mechanisms of EE in healthy mice during rest. values, and a large power distribution results in higher entropy values, with no dependency on the LFP frequency (Valero et al., 2017). 3.2. Environmental enrichment entrains hippocampal gamma oscillations and hippocampal-prefrontal gamma connectivity during memory retrieval 2.6. Statistical analysis We next investigated how enriched or non-enriched environmental We used unpaired t-tests to compare behavioral measures between stimulation early in development impacts memory performance and its genotypes. Paired t-tests and mixed ANOVAs were used to assess the neural correlates in young adulthood. We used the novel object recog­ effects of environmental enrichment on power and firing rate across nition (NOR) task, a well validated memory test that relies on the mice’ brain regions. Two-WAY ANOVAs were used to assess the effects of innate instinct to explore novel objects and depends on hippocampal- environmental enrichment on power and firing rate across genotypes. prefrontal circuits (Warburton and Brown, 2015; Alemany-González Mann-Whitney and Kruskal-Wallis non-parametric tests were used to et al., 2020; Wang et al., 2021). The task was implemented in a T-maze assess the effects of enrichment on phase-amplitude coupling (PAC), and consisted of a habituation phase, familiarization phase, and memory entropy, and directionality of signals (PSI). We used Bonferroni test phase, each lasting ten minutes. During the habituation phase, the correction or Tukey HSD post-hoc tests when appropriate. To identify animals explored the maze without objects. Five minutes later, for the significant correlations between neurophysiological measures and DIs, familiarization phase the mice were put back in the maze where two Pearson or Spearman correlations were used for parametric and non- identical objects had been placed at the end of the lateral arms. Twenty- parametric distributions, respectively. All statistical analyses were per­ four hours later, a familiar object and a new object were placed in the formed in MATLAB R2019a (The Mathworks, MA, USA) and the level of maze for the memory test. We investigated the neural substrates of significance was set to p < 0.05. Results are expressed as mean (± SEM). memory acquisition and retrieval by comparing the neural signals SPSS and JASP 0.15 were used for statistical analyses. recorded during the visits to novel and familiar objects, respectively, during the memory test (Fig. 2A). 3. Results Discrimination indices (DIs) for novel versus familiar objects were positive for all the mice ([time visiting the novel object - time visiting the 3.1. Female mice reared in enriched environments exhibit increased familiar object] / total exploration time; mean DI 0.42 ± 0.05, n = 12 gamma synchrony in the hippocampus during rest mice); that is, mice explored novel objects more than familiar objects, indicating good recognition memory. As observed in a previous study We investigated differential neural dynamics in hippocampal- (De Toma et al., 2019), EE mice tended to display higher DIs than NE prefrontal pathways between mice reared in enriched environments mice (0.47 ± 0.09 vs. 0.37 ± 0.06, n = 6 mice per group), however this (EE group) and non-enriched environments (NE group) for seven weeks difference did not reach significance (unpaired t-test, P = 0.36). This was after weaning. Animals reared in enriched conditions were housed in likely due to ceiling effects on behavioral performance. We quantified groups of 4 to 6 subjects in large cages with toys that promote physical the number of visits to each object and the mean duration of individual activity. The toys were replaced every few days to keep physical and visits per session. Mice of the two groups visited novel objects on more cognitive stimulation constant over weeks. Animals reared in non- occasions than familiar objects during each session (16 vs. 12 visits on enriched conditions were housed in standard cages without toys in average) and the visits lasted longer (840 ± 58 ms to novel objects vs. groups of 2 or 3 individuals. Following the periods with or without 460 ± 66 ms to familiar objects; mixed ANOVA, duration x object enrichment, animals were implanted with electrodes in the CA1 region interaction; F1,10 = 28.95; P < 0.0005; NE mice, P = 0.001; EE mice, P = of the HPC and the prelimbic medial PFC. We then recorded neural 0.013; Fig. 2B). activities during resting episodes, natural sleep, and memory perfor­ Despite the modest beneficial effects of EE on memory performance, mance. Diploid and trisomic mice were mixed in the enriched and non- differential neural activities were detected between the two rearing enriched cages (Fig. 1A). The results for the diploid groups are shown in groups. During the visits to familiar objects, like in resting states, EE Figs. 1 to 3 and those of the trisomic groups are shown in Figs. 4 to 6. mice exhibited enhanced low/high gamma power in the HPC (mixed We first investigated differences in the firing rate of neurons between ANOVA, region x rearing factors; Fig. 2C) and strengthening of theta- wild type enriched and non-enriched mice during resting states. We high gamma coupling (9–14 cycles/s with 50–80 cycles/s; Mann- classified neurons as putative pyramidal and non-pyramidal cells based Whitney test, P = 0.008; Fig. 2D). During the visits to novel objects, on the shape of their action potentials and their firing pattern (Fig. 1B; low/high gamma power also tended to increase in the HPC of EE mice, Alemany-González et al., 2020; Kim et al., 2016). Around 70% and 65% but less consistently than during the visits to familiar objects (mixed of neurons recorded in the HPC and PFC, respectively, were identified as ANOVA; Fig. 2C), while hippocampal theta-gamma coupling was un­ putative pyramidal cells. Presumed pyramidal neurons in the HPC of EE changed (Mann-Whitney test, P > 0.05; Fig. 2D). mice tended to fire more action potentials than in NE mice but the dif­ We further detected a neural information flow (via the phase slope ference did not reach significance (n = 6 mice per group; mixed ANOVA, index or PSI; Nolte et al., 2008) at low gamma frequencies that traveled rearing x region interaction F1,20 = 2.67, P = 0.11; EE and NE: 4.38 and from the HPC to the PFC during the visits to familiar objects that 6 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 Fig. 3. EE enhances pyramidal activity and hippocampal gamma and ripple synchrony in healthy mice during sleep. A: Mean firing rates per animal in the HPC and PFC during REM and NREM sleep. Number of presumed pyramidal neurons isolated in each brain region and brain state in NE and EE wild-type mice: [HPC] REM: n = 17, 15 neurons; NREM: n = 16, 23 neurons; [PFC] REM: n = 15, 17 neurons; NREM: n = 15, 17 neurons. Non-pyramidal neurons: [HPC] REM: n = 14, 13 neurons; NREM: n = 11, 7 neurons; [PFC] REM: n = 9, 11 neurons; NREM: n = 10, 8 neurons. B: Power spectra of signals in HPC and PFC during REM and NREM sleep. In the PFC, delta power was also enhanced (P = 0.028). C: Co-modulation maps of cross-frequency coupling in the HPC during REM sleep and corresponding quantification of theta-high gamma MI. D: Quantification of relevant features of ripple events recorded during NREM sleep. A representative example of a ripple recorded in a mouse of the WT NE group is also shown. 