Advertisement

Postnatal Tshz3 Deletion Drives Altered Corticostriatal Function and Autism Spectrum Disorder–like Behavior

Open AccessPublished:March 27, 2019DOI:https://doi.org/10.1016/j.biopsych.2019.03.974

      Abstract

      Background

      Heterozygous deletion of the TSHZ3 gene, encoding for the teashirt zinc-finger homeobox family member 3 (TSHZ3) transcription factor that is highly expressed in cortical projection neurons (CPNs), has been linked to an autism spectrum disorder (ASD) syndrome. Similarly, mice with Tshz3 haploinsufficiency show ASD-like behavior, paralleled by molecular changes in CPNs and corticostriatal synaptic dysfunctions. Here, we aimed at gaining more insight into “when” and “where” TSHZ3 is required for the proper development of the brain, and its deficiency crucial for developing this ASD syndrome.

      Methods

      We generated and characterized a novel mouse model of conditional Tshz3 deletion, obtained by crossing Tshz3flox/flox with CaMKIIalpha-Cre mice, in which Tshz3 is deleted in CPNs from postnatal day 2 to 3 onward. We characterized these mice by a multilevel approach combining genetics, cell biology, electrophysiology, behavioral testing, and bioinformatics.

      Results

      These conditional Tshz3 knockout mice exhibit altered cortical expression of more than 1000 genes, ∼50% of which have their human orthologue involved in ASD, in particular genes encoding for glutamatergic synapse components. Consistently, we detected electrophysiological and synaptic changes in CPNs and impaired corticostriatal transmission and plasticity. Furthermore, these mice showed strong ASD-like behavioral deficits.

      Conclusions

      Our study reveals a crucial postnatal role of TSHZ3 in the development and functioning of the corticostriatal circuitry and provides evidence that dysfunction in these circuits might be determinant for ASD pathogenesis. Our conditional Tshz3 knockout mouse constitutes a novel ASD model, opening the possibility for an early postnatal therapeutic window for the syndrome linked to TSHZ3 haploinsufficiency.

      Keywords

      The basal ganglia constitute a functional network of subcortical structures playing a crucial role in extrapyramidal motor control and motor learning, as well as in associative learning, planning, working memory, and emotion. The striatum is the main input station of the basal ganglia, receiving major projections coming from layer 5 (L5) cortical projection neurons (CPNs). These corticostriatal glutamatergic afferences contact mainly striatal spiny projection neurons (SPNs), the major neuronal population of this nucleus (
      • DeLong M.R.
      • Wichmann T.
      Circuits and circuit disorders of the basal ganglia.
      ,
      • Levy R.
      • Hazrati L.N.
      • Herrero M.T.
      • Vila M.
      • Hassani O.K.
      • Mouroux M.
      • et al.
      Re-evaluation of the functional anatomy of the basal ganglia in normal and Parkinsonian states.
      ,
      • Redgrave P.
      • Vautrelle N.
      • Reynolds J.N.
      Functional properties of the basal ganglia's re-entrant loop architecture: Selection and reinforcement.
      ). There is accumulating evidence showing altered development and function of corticostriatal circuitry and striatal dysmorphic features in autistic patients (
      • Willsey A.J.
      • Sanders S.J.
      • Li M.
      • Dong S.
      • Tebbenkamp A.T.
      • Muhle R.A.
      • et al.
      Coexpression networks implicate human midfetal deep cortical projection neurons in the pathogenesis of autism.
      ,
      • Parikshak N.N.
      • Luo R.
      • Zhang A.
      • Won H.
      • Lowe J.K.
      • Chandran V.
      • et al.
      Integrative functional genomic analyses implicate specific molecular pathways and circuits in autism.
      ,
      • State M.W.
      • Sestan N.
      Neuroscience. The emerging biology of autism spectrum disorders.
      ,
      • Delmonte S.
      • Gallagher L.
      • O'Hanlon E.
      • McGrath J.
      • Balsters J.H.
      Functional and structural connectivity of frontostriatal circuitry in autism spectrum disorder.
      ,
      • Langen M.
      • Bos D.
      • Noordermeer S.D.
      • Nederveen H.
      • van Engeland H.
      • Durston S.
      Changes in the development of striatum are involved in repetitive behavior in autism.
      ,
      • Langen M.
      • Schnack H.G.
      • Nederveen H.
      • Bos D.
      • Lahuis B.E.
      • de Jonge M.V.
      • et al.
      Changes in the developmental trajectories of striatum in autism.
      ,
      • Shepherd G.M.
      Corticostriatal connectivity and its role in disease.
      ). Autism spectrum disorder (ASD) is a heterogeneous group of neurodevelopmental pathologies characterized by impairments in social communication and interaction, restricted interests, and repetitive behaviors (
      American Psychiatric Association
      Diagnostic and Statistical Manual of Mental Disorders.
      ). ASD etiology has a strong genetic component, with a high number of genes and factors involved (
      • Bourgeron T.
      Current knowledge on the genetics of autism and propositions for future research.
      ,
      • Buxbaum J.D.
      • Daly M.J.
      • Devlin B.
      • Lehner T.
      • Roeder K.
      • State M.W.
      Autism Sequencing C
      The autism sequencing consortium: large-scale, high-throughput sequencing in autism spectrum disorders.
      ,
      • Park H.R.
      • Lee J.M.
      • Moon H.E.
      • Lee D.S.
      • Kim B.N.
      • Kim J.
      • et al.
      A short review on the current understanding of autism spectrum disorders.
      ,
      • Ziats M.N.
      • Rennert O.M.
      The evolving diagnostic and genetic landscapes of autism spectrum disorder.
      ). Accordingly, abnormal corticostriatal function has also been found in ASD mouse models, such as Shank3−/− (
      • Peixoto R.T.
      • Wang W.G.
      • Croney D.M.
      • Kozorovitskiy Y.
      • Sabatini B.L.
      Early hyperactivity and precocious maturation of corticostriatal circuits in Shank3B(−/−) mice.
      ,
      • Peca J.
      • Feliciano C.
      • Ting J.T.
      • Wang W.
      • Wells M.F.
      • Venkatraman T.N.
      • et al.
      Shank3 mutant mice display autistic-like behaviours and striatal dysfunction.
      ), neuroligin-1−/− (
      • Blundell J.
      • Blaiss C.A.
      • Etherton M.R.
      • Espinosa F.
      • Tabuchi K.
      • Walz C.
      • et al.
      Neuroligin-1 deletion results in impaired spatial memory and increased repetitive behavior.
      ), and 16p11+/− mice (
      • Portmann T.
      • Yang M.
      • Mao R.
      • Panagiotakos G.
      • Ellegood J.
      • Dolen G.
      • et al.
      Behavioral abnormalities and circuit defects in the basal ganglia of a mouse model of 16p11.2 deletion syndrome.
      ). In this context, we have recently identified TSHZ3 (teashirt zinc-finger homeobox family member 3, also known as ZNF537) as a susceptibility gene for ASD: patients with heterozygous TSHZ3 deletion show ASD neurodevelopmental phenotypes including autistic behavior and intellectual disability, and in particular, speech disturbance; renal tract abnormalities are also frequently observed in these patients (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ). TSHZ3 gene codes for the highly conserved, zinc-finger homeodomain transcription factor TSHZ3, whose expression starts during prenatal development in both human and mouse (
      • Caubit X.
      • Tiveron M.C.
      • Cremer H.
      • Fasano L.
      Expression patterns of the three Teashirt-related genes define specific boundaries in the developing and postnatal mouse forebrain.
      ). Moreover, TSHZ3 has been identified in human neocortical gene networks with the highest expression during late fetal development, which have been linked to not only ASD but also other neuropsychiatric disorders and IQ (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ,
      • Li M.
      • Santpere G.
      • Imamura Kawasawa Y.
      • Evgrafov O.V.
      • Gulden F.O.
      • Pochareddy S.
      • et al.
      Integrative functional genomic analysis of human brain development and neuropsychiatric risks.
      ). Tshz3-null mutation in mouse (Tshz3−/−) leads to altered expression of cortical layer markers (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ) and is lethal at birth (
      • Caubit X.
      • Thoby-Brisson M.
      • Voituron N.
      • Filippi P.
      • Bevengut M.
      • Faralli H.
      • et al.
      Teashirt 3 regulates development of neurons involved in both respiratory rhythm and airflow control.
      ), suggesting that TSHZ3 plays a critical role during prenatal brain development. Interestingly, heterozygous Tshz3 deletion in mice (Tshz3+/−) faithfully mimics the human disorder, as these animals show reduced sociability, narrowed field of interest, stereotypies, and anxiety, associated with renal tract defects. These ASD-like symptoms are paralleled by altered expression of neocortical gene markers at both embryonic (embryonic day 18.5 [E18.5]) and postnatal (postnatal day 5 [P5] and P20) stages, and by altered function of the corticostriatal pathway (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ). However, L5 CPNs, which project to the ipsi- and/or contralateral striatum, express TSHZ3 not only during fetal development, but also postnatally and in adulthood, an expression pattern that is comparable in human and mouse (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ,
      • Caubit X.
      • Tiveron M.C.
      • Cremer H.
      • Fasano L.
      Expression patterns of the three Teashirt-related genes define specific boundaries in the developing and postnatal mouse forebrain.
      ,
      • Kang H.J.
      • Kawasawa Y.I.
      • Cheng F.
      • Zhu Y.
      • Xu X.
      • Li M.
      • et al.
      Spatio-temporal transcriptome of the human brain.
      ,
      • Workman A.D.
      • Charvet C.J.
      • Clancy B.
      • Darlington R.B.
      • Finlay B.L.
      Modeling transformations of neurodevelopmental sequences across mammalian species.
      ). The postnatal development of the corticostriatal circuitry has been little characterized, but studies in animals have shown that its maturation progresses after birth (
      • Tepper J.M.
      • Sharpe N.A.
      • Koos T.Z.
      • Trent F.
      Postnatal development of the rat neostriatum: Electrophysiological, light- and electron-microscopic studies.
      ,
      • Kwan K.Y.
      • Sestan N.
      • Anton E.S.
      Transcriptional co-regulation of neuronal migration and laminar identity in the neocortex.
      ,
      • Wang H.
      • Cuzon V.C.
      • Pickel V.M.
      Postnatal development of mu-opioid receptors in the rat caudate-putamen nucleus parallels asymmetric synapse formation.
      ,
      • Inaji M.
      • Sato K.
      • Momose-Sato Y.
      • Ohno K.
      Voltage-sensitive dye imaging analysis of functional development of the neonatal rat corticostriatal projection.
      ,
      • Mullah S.H.
      • Inaji M.
      • Nariai T.
      • Momose-Sato Y.
      • Sato K.
      • Ohno K.
      Optical analysis of developmental changes in synaptic potentiation in the neonatal rat corticostriatal projection.
      ). Therefore, the postnatal role of TSHZ3 in the maturation and function of the corticostriatal system needs to be elucidated.
      For this, we generated a novel conditional mutant mouse with Tshz3 deletion in CPNs at an early postnatal stage. We show here that this time- and region-specific loss of Tshz3 results in ASD-relevant behavioral deficits similar to those we previously described in Tshz3+/− mice, which are paralleled by cortical and striatal changes in terms of gene expression, synaptic transmission, and synaptic plasticity. These data suggest that postnatal events might contribute to the TSHZ3-linked ASD syndrome, opening perspectives for a possible early therapeutic window.

