Advertisement

Identification of Vulnerable Interneuron Subtypes in 15q13.3 Microdeletion Syndrome Using Single-Cell Transcriptomics

Open AccessPublished:September 23, 2021DOI:https://doi.org/10.1016/j.biopsych.2021.09.012

      Abstract

      Background

      A number of rare copy number variants (CNVs) have been linked to neurodevelopmental disorders. However, because CNVs encompass many genes, it is often difficult to identify the mechanisms that lead to developmental perturbations.

      Methods

      We used 15q13.3 microdeletion to propose and validate a novel strategy to predict the impact of CNV genes on brain development that could further guide functional studies. We analyzed single-cell transcriptomics datasets containing cortical interneurons to identify their developmental vulnerability to 15q13.3 microdeletion, which was validated in mouse models.

      Results

      We found that Klf13—but not other 15q13.3 genes—is expressed by precursors and neuroblasts in the medial and caudal ganglionic eminences during development, with a peak of expression at embryonic day (E)13.5 and E18.5, respectively. In contrast, in the adult mouse brain, Klf13 expression is negligible. Using Df(h15q13.3)/+ and Klf13+/− embryos, we observed a precursor subtype-specific impairment in proliferation in the medial ganglionic eminence and caudal ganglionic eminence at E13.5 and E17.5, respectively, corresponding to vulnerability predicted by Klf13 expression patterns. Finally, Klf13+/− mice showed a layer-specific decrease in parvalbumin and somatostatin cortical interneurons accompanied by changes in locomotor and anxiety-related behavior.

      Conclusions

      We show that the impact of 15q13.3 microdeletion on precursor proliferation is grounded in a reduction in Klf13 expression. The lack of Klf13 in Df(h15q13.3)/+ cortex might be the major reason for perturbed density of cortical interneurons. Thus, the behavioral defects seen in 15q13.3 microdeletion could stem from a developmental perturbation owing to selective vulnerability of cortical interneurons during sensitive stages of their development.

