Down syndrome (DS), caused by trisomy of chromosome 21, is the leading genetic cause of intellectual disability, yet the mechanisms disrupting fetal brain development remain unclear. We performed single-cell transcriptomic and chromatin accessibility profiling of approximately 250,000 cells from 15 DS and 15 control human fetal cortices (10–20 weeks post-conception). Our analysis revealed a subtype-specific reduction in RORB/FOXP1-expressing excitatory neurons and widespread disruption of neurodevelopmental transcriptional programs. Chromosome 21 transcription factors (TFs) BACH1, PKNOX1, and GABPA emerged as dosage-sensitive hubs regulating genes linked to intellectual disability. Antisense oligonucleotide-mediated normalization of these TFs in human neural progenitors in vitro partially rescued target gene expression. Benchmarking a humanized in vivo model captured additional molecular and cellular signatures of DS, complementing the in vitro model. Together, we present a resource defining the gene-regulatory landscape underlying cortical development in DS and highlight molecular pathways for further investigation.

Aging is a key risk factor for impaired neural repair, yet its effects on axon regeneration and synaptic remodeling in the brain remain unclear. To investigate age-related repair mechanisms in neural circuits, we developed an axonal injury model in the aged (>2 years) mouse somatosensory cortex and tracked fluorescently labeled axons using in vivo multiphoton imaging. Axon degeneration rates were similar to young adults, but regeneration was markedly reduced. Six hours post-lesion, en passant boutons (EPBs)-the most common cortical synapse-showed transient increases in number and size. To assess functional consequences, we used a recurrent neural network model to simulate memory dynamics, revealing distinct changes compared to young adults. Our results suggest that increased synaptic turnover in the aged brain may facilitate partial recovery from injury through synaptic re-wiring, highlighting potential mechanisms supporting neural adaptation in aging.

Down syndrome (DS), caused by trisomy of chromosome 21, is the leading genetic cause of intellectual disability, yet the mechanisms disrupting fetal brain development remain unclear. We performed single-cell transcriptomic and chromatin accessibility profiling of approximately 250,000 cells from 15 DS and 15 control human fetal cortices (10–20 weeks post-conception). Our analysis revealed a subtype-specific reduction in RORB/FOXP1-expressing excitatory neurons and widespread disruption of neurodevelopmental transcriptional programs. Chromosome 21 transcription factors BACH1, PKNOX1, and GABPA emerged as dosage-sensitive hubs regulating genes linked to intellectual disability. Antisense oligonucleotide-mediated normalization of these factors in human neural progenitors in vitro partially rescued target gene expression. Benchmarking a humanized in vivo model captured additional molecular and cellular signatures of DS, complementing the in vitro model. Together, this resource defines the gene-regulatory landscape underlying cortical development in DS and highlights candidate molecular targets and preclinical models for future intervention studies. Single-cell atlas of Down syndrome fetal cortex links transcriptional dysregulation to reduction of layer 4 neurons Chromosome 21 transcription factors PKNOX1, BACH1, and GABPA drive intellectual disability gene dysregulation Transplanted human cells model late-stage DS phenotypes, bypassing scarcity of fetal tissue ASO targeting chromosome 21 transcription factors restores DS-associated molecular signatures

Down syndrome (DS) is the most common genetic cause of intellectual disability, affecting one in 600 live births worldwide, and is caused by trisomy of the human chromosome 21 (Hsa21). Here, we investigated whether trisomy 21 results in changes in excitatory neuron network development, that could contribute to the neurodevelopmental phenotypes of DS. Replaying cerebral cortex development in vitro with Trisomy 21 and control isogenic and non-isogenic human induced pluripotent stem cells (hiPSC) enabled the analysis of the effect of Hsa21 triplication on neural network activity and connectivity specifically in developing excitatory cortical neurons. Network activity analysis revealed a significant decrease in neuronal activity in TS21 neurons early in development. TS21 neurons showed a marked reduction of synchronised bursting activity up to 80 days in vitro and over 5 months in vivo following transplantation into the mouse forebrain. Viral transynaptic tracing identified significant reduction of neuronal connectivity in TS21 neuronal networks in vitro, suggesting that reduced network connectivity contributes to the absence of synchronised bursting. Expression of voltage-gated potassium channels was significantly reduced in TS21 neurons, and single neuron recordings confirmed the lack of hyperpolarization-activated currents, indicating a functional loss of the potassium/sodium hyperpolarization-activated cyclic nucleotide-gated channel 1 (HCN1) in DS TS21 neurons. We conclude that the trisomy of Hsa21 leads to changes in ion channel composition, network activity and connectivity in cortical excitatory neuron networks, all of which collectively are likely to contribute to some of the neurodevelopmental features of DS. ### Competing Interest Statement The authors have declared no competing interest.