      Methods and Materials

      Additional methods and materials are provided in Supplement 1.

      Generating Tshz3-Conditional Knockout Mice

      Animal experimental procedures were approved by the Comité National de Réflexion Ethique sur l’Expérimentation Animale 14' (57-07112012) and in agreement with the European Communities Council Directive (2010/63/EU). Conditional mouse mutants with postnatal loss of Tshz3 in the cortex (hereby referred to as Tshz3-pnCxKO) were generated by crossing Tshz3flox/flox mice with CaMKIIalpha-Cre that express the CRE-recombinase in glutamatergic CPNs, as described in Supplement 1. Control mice were Tshz3flox/flox.

      Morphometric and Dendritic Spine Analysis of L5 CPNs

      We used transgenic mouse lines (age P28) expressing Thy1-green fluorescent protein (GFP) in L5 CPNs (
      • Feng G.
      • Mellor R.H.
      • Bernstein M.
      • Keller-Peck C.
      • Nguyen Q.T.
      • Wallace M.
      • et al.
      Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP.
      ). Thy 1-GFP-M;Tshz3-pnCxKO mice were obtained by crossing CaMKIIalpha-Cre;Tshz3flox/flox male mice with Tshz3flox/flox female mice heterozygous for Thy1-GFP. Stacks from 100-μm vibratome sections (1-μm z-step) were acquired using a Zeiss LSM780 (Carl Zeiss Meditec, Oberkochen, Germany) laser scanning confocal microscope (40× objective, zoom 0.6). Isolated Thy1-GFP-positive L5 CPNs of the primary motor and somatosensory cortex were imaged and reconstructed using the NeuronJ plugin of ImageJ 1.45s (National Institutes of Health, Bethesda, MD) in 2-dimensional projections. All dendrites were semimanually traced and labeled as primary (dendrite originating from the soma), secondary (extending from the primary dendrite), or higher-order dendrites. Fifteen cells were reconstructed for each genotype from 4 littermate pairs of P28 control (Thy1-GFP-M;Tshz3flox/flox) and mutant (Thy1-GFP-M;Tshz3-pnCxKO) mice. Sholl cross-analysis was performed by counting the number of dendrites intersecting concentric spherical radii at 10-μm intervals. Differences at specific radii were analyzed using unpaired Student's t test.
      Analysis of spine density and morphology was performed on 4 littermate pairs (see above). Images were acquired as described above (63× objective, numerical aperture 1.4, 0.03 μm/pixel, voxel size 0.033 μm2 × 0.37 μm). Spine counts were obtained from second- or third-order basal dendritic branches of randomly selected L5 CPNs. Dendrites from five to seven cells were analyzed per animal, providing a cumulated dendrite length >750 μm for each genotype. Spine identification and density were evaluated using NeuronStudio (
      • Rodriguez A.
      • Ehlenberger D.B.
      • Dickstein D.L.
      • Hof P.R.
      • Wearne S.L.
      Automated three-dimensional detection and shape classification of dendritic spines from fluorescence microscopy images.
      ). Statistical analysis was performed using an unpaired Student's t test (BiostaTGV statistical software; http://www.u707.jussieu.fr/biostatgv/) and results considered significant at p < .05. Data are expressed as mean ± SEM.

      Electrophysiology

      Procedures were similar to those described previously (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ,
      • Chassain C.
      • Melon C.
      • Salin P.
      • Vitale F.
      • Couraud S.
      • Durif F.
      • et al.
      Metabolic, synaptic and behavioral impact of 5-week chronic deep brain stimulation in hemiparkinsonian rats.
      ). See also Supplement 1. Acute coronal corticostriatal slices (250-μm thick) from Tshz3-pnCxKO and control mice (P21–28) were obtained as previously described (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ,
      • Chassain C.
      • Melon C.
      • Salin P.
      • Vitale F.
      • Couraud S.
      • Durif F.
      • et al.
      Metabolic, synaptic and behavioral impact of 5-week chronic deep brain stimulation in hemiparkinsonian rats.
      ). Whole-cell patch-clamp recordings were performed in oxygenated artificial cerebrospinal fluid at 34°C to 35°C, flowing at ∼2.5 mL/min. L5 CPNs of the primary motor and somatosensory cortex and SPNs of the dorsolateral striatum were identified by infrared video microscopy and by their electrophysiological properties (
      • Jiang Z.G.
      • North R.A.
      Membrane properties and synaptic responses of rat striatal neurones in vitro.
      ,
      • Hattox A.M.
      • Nelson S.B.
      Layer V neurons in mouse cortex projecting to different targets have distinct physiological properties.
      ), and were recorded in whole-cell patch-clamp by borosilicate micropipettes (5–6 MΩ) filled with an internal solution containing 125-mM K-gluconate, 10-mM NaCl, 1-mM CaCl2, 2-mM MgCl2, 0.5-mM BAPTA, 19-mM HEPES, 0.3-mM Na-guanosine triphosphate, and 1-mM Mg-adenosine triphosphate, pH 7.3. For N-methyl-D-aspartate (NMDA)/alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) ratio experiments, the internal solution contained 140-mM CsCl, 10-mM NaCl, 0.1-mM CaCl2, 10-mM HEPES, 1-mM EGTA, 2-mM Mg-adenosine triphosphate, and 0.5-mM Na-guanosine triphosphate, pH 7.3. For measuring gamma-aminobutyric acidergic (GABAergic) synaptic transmission, the internal solution contained 126-mM mM CsCH3SO3, 10-mM mM HEPES, 1-mM EGTA, 2-mM QX-314 chloride, 0.1-mM CaCl2, 4-mM Mg-adenosine triphosphate, and 0.3-mM Na-guanosine triphosphate, pH 7.3. A stimulating bipolar electrode was placed either in the cortex at the level of L4 to activate local fibers and evoke glutamatergic excitatory postsynaptic currents (EPSCs) and GABAergic inhibitory postsynaptic currents in L5 CPNs, or in the corpus callosum to activate corticostriatal fibers and evoke EPSCs in striatal SPNs. Glutamatergic synaptic transmission was recorded in the presence of 50-μM picrotoxin at a holding potential of −60 mV (L5 CPNs) or −80 mV (SPNs). Miniature EPSCs were recorded the presence of 1-μM tetrodotoxin. GABAergic synaptic transmission was recorded at a holding potential of −35 mV and in the presence of 10-μM cyanquixaline and 10-μM AP-5. Electrophysiological data were acquired by an AxoPatch 200B amplifier and pClamp 10.2 software (Molecular Devices, Wokingham, United Kingdom). Series and input resistance were continuously monitored by sending 5-mV pulses, and neurons showing ≥20% change in these parameters were discarded from the analysis.
      Statistical analysis was performed by Prism 7 (GraphPad Software, San Diego, CA). Two-tailed Mann-Whitney test was used for comparing two datasets, two-tailed Kruskal-Wallis and Dunn's post-test was used for comparing ≥3 datasets, 2-sample Kolmogorov-Smirnov test was used for comparing cumulative distributions, and one- and two-way analysis of variance was used for multiple comparisons. Sample size (n) refers to the number of recorded neurons. The significance threshold was fixed at p < .05. Tests used, p values, and n are indicated in the figures. Data are expressed as mean ± SEM.