7 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 Fig. 4. EE prevents pathological theta and gamma hypersynchrony in Ts65Dn female mice during quiet wakefulness. A: Mean firing rates per animal in the HPC and PFC. Number of pyramidal neurons isolated in each brain region in trisomic NE and EE mice: [HPC] n = 26, 16 neurons; [PFC] n = 28, 26 neurons. Non-pyramidal neurons: [HPC] n = 12, 7 neurons; [PFC] n = 8, 7 neurons. B: Power spectra of signals in the HPC and PFC. Trisomic females exhibited enhanced theta and low/high gamma oscillations in the HPC (two-way ANOVA, genotype x rearing interaction; F1,21 = 8.614, 17.273, 11.34, P = 0.008, <0.0005, 0.003; WT NE vs. TS NE mice, P = 0.011, <0.0005, 0.012) and the PFC (F1,19 = 7.239, 4.436, P = 0.014, 0.048; WT NE vs. TS NE mice, P = 0.043, 0.014, 0.046) that were normalized in the TS EE group (TS EE vs. WT NE mice, [HPC theta, gamma] P = 0.15, 0.88; [PFC theta, gamma] P = 0.51, 0.17). C: Co-modulation maps of cross-frequency coupling in the HPC and corresponding quantification of theta-high gamma MI. D: Correlation between firing rate and gamma power recorded in the HPC. E: Quantification of low gamma entropy in the HPC and PFC for the four experimental groups. F: Proposed model for the effects of EE on gamma synchrony in diploid and trisomic mice reared in enriched and non-enriched environments. For visualization purposes, panels A to C omit the wild-type enriched group, which can be found in Fig. 1. correlated with the DIs. This was consistent with our previous findings in gamma signals recorded during the visits to novel objects did not show a male mice (Alemany-González et al., 2020). The strength of this gamma clear directionality and were not associated with memory performances. communication was similar between rearing groups, in accord with their We conclude that EE entrains hippocampal gamma oscillations and comparable DIs (Mann-Whitney test, P = 0.42; Fig. 2E). Conversely, hippocampal-prefrontal gamma connectivity during memory retrieval. 8 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 WT NE WT EE TS NE TS EE A C Gamma Entropy DI Number of visits Duration of visits 0,6 ******** *** 0,8 HPC 0,8 PFC Memory test 35 0,4 30 *** 0,6 R=-0,25 0,6 TS WT ALL 1,5 ** R=-0,56 0,4 0,4 TS WT ALL 25 P=0,25 ** Time (s) 0,2 P**=0,005 DI 20 0,2 R=-0,01 0,2 0 1 P=0,98 R=0,24 15 DI DI 0 0 R=-0,74 P=0,48 -0,2 10 -0,2 -0,2 0,5 P**=0,01 R=-0,56 -0,4 5 -0,4 -0,4 P=0,06 0 -0,6 -0,6 FAM NOV FAM NOV FAM NOV FAM NOV FAM NOV FAM NOV 30 35 40 45 50 30 35 40 45 50 entropy entropy B Familiar object Novel object x103 HPC x10 3 PFC HPC PFC 18 x103 x10 3 25 12 2 x10 3 x10 3 18 x10 16 20 * 16 x10 3 * 10 50 * * 14 20 20 20 Power (mV/Hz) Power (mV/Hz) 15 15 14 40 12 10 12 15 8 30 10 15 5 10 20 10 10 5 0 5 6 10 8 Theta 8 Theta Theta 10 Theta 6 6 4 4 5 4 2 2 2 0 0 0 0 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 0 5 10 15 20 x10 3 100 ** ** x10 2 ** x10 2 100 * * 2 x10 2 35 80 12 25 35 80 16 x10 35 30 11 20 60 30 60 30 14 25 Power (mV/Hz) Power (mV/Hz) 10 15 40 25 40 20 20 9 10 25 20 12 15 LG 8 LG LG 5 20 20 ** ** 7 60 * * 10 LG 15 50 6 12 15 50 8 14 40 10 40 12 10 5 8 10 30 6 10 30 20 4 6 20 8 5 4 5 6 10 3 HG 4 HG 2 HG 0 20 30 40 50 60 70 80 90 100 20 30 40 50 60 70 80 90 100 HG 20 30 40 50 60 70 80 90 100 20 30 40 50 60 70 80 90 100 Frequency (Hz) Frequency (Hz) Frequency (Hz) Frequency (Hz) 0,8 0,8 0,8 0,8 0,6 R=-0,47 0,6 R=-0,63 0,6 R=-0,40 0,6 R=-0,31 TS WT ALL TS WT ALL TS WT ALL TS WT ALL 0,4 P*=0,03 P**=0,002 P=0,08 P=0,15 0,4 0,4 0,4 R=-0,17 R=-0,34 R=-0,07 R=-0,08 0,2 P=0,61 0,2 P=0,30 0,2 P=0,83 0,2 P=0,81 0 R=-0,64 R=-0,55 R=-0,51 R=-0,27 DI 0 DI 0 DI DI 0 P*=0,03 P=0,08 P=0,13 P=0,39 -0,2 -0,2 -0,2 -0,2 -0,4 -0,4 -0,4 -0,4 2 2 2 -0,6 x102 -0,6 x10 -0,6 x10 -0,6 x10 0 20 40 60 80 100 120 0 5 10 15 20 25 30 0 20 40 60 80 100 120 0 5 10 15 20 25 30 35 HPC Low Gamma Power PFC Low Gamma Power HPC Low Gamma Power PFC Low Gamma Power D Coupling -3 HPC WT NE HPC TS NE HPC TS EE 10 -3 x10-3 HPC WT NE HPC TS NE HPC TS EE 10 x10-3 200 5 200 5 180 4,5 180 4,5 4 10 4 8 160 160 Amp Freq (Hz) Amp Freq (Hz) 140 3,5 140 3,5 8 3 6 120 3 120 2,5 6 2,5 100 100 2 2 4 80 80 1,5 4 1,5 60 60 1 1 2 40 2 40 0,5 0,5 20 20 14 15 16 17 18 19 20 21 14 15 16 17 18 19 20 21 14 15 16 17 18 19 20 21 Theta-HG 14 15 16 17 18 19 20 21 14 15 16 17 18 19 20 21 14 15 16 17 18 19 20 21 Theta-HG Phase freq (Hz) Phase freq (Hz) Phase freq (Hz) Phase freq (Hz) Phase freq (Hz) Phase freq (Hz) E Directionality Familiar Novel 0,8 0,8 HPC-to-PFC *** ** PFC-to-HPC HPC-to-PFC R=0,73 R=0,18 TS WT ALL TS WT ALL 4 0,6 2 0,6 P***=0,00 P=0,42 2 0,4 R=0,67 0,4 R=0,20 0 0,2 P*=0,03 0,2 P=0,55 0 R=0,47 -2 R=0,26 DI 0 0 PSI PSI DI P=0,13 P=0,41 PFC-to-HPC -2 -0,2 -4 -0,2 -4 -0,4 -6 -0,4 -6 -0,6 -0,6 -8 -6 -4 -2 0 2 4 6 8 -8 -10 -8 -6 -4 -2 0 2 4 HPC-to-PFC Directionality HPC-to-PFC Directionality (caption on next page) 9 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 Fig. 5. EE reduces theta and gamma hypersynchrony in Ts65Dn female mice during memory performance. A: Ts65Dn females exhibited poor memory performances that were fully rescued in TS enriched animals. This was due to an increase in the duration of visits to novel objects. B: Power spectra of signals during visits to familiar and novel objects are depicted on the left and right columns, respectively. Correlations between gamma power and DIs are shown below. During the visits to familiar objects, theta and low/high gamma power were increased in the HPC of trisomic females compared to their diploid peers (two-WAY ANOVA, F1,19 = 5.39, 13.41, 16.1, P = 0.03, 0.002, 0.001). During the visits to novel objects, theta and gamma power were also increased in the HPC of trisomic animals (F1,20 = 5.51, 5.56, P = 0.03; TS NE vs. WT NE: P = 0.01, 0.05) whereas in the PFC it was less noticeable (F1,20 = 7.43, 5.44, P = 0.013, 0.03). In trisomic EE mice, the aberrant low/high gamma detected in the HPC was reduced to wild-type levels (TS NE vs. TS EE: P = 0.048, 0.038; TS EE vs. WT NE: P = 0.36, 0.57). C: Correlations between low gamma entropy during visits to familiar objects and DIs. D: Co-modulation maps of cross-frequency coupling in the HPC and corresponding quantification of theta-high gamma MIs. E: HPC-to-PFC directionality (PSI) at low gamma frequencies during visits to the objects. Corresponding correlations between gamma PSI and DIs are also shown. For visualization purposes, we omit the wild-type enriched group in the figure, which can be found in Fig. 2. 