Down syndrome, caused by trisomy 21, affects around six million people worldwide and features learning, memory and language deficits. However, the mechanisms underlying t... More
Down syndrome, caused by trisomy 21, affects around six million people worldwide and features learning, memory and language deficits. However, the mechanisms underlying trisomy 21 neurophenotypes involving human cortical circuitry are unknown. By characterising developing neural network dynamics and single cell excitability profiles, from synaptic and voltage-dependent ion channel behaviour using an isogenic induced pluripotent stem cell-derived neuronal model, we show that trisomy 21 impairs the activity and development of cortical circuitry. This is caused by deficient glutamatergic synaptic connectivity and by aberrant intrinsic membrane properties involving K+ and Na+ channels culminating in spike firing defects that weaken neural network activity and disrupt the synchrony of developing neurons. We also identify transiently activated A-type K+ channels, specifically Kv4.3 channels, as a key orchestrator for Down syndrome during neurodevelopment. Overall, these excitability changes will significantly contribute towards the aberrant neurophenotypes observed later on in life.
Similar content being viewed by others
Abnormal neuronal and synaptic morphology in Down syndrome brains reproduces in human isogenic cellular models
Article Open access 10 June 2026
Consequences of trisomy 21 for brain development in Down syndrome
Article 08 October 2024
Molecular cartography of the human down syndrome and trisomic mouse brain
Article Open access 30 September 2025
Explore related subjects
Discover the latest articles and news in related subjects.
Molecular neuroscience
Neurophysiology
Introduction
Down syndrome is the most prevalent genetic cause of intellectual disability caused by trisomy of human chromosome 21 (hCh21)1,2. Multiple organ systems are affected by the presence of an extra copy of hCh21, but the most debilitating impact of trisomy evident in early life is reflected by the neurophenotype including: neurodevelopmental, psychiatric, neurological and neurodegenerative conditions, which are typically severe and life-changing3. Therefore, identifying the mechanisms that underlie these nervous system phenotypes has been a priority area for Down syndrome research. An increase in gene dosage has long been postulated to cause Down syndrome phenotypes, but in regard to neuronal excitability, there is a paucity of ion channel and neurotransmitter receptor genes located on hCh21, e.g. KCNE1-2, TRPM2, KCNJ6, GRIK1, FNAR1, IFNAR2, IFNGR2, and IL10RB4,5,6,7,8 suggesting that genome wide effects of hCh21 genes are important for establishing the mechanistic basis for Down syndrome.
To date, perturbations to several brain receptors and ion channel signalling pathways have been implicated in Down syndrome, the most prominent to date being the GABAergic signalling pathway9,10,11,12. Reduced cell numbers and depleted synaptic density, and altered neural innervation patterns all feature prominently13,14,15,16,17, but despite this, the development of early human neural networks in Down syndrome, and how cellular and synaptic excitability is affected remains mostly unknown. Here, by using isogenic human cortical neurons derived from induced pluripotent stem cells (iPSCs)18, we investigate neural network activity and functional profiles of single neurons. These cellular models allow critical insights into early developing cortical networks19,20,21 that are impossible to study in vivo. They are beneficial since they are human in origin and contain the same complement of genes with or without increased dosage of those on hCh2118,22. Our study reveals that early Down syndrome developing networks are characterised by widespread changes to excitatory (glutamatergic) synaptic connectivity and aberrant intrinsic excitability involving Na+- and K+-channels. Moreover, by deploying a combination of K+ channel pharmacology, transcriptomics and mathematical modelling, we postulate that Kv4.3 channels could be key mediators of the network and excitability deficits that feature so prominently in trisomy 21.