      Behavioral Testing

      Male mice 101 to 127 days old were maintained in a social enriched environment and tested blind to the genotype for ASD-like criteria (
      American Psychiatric Association
      Diagnostic and Statistical Manual of Mental Disorders.
      ). Stereotypy scores were 1) the burying score obtained with marbles (
      • Thomas A.
      • Burant A.
      • Bui N.
      • Graham D.
      • Yuva-Paylor L.A.
      • Paylor R.
      Marble burying reflects a repetitive and perseverative behavior more than novelty-induced anxiety.
      ) and 2) the number of repeated dips in a hole board (
      • Makanjuola R.O.
      • Hill G.
      • Maben I.
      • Dow R.C.
      • Ashcroft G.W.
      An automated method for studying exploratory and stereotyped behaviour in rats.
      ). A small number of crossed zones in an open field indicated a narrow field of interest (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ). Sociability and interest in social novelty were measured in a two-chamber task (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ,
      • Roubertoux P.L.
      • Carlier M.
      • Tordjman S.
      Deficit in social relationships and reduced field of interest in mice.
      ) adapted from the three-chamber task (
      • Moy S.S.
      • Nadler J.J.
      • Perez A.
      • Barbaro R.P.
      • Johns J.M.
      • Magnuson T.R.
      • et al.
      Sociability and preference for social novelty in five inbred strains: An approach to assess autistic-like behavior in mice.
      ). Although not included in DSM-5 criteria, anxiety, intellectual disability, and motor disorders are observed in 42% to 56%, ∼45%, and ≤79%, respectively, of ASD patients (
      • Lai M.C.
      • Lombardo M.V.
      • Baron-Cohen S.
      Autism.
      ). The presence of these symptoms was assessed by the elevated plus maze (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ), Morris water maze, and notched bar test (
      • Roubertoux P.L.
      • Baril N.
      • Cau P.
      • Scajola C.
      • Ghata A.
      • Bartoli C.
      • et al.
      Differential brain, cognitive and motor profiles associated with partial trisomy. modeling down syndrome in mice.
      ), respectively. For further details and behavioral testing, see Supplement 1.
      The assumption of statistical normality was checked. Analyses of covariance were conducted with genotype as independent variable and activity as covariate, for each score where the activity level must be controlled for. Repeated-measures analysis of covariance was used for testing sociability and social novelty. A two-tailed independent t test was used for the other comparisons. The results are expressed as effect size because the validation of an organism model for a disorder or a disease requires a significant difference and also a large (pathological) difference. Similar to previous articles, effect sizes are expressed as η2 or partial η2 as specified case by case, where values of .30 and .50 correspond to “mild impairment” and “impairment,” respectively, by analogy with the intellectual disability field (
      • Lai M.C.
      • Lombardo M.V.
      • Baron-Cohen S.
      Autism.
      ,
      • Roubertoux P.L.
      • Baril N.
      • Cau P.
      • Scajola C.
      • Ghata A.
      • Bartoli C.
      • et al.
      Differential brain, cognitive and motor profiles associated with partial trisomy. modeling down syndrome in mice.
      ). Data are expressed as mean ± SEM.

      Gene Set Enrichment Analysis

      Preranked gene lists were calculated based on the −log10 of the p value from DESeq2 (https://bioconductor.org/) analysis multiplied by the sign of differential expression (Table S3T in Supplement 3). Gene matrix transposed (.gmt) files for Gene Ontology terms, as well as pathways, were downloaded from the Bader website (http://baderlab.org/GeneSets) (
      • Merico D.
      • Isserlin R.
      • Stueker O.
      • Emili A.
      • Bader G.D.
      Enrichment map: A network-based method for gene-set enrichment visualization and interpretation.
      ). Data from the SynaptomeDB were received by the authors (
      • Pirooznia M.
      • Wang T.
      • Avramopoulos D.
      • Valle D.
      • Thomas G.
      • Huganir R.L.
      • et al.
      SynaptomeDB: An ontology-based knowledgebase for synaptic genes.
      ) and transformed into a .gmt file. Data from the Genes to Cognition (G2C) Synapse Proteomics Database were downloaded from the website (http://www.genes2cognition.org/proteomics/) (
      • Collins M.O.
      • Husi H.
      • Yu L.
      • Brandon J.M.
      • Anderson C.N.
      • Blackstock W.P.
      • et al.
      Molecular characterization and comparison of the components and multiprotein complexes in the postsynaptic proteome.
      ) and transformed into a .gmt file. We performed Gene Set Enrichment Analysis (GSEA) using the software provided by the Broad Institute (
      • Subramanian A.
      • Tamayo P.
      • Mootha V.K.
      • Mukherjee S.
      • Ebert B.L.
      • Gillette M.A.
      • et al.
      Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles.
      ,
      • Mootha V.K.
      • Lindgren C.M.
      • Eriksson K.F.
      • Subramanian A.
      • Sihag S.
      • Lehar J.
      • et al.
      PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes.
      ) with default parameters. For SynaptomeDB, as well as G2C Synapse Proteomics data, we also used the WebGestalt webserver (http://www.webgestalt.org/option.php) (
      • Wang J.
      • Vasaikar S.
      • Shi Z.
      • Greer M.
      • Zhang B.
      WebGestalt 2017: A more comprehensive, powerful, flexible and interactive gene set enrichment analysis toolkit.
      ). Network of GSEA enrichments was produced using Cytoscape (
      • Su G.
      • Morris J.H.
      • Demchak B.
      • Bader G.D.
      Biological network exploration with Cytoscape 3.
      ).

      Accession Codes

      Raw data (FastQ files) from the sequencing experiment (triplicates from wild-type and Tshz3-mutant cortices) and raw abundance measurements for genes (read counts) for each sample are available from Gene Expression Omnibus under accession GSE119791, which should be quoted in any manuscript discussing the data.

      Results

      Morphological Characterization of Conditional Tshz3 Knockout Mice

      Analysis of TSHZ3 expression in Tshz3-pnCxKO mice at P28 (a time allowing strong Cre expression and recombination, as well as brain maturation) showed almost complete loss of this protein in the neocortex (Figure S1A, B in Supplement 1). Using the corticostriatal circuit as a model system, we evidenced it by cholera toxin subunit B retrograde tracing: the ipsi- and contralateral components of the corticostriatal projections were present in Tshz3-pnCxKO mice as in control mice (Figure S1C–F in Supplement 1). Labeling of NeuN (
      • Mullen R.J.
      • Buck C.R.
      • Smith A.M.
      NeuN, a neuronal specific nuclear protein in vertebrates.
      ) and BCL11B/CTIP2 (
      • Arlotta P.
      • Molyneaux B.J.
      • Chen J.
      • Inoue J.
      • Kominami R.
      • Macklis J.D.
      Neuronal subtype-specific genes that control corticospinal motor neuron development in vivo.
      ), a neuron-specific and an L5/6-selective marker, respectively, was similar in both groups, further suggesting that the postnatal loss of Tshz3 affects neither the number of cortical neurons nor their layering (Figure S1G–L in Supplement 1). In Thy1-GFP-M;Tshz3-pnCxKO mice, we did not observe major changes in the overall CPN morphology and projections (Figure S2 in Supplement 1), confirming the lack of major morphological defects after postnatal Tshz3 loss. Thy1-GFP-M;Tshz3-pnCxKO mice were also used for Sholl analysis of L5 CPN dendrite morphology, revealing no major changes between control and mutant mice (Figure S3 in Supplement 1). However, we found a significantly reduced spine density in L5 CPNs (Figure 1), in agreement with other ASD-related mouse models (
      • Lai K.O.
      • Ip N.Y.
      Structural plasticity of dendritic spines: The underlying mechanisms and its dysregulation in brain disorders.
      ). Finally, the density of vesicular glutamate transporter 1 staining was similar in control and Thy1-GFP-M;Tshz3-pnCxKO mice (Figure S4 in Supplement 1), suggesting unchanged density of corticostriatal terminals.
      Figure thumbnail gr1
      Figure 1Reduced spine density in conditional mutant mice with postnatal loss of Tshz3 in the cortex (Tshz3-pnCxKO). (A, B) Three-dimensional reconstruction of representative confocal images showing dendritic spines of green fluorescent protein–positive layer 5 cortical projection neurons of the primary motor and somatosensory cortex from control and Tshz3-pnCxKO mice at postnatal day 28. Scale bar = 5 μm. (C) Density of spines in control (gray) and Tshz3-pnCxKO (black) mice (mean ± SEM). Data are from 1421 spines over 760-μm dendritic segment length and 1092 spines over 788-μm dendritic segment length, respectively. Decrease: total ∼26% (from 19.26 ± 1.38 spines/10 μm in control mice to 14.24 ± 1.42 spines/10 μm in Thy1-GFP-M;Tshz3-pnCxKO mice); stubby ∼24%, (from 6.86 ± 0.36 spines/10 μm in control mice to 5.23 ± 0.46 spines/10 μm in Thy1-GFP-M;Tshz3-pnCxKO mice); mushroom (Mush.) ∼21% (from 6.77 ± 0.42 spines/10 μm in control mice to 5.35 ± 0.66 spines/10 μm in Thy1-GFP-M;Tshz3-pnCxKO mice); thin ∼34%. *p < .05 and **p < .01, control vs. Tshz3-pnCxKO mice, Mann-Whitney test.