      Keywords

      Several rare copy number variants (CNVs) with large effect sizes are implicated in neurodevelopmental disorders (
      International Schizophrenia Consortium
      Rare chromosomal deletions and duplications increase risk of schizophrenia.
      ,
      • Stefansson H.
      • Rujescu D.
      • Cichon S.
      • Pietiläinen O.P.
      • Ingason A.
      • Steinberg S.
      • et al.
      Large recurrent microdeletions associated with schizophrenia.
      ,
      • Mosca S.J.
      • Langevin L.M.
      • Dewey D.
      • Innes A.M.
      • Lionel A.C.
      • Marshall C.C.
      • et al.
      Copy-number variations are enriched for neurodevelopmental genes in children with developmental coordination disorder.
      ,
      • Sebat J.
      Major changes in our DNA lead to major changes in our thinking.
      ,
      • Walsh T.
      • McClellan J.M.
      • McCarthy S.E.
      • Addington A.M.
      • Pierce S.B.
      • Cooper G.M.
      • et al.
      Rare structural variants disrupt multiple genes in neurodevelopmental pathways in schizophrenia.
      ,
      • Mefford H.C.
      • Sharp A.J.
      • Baker C.
      • Itsara A.
      • Jiang Z.
      • Buysse K.
      • et al.
      Recurrent rearrangements of chromosome 1q21.1 and variable pediatric phenotypes.
      ,
      • Helbig I.
      • Mefford H.C.
      • Sharp A.J.
      • Guipponi M.
      • Fichera M.
      • Franke A.
      • et al.
      15q13.3 microdeletions increase risk of idiopathic generalized epilepsy.
      ). These CNVs contain from few to tens of genes and are associated with pleiotropic effects leading to complex phenotype that includes autism, schizophrenia, epilepsy, and intellectual disability (
      International Schizophrenia Consortium
      Rare chromosomal deletions and duplications increase risk of schizophrenia.
      ,
      • Stefansson H.
      • Rujescu D.
      • Cichon S.
      • Pietiläinen O.P.
      • Ingason A.
      • Steinberg S.
      • et al.
      Large recurrent microdeletions associated with schizophrenia.
      ,
      • Mosca S.J.
      • Langevin L.M.
      • Dewey D.
      • Innes A.M.
      • Lionel A.C.
      • Marshall C.C.
      • et al.
      Copy-number variations are enriched for neurodevelopmental genes in children with developmental coordination disorder.
      ,
      • Sebat J.
      Major changes in our DNA lead to major changes in our thinking.
      ,
      • Walsh T.
      • McClellan J.M.
      • McCarthy S.E.
      • Addington A.M.
      • Pierce S.B.
      • Cooper G.M.
      • et al.
      Rare structural variants disrupt multiple genes in neurodevelopmental pathways in schizophrenia.
      ,
      • Mefford H.C.
      • Sharp A.J.
      • Baker C.
      • Itsara A.
      • Jiang Z.
      • Buysse K.
      • et al.
      Recurrent rearrangements of chromosome 1q21.1 and variable pediatric phenotypes.
      ,
      • Helbig I.
      • Mefford H.C.
      • Sharp A.J.
      • Guipponi M.
      • Fichera M.
      • Franke A.
      • et al.
      15q13.3 microdeletions increase risk of idiopathic generalized epilepsy.
      ). Each of these CNVs might have a specific mechanism contributing to developmental disturbance, which is difficult to discern in human context owing to limitations of in vitro approaches reproducing human brain development. One such high-risk variant, the 15q13.3 locus, is associated with wide-ranging clinical outcomes, including intellectual disability, autism spectrum disorder, attention-deficit/hyperactivity disorder, epilepsy, and schizophrenia (
      • Helbig I.
      • Mefford H.C.
      • Sharp A.J.
      • Guipponi M.
      • Fichera M.
      • Franke A.
      • et al.
      15q13.3 microdeletions increase risk of idiopathic generalized epilepsy.
      ,
      • Ben-Shachar S.
      • Lanpher B.
      • German J.R.
      • Qasaymeh M.
      • Potocki L.
      • Nagamani S.C.
      • et al.
      Microdeletion 15q13.3: A locus with incomplete penetrance for autism, mental retardation, and psychiatric disorders.
      ,
      • Dibbens L.M.
      • Mullen S.
      • Helbig I.
      • Mefford H.C.
      • Bayly M.A.
      • Bellows S.
      • et al.
      Familial and sporadic 15q13.3 microdeletions in idiopathic generalized epilepsy: Precedent for disorders with complex inheritance.
      ,
      • Miller D.T.
      • Shen Y.
      • Weiss L.A.
      • Korn J.
      • Anselm I.
      • Bridgemohan C.
      • et al.
      Microdeletion/duplication at 15q13.2q13.3 among individuals with features of autism and other neuropsychiatric disorders.
      ,
      • Van Bon B.W.M.
      • Mefford H.C.
      • Menten B.
      • Koolen D.A.
      • Sharp A.J.
      • Nillesen W.M.
      • et al.
      Further delineation of the 15q13 microdeletion and duplication syndromes: A clinical spectrum varying from non-pathogenic to a severe outcome.
      ). The 15q13.3 microdeletion encompasses seven protein-coding genes (CHRNA7, OTUD7A, KLF13, MTMR10, FAN1, TRPM1, and ARHGAP11B), a human-specific partial duplication of CHRNA7 (CHRFAM7A), a microRNA (miR211), and a putative pseudogene (LOC283710). The hemizygous microdeletion generally has high penetrance for neurodevelopmental disorders, with >80% of total cases diagnosed with at least one neuropsychiatric abnormality (
      • Lowther C.
      • Costain G.
      • Stavropoulos D.J.
      • Melvin R.
      • Silversides C.K.
      • Andrade D.M.
      • et al.
      Delineating the 15q13.3 microdeletion phenotype: A case series and comprehensive review of the literature.
      ). Homozygous 15q13.3 microdeletions are extremely rare and characterized by severe neuropsychiatric phenotypes, suggesting a gene-dosage dependency (
      • Lowther C.
      • Costain G.
      • Stavropoulos D.J.
      • Melvin R.
      • Silversides C.K.
      • Andrade D.M.
      • et al.
      Delineating the 15q13.3 microdeletion phenotype: A case series and comprehensive review of the literature.
      ). However, the etiology underlying the 15q13.3 microdeletion syndrome remains elusive, and exploring mechanisms of 15q13.3 microdeletion–derived developmental impairment should provide insights into the common biology of neurodevelopmental disorders.
      Existing mouse models of 15q13.3 microdeletion show high construct validity (
      • Fejgin K.
      • Nielsen J.
      • Birknow M.R.
      • Bastlund J.F.
      • Nielsen V.
      • Lauridsen J.B.
      • et al.
      A mouse model that recapitulates cardinal features of the 15q13.3 microdeletion syndrome including schizophrenia- and epilepsy-related alterations.
      ,
      • Kogan J.H.
      • Gross A.K.
      • Featherstone R.E.
      • Shin R.
      • Chen Q.
      • Heusner C.L.
      • et al.
      Mouse model of chromosome 15q13.3 microdeletion syndrome demonstrates features related to autism spectrum disorder.
      ). However, only one of them, the Df(h15q13)/+, also shows an increased sensitivity to stress, long-term spatial memory deficits, sensorimotor gating deficits, and high susceptibility to developing myoclonic seizures (
      • Fejgin K.
      • Nielsen J.
      • Birknow M.R.
      • Bastlund J.F.
      • Nielsen V.
      • Lauridsen J.B.
      • et al.
      A mouse model that recapitulates cardinal features of the 15q13.3 microdeletion syndrome including schizophrenia- and epilepsy-related alterations.
      ), thus mirroring key clinical observations seen in patients with the 15q13.3 microdeletion. At the cellular level, Df(h15q13)/+ mice display aberrant brain connectivity, reduced firing of interneurons, and lower sensitivity of pyramidal cells to GABAAR (gamma-aminobutyric acid A receptor) antagonism in the prefrontal cortex (
      • Nilsson S.R.O.
      • Celada P.
      • Fejgin K.
      • Thelin J.
      • Nielsen J.
      • Santana N.
      • et al.
      A mouse model of the 15q13.3 microdeletion syndrome shows prefrontal neurophysiological dysfunctions and attentional impairment.
      ,
      • Thelin J.
      • Halje P.
      • Nielsen J.
      • Didriksen M.
      • Petersson P.
      • Bastlund J.F.
      The translationally relevant mouse model of the 15q13.3 microdeletion syndrome reveals deficits in neuronal spike firing matching clinical neurophysiological biomarkers seen in schizophrenia.
      ), implying that microdeletion causes inhibitory dysfunction. Such a phenotype supports accumulating evidence that excitation-inhibition imbalance in the cortex is a common feature of neurodevelopmental disorders (
      • Foss-Feig J.H.
      • Adkinson B.D.
      • Ji J.L.
      • Yang G.
      • Srihari V.H.
      • McPartland J.C.
      • et al.
      Searching for cross-diagnostic convergence: Neural mechanisms governing excitation and inhibition balance in schizophrenia and autism spectrum disorders.
      ,
      • Marín O.
      Interneuron dysfunction in psychiatric disorders.
      ,
      • Lewis D.A.
      • Hashimoto T.
      • Volk D.W.
      Cortical inhibitory neurons and schizophrenia.
      ,
      • Sohal V.S.
      • Rubenstein J.L.R.
      Excitation-inhibition balance as a framework for investigating mechanisms in neuropsychiatric disorders.
      ). This imbalance has often been traced to reduced interneuron function in animal models (
      • Vasistha N.A.
      • Pardo-Navarro M.
      • Gasthaus J.
      • Weijers D.
      • Müller M.K.
      • García-González D.
      • et al.
      Maternal inflammation has a profound effect on cortical interneuron development in a stage and subtype-specific manner.
      ,
      • Chattopadhyaya B.
      • Cristo G.D.
      GABAergic circuit dysfunctions in neurodevelopmental disorders.
      ,
      • Fazzari P.
      • Paternain A.V.
      • Valiente M.
      • Pla R.
      • Luján R.
      • Lloyd K.
      • et al.
      Control of cortical GABA circuitry development by Nrg1 and ErbB4 signalling.
      ,
      • Canetta S.
      • Bolkan S.
      • Padilla-Coreano N.
      • Song L.J.
      • Sahn R.
      • Harrison N.L.
      • et al.
      Maternal immune activation leads to selective functional deficits in offspring parvalbumin interneurons.
      ,
      • Povysheva N.V.
      • Zaitsev A.V.
      • Rotaru D.C.
      • Gonzalez-Burgos G.
      • Lewis D.A.
      • Krimer L.S.
      Parvalbumin-positive basket interneurons in monkey and rat prefrontal Cortex.
      ,
      • Gonzalez-Burgos G.
      • Cho R.Y.
      • Lewis D.A.
      Alterations in cortical network oscillations and parvalbumin neurons in schizophrenia.
      ,
      • Batista-Brito R.
      • Vinck M.
      • Ferguson K.A.
      • Chang J.T.
      • Laubender D.
      • Lur G.
      • et al.
      Developmental dysfunction of VIP interneurons impairs cortical circuits.
      ). Likewise, postmortem tissue from patients with schizophrenia and autism shows lower numbers of interneurons and decreased expression of GABA and GABA receptors (
      • Benes F.M.
      Evidence for neurodevelopment disturbances in anterior cingulate cortex of post-mortem schizophrenic brain.
      ,
      • Huang H.S.
      • Akbarian S.
      GAD1 mRNA expression and DNA methylation in prefrontal cortex of subjects with schizophrenia.
      ,
      • Mellios N.
      • Huang H.S.
      • Baker S.P.
      • Galdzicka M.
      • Ginns E.
      • Akbarian S.
      Molecular determinants of dysregulated GABAergic gene expression in the prefrontal cortex of subjects with schizophrenia.
      ,
      • Fatemi S.H.
      • Reutiman T.J.
      • Folsom T.D.
      • Thuras P.D.
      GABA(A) receptor downregulation in brains of subjects with autism.
      ,
      • Blatt G.J.
      • Fitzgerald C.M.
      • Guptill J.T.
      • Booker A.B.
      • Kemper T.L.
      • Bauman M.L.
      Density and distribution of hippocampal neurotransmitter receptors in autism: An autoradiographic study.
      ). Moreover, altered gamma oscillations observed in schizophrenia have been linked to deficits in inhibitory transmission (
      • Chen C.M.
      • Stanford A.D.
      • Mao X.
      • Abi-Dargham A.
      • Shungu D.C.
      • Lisanby S.H.
      • et al.
      GABA level, gamma oscillation, and working memory performance in schizophrenia.
      ,
      • Farzan F.
      • Barr M.S.
      • Levinson A.J.
      • Chen R.
      • Wong W.
      • Fitzgerald P.B.
      • Daskalakis Z.J.
      Reliability of long-interval cortical inhibition in healthy human subjects: A TMS-EEG study.
      ).
      The majority of cortical interneurons are generated embryonically in the ganglionic eminences (GEs) and migrate toward the cortex, where they functionally mature into a multitude of subtypes (
      • De Carlos J.A.
      • López-Mascaraque L.
      • Valverde F.
      Dynamics of cell migration from the lateral ganglionic eminence in the rat.
      ,
      • Anderson S.A.
      • Eisenstat D.D.
      • Shi L.
      • Rubenstein J.L.R.
      Interneuron migration from basal forebrain to neocortex: Dependence on Dlx genes.
      ,
      • Gelman D.M.
      • Marín O.
      Generation of interneuron diversity in the mouse cerebral cortex.
      ,
      • Wamsley B.
      • Fishell G.
      Genetic and activity-dependent mechanisms underlying interneuron diversity.
      ). Two main areas generating such interneuron diversity are the medial GE (MGE) and caudal GE (CGE); the MGE produces parvalbumin (PV) and somatostatin (SST) classes of interneurons, and the CGE produces a highly heterogeneous class of interneurons that express vasoactive intestinal peptide, neuropeptide Y, reelin, and several other rarer markers (
      • Nery S.
      • Fishell G.
      • Corbin J.G.
      The caudal ganglionic eminence is a source of distinct cortical and subcortical cell populations.
      ,
      • Butt S.J.
      • Fuccillo M.
      • Nery S.
      • Noctor S.
      • Kriegstein A.
      • Corbin J.G.
      • Fishell G.
      The temporal and spatial origins of cortical interneurons predict their physiological subtype.
      ,
      • Miyoshi G.
      • Butt S.J.
      • Takebayashi H.
      • Fishell G.
      Physiologically distinct temporal cohorts of cortical interneurons arise from telencephalic Olig2-expressing precursors.
      ,
      • Miyoshi G.
      • Hjerling-Leffler J.
      • Karayannis T.
      • Sousa V.H.
      • Butt S.J.
      • Battiste J.
      • et al.
      Genetic fate mapping reveals that the caudal ganglionic eminence produces a large and diverse population of superficial cortical interneurons.
      ). Numerous pathological genetic and environmental insults can affect interneuron development (
      • Vasistha N.A.
      • Khodosevich K.
      The impact of (ab)normal maternal environment on cortical development.
      ). Indeed, we recently showed that one of the most common environmental insults—maternal inflammation—affects interneuron development in a subtype- and developmental stage–dependent manner (
      • Vasistha N.A.
      • Pardo-Navarro M.
      • Gasthaus J.
      • Weijers D.
      • Müller M.K.
      • García-González D.
      • et al.
      Maternal inflammation has a profound effect on cortical interneuron development in a stage and subtype-specific manner.
      ), highlighting selective neuronal vulnerability to pathological insults. Thus, genetic insults such as 15q13.3 microdeletion are likely to also preferentially affect specific neuronal subtypes while sparing others during cortical development, which might underlie behavioral abnormalities in adults.
      Recent efforts in single-cell analysis have uncovered a staggering diversity of interneuron subtypes (
      • Mayer C.
      • Hafemeister C.
      • Bandler R.C.
      • Machold R.
      • Batista Brito R.
      • Jaglin X.
      • et al.
      Developmental diversification of cortical inhibitory interneurons.
      ,
      • Tasic B.
      • Yao Z.
      • Smith K.A.
      • Graybuck L.
      • Nguyen T.N.
      • Bertagnolli D.
      • et al.
      Shared and distinct transcriptomic cell types across neocortical areas.
      ,
      • Mi D.
      • Li Z.
      • Lim L.
      • Li M.
      • Moissidis M.
      • Yang Y.
      • et al.
      Early emergence of cortical interneuron diversity in the mouse embryo.
      ). Therefore, we implemented a strategy to first identify the expression of 15q13.3 locus genes in cortical interneurons in single-cell RNA sequencing (scRNA-seq) data during development through to adulthood (
      • Mayer C.
      • Hafemeister C.
      • Bandler R.C.
      • Machold R.
      • Batista Brito R.
      • Jaglin X.
      • et al.
      Developmental diversification of cortical inhibitory interneurons.
      ,
      • Tasic B.
      • Yao Z.
      • Smith K.A.
      • Graybuck L.
      • Nguyen T.N.
      • Bertagnolli D.
      • et al.
      Shared and distinct transcriptomic cell types across neocortical areas.
      ,
      • Loo L.
      • Simon J.M.
      • Xing L.
      • McCoy E.S.
      • Niehaus J.K.
      • Guo J.
      • et al.
      Single-cell transcriptomic analysis of mouse neocortical development.
      ). We then used enrichment of 15q13.3 locus genes to predict whether and when each gene is involved in development of cortical interneurons, where high expression of a gene during certain developmental period might identify vulnerable subtypes of interneurons and developmental stages. We showed that 15q13.3 locus genes were expressed in precursors of interneurons and immature interneurons, where Klf13 was the only gene with high expression level in precursor cells. Using Df(h15q13.3)/+ mice, we confirmed the impact of the microdeletion on interneuron precursor proliferation. Furthermore, by implementing Klf13 knockout mice, we demonstrated that heterozygous deletion of Klf13 alone had a similar effect on precursor proliferation as the whole 15q13.3 locus. Overall, our study proposes a new strategy to predict the impact of CNV genes on brain development and validates the prediction by gene knockout analyses.