3.3. Environmental enrichment enhances hippocampal gamma synchrony power in the HPC of TS NE mice almost doubled that observed in the WT and sharp wave ripples during sleep EE group (P = 0.014). We also note that excessive theta power in trisomic mice was not due to their inherent hyperlocomotion (Faizi We recorded neural activities during sleep following the familiar­ et al., 2011) because the animals did not engage in activities associated ization phase, aiming to capture the processing of recently acquired with increased mobility while resting. Enhanced theta and gamma os­ memories (Fig. 2A). Rapid eye movement (REM) and non-rapid eye cillations attenuated in the TS EE group and oscillatory power was close movement (NREM) episodes were detected during periods of low to that observed in diploid animals (Fig. 4B). Trisomic females also behavioral activity in an open field as previously reported (Alemany- exhibited broad cross-frequency coupling between theta waves and os­ González et al., 2020). In brief, REM sleep is characterized by prominent cillations at high frequencies (>80 Hz) in the HPC compared to their theta oscillations in the HPC whereas NREM sleep can be easily identi­ non-trisomic peers. In the TS EE group, coupling between theta and high fied by the emergence of slow oscillations (<4 Hz) in the cortex and gamma was selectively enhanced in similar patterns than those observed sharp wave-ripple (SWR) events in the HPC (highly synchronous oscil­ in WT EE mice (Kruskal-Wallis test, rearing factor P < 0.001; TS EE vs. latory activity at 100–250 Hz lasting ~100–200 ms) (Buzsáki, 2002; TS NE, P = 0.026; Fig. 4C), suggesting an improved signal-to-noise ratio Steriade, 2006). Cortical slow waves and hippocampal ripples have been of hippocampal activity in TS EE animals. shown to contribute to offline information processing relevant for Taken together, our results indicated opposite effects of EE on gamma memory retrieval and consolidation (Buzsáki, 2015; Joo and Frank, rhythms in diploid and trisomic animals. That is, enhancement of hip­ 2018) and were therefore investigated here in detail. pocampal gamma in enriched wild-types and reduction of aberrant During REM sleep, presumed pyramidal neurons in the HPC of EE gamma in the HPC and PFC of trisomic mice. Based on recent literature mice tended to fire more action potentials than in NE mice but the dif­ (Guyon et al., 2021), we hypothesized that EE favored “healthy” gamma ference did not reach significance (n = 6 mice per group; mixed ANOVA, rhythms in diploid animals, gamma synchrony emerging from the inter­ rearing x region interaction F1,20 = 0.5, P = 0.68; EE and NE: 2.44 and play between fast-spiking interneurons and pyramidal neurons, and 1.61 spikes/s, respectively). No changes were detected for putative non- reduced “pathological” gamma rhythms in trisomic animals, gamma pyramidal neurons. During NREM sleep, the firing rates of pyramidal synchrony generated by the increase of asynchronous or “noisy” firing of neurons were similar between rearing groups, however the spiking ac­ pyramidal neurons due to inhibitory dysfunction. The increased pyrami­ tivity of non-pyramidal cells tended to decrease (Fig. 3A). Moreover, dal firing in the HPC correlated with low gamma power in trisomic but during REM sleep, as in quiet wake and memory performance, gamma not in wild-type mice (Fig. 4D), in accordance with a LFP “contamination” oscillations were increased in the HPC of EE mice (mixed ANOVA, F1,8 = by asynchronous neuronal firing (Guyon et al., 2021). We next estimated 5.959, P = 0.04; Fig. 3B) along with strengthening of theta-gamma the entropy of the LFP signals, as increased entropy has been linked to coupling (8–10 with 40–80 cycles/s; Mann-Whitney test, P = 0.032; pathological synchronization in epilepsy and schizophrenia (Guyon et al., Fig. 3C). Finally, ripple power was slightly larger in EE mice (unpaired t- 2021; Valero et al., 2017). Trisomic mice exhibited increased low gamma test, P = 0.047) while ripple number per minute, duration, frequency, entropy in the HPC and the PFC (Kruskal-Wallis test, genotype and and entropy were not different between groups (Fig. 3D). rearing as factors; TS NE vs. WT NE: P = 0.057, 0.014) that was reduced in the enriched group (TS NE vs. TS EE: P = 0.008 in both regions). EE did 3.4. Environmental enrichment reduces aberrant spiking and theta/ not produce major effects on gamma entropy in diploid mice (Fig. 4E). gamma hypersynchrony in the HPC and PFC of Ts65Dn female mice Altogether, these results suggested that postnatal EE enhanced “healthy” during rest gamma rhythms in wild-type mice and prevented the emergence of “pathological” gamma rhythms in trisomic animals (Fig. 4F). To further understand the pro-cognitive effects of postnatal EE we used the Ts65Dn trisomic mouse line, a well-established mouse model of 3.5. Environmental enrichment reduces aberrant theta-gamma DS. We investigated differential neural activities between wild-type (WT hypersynchrony in the HPC and PFC of Ts65Dn female mice during NE: n = 6 mice) and trisomic mice and assessed the effects of rearing (TS memory performance NE, EE: n = 7 and 6 mice, respectively). As in trisomic males (Alemany- González et al., 2020), presumed pyramidal neurons in the HPC of Trisomic females performed poorly in the memory test (n = 7 mice, trisomic females displayed higher firing rates compared to their diploid mean DI -0.27 ± 0.06; two-WAY ANOVA, genotype x rearing interac­ counterparts during resting states (two-WAY ANOVA, genotype x region tion, F1,20 = 13.196, P = 0.002; TS NE vs. WT NE, P < 0.001) while interaction: F1,17 = 15.28, P < 0.001; WT NE vs. TS NE, P = 0.007). The trisomic EE females performed like the wild types (n = 6 mice, mean DI trisomic EE group exhibited lower hippocampal spiking (TS EE ns vs. TS 0.35 ± 0.07; TS NE vs TS EE, P < 0.001; TS NE vs. WT NE, P > 0.05; NE, P = 0.002; n = 12, 22 neurons) with firing rates within the range of Fig. 5A). The negative memory indices in trisomic NE mice resulted from the wild types (P = 0.36). In contrast, pyramidal activity in the PFC was the fact that the visits to novel items were shorter than those of diploid similar between genotypes and seemed unaffected by rearing. Further­ mice (509 ms in TS NE mice vs. 840 ms in WT NE mice; F1,19 = 4.469, P more, inconsistent effects were observed in non-pyramidal neurons of = 0.04; post hoc comparison P = 0.03). Trisomic EE mice, like the wild both regions (Fig. 4A). Also consistent with trisomic males, trisomic types, visited novel objects longer than familiar objects (1310 ms vs. females exhibited enhanced theta and low/high gamma oscillations in 677 ms, respectively; paired t-test, P = 0.0005; TS EE mice vs. TS NE the HPC and the PFC (mixed ANOVA; Fig. 4B). Notably, low gamma mice, P = 0.04; Fig. 5A). 