Results
Developmental anomalies in neural network dynamics resulting from trisomy 21
The longitudinal developmental activity profiles of human Down syndrome neurons were probed by studying spike firing properties of two sets of isogenic cell lines derived from a Down syndrome donor mosaic for trisomy 21, which are either euploid controls (C3 and C9) or aneuploid trisomy 21 (C5 and C13) neurons that reproduce and underpin clinical Down syndrome phenotypes18,22,23. To establish the activity profiles of these cell lines, extracellular electrical activity from 2D cultures was recorded using microelectrode arrays (MEAs; Fig. 1A, B) populated with neurons at 4 − 14 weeks in vitro. MEA was preferred over imaging approaches as it is a direct measure of neuronal activity rather than a proxy, and also because early-stage iPSC-derived neurons and immature neurons display developmentally regulated slow Ca2+ oscillations24 which can make the use of Ca2+ to gauge action potential firing unreliable. Thus, the use of MEA recordings ensured that only action potential-dependent changes to excitability were analysed.
Fig. 1: Early developmental trisomy 21 neurons are deficient in spiking and network synchrony.
Fig. 1: Early developmental trisomy 21 neurons are deficient in spiking and network synchrony.
Full size image
A Timeline of electrophysiological characterisation after neural stem cell seeding for terminal differentiation, along with position of 8 ×8 recording (300 µm separation) and ground electrodes (gray). Raster plots showing spikes (white) and bursting of disomic (D21) and trisomic (T21) cells at 4, 10 and 14 weeks (wks in vitro). Scale bar 5 s. B Heat maps of spiking detected by the recording electrodes show greater activity of disomic cells across neural networks. Heat map legends depict spike rates (spikes / second). C Weighted mean spiking rate of disomic and trisomic networks. D Single electrode bursts including spikes per burst, burst duration and inter-burst intervals of disomic and trisomic networks. E Network burst parameters including network burst frequency and spikes per burst of disomic and trisomic networks. F Left, Network synchrony depicted by normalised cross correlation between network bursts at 14-weeks for disomic and trisomic wells. Right, synchrony metric showing average area under the normalised cross-correlation curves of disomic and trisomic cells. The results here are from five different differentiations of two different lines per genotype and at least four wells per time point. n = 4–11 wells. Error bars in this and preceding figures, unless otherwise stated, depict standard error of means (SEM). *P 93%) excitatory (Figure S1H), ascertained by the expression of VGLUT1. They exhibited no overt functional inter-clonal variations in electrophysiological properties between euploid (C3 and C9) or isogenic trisomy 21 (C5 and C13; Figure S2) and therefore we pooled the results according to their genotype. This allowed the comparison of the impact of trisomy on neural circuitry to be assessed during development.
Throughout the course of our developmental timeframe, compared to disomic counterparts, trisomic cells were deficient in action potential spiking. This was manifest by a markedly reduced spiking rate in addition to impaired bursting properties, assessed from single electrode bursts including spikes per burst, burst duration, inter-burst interval and burst frequency, all recorded using MEA (Fig. 1C, D; and Figure S3). Consequently, network bursting, defined as synchronous bursting at a minimum of 20-35% of electrodes, was reduced in number, frequency and synchrony for trisomic neurons. Moreover, network bursts were characterised by fewer spikes in the trisomic cells (Fig. 1E, F; Figure S3) implicating impaired spike firing at the single-cell level as a potential mechanism for the difference between trisomic and disomic cells. The presence of cells in direct contact with the electrode arrays was monitored throughout so we could discount an absence of cells due to cell death or mechanical shearing as an explanation for spiking and network activity deficits (Figure S3).
Network bursts and spike synchrony of disomic and trisomic neurons were blocked by the Na+ channel inhibitor tetrodotoxin (0.5 µM25) and inhibited by the AMPA receptor and NMDA receptor antagonists, CNQX (10 µM)25 and APV (25 µM26; Figure S3G–J) respectively, confirming that intrinsic and synaptic mechanisms were integral for network dynamics. These results suggest that synaptic and cellular anomalies affecting spiking could together underlie the aberrant developmental neural network dynamics due to trisomy 21.