Cohesin and CTCF are major drivers of 3D genome organization, but their role in neurons is still emerging. Here we show a prominent role for cohesin in the expression of genes that facilitate neuronal maturation and homeostasis. Unexpectedly, we observed two major classes of activity-regulated genes with distinct reliance on cohesin in primary cortical neurons. Immediate early genes remained fully inducible by KCl and BDNF, and short-range enhancer-promoter contacts at the Immediate early gene Fos formed robustly in the absence of cohesin. In contrast, cohesin was required for full expression of a subset of secondary response genes characterised by long-range chromatin contacts. Cohesin-dependence of constitutive neuronal genes with key functions in synaptic transmission and neurotransmitter signaling also scaled with chromatin loop length. Our data demonstrate that key genes required for the maturation and activation of primary cortical neurons depend cohesin for their full expression, and that the degree to which these genes rely on cohesin scales with the genomic distance traversed by their chromatin contacts.

Cohesin and CTCF are major drivers of 3D genome organization, but their role in neurons is still emerging. Here, we show a prominent role for cohesin in the expression of genes that facilitate neuronal maturation and homeostasis. Unexpectedly, we observed two major classes of activity-regulated genes with distinct reliance on cohesin in mouse primary cortical neurons. Immediate early genes (IEGs) remained fully inducible by KCl and BDNF, and short-range enhancer-promoter contacts at the IEGs Fos formed robustly in the absence of cohesin. In contrast, cohesin was required for full expression of a subset of secondary response genes characterized by long-range chromatin contacts. Cohesin-dependence of constitutive neuronal genes with key functions in synaptic transmission and neurotransmitter signaling also scaled with chromatin loop length. Our data demonstrate that key genes required for the maturation and activation of primary cortical neurons depend on cohesin for their full expression, and that the degree to which these genes rely on cohesin scales with the genomic distance traversed by their chromatin contacts.

The human brain is a highly complex system comprised of billions of cells, called neurons, that are all interconnected. Normal brain function depends on effective communication between neurons, which requires signals to travel down a nerve cell and then over to the next cell it is connected to. Several diseases result from impairments in the communication between neurons, which is why it is important to study these signals in the living brain. Since the living human brain is difficult to study, scientists use simpler organisms like mice. To see mouse neurons, a piece of skull is replaced with a clear glass window. Powerful microscopes can then be used to capture snapshots of neurons deep inside live brains! Clever use of lasers to excite chemical “tags” inside the neurons makes them glow with fluorescent light. Join us as we explore the inner workings of neuron communication in the living brain.