10 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 (caption on next page) 11 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 Fig. 6. Environmental enrichment rescues enhanced hippocampal-prefrontal gamma synchrony and ripple activity in Ts65Dn mice during sleep. A: Mean firing rates per animal in the HPC and PFC during REM and NREM sleep. Number of presumed pyramidal neurons isolated in each brain region and brain state in trisomic NE and EE mice: [HPC] REM: n = 24, 15 neurons; NREM: n = 28, 16 neurons; [PFC] REM: n = 31, 19 neurons; NREM: n = 25, 20 neurons. Non-pyramidal neurons: [HPC] REM: n = 19, 9 neurons; NREM: n = 9, 13 neurons; [PFC] REM: n = 11, 11 neurons; NREM: n = 10, 7 neurons. B: Power spectra of signals in the HPC and PFC with quantification of relevant bands. During REM sleep, trisomic NE mice showed augmented gamma oscillations in both brain regions (HG HPC, LG PFC: F1,18 = 8.76, 5.86, P = 0.008, 0.026) that was decreased in trisomic EE animals. During NREM sleep, low and high gamma rhythms in the HPC and PFC were augmented in trisomic NE animals (F1,16 = 8.07, 17.52, 11.73, 13.45, P = 0.012, 0.001, 0.008, 0.001) and were reduced in the HPC of the trisomic EE group. C: Co-modulation maps of cross-frequency coupling in the HPC during REM sleep and corresponding quantification of theta-high gamma MI. D: Quantification of relevant features of ripple events recorded during NREM sleep. Lower-right panel: representative example of a ripple recorded in a mouse of the TS NE group (same scale as the color plot shown in Fig. 3D). E: Correlations between ripple power and ripple entropy with discrimination indices (DIs) in the four groups. For visualization purposes, we omit the wild-type enriched group in the figure. The results for this group can be found in Fig. 3. We next investigated neurophysiological divergences in Ts65Dn 3.6. Environmental enrichment attenuates aberrant hippocampal- mice during memory retrieval and acquisition. During the visits to prefrontal gamma synchrony and ripple activity in Ts65Dn mice during familiar objects, theta and low/high gamma power were increased in the sleep HPC of trisomic females compared to their diploid peers (two-WAY ANOVA, Fig. 5B), similar to quiet wakefulness. Also as above, theta During REM sleep trisomic NE mice (n = 7) manifested many of the power increases were not simply due to hyperlocomotion of trisomic neural activity alterations observed during quiet wake and memory mice, as the animals’ mobility was low while they were visiting the performance. However, these were only partially rescued in trisomic objects. Aberrant theta/gamma power lessened in the TS EE group, enriched animals (n = 6). First, many presumed pyramidal neurons in particularly gamma oscillations. In the PFC, trisomic mice showed less the HPC and PFC of TS NE mice exhibited elevated firing rates (two-way pronounced increases of theta and low gamma power that were also ANOVA, genotype x rearing interaction, F1,17 = 5.2, 9.69, P = 0.036, reduced in the EE group. Notably, low gamma power in the HPC of 0.006), and this was only partially prevented in the PFC by EE (P = 0.11; trisomic animals, but not of diploid animals, correlated inversely with Fig. 6A). Second, trisomic mice also showed augmented theta and DIs, while in the PFC these associations were weaker (Fig. 5B). Thus, gamma oscillations in both brain regions. Abnormal gamma oscillations these results put forward the pathological nature of gamma rhythms in were reduced in the HPC and the PFC of enriched animals while trisomic animals, particularly in the HPC, and their strong negative excessive theta oscillations remained enhanced (two-way ANOVA; impact on memory. This was further supported by significant correla­ Fig. 6B). Cross-frequency coupling was indeed improved in EE trisomic tions between gamma entropy in the HPC of TS NE animals and their animals where theta-high gamma coordination was better defined, poor memory performances (R = − 0.74, P = 0.006; Fig. 5C). We note suggesting an improved signal-to-noise ratio within hippocampal mi­ that these correlations were significant for the TS group alone, therefore crocircuits (Kruskal-Wallis, genotype and rearing as factors; P < 0.05; they did not simply emerge from the WT vs TS divergences per se. Power Fig. 6C). changes during the visits to novel objects were comparable to those During NREM sleep trisomic NE mice exhibited similar firing rates of detected during the visits to familiar objects. That is, theta and gamma putative pyramidal neurons in the HPC similar to their trisomic EE peers power were increased in the HPC of trisomic animals whereas in the PFC and the wild types (Fig. 6A). Firing rates of neurons in the PFC were also they were less noticeable. In trisomic EE mice, the aberrant low/high unaffected by genotype or rearing. Moreover, as in all the other brain gamma detected in the HPC was reduced to wild-type levels while no states, low and high gamma rhythms in the HPC and PFC were major effects of enrichment were observed in the PFC at these bands augmented in trisomic NE animals, and they were reduced in enriched (two-WAY ANOVA, Fig. 5B). Moreover, we did not find significant as­ mice only in the HPC (Fig. 6B). In addition, in the PFC, slow waves were sociations between low gamma power and memory indices in either not different between any of the groups. Furthermore, ripple events brain region (Fig. 5B). exhibited increased power (2-way ANOVA, genotype x rearing interac­ Furthermore, like in resting states, trisomic NE mice exhibited broad tion; F1,20 = 5.56, P = 0.029) and reduced entropy (Kruskal-Wallis test; hippocampal coupling between theta and high frequencies (>60 Hz) P < 0.05) in trisomic mice that were normalized in many EE animals (TS both during the visits to familiar and novel items. However, the theta- EE vs. WT NE, P = 0.1, 0.7; Fig. 6D). We note that entropy during ripples high gamma coupling observed in diploid EE animals was not pro­ in TS mice was reduced and not increased as in other brain states. Ripple moted in the trisomic EE group (Fig. 5D). Also, during the visits to power and entropy recorded in the trisomic group correlated inversely familiar objects the selective HPC→PFC low gamma signals observed in with memory indices thus predicting poor memory performances diploid animals reversed in trisomic mice, originating in the PFC and (Fig. 6E). traveling to the HPC (Kruskal-Wallis test, genotype and rearing as fac­ Finally, we performed multiple regression analyses to provide tors; genotype factor P < 0.008; TS NE vs. WT NE, P = 0.003). The further support to the significance of the correlations observed in the circuit’s information flow was rescued in several trisomic EE mice but study. We identified three biomarkers recorded in three different brain was not fully restored in all the animals (TS EE vs. WT NE, P = 0.33; states that seemed relevant to explain memory performances and Fig. 