Defects in synaptic transmission due to trisomy 21
A wide range of synaptic deficits characterise murine models of Down syndrome including changes to GABAergic neurotransmission and synaptic plasticity10. However, the impact of Down syndrome on excitatory neurotransmission, particularly in the context of human neurons, has received little attention. Our iPSC differentiation yielded a majority of excitatory neurons ( > 93%) and consistently these cultures were near-devoid of inhibitory interneuron activity exemplified by the absence of GABAergic synaptic activity in 88-98% of cells despite cell surface expression of functional GABAA receptors (Figure S4). We detected GABAergic postsynaptic currents in only one of eight differentiations that were blocked by the antagonist bicuculline27 (Figure S4). Since inhibitory synaptic activity could arise due to altered cellular differentiation trajectories, for consistency, to avoid any aberrant differentiation artefacts, we excluded this batch of iPSCs from all analyses in this study.
We characterised glutamatergic neurotransmission for iPSC-derived neurons using voltage clamp recordings after they reached the plateau phase of neural network development at 4–6 months. Disomic and trisomic cells were similar in size, deduced by voltage clamp capacitance discharge curves (with trending towards smaller trisomic cells; Figure S4E). Furthermore, these cells produced functional AMPA and NMDA receptor-mediated whole-cell responses (Figure S4). Interestingly, a larger proportion of trisomic neurons (20%; Fig. 2A) did not receive any synaptic inputs compared to disomic neurons (4%) in standard saline solution. Synaptic inputs onto these neurons were exclusively excitatory characterised by their fast rise times and decay kinetics and by charge transfer28,29 (Figure S4G). Definitively, the synaptic currents were abolished by the glutamate receptor blockers, CNQX and APV (Fig. 2B). Importantly, the frequency of trisomic excitatory postsynaptic currents (EPSCs) was only 25% of that recorded in disomic cells without any change to EPSC amplitude or kinetics due to unchanged AMPA receptor expression or clustering (Fig. 2B–E; and Figure S4). These results suggest that during early development trisomic neurons exhibit a severe deficiency in glutamatergic synaptic connectivity without notable changes to cell surface AMPA receptor numbers.
Fig. 2: Synaptic and developmental wiring defects due to trisomy 21.
Fig. 2: Synaptic and developmental wiring defects due to trisomy 21.
Full size image
A Pie charts showing proportion of disomic (D21) and trisomic (T21) cells exhibiting excitatory postsynaptic currents (EPSCs). B EPSCs were blocked by CNQX (10 μM) and APV (25 μM) and had fast kinetics with no residual GABAergic currents in the presence of glutamatergic blockers. C–E Reduced frequency (C) but not amplitude (D) and kinetics (E) of EPSCs in T21 cells in normal saline. F Pie charts showing that all D21 cells received EPSCs in Mg2+-free saline but 10% of T21 cells did not. G Representative EPSC recordings of D21 and T21 cells outside bursts in 0 Mg2+. H–J Reduced frequency (H) but not amplitude (I) or kinetics (J) of EPSCs of T21 cells in 0 Mg2+ saline. K Confocal images and localisation of presynaptic terminal marker synapsin-1 with pan-axonal neurofilament marker SMI-312 and dendritic marker MAP2 counterstained with nuclei marker DAPI in D21 and T21 cells. Scale bars 10 µm. n = 11–59; number of cells/ 3D image stacks. N numbers of 3D image stacks have been depicted in brackets along with p values on graphs; two-tailed unpaired t-test or Mann-Whitney test. Box plots in C, H show median, 25–75% interquartile range and 2-95% whiskers. Bar charts in D, I and K depict means +/- SEM.
To explore the paucity of synaptic currents in disomic and trisomic cells, we increased the excitability of our neurons by removing Mg2+ from the bathing solution (0 Mg2+) to relieve inhibition of NMDA receptors30. As expected, in the absence of Mg2+, neurons showed greater synaptic activity but 10% of trisomic cells (compared to 0% disomic neurons) did not exhibit synaptic currents despite undergoing 0 Mg2+-induced membrane bursting (Fig. 2F) thereby identifying a proportion of cells that fail to integrate into cortical networks during development. Here, the frequency of EPSCs for synaptically connected trisomic cells was approximately 10% of those for disomic cells. The EPSC amplitude, kinetics and AMPA and NMDA current densities were unchanged (Fig. 2G–J; and Figure S4) confirming a likely deficit in synaptic connectivity involving glutamatergic synapses. The absence of NMDA-dependent synaptic transmission, in some neurons, could reflect silent synapses and will impact on synaptic plasticity with consequences for learning and memory.