      Methods and Materials

      Animal Experiments and Genotyping

      All animal experimental procedures were conducted in accordance with guidelines published by the National Animal Ethic Committee of Denmark and Danish legislation. C57BL/6J (Janvier Labs), Klf13−/− (
      • Darwich R.
      • Li W.
      • Yamak A.
      • Komati H.
      • Andelfinger G.
      • Sun K.
      • Nemer M.
      KLF13 is a genetic modifier of the Holt-Oram syndrome gene TBX5.
      ), and Df(h15q13.3)/+ (
      • Fejgin K.
      • Nielsen J.
      • Birknow M.R.
      • Bastlund J.F.
      • Nielsen V.
      • Lauridsen J.B.
      • et al.
      A mouse model that recapitulates cardinal features of the 15q13.3 microdeletion syndrome including schizophrenia- and epilepsy-related alterations.
      ) (Taconic Biosciences) mice used in this study were housed in individually ventilated cages with standard sawdust bedding and environmental enrichment in a 12-hour reversed light/dark cycle with ad libitum access to food and water. Only wild-type (WT) females were used for breeding to minimize the potential maternal effects of Df(h15q13.3)/+. The date of the plug was treated as embryonic day (E)0.5, and the embryos were staged accordingly. Details of genotyping and behavior are provided in the Supplement.

      BrdU Labeling, Immunohistochemistry, Image Acquisition, and Statistical Analysis

      Pregnant dams were injected intraperitoneally at E13.5, E15.5, and E17.5 with 50 mg/kg bromodeoxyuridine (B5002; Sigma-Aldrich), and embryos were collected 2 hours later and fixed in 4% paraformaldehyde overnight. Brains from postnatal mice were collected and processed for immunostaining as previously described (
      • Vasistha N.A.
      • Pardo-Navarro M.
      • Gasthaus J.
      • Weijers D.
      • Müller M.K.
      • García-González D.
      • et al.
      Maternal inflammation has a profound effect on cortical interneuron development in a stage and subtype-specific manner.
      ). Further details can be found in the Supplement.

      scRNA-Seq Data Analysis

      scRNA-seq datasets, codes, and data analysis are detailed in the Supplement.

      Results

      Expression of 15q13 Microdeletion Genes in GEs

      In 15q13.3 microdeletion syndrome, the initial trigger for any developmental abnormalities should arise from reduced expression of 15q13.3 genes. Thus, to identify when during development and in what subtypes of interneurons 15q13.3 microdeletion genes are expressed, we exploited several single-cell transcriptomics datasets characterizing developing and mature interneurons (
      • Mayer C.
      • Hafemeister C.
      • Bandler R.C.
      • Machold R.
      • Batista Brito R.
      • Jaglin X.
      • et al.
      Developmental diversification of cortical inhibitory interneurons.
      ,
      • Tasic B.
      • Yao Z.
      • Smith K.A.
      • Graybuck L.
      • Nguyen T.N.
      • Bertagnolli D.
      • et al.
      Shared and distinct transcriptomic cell types across neocortical areas.
      ,
      • Loo L.
      • Simon J.M.
      • Xing L.
      • McCoy E.S.
      • Niehaus J.K.
      • Guo J.
      • et al.
      Single-cell transcriptomic analysis of mouse neocortical development.
      ).
      To determine the developmental stage of expression for 15q13.3 microdeletion genes, we ordered transcriptomes of single cells at E13.5 from MGE and CGE along differentiation trajectories from neural stem cells to neuron precursors and immature postmitotic neuroblasts (NBs) using a recently developed means of scoring maturation (
      • Mayer C.
      • Hafemeister C.
      • Bandler R.C.
      • Machold R.
      • Batista Brito R.
      • Jaglin X.
      • et al.
      Developmental diversification of cortical inhibitory interneurons.
      ) of interneurons from precursors. As expected, markers of differentiation stage and genes associated with cell proliferation and cell cycle corresponded well with maturation score (Figure 1A). To reconstruct expression trajectories of 15q13.3 microdeletion genes during development, we aligned expression with the aforementioned differentiation markers (Figure 1B, C). Among 15q13.3 microdeletion locus genes, notable expression could be observed only for Klf13, which was confirmed by comparing mean log(counts) for 15q13.3 microdeletion and developmental stage marker genes (Figure 1D, E). Expression levels for Klf13 were higher in mitotic than postmitotic cells in the GEs (Figure 1D–F), which correlated with the majority of Klf13 gene counts detected in cells expressing neural stem cell and neuron precursor markers (Figure 1B, C). Interestingly, at E13.5/E14.5, while only few postmitotic cells expressed Klf13 in the MGE, larger fractions of postmitotic cells expressed Klf13 in the CGE (Figure 1F, G).
      Figure thumbnail gr1
      Figure 1Expression dynamics of 15q13.3 genes during early development in interneuron precursors of the ganglionic eminences. (A) Gene-expression dynamics of developmentally regulated genes in single cells from E13.5/14.5 plotted along maturation scores mirrors the trajectory of interneuron development. (B, C) Heatmaps displaying log-normalized expression values for the 15q13.3 genes along with three known developmentally regulated genes for comparison (Nes, Dlx1, Dcx) at E13.5 in the MGE (B) and E14.5 in the CGE (C) shows Klf13 to have the strongest expression of the 15q13.3 genes. (D, E) Mean log-expression values of all genes plotted against the number of cells expressing them for all mitotic cells (D) and all postmitotic cells (E) at E13.5/E14.5. Barring Klf13, the rest of the 15q13.3 genes show low expression (red circles), with values lower than that for developmentally regulated genes (blue circles) and at the lowest end of the continuum of all genes (gray circles). (F) Klf13 is expressed in mitotic cells (blue) in both MGE and CGE but shows a preference in postmitotic cells (orange) derived from the CGE. Log-normalized expression values are plotted against pseudo-time. Scales are the same for both mitotic and postmitotic cells. (G) Bar plot showing the mean log-normalized expression values for the 15q13.3 genes in mitotic and postmitotic populations from the MGE and CGE at E13.5/E14.5. CGE, caudal ganglionic eminence; E, embryonic day; MGE, medial ganglionic eminence; NB, neuroblast; NP, neuron precursor; NSC, neural stem cell.

      15q13.3 Microdeletion Has Developmental Stage–Specific Effects on MGE and CGE Precursor Proliferation

      Differences in expression levels of Klf13 between GEs and mitotic/postmitotic cells might suggest a differential impact on generation of MGE and CGE interneuron subtypes and selective vulnerability of GEs to 15q13.3 microdeletion. To test this hypothesis, we used Df(h15q13)/+ mice harboring heterozygous microdeletion of 15q13.3 locus (
      • Fejgin K.
      • Nielsen J.
      • Birknow M.R.
      • Bastlund J.F.
      • Nielsen V.
      • Lauridsen J.B.
      • et al.
      A mouse model that recapitulates cardinal features of the 15q13.3 microdeletion syndrome including schizophrenia- and epilepsy-related alterations.
      ) and investigated the effect of 15q13.3 microdeletion on MGE and CGE precursor proliferation at different stages of development. Here, we injected WT females, time-mated with Df(h15q13)/+ males, at E13.5, 15.5, or 17.5 with the proliferation marker BrdU and analyzed neuronal precursor proliferation 2 hours thereafter (Figure 2A). By colabeling for BrdU and MGE or CGE progenitor markers, Nkx2.1, and COUP-TFII, respectively, we found that 15q13.3 microdeletion affects each GE differentially. Specifically, in the MGE, proliferation was affected only at E13.5, whereas in the CGE, proliferation was affected only at E17.5. Proliferation was not affected at E15.5 in either GE (Figure 2B–G). Furthermore, while 15q13.3 microdeletion decreased proliferation in the MGE at E13.5, the microdeletion had the opposite effect on precursor proliferation in the CGE at E17.5. (Figure 2B, G). These data support the notion that different subtypes of interneurons have selective vulnerability to the genetic insult induced by 15q13.3 microdeletion. The effect of the microdeletion on proliferation of neuronal precursors could cause abnormal development of cortical circuitry.
      Figure thumbnail gr2
      Figure 2Developmental stage-dependent perturbation of interneuron progenitor proliferation due to the 15q13.3 microdeletion. (A) Schematic representation of the experimental procedure undertaken to assess interneuron progenitor proliferation during development. Pregnant dams were injected intraperitoneally at E13.5, E15.5, or E17.5 with BrdU (50 mg/kg). Embryos were collected and assessed 2 hours after BrdU injection. Central artwork was modified from (
      • Vasistha N.A.
      • Pardo-Navarro M.
      • Gasthaus J.
      • Weijers D.
      • Müller M.K.
      • García-González D.
      • et al.
      Maternal inflammation has a profound effect on cortical interneuron development in a stage and subtype-specific manner.
      ). (B, C) Quantification of Nkx2.1+ BrdU+ and COUP-TFII+ BrdU+ cells shows a reduction in precursor proliferation in the MGE (B) but not CGE (C) at E13.5 in the 15q13.3 microdeletion embryos (n = 8, 7, p < .001 and n = 6, 8, respectively). Upper panels are from wild-type, and lower ones are from 15q13.3 microdeletion–carrying embryos, respectively. (D, E) Precursor proliferation was not affected in the MGE (D) or CGE (E) at E15.5 (n = 5, 5 and n = 5, 7, respectively). (F, G) In contrast, proliferation of CGE (G) but not MGE (F) precursors was impaired at E17.5 in the 15q13.3 microdeletion embryos (n = 5, 6, p < .0001 and n = 7, 5, respectively), which correlates with the peak neurogenic window of CGE-derived progenitors. White and magenta arrowheads highlight double-positive (Nkx2.1+/BrdU+ or COUP-TFII+/BrdU+) and BrdU-negative (Nkx2.1+ or COUP-TFII+ only) cells, respectively. Comparison of means by unpaired Student’s t test in (B-G) (mean ± SEM are shown, ∗∗∗p < .001, ∗∗∗∗p < .0001); scale bars = 50 μm. BrdU, bromodeoxyuridine; E, embryonic day; WT, wild-type.