      Electrophysiological Properties of L5 CPNs

      The two main subtypes of L5 CPNs, pyramidal tract (PT) and intratelencephalic (IT) CPNs, were distinguished according to their electrophysiological properties (see Supplement 1), in particular the sum of the “sag & rebound” + the afterhyperpolarization responses (Figure 2B), whose value is larger in PT neurons (
      • Lee A.T.
      • Gee S.M.
      • Vogt D.
      • Patel T.
      • Rubenstein J.L.
      • Sohal V.S.
      Pyramidal neurons in prefrontal cortex receive subtype-specific forms of excitation and inhibition.
      ). This “sag & rebound” + afterhyperpolarization sum was similar in control and Tshz3-pnCxKO mice (Figure 2C), indicating no major changes in membrane properties of CPNs and in the proportion of PT versus IT neurons. Other membrane and action potential (AP) properties were similar comparing IT versus PT neurons of the same genotype, as well as between PT and IT neurons of control versus Tshz3-pnCxKO mice (Table S1 in Supplement 1). Moreover, also comparing AP discharge patterns of PT versus IT neurons of the same genotype revealed no significant differences (Figure S6 in Supplement 1). We thus pooled PT and IT neurons of each genotype for further analyses. CPNs from Tshz3-pnCxKO mice showed a lower number of APs in response to depolarizing current pulses compared with control mice, although values were not significantly different (Figure 2D). The inter-AP interval in response to a strong depolarizing current (+300 pA) was also similar (Figure 2E). However, when measuring this parameter in response to a threshold current step (+150 pA, just above the rheobase), we found that the inter-AP intervals were significantly longer in CPNs from Tshz3-pnCxKO mice, suggesting increased accommodation and, possibly, decreased excitability (Figure 2F).
      Figure thumbnail gr2
      Figure 2Electrophysiological properties of layer 5 cortical projection neurons (CPNs). (A) Scheme of a corticostriatal slice with a layer 2/3 neuron (light gray), a gamma-aminobutyric acidergic interneuron (dark gray), and a layer 5 CPN (black), as well as the stimulating and recording electrodes. (B) Sample traces showing the responses to ±200-pA pulses of a pyramidal tract (PT) and an intratelencephalic (IT) CPN recorded from the primary motor and somatosensory cortex [as in panel (A)]: note the prominent sag, rebound, and afterhyperpolarization (AHP) of the PT neuron compared with the IT neuron. (C) Arbitrary classification of CPNs as PT or IT based on the sum of the sag and rebound in response to a −50-pA current pulse, and the AHP following a +50-pA pulse (both for 800 ms). CPNs are considered as PT if sag + rebound + AHP >4.2 mV (dotted line) and as IT if <4.2 mV. The distribution of sag + rebound + AHP values is similar in control mice and conditional mutant mice with postnatal loss of Tshz3 in the cortex (Tshz3-pnCxKO) (n = 14 for each genotype, p = .765, Mann-Whitney test). (D) The number of action potentials (APs) triggered by depolarizing current pulses (800 ms) is similar between control and mutant mice (F1,66 = 0.6076, p = .444, matched 2-way analysis of variance). (E) The graph shows the delay between the beginning of a depolarizing current pulse (+300 pA, 800 ms) and the first 20 APs triggered (see inset) in CPNs: average values are similar in the two groups (n = 14 for each genotype [F1,436 = 0.102, p = .75, two-way analysis of variance]). (F) Same as panel (E), but using a smaller current pulse (+150 pA, 800 ms; see inset) just above the average rheobase: Tshz3-pnCxKO CPNs show significantly longer inter-AP intervals (n = 14 for each genotype [F1,150 = 7.135, p = .008, two-way analysis of variance]). **p = .008. (D–F) Data are expressed as mean ± SEM.

      Cortical Synaptic Transmission

      The paired-pulse ratio of AMPA receptor-mediated EPSCs recorded in L5 CPNs was significantly higher in Tshz3-pnCxKO CPNs compared with control CPNs (Figure 3A), suggesting a decreased probability of AP-dependent glutamate release from L2/3 cortical neurons. In contrast, the paired-pulse ratio of GABAergic inhibitory postsynaptic currents was similar in both genotypes (Figure 3B). Concerning AP-independent spontaneous activity, the distribution of miniature EPSC interevent interval and amplitude, as well as the average frequency and amplitude (Figure 3C), were similar in Tshz3-pnCxKO and control mice. Finally, the NMDA/AMPA ratio was also similar between control and mutant (0.69 ± 0.11, n = 22 vs. 0.51 ± 0.08, n = 20; p = .296, Mann-Whitney test).
      Figure thumbnail gr3
      Figure 3Cortical synaptic transmission. (A) The paired-pulse ratio (PPR) of glutamatergic excitatory postsynaptic currents (EPSCs) is significantly higher in cortical projection neurons (CPNs) of conditional mutant mice with postnatal loss of Tshz3 in the cortex (Tshz3-pnCxKO) (n = 29) compared with control mice (n = 22) (F1,241 = 10.72, p = .0012, two-way analysis of variance). Traces (inset) depict superimposed EPSCs (amplitude is normalized to control EPSC1) triggered by paired-pulse stimulation recorded from CPNs of control and mutant mice. ***p = .0012. (B) The average PPR of gamma-aminobutyric acidergic inhibitory postsynaptic currents (IPSCs) is similar in CPNs of Tshz3-pnCxKO (n = 14) and control (n = 10) mice (F1,109 = 3.38, p = .069). (C) Cumulative distributions of miniature EPSC (mEPSC) interevent interval and amplitude (graphs), as well as frequency and amplitude (insets), are similar in CPNs of control and Tshz3-pnCxKO mice. Traces depict sample mEPSCs. Cumulative distribution data were analyzed by 2-sample Kolmogorov-Smirnov test, average data by Mann-Whitney test; n = 10 per group, p > .05 for all tests. (A–C) Data are expressed as mean ± SEM.

      Corticostriatal Synaptic Transmission and Plasticity

      In the mouse, the bulk of the corticostriatal pathway is constituted by axons of L5 CPNs reaching the striatum around P3 to P4 and forming synapses from P10 onward (
      • Sohur U.S.
      • Padmanabhan H.K.
      • Kotchetkov I.S.
      • Menezes J.R.
      • Macklis J.D.
      Anatomic and molecular development of corticostriatal projection neurons in mice.
      ). Their main targets are SPNs, which constitute >90% of the whole striatal neuronal population and do not express TSHZ3 (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ). These facts and the implication of the corticostriatal pathway in ASD (
      • Shepherd G.M.
      Corticostriatal connectivity and its role in disease.
      ) make this circuit (Figure 4A) a valuable tool for investigating the functional consequences of Tshz3 loss in CPNs.
      Figure thumbnail gr4
      Figure 4Corticostriatal synaptic transmission. (A) Scheme of a corticostriatal slice with a layer 5 cortical projection neuron (black) and a striatal spiny projection neuron (SPN) (gray), as well as the stimulating and recording electrodes. Traces depict the typical response of an SPN to the injection of hyperpolarizing (−200 pA) and depolarizing (+200, +400 pA) currents. (B) The paired-pulse ratio of corticostriatal glutamatergic excitatory postsynaptic currents (EPSCs) is significantly higher in SPNs of conditional mutant mice with postnatal loss of Tshz3 in the cortex (Tshz3-pnCxKO) (n = 15) compared with control mice (n = 21) (F1,87 = 5.7, p = .0182, two-way analysis of variance). Traces depict superimposed EPSCs (amplitude is normalized to control EPSC1) recorded from SPNs of control and mutant mice. *p = .0182. (C) Cumulative distributions of miniature EPSC (mEPSC) interevent interval and amplitude (graphs), as well as frequency and amplitude (insets), are similar in SPNs of control and Tshz3-pnCxKO mice (n = 8 for both; 2-sample Kolmogorov-Smirnov and Mann-Whitney test, p > .05 for all). Traces show sample mEPSCs. (D) The N-methyl-D-aspartate (NMDA)/alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) ratio is significantly higher in SPNs from Tshz3-pnCxKO compared with control mice (n = 8 for both; p < .05, Mann-Whitney test). Traces depict sample EPSCs recorded at −60 and +40 mV to reveal the AMPA and NMDA receptor–mediated component, respectively (gray dots show where currents were measured). *p < .05. (B–D) Data are expressed as mean ± SEM.
      Striatal SPNs recorded from control (n = 16) and Tshz3-pnCxKO (n = 22) mice presented similar resting membrane potential and current–voltage relationship (not shown). Interestingly, the paired-pulse ratio was significantly higher in Tshz3-pnCxKO mice (Figure 4B), suggesting a decreased probability of AP-dependent glutamate release from L5 CPNs. Conversely, miniature EPSC parameters were not changed in Tshz3-pnCxKO compared with control mice (Figure 4C). Furthermore, in Tshz3-pnCxKO mice, the NMDA/AMPA ratio was significantly increased (Figure 4D).
      Concerning synaptic plasticity, long-term potentiation (LTP) was present in both control and mutant mice (Figure 5A). Conversely, long-term depression (LTD) was absent in Tshz3-pnCxKO mice and was even reversed as a transient potentiation (Figure 5B). Interestingly, the blockade of NMDA receptors by AP-5 (40 μM) partially recovered LTD and abolished the transient potentiation (Figure 5B), suggesting that the lack of LTD could be due, at least in part, to increased NMDA receptor–mediated signaling.
      Figure thumbnail gr5
      Figure 5Corticostriatal synaptic plasticity. (A) Long-term potentiation (LTP) is similar in spiny projection neurons from conditional mutant mice with postnatal loss of Tshz3 in the cortex (Tshz3-pnCxKO) and control mice (n = 10 and 7, respectively). The left graph shows the time course of excitatory postsynaptic current (EPSC) amplitude normalized to baseline (black arrow represents LTP induction protocol). The right histogram shows normalized EPSC amplitude after the LTP induction protocol (p < .001 compared with baseline, Mann-Whitney test). Traces depict sample EPSCs at baseline and 15 minutes after the LTP induction protocol (as indicated). ***p < .001. (B) Long-term depression (LTD) is present in control spiny projection neurons (n = 10), while it is absent in Tshz3-pnCxKO mice (n = 11), where a transient potentiation is observed. The application of 40-μM AP-5 abolishes the transient potentiation and partially restores LTD (n = 7). Sample traces depict EPSCs at baseline and 15 minutes after LTD induction protocol in the different experimental conditions (as indicated). Note the absence of LTD in Tshz3-pnCxKO mice and its recovery by AP-5. *p < .05, ***p < .001 compared with baseline, Mann-Whitney test; °p < .05, °°°p < .001 between groups, Kruskal-Wallis and Dunn's post-test. (A, B) Data are expressed as mean ± SEM.