This reduced connectivity facet was confirmed by immunolabelling for the presynaptic marker synapsin-131 and excitatory postsynaptic density protein PSD9532 which had reduced fluorescence in trisomic compared to disomic neurons (Fig. 2K, and Figure S5). These results suggest that impaired integration of glutamatergic cells into developing networks and reduced synaptic connectivity underlies the neural network defects in trisomy 21.
Analysis of single cell bursting in 0 Mg2+ also revealed deficits in network dynamics consistent with our MEA results. Using this approach, all disomic neurons studied exhibited bursting, a feature that is prevalent in mature neocortical neurons33 but approximately a third of trisomic cells did not exhibit membrane current oscillations (Fig. 3A). These oscillations correspond to cellular depolarisation following population action potential spiking34 and can be abolished by blocking AMPA receptors, NMDA receptors or restoring Mg2+ to the saline (Fig. 3B; and Figure S6A–C) suggesting that network bursting is dependent on glutamatergic synaptic activity. Trisomic network bursts recorded in postsynaptic voltage clamp were less frequent (by 33%) compared to disomic bursts and had nearly 66% reduction in current amplitudes and 70% decrease in charge transfer (Fig. 3C–F) similar to our MEA results (Fig. 1).
Fig. 3: Altered bursting of trisomy 21 neurons.
Fig. 3: Altered bursting of trisomy 21 neurons.
Full size image
A Pie charts showing that all disomic (D21) cells had bursts in Mg2+-free saline but around a third of trisomic (T21) cells did not. B Representative traces of membrane bursts of D21 and T21 cells in Mg2+-free saline. Introduction of Mg2+ saline abolishes bursts but not EPSCs (insets). C–F Lower bursting rate (C), burst amplitude (D), burst charge transfer (E) and altered burst kinetics (F) of T21 neurons compared to D21 cells. n = 17–43 cells have been depicted in brackets along with p values on graphs; two-tailed unpaired t-test or Mann-Whitney test. Box plots in (C, D) depict median, 25–75% interquartile range and 2–95% whiskers. Bar chart E depicts means +/- SEM.
Together these results suggest that trisomy 21 results in a reduction in glutamatergic synapses and synaptic activity, and during development a proportion of cells fail to connect into neural networks. This reduced connectivity is also manifest by the impaired bursting properties of trisomic neurons.
Intrinsic membrane and Na+ channel defects in human Down syndrome cells
Glutamatergic activity is important for network development, but since intrinsic membrane properties can determine the overall state of excitability, strength and synchronicity of neural networks, the excitability and spike firing properties of single neurons was explored using current clamp recordings. Trisomic neurons displayed a trend towards more depolarised resting membrane potentials even though the input resistance, membrane time constant and capacitance were unchanged (Fig. 4A; and Figure S6D). Although a difference of 6 mV in median resting potentials between disomic and trisomic cells could be relevant physiologically, injecting constant current steps to elicit spikes, revealed that the rheobase of the two cell types was unchanged (Fig. 4B). However, injecting current steps as increments of the rheobase revealed spike firing differences. Trisomic cells fired fewer action potentials compared disomic cells (Fig. 4C, D) in the absence of changes to the spike firing threshold (Fig. 4E).
Fig. 4: Spike firing anomalies due to trisomy 21.
Fig. 4: Spike firing anomalies due to trisomy 21.