Background: Astrocytes provide a vital support to neurons in normal and pathological conditions. In Alzheimer’s disease (AD) brains, reactive astrocytes have been found surrounding amyloid plaques, forming an astrocytic scar. However, their role and potential mechanisms whereby they affect neuroinflammation, amyloid pathology, and synaptic density in AD remain unclear. Methods: To explore the role of astrocytes on Aβ pathology and neuroinflammatory markers, we pharmacologically ablated them in organotypic brain culture slices (OBCSs) from 5XFAD mouse model of AD and wild-type (WT) littermates with the selective astrocytic toxin L-alpha-aminoadipate (L-AAA). To examine the effects on synaptic circuitry, we measured dendritic spine number and size in OBCSs from Thy-1-GFP transgenic mice incubated with synthetic Aβ42 or double transgenics Thy-1-GFP/5XFAD mice treated with LAAA or vehicle for 24 h. Results: Treatment of OBCSs with L-AAA resulted in an increased expression of pro-inflammatory cytokine IL-6 in conditioned media of WTs and 5XFAD slices, associated with changes in microglia morphology but not in density. The profile of inflammatory markers following astrocytic loss was different in WT and transgenic cultures, showing reductions in inflammatory mediators produced in astrocytes only in WT sections. In addition, pharmacological ablation of astrocytes led to an increase in Aβ levels in homogenates of OBCS from 5XFAD mice compared with vehicle controls, with reduced enzymatic degradation of Aβ due to lower neprilysin and insulin-degrading enzyme (IDE) expression. Furthermore, OBSCs from wild-type mice treated with L-AAA and synthetic amyloid presented 56% higher levels of Aβ in culture media compared to sections treated with Aβ alone, concomitant with reduced expression of IDE in culture medium, suggesting that astrocytes contribute to Aβ clearance and degradation. Quantification of hippocampal dendritic spines revealed a reduction in their density following L-AAA treatment in all groups analyzed. In addition, pharmacological ablation of astrocytes resulted in a decrease in spine size in 5XFAD OBCSs but not in OBCSs from WT treated with synthetic Aβ compared to vehicle control. Conclusions: Astrocytes play a protective role in AD by aiding Aβ clearance and supporting synaptic plasticity.

Cohesin and CTCF are major drivers of 3D genome organization. Even though human mutations underscore the importance of cohesin and CTCF for neurodevelopment, their role in neurons is only just beginning to be addressed. Here we conditionally ablate Rad21 in cortical neurons, revealing a prominent role for cohesin in the expression of genes that facilitate neuronal maturation, homeostasis, and activation. In agreement with recent reports, activity-dependent genes were downregulated at baseline. However, in contrast to current models that attribute impaired activity-dependent gene expression to a role for cohesin and CTCF in anchoring enhancer-promoter contacts, we show that nearly all activity-dependent genes remain inducible in the absence of cohesin. While CTCF-based chromatin loops were substantially weakened, long-range contacts still formed robustly between activity-dependent enhancers and their target immediate early gene promoters. We suggest a model where neuronal cohesin facilitates the precise level of activity-dependent gene expression, rather than inducibility per se, and where inducibility of activity-dependent gene expression is linked to cohesin-independent enhancer-promoter contacts. These data expand our understanding of the importance of cohesin-independent enhancer-promoter contacts in regulating gene expression.

A set of prepared datasets for running experiments to replicate paper (see below). List of data: Training capspix2pix: crops256.zip - folder containing 256x256 crops from the original dataset for training capspix2pix. Images are in the "train/original" folder, and labels are in the "train/mask" folder. syn256_x_data_val.npy + syn256_y_data_val.npy + syn256_y_points_data_val.npy (images + labels + centrelines) - validation synthetic dataset, used while training capspix2pix for plotting Training u-net: capspix2pix_AR_data_train.npy + capspix2pix_AR_mask_train.npy (images + labels) - data generated from a capspix2pix model from real labels capspix2pix_SSM_data_train.npy + capspix2pix_AR_mask_train.npy (images + labels) - data generated from a capspix2pix model from synthetic labels PBAM_SSM_data_train.npy + PBAM_SSM_mask_train.npy (images + labels) - data generated from PBAM (Physics-based model) for training u-net pix2pix_AR_data_train.npy + pix2pix_AR_mask_train.npy (images + labels) - data generated from a pix2pix model from real labels for training u-net pix2pix_SSM_data_train.npy + pix2pix_SSM_mask_train.npy (images + labels) - data generated from a pix2pix model from synthetic labels for training u-net real_data_data_train.npy + real_data_mask_train.npy (images + labels) - augmented real dataset for training u-net Testing u-net: org64_data_test.npy + org64_mask_test.npy (images + labels) - crops from original test dataset for testing u-net Interpolation: crops256_inter_data_train.npy + crops256_inter_mask_train.npy (images + labels) - example data for interpolation Please cite the following paper when using this dataset: Bass, C., Dai, T., Billot, B., Arulkumaran, K., Creswell, A., Clopath, C., De Paola, V., and Bharath, A. A., 2019. “Image synthesis with a convolutional capsule generative adversarial network,” Medial Imaging with Deep Learning. See Github page for further instructions: https://github.com/CherBass/CapsPix2Pix