      ASD-Relevant Behavior

      Tshz3-pnCxKO mice displayed more stereotyped behaviors than control mice: they buried more marbles and dipped more repeatedly in the hole board (Figure 6A and Figure S7A, B in Supplement 1). Tshz3-pnCxKO mice showed reduced field of interest with smaller number of zone crossing in the open field (Figure 6B, C) and also had lower sociability and lower interest in social novelty, as revealed by the two-chamber test (Figure 6D–F). In addition, Tshz3-pnCxKO mice showed increased anxiety-like behavior: they avoided the central area more in the open field (Figure 6B, G), as well as the open arms of the elevated plus maze (Figure 6H). Mutant and control mice learned equally in the Morris water maze (see Supplement 1): neither the learning slopes nor the probe scores differed, and the visible platform values were identical in the two groups (Figure 6I). The loss of Tshz3 did not modify the hind paw coordination (2.0 ± 0.36 vs. 1.78 ± 0.52 hind paw slips for control mice vs. Tshz3-pnCxKO mice; n = 9 per group [t16 = 0.87, p = .39]). Finally, mutant mice did not show visual, auditory, and olfactory deficits (Figure S7C–E in Supplement 1).
      Figure thumbnail gr6
      Figure 6Stereotyped behavior, field of interest, sociability, interest in social novelty, anxiety-like behavior, and spatial learning. Conditional mutant mice with postnatal loss of Tshz3 in the cortex (Tshz3-pnCxKO) displayed more stereotyped behaviors than control mice: (A) they dipped more repeatedly in the hole board (F1,17 = 45.98, p = 6 × 10−6; partial η2 = .73) when controlled for nonstereotyped dips (see in ). (B, C) Tshz3-pnCxKO mice showed reduced field of interest with smaller number of zone crossing in the open field when controlled for distance walked (F1,17 = 124.59, p = 3 × 10−9, partial η2 = .88; covariate: distance walked, which was similar in the two groups [t18 = 0.39, p = .70, Cohen's d = 0.17]). Concerning sociability, Tshz3-pnCxKO mice did not differ in exploration from control mice (D), but they did not show more contacts with conspecific C57BL/6J mice than (E) with a lure or (F) with a new SWR mouse vs. the familiar C57BL/6J mouse. Two repeated-measures analyses of covariance were designed. The interactions between the genotype and the content of the pencil box were significant in each condition (D–F) (condition B6 vs. lure [F1,17= 31.75, p = .00003, partial η2 = .65], condition B6 vs. SWR mouse [F1,17 = 30.83, p = .00003, partial η2 = .65; covariate: exploration level, panel (D)]). Control mice explored more the box containing a C57BL/6J than that containing a lure (dependent [t9 = 12.48; p = 3 × 10−7; η2 = .52]), and the box containing the SWR mouse more than that containing the familiar C57BL/6J animal (dependent [t9 = 8.10; p = .00002; η2 = .58]). In contrast, Tshz3-pnCxKO mice did not show more contacts with a conspecific C57BL/6J mouse than with a lure (dependent [t9 = 0.39, p = .71]), or with the new SWR vs. the familiar C57BL/6J mouse (dependent [t9 = 0.41, p = .69]). (G) Tshz3-pnCxKO mice showed more anxiety-like behavior: they avoided more the central area in the open field (F1,17 = 124.59, p = 3 × 10−9; partial η2 = .88; covariate: distance walked) when controlled for distance walked. (H) Tshz3-pnCxKO mice avoided also the open arms of the elevated plus maze when controlled for distance walked in the closed arms (F1,17 = 36.41, p = .00001, partial η2 = .68; covariate: activity in closed arms). (I) Both groups learned in a spatial task (Morris water maze): neither the learning slopes nor the probe score differed (t19 = 0.53, p = .60, and t19 = 0.30, p = .77, respectively); note that “visible platform” dots partially overlap. In all cases except one (marble burying), the effect size of the differences is large (see Results). (A) n = 11 per group. (B–H) n = 10 per group. (K) n = 9 per Tshz3-pnCxKO mouse and n = 12 per control mouse. (K) *p < .01, ***p < .0001; data are expressed as mean ± SEM.

      Genes Differentially Expressed in the Tshz3-pnCxKO Cerebral Cortex Are Strongly Associated With ASD

      RNA sequencing, performed at P28 in the cortex of Tshz3-pnCxKO and control mice, identified 1025 differentially expressed genes (DEGs), among which 767 were upregulated and 258 downregulated (p < .05) (Table S2A in Supplement 2). A total of 993 of these DEGs have nonambiguous human orthologues (Table S2B in Supplement 2). GSEA showed consistent negative or positive enrichment of Gene Ontology terms associated with neurological and synaptic functions/pathways, as well as neurological disorders, such as ASD and Alzheimer’s disease (Figure 7A, B; Table S3A–X in Supplement 3). GSEA from the SynaptomeDB [http://metamoodics.org/SynaptomeDB/index.php (
      • Pirooznia M.
      • Wang T.
      • Avramopoulos D.
      • Valle D.
      • Thomas G.
      • Huganir R.L.
      • et al.
      SynaptomeDB: An ontology-based knowledgebase for synaptic genes.
      )] and from the G2C postsynaptic proteome datasets [http://www.genes2cognition.org/proteomics (
      • Collins M.O.
      • Husi H.
      • Yu L.
      • Brandon J.M.
      • Anderson C.N.
      • Blackstock W.P.
      • et al.
      Molecular characterization and comparison of the components and multiprotein complexes in the postsynaptic proteome.
      )] revealed positive enrichment of postsynaptic, presynaptic, and presynaptic active zone and negative enrichment for vesicle Gene Ontology terms (Figure 7C; Table S3L–O in Supplement 3). Interestingly, we found a strong enrichment for the NMDA receptor pathway, while few genes were related to the AMPA receptor pathway (Figure 7C). Of the 1025 DEGs, 173 (16.7%) encode for postsynaptic density (PSD) proteins from the adult mouse cerebral cortex, among which 167 have human orthologues (
      • Bayes A.
      • Collins M.O.
      • Croning M.D.
      • van de Lagemaat L.N.
      • Choudhary J.S.
      • Grant S.G.
      Comparative study of human and mouse postsynaptic proteomes finds high compositional conservation and abundance differences for key synaptic proteins.
      ) (Table S2B–D in Supplement 2). Interestingly, among these 167 proteins, 28 were identified as components of the DLG4, DLGAP1, and/or SHANK3 postsynaptic protein-interaction networks of the adult mouse cortex (
      • Li J.
      • Zhang W.
      • Yang H.
      • Howrigan D.P.
      • Wilkinson B.
      • Souaiaia T.
      • et al.
      Spatiotemporal profile of postsynaptic interactomes integrates components of complex brain disorders.
      ) (Table S2E in Supplement 2).
      Figure thumbnail gr7
      Figure 7Gene Set Enrichment Analysis (GSEA) enrichment. (A) Positive and (B) negative GSEA enrichment of the Gene Ontology category “Biological Process” using the preranked Tshz3 gene list. The normalized enriched score (NES) value from the gene set enrichment is shown. Red color indicates neuro-related terms, orange color designates kidney-related terms, and all other terms are shown in light blue. (C) Network of GSEA enrichments of different synaptome areas and postsynaptic neurotransmitter pathways. The color of the nodes represents the −log10 of the p value (pval) from GSEA for each term, the size of the nodes roughly indicates the total number of genes found for each term based on the formula (number of genes/10) × 2. The strength of the interactions (line thickness) indicates shared genes between the terms. (D) Histogram showing the most represented human brain and nervous system pathologies associated with the 357 human orthologs of the 382 Tshz3-regulated differentially expressed genes (DEGs) with log2 fold change >0.5 and p value (or false discovery rate) <.05. Each gene is scored on the basis of the number of relevant publications that associate it with a pathology. Scores were given as follows: 1, one publication; 2, two publications; 3, >2 publications. AMPAR, alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor; ASD, autism spectrum disorder; GF, growth factor; mGluR5, metabotropic glutamate receptor 5; NADH, nicotinamide adenine dinucleotide; NADP, nicotinamide adenine dinucleotide phosphate; ncRNA, noncoding RNA; NMDAR, N-methyl-D-aspartate receptor; PSD, postsynaptic density; PSP, postsynaptic potential.
      Last, of the 993 human orthologues, 741 (74.6%) are involved in brain and nervous system disorders. Among these 741 genes, 489 (66%) are known or proposed to be involved in ASD (Table S2F, G in Supplement 2). Note that the percentage of ASD-associated genes remains high when restricting the analysis to the 357 human orthologues of the 382 DEGs having an absolute value of log2 fold change >0.5 and the adjusted p value (or false discovery rate) <.05 (Table S2H in Supplement 2). In this condition, 155 (43.4%) genes of 357 are associated with ASD, among which 13 (GRIN2A, GRIN2B, MIF, MYH6, MYH14, NDUFA13, NOS1, PHGDH, PRR7, PURA, RFTN1, SYT2, VGF) encode for PSD proteins (Figure 7D; Table S2H–K in Supplement 2).