      Klf13 Heterozygous Mice Recapitulate Developmental Impairment of 15q13.3 Microdeletion

      As our findings suggested that the developmental impairment in 15q13.3 mice stemmed largely from the lack of Klf13, we sought to confirm whether the lack of Klf13 on its own affects interneuron production and reproduces the effect of 15q13.3 microdeletion on precursor cells. First, we confirmed that both MGE and CGE progenitors express KLF13 (Figure S1A, B). Then, we studied the proliferation of MGE and CGE precursors in a previously described Klf13 knockout mouse model (
      • Darwich R.
      • Li W.
      • Yamak A.
      • Komati H.
      • Andelfinger G.
      • Sun K.
      • Nemer M.
      KLF13 is a genetic modifier of the Holt-Oram syndrome gene TBX5.
      ). We found a reduced proliferation of MGE progenitors in Klf13 heterozygous embryos at E13.5 as evidenced by the proportion of Nkx2.1+ progenitors that were also labeled for BrdU (Figure 3A, B). In contrast, at E17.5, an increased proportion of COUPT-TFII+ CGE progenitors were labeled for BrdU in the Klf13+/− embryos (Figure 3C, D). Therefore, Klf13 heterozygous deletion shows similar effect on precursor proliferation as heterozygous 15q13.3 microdeletion.
      Figure thumbnail gr3
      Figure 3Klf13 mutant mice recapitulate proliferation deficits seen in 15q13.3 microdeletion, which result in a reduced number of interneurons. (A–D) Quantification of Nkx2.1+ BrdU+ MGE precursors shows reduced proliferation in Klf13+/− embryos at E13.5 (A, B) but an increase proliferation in COUP-TFII+ BrdU+ CGE precursors at E17.5 (C, D) (n = 7 each at E13.5 and 6 each at E17.5, p < .005 at both stages). (E–J) Somatosensory cortices of Klf13+/− pups at P15 show a reduced density of total PV+ but not SST+ interneurons (E–H) (n = 9 for wild-type, and n = 8 and Klf13+/−, p < .005). (I–J) Layer-wise analysis of PV and SST interneuron densities, however, showed a specific reduction in PV+ interneurons in layer 2/3 (I) and in SST+ interneurons in layer 1 (J) (n = 8 for each genotype, p < .0005 in (I), and p < .005 in (J). White and magenta arrowheads highlight double-positive (Nkx2.1+/BrdU+ or COUP-TFII+/BrdU+) and BrdU-negative (Nkx2.1+ or COUP-TFII+ only) cells, respectively. Comparison of means by unpaired Student’s t test in (B, D, G, H) and multiple unpaired t tests with Holm-Sidak correction in (I, J) (mean ± SEM are shown, ∗∗p < .005,∗∗∗p < .001); scale bars = 50 μm (A, C) and 10 μm (E, F). BrdU, bromodeoxyuridine; CGE, caudal ganglionic eminence; E, embryonic day; MGE, medial ganglionic eminence; n.s., not significant; PV, parvalbumin; SST, somatostatin; WT, wild-type.
      To explore the consequences of impaired precursor proliferation on postnatal distribution of cortical interneurons, we compared the proportion of PV and SST interneurons in the anterior cingulate (ACC) and primary somatosensory (S1) cortices of WT, Klf13+/− and Klf13−/− mice. As not all PV interneurons express detectable levels of PV in the ACC at postnatal day (P)15, we used COX6A2 that we have previously shown to be specific for PV+ interneurons with earlier onset of expression (
      • Sanz-Morello B.
      • Pfisterer U.
      • Winther Hansen N.
      • Demharter S.
      • Thakur A.
      • Fujii K.
      • et al.
      Complex IV subunit isoform COX6A2 protects fast-spiking interneurons from oxidative stress and supports their function.
      ) to identify putative PV interneurons in the ACC. The density of PV interneurons was decreased in both S1 and ACC (Figure 3E–I; Figures S2A–C and S3A–C). Similarly, SST interneurons showed a decreased density in S1 (Figure 3E–J; Figures S2A, B, D, F) and ACC (Figures S3A, B, D). Importantly, decrease in proliferation of interneuron precursors and subsequent reduction in the density of PV and SST interneurons correlated with behavioral impairment in Klf13+/− mice. Thus, in the open field test, Klf13+/− mice showed reduced locomotor activity and speed in both females and males (Figure 4A–E). In addition, male mice showed anxiety-related behavior observed as increased distance traveled in wall-near zone in the open field as compared with distance traveled in the inner zone of the open field (Figure 4F–I). This thigmotaxis behavior was less clear or absent in females (Figure 4F–I). The decrease in Klf13 expression was hence sufficient to recapitulate the developmental impairment seen in 15q13.3 microdeletion mice in this and other studies (
      • Steullet P.
      • Cabungcal J.H.
      • Coyle J.
      • Didriksen M.
      • Gill K.
      • Grace A.A.
      • et al.
      Oxidative stress-driven parvalbumin interneuron impairment as a common mechanism in models of schizophrenia.
      ,
      • Funk M.
      • Schuelert N.
      • Jaeger S.
      • Dorner-Ciossek C.
      • Rosenbrock H.
      • Mack V.
      Activation of group II metabotropic receptors attenuates cortical E-I imbalance in a 15q13.3 microdeletion mouse model.
      ,
      • Al-Absi A.-R.
      • Qvist P.
      • Glerup S.
      • Sanchez C.
      • Nyengaard J.R.
      Df(h15q13)/+ mouse model reveals loss of astrocytes and synaptic-related changes of the excitatory and inhibitory circuits in the medial prefrontal cortex.
      ). Moreover, the absence of Klf13 had a behavioral imprint related to cortical impairment.
      Figure thumbnail gr4
      Figure 4Klf13+/− mice display decreased locomotor and increased anxiety-like behavior in an open field test. (A) Tracking of movement over a 5-minute period in an open field chamber from representative WT and Klf13+/− male and female mice. (B, C) Decreased total distance traveled in open field in Klf13+/− as compared to WT mice across all mice (B) and when separated by sex (C) (n = 20 for WT, and n = 15 Klf13+/−, where 12 females and 8 males were WT, and 8 females and 7 males were Klf13+/−). (D, E) Decreased speed of movement in open field in Klf13+/− as compared to WT mice across all mice (D) and when separated by sex (E). (F–I) Klf13+/− mice show increased thigmotaxis behavior, suggesting increased anxiety compared to WT mice (F). This phenotype was only found in male Klf13+/− mice (G). Male Klf13+/− but not female Klf13+/− mice traveled greater distance in the outer zone and reciprocally lesser distance in the inner zone as compared to WT mice (H, I). Comparison of means by unpaired Student’s t test with Welch correction in (B, D, F) and multiple unpaired t tests with Holm-Sidak correction in (C, E, G–I) (mean ± SD is shown, ∗∗p < .005,∗∗∗p < .001). m, meters; n.s., not significant; WT, wild-type.

      The Effect of 15q13.3 Microdeletion on MGE and CGE Precursors Is Likely Driven by KLF13-Associated Signaling Network

      As both MGE and CGE express Klf13 gene, and expression levels of other 15q13.3 microdeletion genes are very low, selective vulnerability of MGE and CGE cells at E13.5 and E17.5, respectively, is likely to be mediated by proteins that are associated with KLF13. Thus, we used STRING protein-protein interaction network analysis (string-db.org) to identify proteins that are associated with KLF13. In addition to the proteins coded by the 15q13.3 locus, STRING analysis identified 11 other proteins that either interact with KLF13, are coexpressed, or are mentioned together in research literature (Figure 5A) (
      • Kaczynski J.
      • Zhang J.S.
      • Ellenrieder V.
      • Conley A.
      • Duenes T.
      • Kester H.
      • et al.
      The Sp1-like protein BTEB3 inhibits transcription via the basic transcription element box by interacting with mSin3A and HDAC-1 co-repressors and competing with Sp1.
      ,
      • Kislinger T.
      • Cox B.
      • Kannan A.
      • Chung C.
      • Hu P.
      • Ignatchenko A.
      • et al.
      Global survey of organ and organelle protein expression in mouse: Combined proteomic and transcriptomic profiling.
      ). We also included Fbxw7, Sp1, Crebbp, and Gsk3b, which were not identified by STRING analysis but have been shown to affect stability and expression of KLF13 (
      • Kim D.S.
      • Zhang W.
      • Millman S.E.
      • Hwang B.J.
      • Kwon S.J.
      • Clayberger C.
      • et al.
      Fbw7γ-mediated degradation of KLF13 prevents RANTES expression in resting human but not murine T lymphocytes.
      ,
      • Song C.Z.
      • Keller K.
      • Murata K.
      • Asano H.
      • Stamatoyannopoulos G.
      Functional interaction between coactivators CBP/p300, PCAF, and transcription factor FKLF2.
      ). To determine whether expression levels of Klf13-associated genes can explain the effect of 15q13.3 microdeletion on precursor proliferation at E17.5, we explored an scRNA-seq dataset generated at later stages of neurogenesis (P0) (
      • Loo L.
      • Simon J.M.
      • Xing L.
      • McCoy E.S.
      • Niehaus J.K.
      • Guo J.
      • et al.
      Single-cell transcriptomic analysis of mouse neocortical development.
      ). We first extracted cells that are annotated to belong to the GEs and then split those cells based on the MGE and CGE markers (Figure 5B, C).
      Figure thumbnail gr5
      Figure 5Analysis of Klf13 partners in proliferating interneuron precursors. (A) Summarized network of known and predicted protein-protein interactions of KLF13 analyzed from the STRING database (v11). The network nodes are proteins, while the edges represent predicted functional associations. The colored lines denote the nature of evidence used in predicting the associations. (B, C) UMAP plots of mitotic interneuron lineage cells identified from (
      • Loo L.
      • Simon J.M.
      • Xing L.
      • McCoy E.S.
      • Niehaus J.K.
      • Guo J.
      • et al.
      Single-cell transcriptomic analysis of mouse neocortical development.
      ) at P0 showing at least 5 distinct clusters (B) derived from both CGE and MGE and showing typical marker gene expression profiles (C). (D, E) Dot plots displaying the average log-normalized expression values of potential Klf13 partners in proliferating (mitotic) interneuron precursors (D) during early development (E13.5/E14.5) (D), and late development (P0, obtained from Dataset 3) (E). The size of the dots represents the percentage of cells expressing the genes of all proliferating cells at the respective developmental age. Mean log-normalized counts are indicated for selected interacting genes. CGE, caudal ganglionic eminence; E, embryonic day; MGE, medial ganglionic eminence; P, postnatal day.
      Although several Klf13-associated genes are expressed at equal levels in MGE and CGE precursors at E13.5/14.5, Ccnd1 and Gsk3b expression is higher in MGE than CGE precursors (Figure 5D). Thus, at P0, among Klf13-associated genes, Ccnd1 and Serpinh1 were preferentially expressed in the CGE in addition to being expressed in a higher percentage of cells (Figure 5E). Both CCND1 and GSK3B are crucial for the canonical Wnt signaling pathway and have been implicated in regulating neural precursor proliferation (
      • Glickstein S.B.
      • Alexander S.
      • Ross M.E.
      Differences in cyclin D2 and D1 protein expression distinguish forebrain progenitor subsets.
      ,
      • Lange C.
      • Huttner W.B.
      • Calegari F.
      Cdk4/CyclinD1 overexpression in neural stem cells shortens G1, delays neurogenesis, and promotes the generation and expansion of basal progenitors.
      ,
      • Kim W.Y.
      • Wang X.
      • Wu Y.
      • Doble B.W.
      • Patel S.
      • Woodgett J.R.
      • Snider W.D.
      GSK-3 is a master regulator of neural progenitor homeostasis.
      ). Hence, differential expression of Ccnd1 and Gsk3b between the GEs at E13.5/14.5 and P0 might underlie previously identified selective vulnerability of MGE and CGE precursors at E13.5 and E17.5 (Figure 2B, G).