Automatically tracing elongated structures, such as axons and blood vessels, is a challenging problem in the field of biomedical imaging, but one with many downstream applications. Real, labelled data is sparse, and existing algorithms either lack robustness to different datasets, or otherwise require significant manual tuning. Here, we instead learn a tracking algorithm in a synthetic environment, and apply it to tracing axons. To do so, we formulate tracking as a reinforcement learning problem, and apply deep reinforcement learning techniques with a continuous action space to learn how to track at the subpixel level. We train our model on simple synthetic data and test it on mouse cortical two-photon microscopy images. Despite the domain gap, our model approaches the performance of a heavily engineered tracker from a standard analysis suite for neuronal microscopy. We show that fine-tuning on real data improves performance, allowing better transfer when real labelled data is available. Finally, we demonstrate that our model’s uncertainty measure—a feature lacking in hand-engineered trackers—corresponds with how well it tracks the structure.

Machine learning for biomedical imaging often suffers from a lack of labelled training data. One solution is to use generative models to synthesise more data. To this end, we introduce CapsPix2Pix, which combines convolutional capsules with the pix2pix framework, to synthesise images conditioned on class segmentation labels. We apply our approach to a new biomedical dataset of cortical axons imaged by two-photon microscopy, as a method of data augmentation for small datasets. We evaluate performance both qualitatively and quantitatively. Quantitative evaluation is performed by using image data generated by either CapsPix2Pix or pix2pix to train a U-net on a segmentation task, then testing on real microscopy data. Our method quantitatively performs as well as pix2pix, with an order of magnitude fewer parameters. Additionally, CapsPix2Pix is far more capable at synthesising images of different appearance, but the same underlying geometry. Finally, qualitative analysis of the features learned by CapsPix2Pix suggests that individual capsules capture diverse and often semantically meaningful groups of features, covering structures such as synapses, axons and noise.

In this study, we investigated whether intrinsic glial dysfunction contributes to the pathogenesis of schizophrenia (SCZ).Our approach was to establish humanized glial chimeric mice using glial progenitor cells (GPCs) produced from induced pluripotent stem cells derived from patients with childhood-onset SCZ.After neonatal implantation into myelin-deficient shiverer mice, SCZ GPCs showed premature migration into the cortex, leading to reduced white matter expansion and hypomyelination relative to controls.The SCZ glial chimeras also showed delayed astrocytic differentiation and abnormal astrocytic morphologies.When established in myelin wild-type hosts, SCZ glial mice showed reduced prepulse inhibition and abnormal behavior, including excessive anxiety, antisocial traits and disturbed sleep.RNAseq of cultured SCZ hGPCs revealed disrupted glial differentiation-associated and synaptic gene expression, indicating that glial pathology was cell-autonomous.Our data therefore suggest a causal role for impaired glial maturation in the development of schizophrenia, and provide a humanized model for its in vivo assessment.

Background: Altered microglial markers and morphology have been demonstrated in patients with schizophrenia in post-mortem and in vivo studies. However, it is unclear if changes are due to antipsychotic treatment. Aims: Here we aimed to determine whether antipsychotic medication affects microglia in vivo. Methods: To investigate this we administered two clinically relevant doses (0.05 mg n=12 and 2.5 mg n=7 slow-release pellets, placebo n=20) of haloperidol, over 2 weeks, to male Sprague Dawley rats to determine the effect on microglial cell density and morphology (area occupied by processes and microglial cell area). We developed an analysis pipeline for the automated assessment of microglial cells and used lipopolysaccharide (LPS) treatment (n=13) as a positive control for analysis. We also investigated the effects of haloperidol (n=9) or placebo (n=10) on the expression of the translocator protein 18 kDa (TSPO) using autoradiography with [H-3]PBR28, a TSPO ligand used in human positron emission tomography (PET) studies. Results: Here we demonstrated that haloperidol at either dose does not alter microglial measures compared with placebo control animals (p > 0.05). Similarly there was no difference in [H-3]PBR28 binding between placebo and haloperidol tissue (p > 0.05). In contrast, LPS was associated with greater cell density (p = 0.04) and larger cell size (p = 0.01). Conclusion: These findings suggest that haloperidol does not affect microglial cell density, morphology or TSPO expression, indicating that clinical study alterations are likely not the consequence of antipsychotic treatment. The automated cell evaluation pipeline was able to detect changes in microglial morphology induced by LPS and is made freely available for future use.