      Discussion

      In our previous work (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ), we showed a direct link between TSHZ3 loss and ASD and reported dramatic changes in gene expression in the fetal neocortex (E18.5) of Tshz3−/− mice, with 243 DEGs.
      Here, transcriptomic analysis of Tshz3-pnCxKO mice at P28 revealed a very different pattern and number of DEGs. Notably, there were 1025 DEGs, among which only 38 previously identified in Tshz3−/− (see Table S4 in Supplement 4), showing that TSHZ3 has different regulatory functions at prenatal versus postnatal stages. Interestingly, postnatal Tshz3 loss also clearly leads to abnormal expression of 173 genes of the PSD, as well as 59 “presynaptic” genes such as Nrxn2, Syt2, and Syn2 (Table S2L in Supplement 2) implicated in synapse adhesion and neurotransmitter release (
      • Weingarten J.
      • Lassek M.
      • Mueller B.F.
      • Rohmer M.
      • Lunger I.
      • Baeumlisberger D.
      • et al.
      The proteome of the presynaptic active zone from mouse brain.
      ,
      • Zhang W.
      • Zhang Y.
      • Zheng H.
      • Zhang C.
      • Xiong W.
      • Olyarchuk J.G.
      • et al.
      SynDB: A Synapse protein DataBase based on synapse ontology.
      ). Altogether, these results highlight a fascinating dual role of TSHZ3 in prenatal versus postnatal stages, confirming that it participates in different transcriptional programs dependent on the developmental stage. All these DEGs belong to a converging network in terms of human brain diseases: notably, in both cases, ASD first, and schizophrenia second, are the most represented pathologies associated to these genes. Consistently, we show here that Tshz3-pnCxKO mice display the whole set of ASD-like behavioral abnormalities, namely social interaction deficits, restricted fields of interest, and stereotypies, similarly to Tshz3+/− mice, a model mimicking the human pathology (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ). Therefore, the alteration of different transcriptional programs resulting from Tshz3 deficiency at embryonic or postnatal stages converges into a similar ASD phenotype. These data raise the issue of the contribution of postnatal events to the ASD phenotype of Tshz3+/− mice and call for future rescue experiments to test the hypothesis of an early postnatal therapeutic window for the ASD-like syndrome linked to TSHZ3 haploinsufficiency.
      Mutations linked to ASD induce structural and functional changes in brain circuitry, including spine morphology/density, synaptic transmission, and plasticity (
      • Bourgeron T.
      From the genetic architecture to synaptic plasticity in autism spectrum disorder.
      ), strongly involving the corticostriatal circuitry (
      • Delmonte S.
      • Gallagher L.
      • O'Hanlon E.
      • McGrath J.
      • Balsters J.H.
      Functional and structural connectivity of frontostriatal circuitry in autism spectrum disorder.
      ,
      • Shepherd G.M.
      Corticostriatal connectivity and its role in disease.
      ,
      • Maloney S.E.
      • Rieger M.A.
      • Dougherty J.D.
      Identifying essential cell types and circuits in autism spectrum disorders.
      ), suggesting that this pathway constitutes a main target for studying brain dysfunctions associated to this pathology. Accordingly, we reported functional abnormalities in the corticostriatal pathway associated with the ASD-related phenotype in Tshz3+/− mice (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ). Consistently, here, we found that cortical AP-dependent glutamate release is significantly reduced in Tshz3-pnCxKO mice and that L5 CPNs have decreased spine density, which is usually associated to diminished synaptic activity (
      • Sekino Y.
      • Kojima N.
      • Shirao T.
      Role of actin cytoskeleton in dendritic spine morphogenesis.
      ), as well as slightly reduced excitability. L2/3 and L5 neurons normally express TSHZ3, and thus these changes could be attributed to a “direct” effect of postnatal Tshz3 loss on their maturation and functioning as well as on the establishment of their synaptic network. In turn, the reduced AP-dependent glutamate release in the striatum could be interpreted as a consequence of the above-mentioned modifications at cortical level. In parallel, the increased NMDA/AMPA ratio measured in striatal SPNs could be interpreted as a compensatory mechanism: NMDA receptor signaling, which is crucial for synaptic plasticity (
      • Fino E.
      • Paille V.
      • Cui Y.
      • Morera-Herreras T.
      • Deniau J.M.
      • Venance L.
      Distinct coincidence detectors govern the corticostriatal spike timing-dependent plasticity.
      ), could be enhanced to compensate for the lower corticostriatal input. In this respect, the abnormal corticostriatal LTD reported here is particularly striking and fitting: LTD in the striatum can be induced in conditions limiting NMDA receptor activation; conversely, LTP induction requires their full activation (
      • Chassain C.
      • Melon C.
      • Salin P.
      • Vitale F.
      • Couraud S.
      • Durif F.
      • et al.
      Metabolic, synaptic and behavioral impact of 5-week chronic deep brain stimulation in hemiparkinsonian rats.
      ,
      • Calabresi P.
      • Maj R.
      • Pisani A.
      • Mercuri N.B.
      • Bernardi G.
      Long-term synaptic depression in the striatum: Physiological and pharmacological characterization.
      ,
      • Calabresi P.
      • Pisani A.
      • Mercuri N.B.
      • Bernardi G.
      Long-term potentiation in the striatum is unmasked by removing the voltage-dependent magnesium block of Nmda receptor channels.
      ,
      • Lovinger D.M.
      • Tyler E.C.
      • Merritt A.
      Short- and long-term synaptic depression in rat neostriatum.
      ,
      • Partridge J.G.
      • Tang K.C.
      • Lovinger D.M.
      Regional and postnatal heterogeneity of activity-dependent long-term changes in synaptic efficacy in the dorsal striatum.
      ,
      • Lovinger D.M.
      Neurotransmitter roles in synaptic modulation, plasticity and learning in the dorsal striatum.
      ,
      • Bagetta V.
      • Picconi B.
      • Marinucci S.
      • Sgobio C.
      • Pendolino V.
      • Ghiglieri V.
      • et al.
      Dopamine-dependent long-term depression is expressed in striatal spiny neurons of both direct and indirect pathways: Implications for Parkinson's disease.
      ,
      • Paille V.
      • Picconi B.
      • Bagetta V.
      • Ghiglieri V.
      • Sgobio C.
      • Di Filippo M.
      • et al.
      Distinct levels of dopamine denervation differentially alter striatal synaptic plasticity and NMDA receptor subunit composition.
      ). In this context, the exacerbated NMDA receptor–mediated signaling might lead to the observed loss of LTD and the transient LTP-like phenomenon; the partial recovery of LTD by NMDA receptor blockade strongly supports this hypothesis. Alternatively, or concomitantly, LTD loss could be attributed to an occlusion effect: as basal corticostriatal synaptic transmission is downregulated, it cannot be further depressed. Interestingly, changes in corticostriatal function have been evidenced in several mouse models of ASD (
      • Peca J.
      • Feliciano C.
      • Ting J.T.
      • Wang W.
      • Wells M.F.
      • Venkatraman T.N.
      • et al.
      Shank3 mutant mice display autistic-like behaviours and striatal dysfunction.
      ,
      • Blundell J.
      • Blaiss C.A.
      • Etherton M.R.
      • Espinosa F.
      • Tabuchi K.
      • Walz C.
      • et al.
      Neuroligin-1 deletion results in impaired spatial memory and increased repetitive behavior.
      ,
      • Portmann T.
      • Yang M.
      • Mao R.
      • Panagiotakos G.
      • Ellegood J.
      • Dolen G.
      • et al.
      Behavioral abnormalities and circuit defects in the basal ganglia of a mouse model of 16p11.2 deletion syndrome.
      ,
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ). These alterations are heterogeneous, possibly owing to the different models, experimental protocols, and genetic backgrounds. However, they all strongly suggest that altered corticostriatal synaptic transmission/plasticity is a hallmark of ASD.
      Long-lasting changes of synaptic strength can reinforce (LTP) or depress (LTD) specific neural circuits, thus gating salient information while suppressing unwanted ones to optimize behavioral responses (
      • Berretta N.
      • Nistico R.
      • Bernardi G.
      • Mercuri N.B.
      Synaptic plasticity in the basal ganglia: A similar code for physiological and pathological conditions.
      ). If, as in our case, one of the two forms of plasticity is altered or absent, these processes would be unpaired, contributing to the observed behavioral abnormalities. Whereas the molecular and anatomo-functional defects in corticostriatal circuitry could underlie at least part of the ASD-like features of Tshz3-pnCxKO mice, other brain circuits and structures might also be involved, such as those involving the medial prefrontal cortex and the basolateral amygdala (
      • Wang X.
      • Kery R.
      • Xiong Q.
      Synaptopathology in autism spectrum disorders: Complex effects of synaptic genes on neural circuits.
      ), as TSHZ3 and CaMKIIalpha are expressed there (
      • Caubit X.
      • Gubellini P.
      • Andrieux J.
      • Roubertoux P.L.
      • Metwaly M.
      • Jacq B.
      • et al.
      TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
      ,
      • Caubit X.
      • Tiveron M.C.
      • Cremer H.
      • Fasano L.
      Expression patterns of the three Teashirt-related genes define specific boundaries in the developing and postnatal mouse forebrain.
      ,
      • Casanova E.
      • Fehsenfeld S.
      • Mantamadiotis T.
      • Lemberger T.
      • Greiner E.
      • Stewart A.F.
      • Schutz G.
      A CamKIIalpha iCre BAC allows brain-specific gene inactivation.
      ). In agreement with the very low Tshz3 expression in the hippocampus (
      • Caubit X.
      • Tiveron M.C.
      • Cremer H.
      • Fasano L.
      Expression patterns of the three Teashirt-related genes define specific boundaries in the developing and postnatal mouse forebrain.
      ), Tshz3-pnCxKO mice have no memory deficit.
      Overall, the above-mentioned molecular, functional, and morphological changes in CPNs due to Tshz3 loss further support the current idea that ASD can be considered as a synaptopathy (
      • Wang X.
      • Kery R.
      • Xiong Q.
      Synaptopathology in autism spectrum disorders: Complex effects of synaptic genes on neural circuits.
      ,
      • Reig-Viader R.
      • Sindreu C.
      • Bayes A.
      Synaptic proteomics as a means to identify the molecular basis of mental illness: Are we getting there?.
      ). Here, we show that several DEGs in Tshz3-pnCxKO mice are actually involved in glutamatergic synaptic transmission at both pre- and postsynaptic level, including genes involved in NMDA receptor signaling pathway, PSD-95, and SH3 and multiple ankyrin repeat domains protein complex. There is increasing evidence that NMDA receptor–dependent signaling pathway is involved not only in ASD, but also in Alzheimer’s disease, while AMPA receptor–mediated signaling is relatively spared (
      • Lepeta K.
      • Lourenco M.V.
      • Schweitzer B.C.
      • Martino Adami P.V.
      • Banerjee P.
      • Catuara-Solarz S.
      • et al.
      Synaptopathies: Synaptic dysfunction in neurological disorders - A review from students to students.
      ), leading to the hypothesis that TSHZ3-related ASD can be considered as a synaptopathy linked to NMDA receptor pathway abnormalities.
      Here, we demonstrate that Tshz3 plays an essential role in the cortex and in corticostriatal circuitry during postnatal development, which is different from that played at prenatal stages in terms of gene expression modulation. This supports the position of TSHZ3 as a critical neurodevelopmental regulator participating at differential age-related processes of gene transcription and confirms its role as an ASD-risk gene, providing new ASD models in rodents and pointing to dysfunctional corticostriatal circuitry as a substrate for this group of neurodevelopmental disorders.