      Klf13 Shows Dynamic Expression in Postmitotic Cortical Interneurons During Development

      Cortical interneurons are a very diverse class of neurons, and both MGE and CGE produce multiple neuronal subtypes. Recent scRNA-seq analysis identified approximately 60 transcriptomic subtypes of cortical interneurons (
      • Tasic B.
      • Yao Z.
      • Smith K.A.
      • Graybuck L.
      • Nguyen T.N.
      • Bertagnolli D.
      • et al.
      Shared and distinct transcriptomic cell types across neocortical areas.
      ). However, such diversity cannot be distinguished during brain development, and only large families of subtypes are identified in single-cell transcriptomics data during embryogenesis and early postnatal maturation (
      • Mayer C.
      • Hafemeister C.
      • Bandler R.C.
      • Machold R.
      • Batista Brito R.
      • Jaglin X.
      • et al.
      Developmental diversification of cortical inhibitory interneurons.
      ,
      • Mi D.
      • Li Z.
      • Lim L.
      • Li M.
      • Moissidis M.
      • Yang Y.
      • et al.
      Early emergence of cortical interneuron diversity in the mouse embryo.
      ). In the absence of robust markers distinguishing neuronal subtypes during development, we again exploited scRNA-seq datasets for developing interneurons (
      • Mayer C.
      • Hafemeister C.
      • Bandler R.C.
      • Machold R.
      • Batista Brito R.
      • Jaglin X.
      • et al.
      Developmental diversification of cortical inhibitory interneurons.
      ) to gain further insights into potential vulnerability of interneuron subtypes to 15q13.3 microdeletion. We extracted transcriptomes of postmitotic cells (the majority of which are NBs according to marker gene expression—Figure 1A, D, E) from the MGE and CGE at E13.5 and from cortical NBs and immature interneurons at E18.5 and P10, respectively, and analyzed the expression of 15q13.3 genes at each developmental period. We plotted normalized average gene expression counts for all postmitotic cells studied at E13.5, E18.5, and P10. Interestingly, Klf13 was again the only gene presenting notable expression levels up to P10 (Figure 6A–F). Furthermore, we show a robust detection of Klf13 counts in various subtypes of interneurons (Figure 6D–F). To predict vulnerability of different subtypes of interneurons to heterozygous deletion of Klf13, we compared Klf13 expression levels across subtypes for each developmental stage. To this end, at E13.5, Klf13 showed the highest expression in cardinal CGE-derived interneuron subtypes Vip and Id2, whereas expression in cardinal MGE-derived interneuron subtypes Pvalb and Sst was lower, particularly for Pvalb (Figure 6A). At E18.5, the trend for stronger expression of Klf13 in cardinal CGE-derived interneuron subtypes persisted (Figure 6B). However, whereas Vip interneurons continued expressing high levels of Klf13, the level of Klf13 expression in Id2 interneurons dropped, and by P10, expression of Klf13 was dramatically decreased across all CGE-derived subtypes.
      Figure thumbnail gr6
      Figure 6Transcriptomic expression profiles of 15q13.3 microdeletion genes in postmitotic interneuron precursors during various developmental stages. (AC) Matrix plots of log-normalized expression of 15q13.3 microdeletion genes in postmitotic interneuron subtypes at E13.5 (A), E18.5 (B), and P10 (C). Klf13 is expressed broadly across postmitotic interneuron subtypes at all 3 time points. Expression of Mtmr10 can be seen in Pvalb chandelier cells at E18 (B) and in Nos1 cells at P10 (C). Fan1 expression can be noticed in Sst non-Martinotti cells at P10 (C). The subtype-specific precursors were identified upon integration and alignment of developmental datasets to adult datasets containing mature interneuron subtypes. (DF) Dot plots visualizing both relative strength and percentage of expressing cells expressing 15q13.3 microdeletion genes in postmitotic interneuron subtypes at E13.5 (D), E18.5 (E), and P10 (F). CGE, caudal ganglionic eminence; E, embryonic day; MGE, medial ganglionic eminence; P, postnatal day.
      In general, cardinal MGE-derived subtypes of interneurons exhibited lower Klf13 expression than cardinal CGE-derived subtypes. Nevertheless, Klf13 expression was relatively high in postmitotic Sst interneurons at E13.5, soon after they become NBs, which then declined by P10. This also correlates with the decrease in MGE precursor proliferation in Df(h15q13)/+ mice at E13.5, but not later (Figure 2B, D, F). Thus, the effect of Klf13 deletion on Sst interneurons might be more pronounced in the middle than the late period of embryonic neurogenesis.
      We also found restricted expression of other 15q13.3 microdeletion genes in postmitotic interneurons during development. At E13.5, barring Klf13, expression of other 15q13.3 genes was very low. However, at E18.5, Mtmr10 showed specific expression in Pvalb chandelier cells, whereas at P10, some Fan1 expression could be detected in Sst non-Martinotti cells (Figure 6B, C). In addition, we observed low expression of Chrna7, Otud7a, and Mtmr10 in the Igfbp6 subtype at P10 (Figure 6C). Nevertheless, despite the low-scale expression of Fan1 and Mtmr10 during NB maturation, misexpression of Klf13 should contribute to the majority of embryonic impairments due to CNVs in the 15q13.3 locus.

      In the Mature Cortex, Expression of 15q13.3 Genes Is Strong in CGE-Derived Cortical Interneuron Subtypes

      Some mental disorders, for instance, schizophrenia, have an onset around adolescence and early adulthood. By this time, cortical circuit maturation is almost complete. To investigate the expression of 15q13.3 genes at the prodromal stages of the disorder, we used single-cell transcriptomics data from P56 mouse cortex (
      • Tasic B.
      • Yao Z.
      • Smith K.A.
      • Graybuck L.
      • Nguyen T.N.
      • Bertagnolli D.
      • et al.
      Shared and distinct transcriptomic cell types across neocortical areas.
      ). Expression levels were negligible for Mtmr10, Fan1, and Trpm1 across the delineated subtypes at this stage. Interestingly, in contrast to broad expression at embryonic and neonatal stages, in the mature cortex, Klf13 expression was restricted to a few specific subtypes of CGE-derived Vip and Sncg interneurons with low to moderate expression levels (Figure 7). All MGE-derived subtypes showed very low expression of Klf13. On the other hand, Otud7a was now robustly and broadly expressed in many interneuron subtypes, being highest in Vip subtypes (especially in Vip_Ptprt_Pkp2, Vip_Rspo4_Rxfp1_Chat, and Vip_Rspo1_Itga4). Importantly, the expression of Chrna7, which was also very low at embryonic and early postnatal stages, was now expressed at high levels in specific subtypes of Lamp5 interneurons (Lamp5_Fam19a1_Pax6, Lamp5_Krt73, Lamp5_Fam19a1_Tmem182), of which Lamp5_Fam19a1_Tmem182 was predicted to correspond to single bouquet cells (
      • Tasic B.
      • Yao Z.
      • Smith K.A.
      • Graybuck L.
      • Nguyen T.N.
      • Bertagnolli D.
      • et al.
      Shared and distinct transcriptomic cell types across neocortical areas.
      ). Similar to other 15q13.3 microdeletion genes, Chrna7 expression was almost absent from Pvalb and Sst interneuron subtypes. Together, this suggests that CGE-derived Vip and Lamp5 subtypes and their associated microcircuits are vulnerable to 15q13.3 microdeletion in late adolescent stages.
      Figure thumbnail gr7
      Figure 7Transcriptomic expression profiles of 15q13.3 microdeletion genes in mature interneuron subtypes in adult mouse cortex. Matrix plot of log-normalized expression of 15q13.3 genes in the major MGE- and CGE-derived interneuron subtypes at P56 in mouse visual and motor cortex. CGE, caudal ganglionic eminence; MGE, medial ganglionic eminence; P, postnatal day.

      Dynamic Expression of 15q13.3 Microdeletion Genes During Interneuron Development

      To visualize the expression dynamics of 15q13.3 microdeletion genes across interneuron subtypes and developmental stages, we converged the high-resolution interneuron subtypes into cardinal identities (Pvalb, Sst, Vip, Id2, and Nos1) and compared the expression values of the six 15q13.3 microdeletion genes during development.
      Klf13 expression in Sst and Nos1 subtypes remained nearly constant throughout the developmental stages. In contrast, Klf13 expression was higher in Vip and Id2 clusters at E13, peaking at E18 before being downregulated in both subtypes by P10 (Figure 8A). Expression in the Pvalb subtype followed a similar pattern to Sst but showed the lowest expression among all subtypes by P10. By P56, however, Klf13 expression was downregulated in all subtypes. Interestingly, Klf13 expression in basket and chandelier Pvalb cells followed reciprocal trajectories. While Klf13 expression in chandelier cells peaked at E18 followed by decline to P10, in basket cells, Klf13 expression dropped at E18 and peaked at P10 (Figure 8B). Expression of three genes—Mtmr10, Fan1, and Trpm1—was almost negligible across the developmental stages and mature interneurons, and for two other genes—Chrna7 and Otud7a—significant expression could be detected only at P56 (Figure 8C–G). Accordingly, Vip and Id2 subtypes are likely to be affected by reduced expression of Chrna7 due to hemizygous 15q13.3 microdeletion, while the decrease in Otud7a expression should have a rather broad effect on mature interneuron subtypes (Figure 8C, E).
      Figure thumbnail gr8
      Figure 8Overview of the dynamic expression 15q13.3 genes across major interneuron subtypes throughout cortical development. (A–G) Scatter plots with average log-normalized expression values vs. developmental stage in five cardinal interneuron classes (Pvalb-, Sst-, Vip-, Id2-, and Nos1-expressing interneurons) for Klf13 (A), Chrna7 (C), Mtmr10 (D), Otud7a (E), Fan1 (F), and Trpm1 (G). Klf13 exhibits highest expression in Vip interneuron subtypes at E18 (A) and a contrasting expression profile in the two subclasses of Pvalb-expressing interneurons (basket and chandelier cells) at E18 and P10 during development (B). Data points at P56 have been marked separately as the adult dataset represents non-UMI read counts, while the developmental datasets represent UMI-based counts. Avg, average; E, embryonic day; P, postnatal day; UMI, unique molecular identifier.