5E). As we previously reported for male mice, HPC→PFC low correlated with memory indices: low gamma power in the HPC during gamma signals correlated strongly with memory indices in the diploid resting, HPC→PFC gamma signals during the visits to familiar objects, group but not in the trisomic group, implicating HPC→PFC gamma and ripple power during NREM sleep. Multiple regression models could communication in the neural correlates of recognition memory in both predict the DIs when combining the three biomarkers in all the mice genders. This was further supported by multiple regression analyses. A (F3,17 = 8.36, P = 0.001, R2 = 0.593), HPC→PFC gamma signals and combination of HPC→PFC low gamma signals and low gamma power in ripple power being main contributors to the prediction (P = 0.015 and the HPC and PFC during the visits to familiar objects in all animals could 0.023, respectively) whereas HPC gamma power only had marginal predict DIs (F3,17 = 8.546, P = 0.001, R2 = 0.58), HPC→PFC gamma contribution (P = 0.819). This is in line with our previous observations signals being the main contributors to the prediction (P = 0.013). During in males, where circuit connectivity and ripple amplitude seemed crit­ the visits to novel objects, as in the wild types, we did not detect a clear ical for memory performance (Alemany-González et al., 2020). directionality of signals and therefore there were no differences between genotypes or rearing conditions and no associations with memory indices (Fig. 5E). 12 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 4. Discussion One possibility is that the hypersynchrony observed in DS animals was well above the levels of those observed in enriched diploid animals. We We report that physical, social, and cognitive stimulation during found that this was the case for hippocampal gamma, but not for pre­ critical developmental periods shapes hippocampal neural dynamics in frontal gamma. This implies an inverted-U shaped range of gamma wild-type mice and prevents the disruption of hippocampal-prefrontal synchrony in the HPC for optimal memory performance, whereby pathways in a mouse model of intellectual disability. Diploid animals insufficient or excessive gamma is detrimental to memory. Another reared in enriched environments exhibited enhanced pyramidal spiking, plausible explanation is that distinct neural mechanisms underlie gamma rhythms and sleep ripples in the HPC. Conversely, in trisomic “healthy” and “pathological” gamma synchronization. A widely mice theta and gamma power were increased in the HPC and PFC across accepted hypothesis is that elevation in the ratio of cellular excitation to distinct brain states, sleep ripples were abnormal, and circuit gamma inhibition could give rise to increased gamma oscillations that would communication was disrupted during memory retrieval. These alter­ result in social and cognitive deficits like the ones observed in patients ations were attenuated in their environmentally enriched peers. with autism and schizophrenia (Sohal and Rubenstein, 2019; Yizhar Wild-type mice reared in stimulating environments exhibited et al., 2011). In fact, enhanced broadband gamma can arise from enhanced pyramidal activity and gamma synchrony within hippocam­ asynchronous activities that result from deficient inhibition and pal microcircuits across different brain states. We note that non- increased, but desynchronized, pyramidal activity (Guyon et al., 2021). pyramidal neurons likely included a variety of interneuron subtypes “Healthy” gamma would arise from a true synchronization of neural for which it was difficult to obtain consistent effects. Concomitantly, activity whereas “pathological” gamma would arise from asynchronous hippocampal gamma synchrony (power and coupling with theta waves) hyperactivity. In support of this, in this study pyramidal spiking and LFP was consistently enhanced during resting, memory performance, and entropy, a measure of asynchrony, in the HPC and PFC were increased in REM sleep. These findings are in line with previous observations trisomic mice and predicted pathological gamma power. Likewise, rip­ reporting increased gamma oscillations in the HPC of anesthetized ro­ ple power was augmented in diploid enriched mice but was even larger dents reared in enriched environments (Shinohara et al., 2013; Tanaka in trisomic animals. Ripples arise from the excitatory recurrent et al., 2017) and provide new evidence for experience-dependent CA3→CA1 system and their activity is coordinated by a consortium of enhancement of hippocampal gamma synchrony during alertness, interneurons (Buzsáki, 2015). Pathological ripples may emerge from sleep and memory performance. Furthermore, our observations of unbalanced inhibition of neural networks that cause broad disruptions enhanced sleep ripples in the EE group were also reported previously in of sleep patterns in individuals with DS and in DS mouse models (Raveau anesthetized mice (Tanaka et al., 2017). Here, ripples were recorded et al., 2018; Ruiz-Mejias, 2019). Together, these findings provide new during episodes of NREM sleep immediately after the familiarization insight into the GABAergic dysfunction observed in DS. Although our phase, thus their enhancement may reflect more efficacious memory study failed to detect consistent alterations in the neural activities of acquisition and consolidation in enriched animals (Buzsáki, 2015; Joo GABAergic interneurons in trisomic animals, our findings challenge the and Frank, 2018). Moreover, our findings unraveled again strong asso­ widely accepted hypothesis of overinhibition of principal cells in DS. ciations between HPC→PFC low gamma signals recorded during the They indeed suggest deficient inhibition of hippocampal-prefrontal visits to familiar objects and discrimination indices, confirming that neural networks in vivo in DS that result in increased and disorga­ both male and female mice with robust HPC→PFC gamma connectivity nized spiking activity of principal neurons. These results are in line with exhibit superior memory performances in the NOR task (Alemany- recent investigations that propose that a disturbed inhibitory tone rather González et al., 2020). Conflicting results exist on whether recognition than a generalized overinhibition underlies some of the characteristic memory for objects requires the PFC (Spanswick and Dyck, 2012; Morici cognitive deficits in DS (Valbuena et al., 2021). et al., 2015). A recent study however implicates the medial PFC in this In a previous study conducted in Ts65Dn males we investigated the type of memory via optogenetic interrogation and identifies HPC→PFC neural substrates underlying the rescuing abilities of a green tea extract coupling as a major neural mechanism involved (Wang et al., 2021). In containing epigallocatechin-3-gallate (EGCG) (Alemany-González et al., summary, the neurophysiological changes identified here