      Acknowledgments and Disclosures

      This work was supported by the French National Research Agency (ANR) “TSHZ3inASD” project Grant No. ANR-17-CE16-0030-01 (to LF, LK-LG, and BH); a grant from the Fédération pour la Recherche sur le Cerveau (FRC) (to LF); the Centre National de la Recherche Scientifique (CNRS), the Institut National de la Santé et de la Recherche Médicale (INSERM) , and Aix Marseille Univ ; “Investments for the Future” program (France-BioImaging) Grant No. ANR-10-INSB-04-01 (to the imaging platform of the IBDM); and "France Génomique" national infrastructure “Investissement d’Avenir” Grant No. ANR-10-INSB-09 (to DS and ED). IBDM and LPC are affiliated with Marseille NeuroSchool, a graduate school supported by the A*MIDEX foundation and the French ANR funded by the French Government "Investissements d’Avenir" program (nEURo*AMU, ANR-17-EURE-0029 grant).
      We wish to thank Dr. Francis Castet and Dr. Jean-Pierre Kessler (IBDM) for, respectively, performing some experiments and his kind advice on NMDA receptors.
      The authors report no biomedical financial interests or potential conflicts of interest.

      Supplementary Material

      References

        • DeLong M.R.
        • Wichmann T.
        Circuits and circuit disorders of the basal ganglia.
        Arch Neurol. 2007; 64: 20-24
        • Levy R.
        • Hazrati L.N.
        • Herrero M.T.
        • Vila M.
        • Hassani O.K.
        • Mouroux M.
        • et al.
        Re-evaluation of the functional anatomy of the basal ganglia in normal and Parkinsonian states.
        Neuroscience. 1997; 76: 335-343
        • Redgrave P.
        • Vautrelle N.
        • Reynolds J.N.
        Functional properties of the basal ganglia's re-entrant loop architecture: Selection and reinforcement.
        Neuroscience. 2011; 198: 138-151
        • Willsey A.J.
        • Sanders S.J.
        • Li M.
        • Dong S.
        • Tebbenkamp A.T.
        • Muhle R.A.
        • et al.
        Coexpression networks implicate human midfetal deep cortical projection neurons in the pathogenesis of autism.
        Cell. 2013; 155: 997-1007
        • Parikshak N.N.
        • Luo R.
        • Zhang A.
        • Won H.
        • Lowe J.K.
        • Chandran V.
        • et al.
        Integrative functional genomic analyses implicate specific molecular pathways and circuits in autism.
        Cell. 2013; 155: 1008-1021
        • State M.W.
        • Sestan N.
        Neuroscience. The emerging biology of autism spectrum disorders.
        Science. 2012; 337: 1301-1303
        • Delmonte S.
        • Gallagher L.
        • O'Hanlon E.
        • McGrath J.
        • Balsters J.H.
        Functional and structural connectivity of frontostriatal circuitry in autism spectrum disorder.
        Front Hum Neurosci. 2013; 7: 430
        • Langen M.
        • Bos D.
        • Noordermeer S.D.
        • Nederveen H.
        • van Engeland H.
        • Durston S.
        Changes in the development of striatum are involved in repetitive behavior in autism.
        Biol Psychiatry. 2014; 76: 405-411
        • Langen M.
        • Schnack H.G.
        • Nederveen H.
        • Bos D.
        • Lahuis B.E.
        • de Jonge M.V.
        • et al.
        Changes in the developmental trajectories of striatum in autism.
        Biol Psychiatry. 2009; 66: 327-333
        • Shepherd G.M.
        Corticostriatal connectivity and its role in disease.
        Nat Rev Neurosci. 2013; 14: 278-291
        • American Psychiatric Association
        Diagnostic and Statistical Manual of Mental Disorders.
        5th ed. American Psychiatric Press, Washington, DC2013
        • Bourgeron T.
        Current knowledge on the genetics of autism and propositions for future research.
        C R Biol. 2016; 339: 300-307
        • Buxbaum J.D.
        • Daly M.J.
        • Devlin B.
        • Lehner T.
        • Roeder K.
        • State M.W.
        • Autism Sequencing C
        The autism sequencing consortium: large-scale, high-throughput sequencing in autism spectrum disorders.
        Neuron. 2012; 76: 1052-1056
        • Park H.R.
        • Lee J.M.
        • Moon H.E.
        • Lee D.S.
        • Kim B.N.
        • Kim J.
        • et al.
        A short review on the current understanding of autism spectrum disorders.
        Exp Neurobiol. 2016; 25: 1-13
        • Ziats M.N.
        • Rennert O.M.
        The evolving diagnostic and genetic landscapes of autism spectrum disorder.
        Front Genet. 2016; 7: 65
        • Peixoto R.T.
        • Wang W.G.
        • Croney D.M.
        • Kozorovitskiy Y.
        • Sabatini B.L.
        Early hyperactivity and precocious maturation of corticostriatal circuits in Shank3B(−/−) mice.
        Nature Neuroscience. 2016; 19: 716-724
        • Peca J.
        • Feliciano C.
        • Ting J.T.
        • Wang W.
        • Wells M.F.
        • Venkatraman T.N.
        • et al.
        Shank3 mutant mice display autistic-like behaviours and striatal dysfunction.
        Nature. 2011; 472: 437-442
        • Blundell J.
        • Blaiss C.A.
        • Etherton M.R.
        • Espinosa F.
        • Tabuchi K.
        • Walz C.
        • et al.
        Neuroligin-1 deletion results in impaired spatial memory and increased repetitive behavior.
        J Neurosci. 2010; 30: 2115-2129
        • Portmann T.
        • Yang M.
        • Mao R.
        • Panagiotakos G.
        • Ellegood J.
        • Dolen G.
        • et al.
        Behavioral abnormalities and circuit defects in the basal ganglia of a mouse model of 16p11.2 deletion syndrome.
        Cell Rep. 2014; 7: 1077-1092
        • Caubit X.
        • Gubellini P.
        • Andrieux J.
        • Roubertoux P.L.
        • Metwaly M.
        • Jacq B.
        • et al.
        TSHZ3 deletion causes an autism syndrome and defects in cortical projection neurons.
        Nat Genet. 2016; 48: 1359-1369
        • Caubit X.
        • Tiveron M.C.
        • Cremer H.
        • Fasano L.
        Expression patterns of the three Teashirt-related genes define specific boundaries in the developing and postnatal mouse forebrain.
        J Comp Neurol. 2005; 486: 76-88
        • Li M.
        • Santpere G.
        • Imamura Kawasawa Y.
        • Evgrafov O.V.
        • Gulden F.O.
        • Pochareddy S.
        • et al.
        Integrative functional genomic analysis of human brain development and neuropsychiatric risks.
        Science. 2018; 362: eaat7615
        • Caubit X.
        • Thoby-Brisson M.
        • Voituron N.
        • Filippi P.
        • Bevengut M.
        • Faralli H.
        • et al.
        Teashirt 3 regulates development of neurons involved in both respiratory rhythm and airflow control.
        J Neurosci. 2010; 30: 9465-9476
        • Kang H.J.
        • Kawasawa Y.I.
        • Cheng F.
        • Zhu Y.
        • Xu X.
        • Li M.
        • et al.
        Spatio-temporal transcriptome of the human brain.
        Nature. 2011; 478: 483-489
        • Workman A.D.
        • Charvet C.J.
        • Clancy B.
        • Darlington R.B.
        • Finlay B.L.
        Modeling transformations of neurodevelopmental sequences across mammalian species.
        J Neurosci. 2013; 33: 7368-7383
        • Tepper J.M.
        • Sharpe N.A.
        • Koos T.Z.
        • Trent F.
        Postnatal development of the rat neostriatum: Electrophysiological, light- and electron-microscopic studies.
        Dev Neurosci. 1998; 20: 125-145
        • Kwan K.Y.
        • Sestan N.
        • Anton E.S.
        Transcriptional co-regulation of neuronal migration and laminar identity in the neocortex.
        Development. 2012; 139: 1535-1546
        • Wang H.
        • Cuzon V.C.
        • Pickel V.M.
        Postnatal development of mu-opioid receptors in the rat caudate-putamen nucleus parallels asymmetric synapse formation.
        Neuroscience. 2003; 118: 695-708
        • Inaji M.
        • Sato K.
        • Momose-Sato Y.
        • Ohno K.
        Voltage-sensitive dye imaging analysis of functional development of the neonatal rat corticostriatal projection.
        Neuroimage. 2011; 54: 1831-1839
        • Mullah S.H.
        • Inaji M.
        • Nariai T.
        • Momose-Sato Y.
        • Sato K.
        • Ohno K.
        Optical analysis of developmental changes in synaptic potentiation in the neonatal rat corticostriatal projection.
        Neuroscience. 2012; 201: 338-348
        • Feng G.
        • Mellor R.H.
        • Bernstein M.
        • Keller-Peck C.
        • Nguyen Q.T.
        • Wallace M.
        • et al.
        Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP.
        Neuron. 2000; 28: 41-51
        • Rodriguez A.
        • Ehlenberger D.B.
        • Dickstein D.L.
        • Hof P.R.
        • Wearne S.L.
        Automated three-dimensional detection and shape classification of dendritic spines from fluorescence microscopy images.
        PLoS One. 2008; 3: e1997
        • Chassain C.
        • Melon C.
        • Salin P.
        • Vitale F.
        • Couraud S.
        • Durif F.
        • et al.
        Metabolic, synaptic and behavioral impact of 5-week chronic deep brain stimulation in hemiparkinsonian rats.
        J Neurochem. 2016; 136: 1004-1016
        • Jiang Z.G.
        • North R.A.
        Membrane properties and synaptic responses of rat striatal neurones in vitro.
        J Physiol. 1991; 443: 533-553
        • Hattox A.M.
        • Nelson S.B.
        Layer V neurons in mouse cortex projecting to different targets have distinct physiological properties.