      Discussion

      The mammalian brain is composed of diverse cell types that contribute to behavior in different ways. Transcriptomics studies involving whole-tissue RNA-seq, while useful, are restricted in associating specific cell types with disease, and we still lack knowledge to associate complex behavioral outcomes such as schizophrenia with dysfunctions of specific neuronal subtypes. The development of single-cell transcriptomics approaches offers a novel means of studying the contribution of individual neuronal subtypes to abnormalities in neuronal circuitry underlying mental disorders. Previous efforts in this direction sought common variants from genome-wide association studies that were enriched in transcriptomes of classes of neural cells in the brain (
      • Skene N.G.
      • Bryois J.
      • Bakken T.E.
      • Breen G.
      • Crowley J.J.
      • Gaspar H.A.
      • et al.
      Genetic identification of brain cell types underlying schizophrenia.
      ,
      • Watanabe K.
      • Umićević Mirkov M.
      • de Leeuw C.A.
      • van den Heuvel M.P.
      • Posthuma D.
      Genetic mapping of cell type specificity for complex traits.
      ). Importantly, these studies identified convergent cell types that potentially underlie symptoms of psychiatric disorders. It is expected that studying genes residing in rare CNVs will have greater power in identifying neuronal subtypes contributing to psychiatric disorders because the association of CNVs to mental illnesses is stronger than for common variants. Thus, in our study, we took a further step in resolution and analyzed enrichment of 15q13.3 CNV genes in individual interneuron subtypes throughout brain development and validated our results in mouse models.
      The etiology of rare variants contributing to psychiatric disorders, such as 15q13.3 microdeletion syndrome, remains elusive, and no genes driving pathogenesis have been clearly identified. CHRNA7 and OTUD7A were previously proposed as candidate driver genes for the disorder as most cases show overlapping deletions for these two genes. However, CHRNA7 exhibits haploinsufficiency in humans, and Chrna7−/− mice do not consistently replicate the phenotype of Df(h15q13.3)/+ mice (
      • Freund R.K.
      • Graw S.
      • Choo K.S.
      • Stevens K.E.
      • Leonard S.
      • Dell’Acqua M.L.
      Genetic knockout of the α7 nicotinic acetylcholine receptor gene alters hippocampal long-term potentiation in a background strain-dependent manner.
      ,
      • Yin J.
      • Chen W.
      • Yang H.
      • Xue M.
      • Schaaf C.P.
      Chrna7 deficient mice manifest no consistent neuropsychiatric and behavioral phenotypes.
      ), whereas Otud7a plays an important role only during postnatal maturation (
      • Uddin M.
      • Unda B.K.
      • Kwan V.
      • Holzapfel N.T.
      • White S.H.
      • Chalil L.
      • et al.
      OTUD7A regulates neurodevelopmental phenotypes in the 15q13.3 microdeletion syndrome.
      ). Here, we show that the expression of Chrna7 and Otud7a is negligible in interneurons and their precursors during embryonic period when precursor proliferation in affected by 15q13.3 microdeletion. Hence, we suggest that while important for cortical maturation, neither gene is sufficient to explain the developmental impairments seen in 15q13.3 microdeletion syndrome.
      Instead, our analysis points toward a major role for Klf13, which might underlie the effect of 15q13.3 microdeletion on interneuron development. Here, Klf13 is expressed both by a subset of dividing interneuron progenitors and by postmitotic neuroblasts in the GEs. Importantly, while MGE-derived interneuron subtypes express higher levels of Klf13 at E13.5 followed by declining of Klf13 expression, high expression of Klf13 in the CGE persists to E17.5. Importantly, using tissue from Df(h15q13.3)/+ and Klf13+/− embryonic mice (
      • Darwich R.
      • Li W.
      • Yamak A.
      • Komati H.
      • Andelfinger G.
      • Sun K.
      • Nemer M.
      KLF13 is a genetic modifier of the Holt-Oram syndrome gene TBX5.
      ), we show that the microdeletion affects proliferation in both MGE and CGE precursors but at different developmental stages—namely, MGE precursor proliferation is affected at E13.5 and that of CGE at E17.5. This emphasizes the role of Klf13 in embryonic neurodevelopment and supports our findings that Klf13 gene explains the impairments seen in mice carrying the 15q13.3 microdeletion. Furthermore, the difference in impact on MGE and CGE precursors in both mice is likely due to the differential expression of the Klf13-associated gene network, where Wnt signaling components Ccnd1 and Gsk3b might play the main role. By studying the expression of 15q13.3 genes in mitotic and postmitotic interneuron subtypes, we propose a significant effect on CGE-derived Vip and Id2 expressing interneurons starting from embryonic stages and persisting through early adulthood. Both subtypes have been described to play a role in cortical microcircuits and to mediate complex behaviors in mouse models (
      • Luongo F.J.
      • Horn M.E.
      • Sohal V.S.
      Putative microcircuit-level substrates for attention are disrupted in mouse models of autism.
      ,
      • Khoshkhoo S.
      • Vogt D.
      • Sohal V.S.
      Dynamic, cell-type-specific roles for GABAergic interneurons in a mouse model of optogenetically inducible seizures.
      ,
      • Ayzenshtat I.
      • Karnani M.M.
      • Jackson J.
      • Yuste R.
      Cortical control of spatial resolution by VIP + interneurons.
      ,
      • Lee A.T.
      • Cunniff M.M.
      • See J.Z.
      • Wilke S.A.
      • Luongo F.J.
      • Ellwood I.T.
      • et al.
      VIP interneurons contribute to avoidance behavior by regulating information flow across hippocampal-prefrontal networks.
      ). Moreover, the functional impairment of Vip interneuron subtype has been shown to cause deficits in cortical circuits (
      • Batista-Brito R.
      • Vinck M.
      • Ferguson K.A.
      • Chang J.T.
      • Laubender D.
      • Lur G.
      • et al.
      Developmental dysfunction of VIP interneurons impairs cortical circuits.
      ,
      • Khoshkhoo S.
      • Vogt D.
      • Sohal V.S.
      Dynamic, cell-type-specific roles for GABAergic interneurons in a mouse model of optogenetically inducible seizures.
      ,
      • Stanco A.
      • Pla R.
      • Vogt D.
      • Chen Y.
      • Mandal S.
      • Walker J.
      • et al.
      NPAS1 represses the generation of specific subtypes of cortical interneurons.
      ,
      • Goff K.M.
      • Goldberg E.M.
      Vasoactive intestinal peptide-expressing interneurons are impaired in a mouse model of Dravet syndrome.
      ). Finally, the differential impact on MGE and CGE progenitor proliferation in Df(h15q13.3)/+ embryos bears similarity with our observations in a model of maternal immune activation (
      • Vasistha N.A.
      • Pardo-Navarro M.
      • Gasthaus J.
      • Weijers D.
      • Müller M.K.
      • García-González D.
      • et al.
      Maternal inflammation has a profound effect on cortical interneuron development in a stage and subtype-specific manner.
      ). In both maternal immune activation and Df(h15q13.3)/+ mice, we show a differential impact on interneuron progenitors that depends on the developmental stage, i.e., MGE and CGE precursors show impaired proliferation, with MGE proliferation being affected during early neurogenesis and CGE proliferation at late neurogenesis, respectively. This translates into an impact on specific interneuron subtypes in the postnatal brain, suggesting a common etiology for these distinct models replicating genetic (15q13.3) or environmental (maternal inflammation) insults. This differential impact on interneuron subtypes was also seen in Klf13 knockout mice, where a reduction in SST+ and PV+ interneurons could be observed in cingulate and somatosensory cortex. Our study hence identified Klf13 as being important for normal interneuron development. Furthermore, we show that absence of Klf13 results in marked behavioral impairment reminiscent of 15q13.3 microdeletion mice and other genetic models of neurodevelopmental disorders. Genome-wide association studies suggest KLF13 to be associated with schizophrenia and drug dose response in patients with schizophrenia (
      • Stephens S.H.
      • Franks A.
      • Berger R.
      • Palionyte M.
      • Fingerlin T.E.
      • Wagner B.
      • et al.
      Multiple genes in the 15q13-q14 chromosomal region are associated with schizophrenia.
      ,
      • Koga A.T.
      • Strauss J.
      • Zai C.
      • Remington G.
      • De Luca V.
      Genome-wide association analysis to predict optimal antipsychotic dosage in schizophrenia: A pilot study.
      ) and the KLF13 locus with significant DNA methylation differences in individuals with schizophrenia (
      • Montano C.
      • Taub M.A.
      • Jaffe A.
      • Briem E.
      • Feinberg J.I.
      • Trygvadottir R.
      • et al.
      Association of DNA methylation differences with schizophrenia in an epigenome-wide association study.
      ). This calls for further studies into the cellular and molecular role of KLF13 in neurodevelopment.
      The expression of Chrna7 and Otud7a in the adult but not developing neocortex points toward their role in circuit maturation rather than in the generation and migration of interneurons. Specifically, disturbed expression of these genes might impair Vip and Id2 subtypes at adult stages. In addition to the six protein-coding genes, the 15q13.3 locus also houses a microRNA, miR-211. Interestingly, a mouse model overexpressing miR-211 shows increased epileptiform activity, thereby indicating a crucial role for this microRNA (
      • Bekenstein U.
      • Mishra N.
      • Milikovsky D.Z.
      • Hanin G.
      • Zelig D.
      • Sheintuch L.
      • et al.
      Dynamic changes in murine forebrain miR-211 expression associate with cholinergic imbalances and epileptiform activity.
      ). However, selection for polyadenylated RNA molecules in the preparation of the single-cell transcriptomics datasets precluded us from studying miR-211 expression.
      Our strategy of using scRNA-seq datasets allows to predict the cell-type–specific effect of individual genes and CNVs on brain development and maturation. By implementing this strategy, we identified Klf13 as a novel regulator of interneuron development and major contributor to the developmental pathology of 15q13.3 CNV. Furthermore, our study shows that selective vulnerability of interneuron subtypes to 15q13.3 CNV during development is based on the reduction in expression of Klf13 gene, where Klf13-controlled set of genes might work as a “receiving set” for vulnerability to a genetic insult, similar to recently proposed mechanisms of environmental impact on cortical development (
      • Vasistha N.A.
      • Khodosevich K.
      The impact of (ab)normal maternal environment on cortical development.
      ). Such delineation of genes responsible for selective vulnerability is crucial for directing future efforts in studying the neurobiological basis of phenotypes observed in neurodevelopmental disorders.