in enriched 2020), a DYRK1A kinase inhibitor with pro-cognitive effects in in­ wild-type animals combined task-independent adjustments (augmented dividuals with DS (de la Torre et al., 2016). Considering that equivalent hippocampal pyramidal activity and gamma synchrony across different pathological synchrony was detected in Ts65Dn females and males, this brain states) and memory-dependent adjustments (enhanced theta- study provides unique insights into the comparative rescuing effects of gamma coupling and ripples in the HPC associated with memory EE and EGCG in the same mouse model of DS. We found that both pro- retrieval). cognitive treatments attenuated aberrant theta, gamma, and ripple Non-enriched trisomic females exhibited robust increases in theta power within the circuit and corrected HPC→PFC gamma directionality and gamma power in the HPC and PFC across all brain states. This is in during memory retrieval. Green tea extracts, EE and their combination line with previous studies reporting increased theta-gamma oscillations restored >70% of the phosphoprotein deregulation in Ts65Dn mice and synchrony within hippocampal-prefrontal pathways in mouse while re-establishment of a proper epigenetic state and rescue of the models of DS and other intellectual disabilities (Arbab et al., 2018; kinome deregulation may contribute to the cognitive rescue induced by Chang et al., 2020; Ramon-Duaso et al., 2019). Here, we also detected green tea extracts (De Toma et al., 2020). Future studies should inves­ broad theta-high frequency coupling and enlarged ripples in the HPC. tigate the neural substrates of a combination of EE and EGCG in diploid Finally, trisomic mice exhibited reversed HPC→PFC gamma connectiv­ and trisomic animals. We also note that the Ts65Dn model is known to ity during memory retrieval. Aberrant theta, gamma, ripple power and contain additional genes not triplicated in human DS (Constestabile circuit miscommunication were also observed in trisomic male mice of et al., 2017) that might impact the phenotypes observed here. Therefore, the same transgenic line (Alemany-González et al., 2020), demon­ it would be important to confirm the findings presented here in other strating similar aberrant synchrony in DS subjects of both genders. In mouse lines of DS. fact, they predicted poor memory performances and were attenuated in trisomic EE animals, unraveling their critical involvement in memory 5. Concluding remarks processing. These results suggest that normalizing hippocampal- prefrontal hypersynchrony and miscommunication are major neural In closing, neural activities associated with brain state adjustments mechanisms underlying the beneficial effects of EE in intellectual and memory-associated adjustments in the HPC are good candidates to disability. underlie the beneficial effects of EE on cognition in wild-type mice. EE enhanced gamma synchrony in diploid mice and reduced aber­ Concomitantly, EE attenuates hippocampal and prefrontal pathological rant gamma synchrony in trisomic mice. How can this be reconciled? hypersynchrony in trisomic mice. This strongly suggests distinct neural 13 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 mechanisms for the generation and rescue of healthy and pathological Constestabile, A., Magara, S., Cancedda, L., 2017. The GABAergic hypothesis for cognitive disabilities in Down syndrome. Front. Cell. Neurosci. 11, 54. https://doi. brain synchrony, respectively, by EE. org/10.3389/fncel.2017.00054. Cramer, N., Galdzicki, Z., 2012. From abnormal hippocampal synaptic plasticity in down CRediT authorship contribution statement syndrome mouse models to cognitive disability in Down syndrome. Neural Plast. 2012, 101542. https://doi.org/10.1155/2012/101542. De la Torre, R., de Sola, S., Hernandez, G., Farré, M., Pujol, J., Rodriguez, J., Espadaler, J. Maria Alemany-González: Conceptualization, Investigation, M., Langohr, K., Cuenca-Royo, A., Principe, A., Xicota, L., Janel, N., Catuara- Formal analysis, Visualization, Writing – original draft, Writing – review Solarz, S., Sanchez-Benavides, G., Bléhaut, H., Dueñas-Espín, I., del Hoyo, L., & editing. Marta Vilademunt: Formal analysis, Visualization, Writing – Benejam, B., Blanco-Hinojo, L., Videla, S., Fitó, M., Delabar, J.M., Dierssen, M., TESDAD study group, 2016. Safety and efficacy of cognitive training plus review & editing. Thomas Gener: Conceptualization, Investigation, epigallocatechin-3-gallate in young adults with Down’s syndrome (TESDAD): a Visualization, Supervision, Writing – review & editing. M. Victoria double-blind, randomised, placebo-controlled, phase 2 trial. Lancet Neurol. 15, Puig: Conceptualization, Supervision, Formal analysis, Writing – orig­ 801–810. https://doi.org/10.1016/S1474-4422(16)30034-5. De Toma, I., Ortega, M., Aloy, P., Sabidó, E., Dierssen, M., 2019. DYRK1A overexpression inal draft, Writing – review & editing, Funding acquisition. alters cognition and neural-related proteomic pathways in the hippocampus that are rescued by green tea extract and/or environmental enrichment. Front. Mol. Neurosci. 12 https://doi.org/10.3389/fnmol.2019.00272. Declaration of Competing Interest De Toma, I., Ortega, M., Catuara-Solarz, S., Sierra, C., Sabidó, E., Dierssen, M., 2020. Re- establishment of the epigenetic state and rescue of kinome deregulation in Ts65Dn None. mice upon treatment with green tea extract and environmental enrichment. Sci. Rep. 10, 16023. https://doi.org/10.1038/s41598-020-72625-z. Delgado-Sallent, C., Nebot, P., Gener, T., Fath, A.B., Timplalexi, M., Puig, M.V., 2021. Data availability Atypical, but not typical, antipsychotic drugs reduce hypersynchronized prefrontal- hippocampal circuits during psychosis-like states in mice: contribution of 5-HT2A Data will be made available on request. and 5-HT1A receptors. Cereb. Cortex N Y N 1991 bhab427. https://doi.org/ 10.1093/cercor/bhab427. Faizi, M., Bader, P.L., Tun, C., Encarnacion, A., Kleschevnikov, A., Belichenko, P., Acknowledgements Saw, N., Priestley, M., Tsien, R.W., Mobley, W.C., Shamloo, M., 2011. Comprehensive behavioral phenotyping of Ts65Dn mouse model of Down syndrome: activation of β1-adrenergic receptor by xamoterol as a potential cognitive enhancer. We are grateful to Mara Dierssen and Jose A. Garrido for providing Neurobiol. Dis. 43, 397–413. https://doi.org/10.1016/j.nbd.2011.04.011. scientific insight and to Manuel Valero for providing scripts to analyze Figueroa-Jimenez, M.D., Cañete-Massé, C., Carbó-Carreté, M., Zarabozo-Hurtado, D., ripple activity and entropy. We also thank Pau Nebot for excellent Peró-Cebollero, M., Salazar-Estrada, J.G., Guàrdia-Olmos, J., 2021. Resting-state default mode network connectivity in young individuals with Down syndrome. Brain technical support. This work was supported by the Jerome Lejeune Behav. 