        J Neurophysiol. 2007; 98: 3330-3340
        • Thomas A.
        • Burant A.
        • Bui N.
        • Graham D.
        • Yuva-Paylor L.A.
        • Paylor R.
        Marble burying reflects a repetitive and perseverative behavior more than novelty-induced anxiety.
        Psychopharmacology. 2009; 204: 361-373
        • Makanjuola R.O.
        • Hill G.
        • Maben I.
        • Dow R.C.
        • Ashcroft G.W.
        An automated method for studying exploratory and stereotyped behaviour in rats.
        Psychopharmacology (Berl). 1977; 52: 271-277
        • Roubertoux P.L.
        • Carlier M.
        • Tordjman S.
        Deficit in social relationships and reduced field of interest in mice.
        in: Roubertoux P.L. Deficit in Social Relationships and Reduced Field of Interest in Mice. Springer, New York2015: 335-370
        • Moy S.S.
        • Nadler J.J.
        • Perez A.
        • Barbaro R.P.
        • Johns J.M.
        • Magnuson T.R.
        • et al.
        Sociability and preference for social novelty in five inbred strains: An approach to assess autistic-like behavior in mice.
        Genes Brain Behav. 2004; 3: 287-302
        • Lai M.C.
        • Lombardo M.V.
        • Baron-Cohen S.
        Autism.
        Lancet. 2014; 383: 896-910
        • Roubertoux P.L.
        • Baril N.
        • Cau P.
        • Scajola C.
        • Ghata A.
        • Bartoli C.
        • et al.
        Differential brain, cognitive and motor profiles associated with partial trisomy. modeling down syndrome in mice.
        Behav Genet. 2017; 47: 305-322
        • Merico D.
        • Isserlin R.
        • Stueker O.
        • Emili A.
        • Bader G.D.
        Enrichment map: A network-based method for gene-set enrichment visualization and interpretation.
        PLoS One. 2010; 5: e13984
        • Pirooznia M.
        • Wang T.
        • Avramopoulos D.
        • Valle D.
        • Thomas G.
        • Huganir R.L.
        • et al.
        SynaptomeDB: An ontology-based knowledgebase for synaptic genes.
        Bioinformatics. 2012; 28: 897-899
        • Collins M.O.
        • Husi H.
        • Yu L.
        • Brandon J.M.
        • Anderson C.N.
        • Blackstock W.P.
        • et al.
        Molecular characterization and comparison of the components and multiprotein complexes in the postsynaptic proteome.
        J Neurochem. 2006; 97: 16-23
        • Subramanian A.
        • Tamayo P.
        • Mootha V.K.
        • Mukherjee S.
        • Ebert B.L.
        • Gillette M.A.
        • et al.
        Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles.
        Proc Natl Acad Sci U S A. 2005; 102: 15545-15550
        • Mootha V.K.
        • Lindgren C.M.
        • Eriksson K.F.
        • Subramanian A.
        • Sihag S.
        • Lehar J.
        • et al.
        PGC-1alpha-responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes.
        Nat Genet. 2003; 34: 267-273
        • Wang J.
        • Vasaikar S.
        • Shi Z.
        • Greer M.
        • Zhang B.
        WebGestalt 2017: A more comprehensive, powerful, flexible and interactive gene set enrichment analysis toolkit.
        Nucleic Acids Res. 2017; 45: W130-W137
        • Su G.
        • Morris J.H.
        • Demchak B.
        • Bader G.D.
        Biological network exploration with Cytoscape 3.
        Curr Protoc Bioinformatics. 2014; 47: 8.13.1-8.13.24
        • Mullen R.J.
        • Buck C.R.
        • Smith A.M.
        NeuN, a neuronal specific nuclear protein in vertebrates.
        Development. 1992; 116: 201-211
        • Arlotta P.
        • Molyneaux B.J.
        • Chen J.
        • Inoue J.
        • Kominami R.
        • Macklis J.D.
        Neuronal subtype-specific genes that control corticospinal motor neuron development in vivo.
        Neuron. 2005; 45: 207-221
        • Lai K.O.
        • Ip N.Y.
        Structural plasticity of dendritic spines: The underlying mechanisms and its dysregulation in brain disorders.
        Biochim Biophys Acta. 2013; 1832: 2257-2263
        • Lee A.T.
        • Gee S.M.
        • Vogt D.
        • Patel T.
        • Rubenstein J.L.
        • Sohal V.S.
        Pyramidal neurons in prefrontal cortex receive subtype-specific forms of excitation and inhibition.
        Neuron. 2014; 81: 61-68
        • Sohur U.S.
        • Padmanabhan H.K.
        • Kotchetkov I.S.
        • Menezes J.R.
        • Macklis J.D.
        Anatomic and molecular development of corticostriatal projection neurons in mice.
        Cereb Cortex. 2014; 24: 293-303
        • Bayes A.
        • Collins M.O.
        • Croning M.D.
        • van de Lagemaat L.N.
        • Choudhary J.S.
        • Grant S.G.
        Comparative study of human and mouse postsynaptic proteomes finds high compositional conservation and abundance differences for key synaptic proteins.
        PLoS One. 2012; 7: e46683
        • Li J.
        • Zhang W.
        • Yang H.
        • Howrigan D.P.
        • Wilkinson B.
        • Souaiaia T.
        • et al.
        Spatiotemporal profile of postsynaptic interactomes integrates components of complex brain disorders.
        Nat Neurosci. 2017; 20: 1150-1161
        • Weingarten J.
        • Lassek M.
        • Mueller B.F.
        • Rohmer M.
        • Lunger I.
        • Baeumlisberger D.
        • et al.
        The proteome of the presynaptic active zone from mouse brain.
        Mol Cell Neurosci. 2014; 59: 106-118
        • Zhang W.
        • Zhang Y.
        • Zheng H.
        • Zhang C.
        • Xiong W.
        • Olyarchuk J.G.
        • et al.
        SynDB: A Synapse protein DataBase based on synapse ontology.
        Nucleic Acids Res. 2007; 35: D737-D741
        • Bourgeron T.
        From the genetic architecture to synaptic plasticity in autism spectrum disorder.
        Nat Rev Neurosci. 2015; 16: 551-563
        • Maloney S.E.
        • Rieger M.A.
        • Dougherty J.D.
        Identifying essential cell types and circuits in autism spectrum disorders.
        Int Rev Neurobiol. 2013; 113: 61-96
        • Sekino Y.
        • Kojima N.
        • Shirao T.
        Role of actin cytoskeleton in dendritic spine morphogenesis.
        Neurochem Int. 2007; 51: 92-104
        • Fino E.
        • Paille V.
        • Cui Y.
        • Morera-Herreras T.
        • Deniau J.M.
        • Venance L.
        Distinct coincidence detectors govern the corticostriatal spike timing-dependent plasticity.
        J Physiol. 2010; 588: 3045-3062
        • Calabresi P.
        • Maj R.
        • Pisani A.
        • Mercuri N.B.
        • Bernardi G.
        Long-term synaptic depression in the striatum: Physiological and pharmacological characterization.
        J Neurosci. 1992; 12: 4224-4233
        • Calabresi P.
        • Pisani A.
        • Mercuri N.B.
        • Bernardi G.
        Long-term potentiation in the striatum is unmasked by removing the voltage-dependent magnesium block of Nmda receptor channels.
        Eur J Neurosci. 1992; 4: 929-935
        • Lovinger D.M.
        • Tyler E.C.
        • Merritt A.
        Short- and long-term synaptic depression in rat neostriatum.
        J Neurophysiol. 1993; 70: 1937-1949
        • Partridge J.G.
        • Tang K.C.
        • Lovinger D.M.
        Regional and postnatal heterogeneity of activity-dependent long-term changes in synaptic efficacy in the dorsal striatum.
        J Neurophysiol. 2000; 84: 1422-1429
        • Lovinger D.M.
        Neurotransmitter roles in synaptic modulation, plasticity and learning in the dorsal striatum.
        Neuropharmacology. 2010; 58: 951-961
        • Bagetta V.
        • Picconi B.
        • Marinucci S.
        • Sgobio C.
        • Pendolino V.
        • Ghiglieri V.
        • et al.
        Dopamine-dependent long-term depression is expressed in striatal spiny neurons of both direct and indirect pathways: Implications for Parkinson's disease.
        J Neurosci. 2011; 31: 12513-12522
        • Paille V.
        • Picconi B.
        • Bagetta V.
        • Ghiglieri V.
        • Sgobio C.
        • Di Filippo M.
        • et al.
        Distinct levels of dopamine denervation differentially alter striatal synaptic plasticity and NMDA receptor subunit composition.
        J Neurosci. 2010; 30: 14182-14193
        • Berretta N.
        • Nistico R.
        • Bernardi G.
        • Mercuri N.B.
        Synaptic plasticity in the basal ganglia: A similar code for physiological and pathological conditions.
        Prog Neurobiol. 2008; 84: 343-362
        • Wang X.
        • Kery R.
        • Xiong Q.
        Synaptopathology in autism spectrum disorders: Complex effects of synaptic genes on neural circuits.
        Prog Neuropsychopharmacol Biol Psychiatry. 2018; 84: 398-415
        • Casanova E.
        • Fehsenfeld S.
        • Mantamadiotis T.
        • Lemberger T.
        • Greiner E.
        • Stewart A.F.
        • Schutz G.
        A CamKIIalpha iCre BAC allows brain-specific gene inactivation.
        Genesis. 2001; 31: 37-42
        • Reig-Viader R.
        • Sindreu C.
        • Bayes A.
        Synaptic proteomics as a means to identify the molecular basis of mental illness: Are we getting there?.
        Prog Neuropsychopharmacol Biol Psychiatry. 2018; 84: 353-361
        • Lepeta K.
        • Lourenco M.V.
        • Schweitzer B.C.
        • Martino Adami P.V.
        • Banerjee P.
        • Catuara-Solarz S.
        • et al.
        Synaptopathies: Synaptic dysfunction in neurological disorders - A review from students to students.
        J Neurochem. 2016; 138: 785-805