      Acknowledgments and Disclosures

      This work was supported by the Novo Nordisk Foundation Hallas-Møller Investigator grant (Grant No. NNF16OC0019920 [to KK]), Lundbeck-NIH Brain Initiative grant (Grant No. 2017-2241 [to KK]), Lundbeck Ascending Investigator grant (Grant No. 2020-1025 [to KK]), and DFF-Forskningsprojekt1 (Grant No. 8020-00083B [to KK]). NAV is supported by a BRIDGE-Translational Excellence Programme fellowship funded by Novo Nordisk Foundation (Grant No. NNF18SA0034956). CB and OK are supported by the Novo Nordisk Foundation Laureate Program (Grant No. NNF15OC0014186) and DFF-Forskningsprojekt2. SM was supported by the SMP-Erasmus+ Traineeship Grant by the European Commission.
      We thank Dr. Yasuko Antoku and the BRIC microscopy core facility for assistance with microscopy, Viktor Petukhov for generously overseeing the computation, and members of the Khodosevich lab for constructive discussions relating to this study. We are grateful to Professor Mona Nemer (University of Ottawa, Canada) for generously providing the Klf13+/− mouse model and to Megan Fortier for help with breeding and shipment. We also thank the research labs associated to the data analyzed above for making their datasets and code publicly available.
      The authors report no biomedical financial interests or potential conflicts of interest.