11, e01905 https://doi.org/10.1002/brb3.1905. Foundation (grant #1419) and by the Spanish government grants Gener, T., Tauste-Campo, A., Alemany-González, M., Nebot, P., Delgado-Sallent, C., SAF2016-80726-R and PID2019-104683RB-I00 funded by MCIN/AEI/ Chanovas, J., Puig, M.V., 2019. Serotonin 5-HT1A, 5-HT2A and dopamine D2 receptors strongly influence prefronto-hippocampal neural networks in alert mice: 10.13039/501100011033 and by ERDF “A way of making Europe”. contribution to the actions of risperidone. Neuropharmacology 158, 107743. https://doi.org/10.1016/j.neuropharm.2019.107743. References Gulinello, M., Mitchell, H.A., Chang, Q., Timothy O’brien, W., Zhou, Z., Abel, T., Wang, L., Corbin, J.G., Veeraragavan, S., Samaco, R.C., Andrews, N.A., Fagiolini, M., Cole, T.B., Burbacher, T.M., Crawley, J.N., 2018. Rigor and reproducibility in rodent Alemany-González, M., Gener, T., Nebot, P., Vilademunt, M., Dierssen, M., Puig, M.V., behavioral research. Neurobiol. Learn. Mem. S1074-7427, 30001–30007. https:// 2020. Prefrontal-hippocampal functional connectivity encodes recognition memory doi.org/10.1016/j.nlm.2018.01.001. and is impaired in intellectual disability. Proc. Natl. Acad. Sci. 117, 11788–11798. Guyon, N., Zacharias, L.R., de Oliveira, E.F., Kim, H., Leite, J.P., Lopes-Aguiar, C., https://doi.org/10.1073/pnas.1921314117. Carlén, M., 2021. Network asynchrony underlying increased broadband gamma Anagnostopoulou, A., Styliadis, C., Kartsidis, P., Romanopoulou, E., Zilidou, V., power. J. Neurosci. https://doi.org/10.1523/JNEUROSCI.2250-20.2021. Karali, C., Karagianni, M., Klados, M., Paraskevopoulos, E., Bamidis, P.D., 2021. Hirase, H., Shinohara, Y., 2014. Transformation of cortical and hippocampal neural Computerized physical and cognitive training improves the functional architecture circuit by environmental enrichment. Neuroscience 280, 282–298. https://doi.org/ of the brain in adults with Down syndrome: a network science EEG study. Netw. 10.1016/J.NEUROSCIENCE.2014.09.031. Neurosci. Camb. Mass. 5, 274–294. https://doi.org/10.1162/netn_a_00177. Jensen, O., Colgin, L.L., 2007. Cross-frequency coupling between neuronal oscillations. Anderson, J.S., Nielsen, J.A., Ferguson, M.A., Burback, M.C., Cox, E.T., Dai, L., Gerig, G., Trends Cogn. Sci. 11, 267–269. https://doi.org/10.1016/j.tics.2007.05.003. Edgin, J.O., Korenberg, J.R., 2013. Abnormal brain synchrony in Down syndrome. Joo, H.R., Frank, L.M., 2018. The hippocampal sharp wave–ripple in memory retrieval NeuroImage Clin. 2, 703–715. https://doi.org/10.1016/j.nicl.2013.05.006. for immediate use and consolidation. Nat. Rev. Neurosci. 19, 744–757. https://doi. Arbab, T., Battaglia, F.P., Pennartz, C.M.A., Bosman, C.A., 2018. Abnormal hippocampal org/10.1038/s41583-018-0077-1. theta and gamma hypersynchrony produces network and spike timing disturbances Karaaslan, O., Mahoney, G., 2013. Effectiveness of responsive teaching with children in the Fmr1-KO mouse model of Fragile X syndrome. Neurobiol. Dis. 114, 65–73. with Down syndrome. Intellect. Dev. Disabil. 51, 458–469. https://doi.org/10.1352/ https://doi.org/10.1016/j.nbd.2018.02.011. 1934-9556-51.6.458. Begenisic, T., Spolidoro, M., Braschi, C., Baroncelli, L., Milanese, M., Pietra, G., Kim, H., Ährlund-Richter, S., Wang, X., Deisseroth, K., Carlén, M., 2016. Prefrontal Fabbri, M.E., Bonanno, G., Cioni, G., Maffei, L., Sale, A., 2011. Environmental parvalbumin neurons in control of attention. Cell 164, 208–218. https://doi.org/ enrichment decreases GABAergic inhibition and improves cognitive abilities, 10.1016/j.cell.2015.11.038. synaptic plasticity, and visual functions in a mouse model of Down syndrome. Front. Kleschevnikov, A.M., Belichenko, P.V., Villar, A.J., Epstein, C.J., Malenka, R.C., Cell. Neurosci. 5, 29. https://doi.org/10.3389/fncel.2011.00029. Mobley, W.C., 2004. Hippocampal long-term potentiation suppressed by increased Begenisic, T., Sansevero, G., Baroncelli, L., Cioni, G., Sale, A., 2015. Early environmental inhibition in the Ts65Dn mouse, a genetic model of Down syndrome. J. Neurosci. 24, therapy rescues brain development in a mouse model of Down syndrome. Neurobiol. 8153–8160. https://doi.org/10.1523/JNEUROSCI.1766-04.2004. Dis. 82, 409–419. https://doi.org/10.1016/j.nbd.2015.07.014. Leger, M., Quiedeville, A., Bouet, V., Haelewyn, B., Boulouard, M., Schumann-Bard, P., Belichenko, P.V., Kleschevnikov, A.M., Salehi, A., Epstein, C.J., Mobley, W.C., 2007. Freret, T., 2013. Object recognition test in mice. Nat. Protoc. 8, 2531–2537. https:// Synaptic and cognitive abnormalities in mouse models of Down syndrome: exploring doi.org/10.1038/nprot.2013.155. genotype-phenotype relationships. J. Comp. Neurol. 504, 329–345. https://doi.org/ Leger, M., Paizanis, E., Dzahini, K., Quiedeville, A., Bouet, V., Cassel, J.-C., Freret, T., 10.1002/cne.21433. Schumann-Bard, P., Boulouard, M., 2015. Environmental enrichment duration Bokil, H., Andrews, P., Kulkarni, J.E., Mehta, S., Mitra, P.P., 2010. Chronux: a platform differentially affects behavior and neuroplasticity in adult mice. Cereb. Cortex 25, for analyzing neural signals. J. Neurosci. Methods 192, 146–151. https://doi.org/ 4048–4061. https://doi.org/10.1093/cercor/bhu119. 10.1016/j.jneumeth.2010.06.020. Livingston, G., Sommerlad, A., Orgeta, V., Costafreda, S.G., Huntley, J., Ames, D., Brenes, J.C., Lackinger, M., Höglinger, G.U., Schratt, G., Schwarting, R.K.W., Wöhr, M., Ballard, C., Banerjee, S., Burns, A., Cohen-Mansfield, J., Cooper, C., Fox, N., Gitlin, L. 2016. Differential effects of social and physical environmental enrichment on brain N., Howard, R., Kales, H.C., Larson, E.B., Ritchie, K., Rockwood, K., Sampson, E.L., plasticity, cognition, and ultrasonic communication in rats. J. Comp. Neurol. 524, Samus, Q., Schneider, L.S., Selbæk, G., Teri, L., Mukadam, N., 2017. Dementia 1586–1607. https://doi.org/10.1002/cne.23842. prevention, intervention, and care. Lancet Lond. Engl. 390, 2673–2734. https://doi. Buzsáki, G., 2002. Theta oscillations in the hippocampus. Neuron 33, 325–340. org/10.1016/S0140-6736(17)31363-6. Buzsáki, G., 2015. Hippocampal sharp wave-ripple: a cognitive biomarker for episodic Martínez Cué, C., Dierssen, M., 2020. Plasticity as a therapeutic target for improving memory and planning. Hippocampus 25, 1073–1188. https://doi.org/10.1002/ cognition and behavior in Down syndrome. In: Progress in Brain Research. Elsevier hipo.22488. B.V, pp. 269–302. https://doi.org/10.1016/bs.pbr.2019.11.001. Chang, P., Bush, D., Schorge, S., Good, M., Canonica, T., Shing, N., Noy, S., Wiseman, F. Martínez-Cué, C., Baamonde, C., Lumbreras, M., Paz, J., Davisson, M.T., Schmidt, C., K., Burgess, N., Tybulewicz, V.L.J., Walker, M.C., Fisher, E.M.C., 2020. Altered Dierssen, M., Flórez, J., 2002. Differential effects of environmental enrichment on hippocampal-prefrontal neural dynamics in mouse models of Down syndrome. Cell behavior and learning of male and female Ts65Dn mice, a model for Down Rep. 30, 1152–1163 e4. https://doi.org/10.1016/j.celrep.2019.12.065. 