Full size image
A Resting membrane potential of disomic (D21) and trisomic (T21) neurons. B Rheobase of D21 and T21 neurons. C Spike outputs elicited by injecting depolarising steps of currents as increments of rheobase (rheo). D Input-output relationships of spikes and current injection. E Threshold potential at which neurons fire action potentials at rheobase. F Current at which maximum spikes are elicited in a step protocol of fixed current increments. G Average latency of first spike for the rheobase step current injection protocol. H Example of spike jitter traces and average spike jitter at rheobase for D21 (n = 65) and T21 (n = 54) cells. I–M Average action potential waveform (I), peak potential (J), area (K), rise time (L) and T50 (M) of D21 and T21 spikes at rheobase. n = 54-112 cells. *P 1.3 and P Less
NAD⁺ homeostasis is vital for neuronal health, as demonstrated by the opposing roles of nicotinamide mononucleotide adenylyltransferase 2 (NMNAT2), a NAD⁺-synthesizin... More
NAD⁺ homeostasis is vital for neuronal health, as demonstrated by the opposing roles of nicotinamide mononucleotide adenylyltransferase 2 (NMNAT2), a NAD⁺-synthesizing enzyme, and sterile alpha and TIR motif-containing protein 1 (SARM1), a NAD⁺ hydrolase. Neurodegenerative insults that decrease NMNAT2 activate SARM1, leading to axon loss. To understand how the NMNAT2–SARM1 axis influences brain energy metabolism, multi-omics approaches are used to investigate the metabolic changes resulting from neuronal NMNAT2 loss. Loss of NMNAT2 in glutamatergic neurons leads to a significant metabolic shift in the cerebral cortex from glucose to lipid catabolism, reduced lipid abundance, and pronounced neurodegenerative phenotypes and motor behavioral deficits. These metabolic disturbances are accompanied by altered glial expression of enzymes regulating glucose and lipid metabolism, enhanced inflammatory signaling, and disrupted astrocytic transcriptomic profiles related to cholesterol synthesis and immune activation. Notably, SARM1 deletion in NMNAT2-deficient mice restored lipid metabolism, astrocyte transcriptomic profiles, and mitigated neurodegeneration and motor behaviors. These findings suggest that neuronal NAD⁺ depletion triggers maladaptive, SARM1-dependent metabolic reprogramming, shifting energy use from glucose to lipids, which in turn promotes inflammation and neurodegeneration. Less
Microglial cells (MGs), originally derived from progenitor cells in a yolk sac during early development, are glial cells located in a physiological and pathological brain... More
Microglial cells (MGs), originally derived from progenitor cells in a yolk sac during early development, are glial cells located in a physiological and pathological brain. Since the brain contains various cell types, MGs could frequently interact with different cells, such as astrocytes (ACs), pericytes (PCs), and endothelial cells (ECs). However, how microglial traits are regulated via cell–cell interactions by ACs, PCs, or ECs and how they are different depending on the contacted cell types is unclear. This study aimed to clarify these questions by coculturing MGs with ACs, PCs, or ECs using mouse brain-derived cells, and microglial phenotypic changes were investigated under culture conditions that enabled direct cell–cell contact. Our results showed that ACs or PCs dose-dependently increased the number of MG, while ECs decreased it. Microarray and gene ontology analysis showed that cell fate-related genes (e.g., cell cycle, proliferation, growth, death, and apoptosis) of MGs were altered after a cell–cell contact with ACs, PCs, and ECs. Notably, microarray analysis showed that several genes, such as gap junction protein alpha 1 (Gja1), were prominently upregulated in MGs after coincubation with ACs, PCs, or ECs, regardless of cell types. Similarly, immunohistochemistry showed that an increased Gja1 expression was observed in MGs after coincubation with ACs, PCs, or ECs. Immunofluorescent and fluorescence-activated cell sorting analysis also showed that calcein-AM was transferred into MGs after coincubation with ACs, PCs, or ECs, confirming that intercellular interactions occurred between these cells. However, while Gja1 inhibition reduced the number of MGs after coincubation with ACs and PCs, this was increased after coincubation with ECs; this indicates that ACs and PCs positively regulate microglial numbers via Gja1, while ECs decrease it. Results show that ACs, PCs, or ECs exert both common and specific cell type-dependent effects on MGs through intercellular interactions. These findings also suggest that brain microglial phenotypes are different depending on their surrounding cell types, such as ACs, PCs, or ECs. Less
,-1-mg-ml--2.jpg)