This study has used dense reconstructions from serial EM images to compare the neuropil ultrastructure and connectivity of aged and adult mice. The analysis used models of axons, dendrites, and their synaptic connections, reconstructed from volumes of neuropil imaged in layer 1 of the somatosensory cortex. This shows the changes to neuropil structure that accompany a general loss of synapses in a well-defined brain region. The loss of excitatory synapses was balanced by an increase in their size such that the total amount of synaptic surface, per unit length of axon, and per unit volume of neuropil, stayed the same. There was also a greater reduction of inhibitory synapses than excitatory, particularly those found on dendritic spines, resulting in an increase in the excitatory/inhibitory balance. The close correlations, that exist in young and adult neurons, between spine volume, bouton volume, synaptic size, and docked vesicle numbers are all preserved during aging. These comparisons display features that indicate a reduced plasticity of cortical circuits, with fewer, more transient, connections, but nevertheless an enhancement of the remaining connectivity that compensates for a generalized synapse loss.

Harnessing the potential of human stem cells for modeling the physiology and diseases of cortical circuitry requires monitoring cellular dynamics in vivo. We show that human induced pluripotent stem cell (iPSC)-derived cortical neurons transplanted into the adult mouse cortex consistently organized into large (up to ~100 mm3) vascularized neuron-glia territories with complex cytoarchitecture. Longitudinal imaging of >4000 grafted developing human neurons revealed that neuronal arbors refined via branch-specific retraction; human synaptic networks substantially restructured over 4 months, with balanced rates of synapse formation and elimination; and oscillatory population activity mirrored the patterns of fetal neural networks. Lastly, we found increased synaptic stability and reduced oscillations in transplants from two individuals with Down syndrome, demonstrating the potential of in vivo imaging in human tissue grafts for patient-specific modeling of cortical development, physiology, and pathogenesis.

Studies of structural plasticity in the brain often require the detection and analysis of axonal synapses (boutons). To date, bouton detection has been largely manual or semi-automated, relying on a step that traces the axons before detection the boutons. If tracing the axon fails, the accuracy of bouton detection is compromised. In this paper, we propose a new algorithm that does not require tracing the axon to detect axonal boutons in 3D two-photon images taken from the mouse cortex. To find the most appropriate techniques for this task, we compared several well-known algorithms for interest point detection and feature descriptor generation. The final algorithm proposed has the following main steps: (1) a Laplacian of Gaussian (LoG) based feature enhancement module to accentuate the appearance of boutons; (2) a Speeded Up Robust Features (SURF) interest point detector to find candidate locations for feature extraction; (3) non-maximum suppression to eliminate candidates that were detected more than once in the same local region; (4) generation of feature descriptors based on Gabor filters; (5) a Support Vector Machine (SVM) classifier, trained on features from labelled data, and was used to distinguish between bouton and non-bouton candidates. We found that our method achieved a Recall of 95%, Precision of 76%, and F1 score of 84% within a new dataset that we make available for accessing bouton detection. On average, Recall and F1 score were significantly better than the current state-of-the-art method, while Precision was not significantly different. In conclusion, in this article we demonstrate that our approach, which is independent of axon tracing, can detect boutons to a high level of accuracy, and improves on the detection performance of existing approaches. The data and code (with an easy to use GUI) used in this article are available from open source repositories.

optical imaging has emerged as a powerful tool with which to study cellular responses to injury and disease in the mammalian CNS. Important new insights have emerged regarding axonal degeneration and regeneration, glial responses and neuroinflammation, changes in the neurovascular unit, and, more recently, neural transplantations. Accompanying a 2017 SfN Mini-Symposium, here, we discuss selected recent advances in understanding the neuronal, glial, and other cellular responses to CNS injury and disease with imaging of the rodent brain or spinal cord. We anticipate that optical imaging will continue to be at the forefront of breakthrough discoveries of fundamental mechanisms and therapies for CNS injury and disease.