14 M. Alemany-González et al. Neurobiology of Disease 173 (2022) 105841 syndrome. Behav. Brain Res. 134, 185–200. https://doi.org/10.1016/s0166-4328 Shinohara, Y., Hosoya, A., Hirase, H., 2013. Experience enhances gamma oscillations and (02)00026-8. interhemispheric asymmetry in the hippocampus. Nat. Commun. 4 https://doi.org/ Martínez-Cué, C., Rueda, N., García, E., Davisson, M.T., Schmidt, C., Flórez, J., 2005. 10.1038/ncomms2658. Behavioral, cognitive and biochemical responses to different environmental Sohal, V.S., Rubenstein, J.L.R., 2019. Excitation-inhibition balance as a framework for conditions in male Ts65Dn mice, a model of Down syndrome. Behav. Brain Res. 163, investigating mechanisms in neuropsychiatric disorders. Mol. Psychiatry 24, 174–185. https://doi.org/10.1016/j.bbr.2005.04.016. 1248–1257. https://doi.org/10.1038/s41380-019-0426-0. McShane, B.B., Galante, R.J., Biber, M., Jensen, S.T., Wyner, A.J., Pack, A.I., 2012. Spanswick, S.C., Dyck, R.H., 2012. Object/context specific memory deficits following Assessing REM sleep in mice using video data. Sleep 35, 433–442. https://doi.org/ medial frontal cortex damage in mice. PLoS One 7, e43698. https://doi.org/ 10.5665/sleep.1712. 10.1371/journal.pone.0043698. Mesa-Gresa, P., Pérez-Martinez, A., Redolat, R., 2013. Environmental enrichment Steriade, M., 2006. Grouping of brain rhythms in corticothalamic systems. Neuroscience improves novel object recognition and enhances agonistic behavior in male mice. 137, 1087–1106. https://doi.org/10.1016/j.neuroscience.2005.10.029. Aggress. Behav. 39, 269–279. https://doi.org/10.1002/ab.21481. Tanaka, M., Wang, X., Mikoshiba, K., Hirase, H., Shinohara, Y., 2017. Rearing- Morici, F.J., Bekinschtein, P., Weisstaub, N.V., 2015. Medial prefrontal cortex role in environment-dependent hippocampal local field potential differences in wild-type recognition memory in rodents. Behav. Brain Res. 292, 241–251. and inositol trisphosphate receptor type 2 knockout mice. J. Physiol. 595, Nithianantharajah, J., Hannan, A.J., 2006. Enriched environments, experience- 6557–6568. https://doi.org/10.1113/JP274573. dependent plasticity and disorders of the nervous system. Nat. Rev. Neurosci. 7, Tort, A.B.L., Komorowski, R., Eichenbaum, H., Kopell, N., 2010. Measuring phase- 697–709. https://doi.org/10.1038/nrn1970. amplitude coupling between neuronal oscillations of different frequencies. Nolte, G., Ziehe, A., Nikulin, V.V., Schlögl, A., Krämer, N., Brismar, T., Müller, K.-R., J. Neurophysiol. 104, 1195–1210. https://doi.org/10.1152/jn.00106.2010. 2008. Robustly estimating the flow direction of information in complex physical Valbuena, S., García, A., Mazier, W., Paternain, A.V., Lerma, J., 2021. Unbalanced systems. Phys. Rev. Lett. 100, 234101. https://doi.org/10.1103/ dendritic inhibition of CA1 neurons drives spatial-memory deficits in the Ts2Cje PhysRevLett.100.234101. Down syndrome model. Nat. Commun. 10, 4991. https://doi.org/10.1038/s41467- Pons-espinal, M., Lagran, M.M.D., Dierssen, M., 2013. Environmental enrichment rescues 019-13004-9. DYRK1A activity and hippocampal adult neurogenesis in TgDyrk1A. Neurobiol. Dis. Valero, M., Averkin, R.G., Fernandez-lamo, I., Cid, E., Tamas, G., Valero, M., Averkin, R. 60, 18–31. https://doi.org/10.1016/j.nbd.2013.08.008. G., Fernandez-lamo, I., Aguilar, J., Lopez-pigozzi, D., Brotons-mas, J.R., Cid, E., Puig, M.V., Alemany-González, M., Gener, T., 2020. Data for prefrontal-hippocampal Tamas, G., Menendez de la Prida, L., 2017. Mechanisms for selective single-cell functional connectivity encodes recognition memory and is impaired in intellectual reactivation during offline sharp-wave ripples and their distortion by fast ripples. disability 1. https://doi.org/10.17632/wg4zm32gsb.1. Neuron 94, 1234–1247.e7. https://doi.org/10.1016/j.neuron.2017.05.032. Pujol, J., del Hoyo, L., Blanco-Hinojo, L., de Sola, S., Macià, D., Martínez-Vilavella, G., van Praag, H., Kempermann, G., Gage, F.H., 2000. Neural consequences of Amor, M., Deus, J., Rodríguez, J., Farré, M., Dierssen, M., de la Torre, R., 2015. environmental enrichment. Nat. Rev. Neurosci. 1, 191–198. https://doi.org/ Anomalous brain functional connectivity contributing to poor adaptive behavior in 10.1038/35044558. Down syndrome. Cortex 64, 148–156. https://doi.org/10.1016/j. Velikova, S., Magnani, G., Arcari, C., Falautano, M., Franceschi, M., Comi, G., Leocani, L., cortex.2014.10.012. 2011. Cognitive impairment and EEG background activity in adults with Down’s Ramírez-Toraño, F., García-Alba, J., Bruña, R., Esteba-Castillo, S., Vaquero, L., syndrome: a topographic study. Hum. Brain Mapp. 32, 716–729. https://doi.org/ Pereda, E., Maestú, F., Fernández, A., 2021. Hypersynchronized 10.1002/hbm.21061. magnetoencephalography brain networks in patients with mild cognitive Wang, H., Xu Xiaxia, Xu, Xinxin, Gao J., Zhang, T., 2020. Enriched environment and impairment and Alzheimer’s disease in Down syndrome. Brain Connect. https://doi. social isolation affect cognition ability via altering excitatory and inhibitory synaptic org/10.1089/brain.2020.0897. density in mice hippocampus. Neurochem. Res. 45, 2417–2432. https://doi.org/ Ramon-Duaso, C., Gener, T., Consegal, M., Fernández-Avilés, C., Gallego, J.J., 10.1007/s11064-020-03102-2. Castarlenas, L., Swanson, M.S., de la Torre, R., Maldonado, R., Puig, M.V., Wang, C., Furlong, T.M., Stratton, P.G., Lee, C.C.Y., Xu, L., Merlin, S., Nolan, C., Robledo, P., 2019. Methylphenidate attenuates the cognitive and mood alterations Arabzadeh, E., Marek, R., Sah, P., 2021. Hippocampus–prefrontal coupling regulates observed in Mbnl2 knockout mice and reduces microglia overexpression. Cereb. recognition memory for novelty discrimination. J. Neurosci. 41, 9617–9632. https:// Cortex 29, 2978–2997. https://doi.org/10.1093/cercor/bhy164. doi.org/10.1523/JNEUROSCI.1202-21.2021. Raveau, M., Polygalov, D., Boehringer, R., Amano, K., Yamakawa, K., Mchugh, T.J., Warburton, E.C., Brown, M.W., 2015. Neural circuitry for rat recognition memory. 2018. Alterations of in vivo CA1 network activity in Dp(16)1Yey Down syndrome Behav. Brain Res. 285, 131–139. https://doi.org/10.1016/j.bbr.2014.09.050. model mice. Elife 7, e31543. https://doi.org/10.7554/eLife.31543.001. Yizhar, O., Fenno, L.E., Prigge, M., Schneider, F., Davidson, T.J., O’Shea, D.J., Sohal, V. Ruiz-Mejias, M., 2019. Outer brain oscillations in Down syndrome. Front. Syst. Neurosci. S., Goshen, I., Finkelstein, J., Paz, J.T., Stehfest, K., Fudim, R., Ramakrishnan, C., 13, 17. https://doi.org/10.3389/fnsys.2019.00017. Huguenard, J.R., Hegemann, P., Deisseroth, K., 2011. Neocortical excitation/ Scheffer-Teixeira, R., Tort, A.B.L., 2017. Unveiling fast field oscillations through inhibition balance in information processing and social dysfunction. Nature 477, comodulation. eNeuro. https://doi.org/10.1523/ENEURO.0079-17.2017, 4: 171–178. https://doi.org/10.1038/nature10360. ENEURO.0079-17.2017. 15