      Supplementary Material

      References

        • International Schizophrenia Consortium
        Rare chromosomal deletions and duplications increase risk of schizophrenia.
        Nature. 2008; 455: 237-241
        • Stefansson H.
        • Rujescu D.
        • Cichon S.
        • Pietiläinen O.P.
        • Ingason A.
        • Steinberg S.
        • et al.
        Large recurrent microdeletions associated with schizophrenia.
        Nature. 2008; 455: 232-236
        • Mosca S.J.
        • Langevin L.M.
        • Dewey D.
        • Innes A.M.
        • Lionel A.C.
        • Marshall C.C.
        • et al.
        Copy-number variations are enriched for neurodevelopmental genes in children with developmental coordination disorder.
        J Med Genet. 2016; 53: 812-819
        • Sebat J.
        Major changes in our DNA lead to major changes in our thinking.
        Nat Genet. 2007; 39: S3-S5
        • Walsh T.
        • McClellan J.M.
        • McCarthy S.E.
        • Addington A.M.
        • Pierce S.B.
        • Cooper G.M.
        • et al.
        Rare structural variants disrupt multiple genes in neurodevelopmental pathways in schizophrenia.
        Science. 2008; 320: 539-543
        • Mefford H.C.
        • Sharp A.J.
        • Baker C.
        • Itsara A.
        • Jiang Z.
        • Buysse K.
        • et al.
        Recurrent rearrangements of chromosome 1q21.1 and variable pediatric phenotypes.
        N Engl J Med. 2008; 359: 1685-1699
        • Helbig I.
        • Mefford H.C.
        • Sharp A.J.
        • Guipponi M.
        • Fichera M.
        • Franke A.
        • et al.
        15q13.3 microdeletions increase risk of idiopathic generalized epilepsy.
        Nat Genet. 2009; 41: 160-162
        • Ben-Shachar S.
        • Lanpher B.
        • German J.R.
        • Qasaymeh M.
        • Potocki L.
        • Nagamani S.C.
        • et al.
        Microdeletion 15q13.3: A locus with incomplete penetrance for autism, mental retardation, and psychiatric disorders.
        J Med Genet. 2009; 46: 382-388
        • Dibbens L.M.
        • Mullen S.
        • Helbig I.
        • Mefford H.C.
        • Bayly M.A.
        • Bellows S.
        • et al.
        Familial and sporadic 15q13.3 microdeletions in idiopathic generalized epilepsy: Precedent for disorders with complex inheritance.
        Hum Mol Genet. 2009; 18: 3626-3631
        • Miller D.T.
        • Shen Y.
        • Weiss L.A.
        • Korn J.
        • Anselm I.
        • Bridgemohan C.
        • et al.
        Microdeletion/duplication at 15q13.2q13.3 among individuals with features of autism and other neuropsychiatric disorders.
        J Med Genet. 2009; 46: 242-248
        • Van Bon B.W.M.
        • Mefford H.C.
        • Menten B.
        • Koolen D.A.
        • Sharp A.J.
        • Nillesen W.M.
        • et al.
        Further delineation of the 15q13 microdeletion and duplication syndromes: A clinical spectrum varying from non-pathogenic to a severe outcome.
        J Med Genet. 2009; 46: 511-523
        • Lowther C.
        • Costain G.
        • Stavropoulos D.J.
        • Melvin R.
        • Silversides C.K.
        • Andrade D.M.
        • et al.
        Delineating the 15q13.3 microdeletion phenotype: A case series and comprehensive review of the literature.
        Genet Med. 2015; 17: 149-157
        • Fejgin K.
        • Nielsen J.
        • Birknow M.R.
        • Bastlund J.F.
        • Nielsen V.
        • Lauridsen J.B.
        • et al.
        A mouse model that recapitulates cardinal features of the 15q13.3 microdeletion syndrome including schizophrenia- and epilepsy-related alterations.
        Biol Psychiatry. 2014; 76: 128-137
        • Kogan J.H.
        • Gross A.K.
        • Featherstone R.E.
        • Shin R.
        • Chen Q.
        • Heusner C.L.
        • et al.
        Mouse model of chromosome 15q13.3 microdeletion syndrome demonstrates features related to autism spectrum disorder.
        J Neurosci. 2015; 35: 16282-16294
        • Nilsson S.R.O.
        • Celada P.
        • Fejgin K.
        • Thelin J.
        • Nielsen J.
        • Santana N.
        • et al.
        A mouse model of the 15q13.3 microdeletion syndrome shows prefrontal neurophysiological dysfunctions and attentional impairment.
        Psychopharmacol (Berl). 2016; 233: 2151-2163
        • Thelin J.
        • Halje P.
        • Nielsen J.
        • Didriksen M.
        • Petersson P.
        • Bastlund J.F.
        The translationally relevant mouse model of the 15q13.3 microdeletion syndrome reveals deficits in neuronal spike firing matching clinical neurophysiological biomarkers seen in schizophrenia.
        Acta Physiol (Oxf). 2017; 220: 124-136
        • Foss-Feig J.H.
        • Adkinson B.D.
        • Ji J.L.
        • Yang G.
        • Srihari V.H.
        • McPartland J.C.
        • et al.
        Searching for cross-diagnostic convergence: Neural mechanisms governing excitation and inhibition balance in schizophrenia and autism spectrum disorders.
        Biol Psychiatry. 2017; 81: 848-861
        • Marín O.
        Interneuron dysfunction in psychiatric disorders.
        Nat Rev Neurosci. 2012; 13: 107-120
        • Lewis D.A.
        • Hashimoto T.
        • Volk D.W.
        Cortical inhibitory neurons and schizophrenia.
        Nat Rev Neurosci. 2005; 6: 312-324
        • Sohal V.S.
        • Rubenstein J.L.R.
        Excitation-inhibition balance as a framework for investigating mechanisms in neuropsychiatric disorders.
        Mol Psychiatry. 2019; 24: 1248-1257
        • Vasistha N.A.
        • Pardo-Navarro M.
        • Gasthaus J.
        • Weijers D.
        • Müller M.K.
        • García-González D.
        • et al.
        Maternal inflammation has a profound effect on cortical interneuron development in a stage and subtype-specific manner.
        Mol Psychiatry. 2020; 25: 2313-2329
        • Chattopadhyaya B.
        • Cristo G.D.
        GABAergic circuit dysfunctions in neurodevelopmental disorders.
        Front Psychiatry. 2012; 3: 51
        • Fazzari P.
        • Paternain A.V.
        • Valiente M.
        • Pla R.
        • Luján R.
        • Lloyd K.
        • et al.
        Control of cortical GABA circuitry development by Nrg1 and ErbB4 signalling.
        Nature. 2010; 464: 1376-1380
        • Canetta S.
        • Bolkan S.
        • Padilla-Coreano N.
        • Song L.J.
        • Sahn R.
        • Harrison N.L.
        • et al.
        Maternal immune activation leads to selective functional deficits in offspring parvalbumin interneurons.
        Mol Psychiatry. 2016; 21: 956-968
        • Povysheva N.V.
        • Zaitsev A.V.
        • Rotaru D.C.
        • Gonzalez-Burgos G.
        • Lewis D.A.
        • Krimer L.S.
        Parvalbumin-positive basket interneurons in monkey and rat prefrontal Cortex.
        —PubMed. J Neurophysiol. 2008; 100: 2348-2360
        • Gonzalez-Burgos G.
        • Cho R.Y.
        • Lewis D.A.
        Alterations in cortical network oscillations and parvalbumin neurons in schizophrenia.
        Biol Psychiatry. 2015; 77: 1031-1040
        • Batista-Brito R.
        • Vinck M.
        • Ferguson K.A.
        • Chang J.T.
        • Laubender D.
        • Lur G.
        • et al.
        Developmental dysfunction of VIP interneurons impairs cortical circuits.
        Neuron. 2017; 95: 884-895.e9
        • Benes F.M.
        Evidence for neurodevelopment disturbances in anterior cingulate cortex of post-mortem schizophrenic brain.
        Schizophr Res. 1991; 5: 187-188
        • Huang H.S.
        • Akbarian S.
        GAD1 mRNA expression and DNA methylation in prefrontal cortex of subjects with schizophrenia.
        PLOS ONE Imhof A, editor. 2007; 2e809
        • Mellios N.
        • Huang H.S.
        • Baker S.P.
        • Galdzicka M.
        • Ginns E.
        • Akbarian S.
        Molecular determinants of dysregulated GABAergic gene expression in the prefrontal cortex of subjects with schizophrenia.
        Biol Psychiatry. 2009; 65: 1006-1014
        • Fatemi S.H.
        • Reutiman T.J.
        • Folsom T.D.
        • Thuras P.D.
        GABA(A) receptor downregulation in brains of subjects with autism.
        J Autism Dev Disord. 2009; 39: 223-230
        • Blatt G.J.
        • Fitzgerald C.M.
        • Guptill J.T.
        • Booker A.B.
        • Kemper T.L.
        • Bauman M.L.
        Density and distribution of hippocampal neurotransmitter receptors in autism: An autoradiographic study.
        J Autism Dev Disord. 2001; 31: 537-543
        • Chen C.M.
        • Stanford A.D.
        • Mao X.
        • Abi-Dargham A.
        • Shungu D.C.
        • Lisanby S.H.
        • et al.
        GABA level, gamma oscillation, and working memory performance in schizophrenia.
        NeuroImage Clin. 2014; 4: 531-539
        • Farzan F.
        • Barr M.S.
        • Levinson A.J.
        • Chen R.
        • Wong W.
        • Fitzgerald P.B.
        • Daskalakis Z.J.
        Reliability of long-interval cortical inhibition in healthy human subjects: A TMS-EEG study.
        J Neurophysiol. 2010; 104: 1339-1346
        • De Carlos J.A.
        • López-Mascaraque L.
        • Valverde F.
        Dynamics of cell migration from the lateral ganglionic eminence in the rat.
        J Neurosci. 1996; 16: 6146-6156
        • Anderson S.A.
        • Eisenstat D.D.
        • Shi L.
        • Rubenstein J.L.R.
        Interneuron migration from basal forebrain to neocortex: Dependence on Dlx genes.
        Science. 1997; 278: 474-476
        • Gelman D.M.
        • Marín O.
        Generation of interneuron diversity in the mouse cerebral cortex.
        Eur J Neurosci. 2010; 31: 2136-2141
        • Wamsley B.
        • Fishell G.
        Genetic and activity-dependent mechanisms underlying interneuron diversity.
        Nat Rev Neurosci. 2017; 18: 299-309
        • Nery S.
        • Fishell G.
        • Corbin J.G.
        The caudal ganglionic eminence is a source of distinct cortical and subcortical cell populations.
        Nat Neurosci. 2002; 5: 1279-1287
        • Butt S.J.
        • Fuccillo M.
        • Nery S.
        • Noctor S.
        • Kriegstein A.
        • Corbin J.G.
        • Fishell G.
        The temporal and spatial origins of cortical interneurons predict their physiological subtype.
        Neuron. 2005; 48: 591-604
        • Miyoshi G.
        • Butt S.J.
        • Takebayashi H.
        • Fishell G.
        Physiologically distinct temporal cohorts of cortical interneurons arise from telencephalic Olig2-expressing precursors.
        J Neurosci. 2007; 27: 7786-7798
        • Miyoshi G.
        • Hjerling-Leffler J.
        • Karayannis T.
        • Sousa V.H.
        • Butt S.J.
        • Battiste J.
        • et al.
        Genetic fate mapping reveals that the caudal ganglionic eminence produces a large and diverse population of superficial cortical interneurons.
        J Neurosci. 2010; 30: 1582-1594
        • Vasistha N.A.
        • Khodosevich K.
        The impact of (ab)normal maternal environment on cortical development.
        Prog Neurobiol. 2021; 202: 102054
        • Mayer C.
        • Hafemeister C.
        • Bandler R.C.
        • Machold R.
        • Batista Brito R.
        • Jaglin X.
        • et al.
        Developmental diversification of cortical inhibitory interneurons.
        Nature. 2018; 555: 457-462
        • Tasic B.
        • Yao Z.
        • Smith K.A.
        • Graybuck L.
        • Nguyen T.N.
        • Bertagnolli D.
        • et al.
        Shared and distinct transcriptomic cell types across neocortical areas.
        Nature. 2018; 563: 72-78
        • Mi D.
        • Li Z.
        • Lim L.
        • Li M.
        • Moissidis M.
        • Yang Y.
        • et al.
        Early emergence of cortical interneuron diversity in the mouse embryo.
        Science. 2018; 360: 81-85
        • Loo L.
        • Simon J.M.
        • Xing L.
        • McCoy E.S.
        • Niehaus J.K.
        • Guo J.
        • et al.
        Single-cell transcriptomic analysis of mouse neocortical development.
        Nat Commun. 2019; 10: 134
        • Darwich R.
        • Li W.
        • Yamak A.
        • Komati H.
        • Andelfinger G.
        • Sun K.
        • Nemer M.
        KLF13 is a genetic modifier of the Holt-Oram syndrome gene TBX5.
        Hum Mol Genet. 2017; 26: 942-954
        • Sanz-Morello B.
        • Pfisterer U.
        • Winther Hansen N.
        • Demharter S.
        • Thakur A.
        • Fujii K.
        • et al.
        Complex IV subunit isoform COX6A2 protects fast-spiking interneurons from oxidative stress and supports their function.
        EMBO J. 2020; 39e105759
        • Steullet P.
        • Cabungcal J.H.
        • Coyle J.
        • Didriksen M.
        • Gill K.
        • Grace A.A.
        • et al.
        Oxidative stress-driven parvalbumin interneuron impairment as a common mechanism in models of schizophrenia.
        Mol Psychiatry. 2017; 22: 936-943
        • Funk M.
        • Schuelert N.
        • Jaeger S.
        • Dorner-Ciossek C.
        • Rosenbrock H.
        • Mack V.
        Activation of group II metabotropic receptors attenuates cortical E-I imbalance in a 15q13.3 microdeletion mouse model.
        bioRxiv. 2020; https://doi.org/10.1101/2020.09.17.301259
        • Al-Absi A.-R.
        • Qvist P.
        • Glerup S.
        • Sanchez C.
        • Nyengaard J.R.
        Df(h15q13)/+ mouse model reveals loss of astrocytes and synaptic-related changes of the excitatory and inhibitory circuits in the medial prefrontal cortex.
        Cereb Cortex. 2020; 31: 1-13
        • Kaczynski J.
        • Zhang J.S.
        • Ellenrieder V.
        • Conley A.
        • Duenes T.
        • Kester H.
        • et al.
        The Sp1-like protein BTEB3 inhibits transcription via the basic transcription element box by interacting with mSin3A and HDAC-1 co-repressors and competing with Sp1.
        J Biol Chem. 2001; 276: 36749-36756
        • Kislinger T.
        • Cox B.
        • Kannan A.
        • Chung C.
        • Hu P.
        • Ignatchenko A.
        • et al.
        Global survey of organ and organelle protein expression in mouse: Combined proteomic and transcriptomic profiling.
        Cell. 2006; 125: 173-186
        • Kim D.S.
        • Zhang W.
        • Millman S.E.
        • Hwang B.J.
        • Kwon S.J.
        • Clayberger C.
        • et al.
        Fbw7γ-mediated degradation of KLF13 prevents RANTES expression in resting human but not murine T lymphocytes.
        Blood. 2012; 120: 1658-1667
        • Song C.Z.
        • Keller K.
        • Murata K.
        • Asano H.
        • Stamatoyannopoulos G.
        Functional interaction between coactivators CBP/p300, PCAF, and transcription factor FKLF2.
        J Biol Chem. 2002; 277: 7029-7036
        • Glickstein S.B.
        • Alexander S.
        • Ross M.E.
        Differences in cyclin D2 and D1 protein expression distinguish forebrain progenitor subsets.
        Cereb Cortex. 2007; 17: 632-642
        • Lange C.
        • Huttner W.B.
        • Calegari F.
        Cdk4/CyclinD1 overexpression in neural stem cells shortens G1, delays neurogenesis, and promotes the generation and expansion of basal progenitors.
        Cell Stem Cell. 2009; 5: 320-331
        • Kim W.Y.
        • Wang X.
        • Wu Y.
        • Doble B.W.
        • Patel S.
        • Woodgett J.R.
        • Snider W.D.
        GSK-3 is a master regulator of neural progenitor homeostasis.
        Nat Neurosci. 2009; 12: 1390-1397
        • Skene N.G.
        • Bryois J.
        • Bakken T.E.
        • Breen G.
        • Crowley J.J.
        • Gaspar H.A.
        • et al.
        Genetic identification of brain cell types underlying schizophrenia.
        Nat Genet. 2018; 50: 825-833
        • Watanabe K.
        • Umićević Mirkov M.
        • de Leeuw C.A.
        • van den Heuvel M.P.
        • Posthuma D.
        Genetic mapping of cell type specificity for complex traits.
        Nat Commun. 2019; 10: 3222
        • Freund R.K.
        • Graw S.
        • Choo K.S.
        • Stevens K.E.
        • Leonard S.
        • Dell’Acqua M.L.
        Genetic knockout of the α7 nicotinic acetylcholine receptor gene alters hippocampal long-term potentiation in a background strain-dependent manner.
        Neurosci Lett. 2016; 627: 1-6
        • Yin J.
        • Chen W.
        • Yang H.
        • Xue M.
        • Schaaf C.P.
        Chrna7 deficient mice manifest no consistent neuropsychiatric and behavioral phenotypes.
        Sci Rep. 2017; 7: 39941
        • Uddin M.
        • Unda B.K.
        • Kwan V.
        • Holzapfel N.T.
        • White S.H.
        • Chalil L.
        • et al.
        OTUD7A regulates neurodevelopmental phenotypes in the 15q13.3 microdeletion syndrome.
        Am J Hum Genet. 2018; 102: 278-295
        • Luongo F.J.
        • Horn M.E.
        • Sohal V.S.
        Putative microcircuit-level substrates for attention are disrupted in mouse models of autism.
        Biol Psychiatry. 2016; 79: 667-675
        • Khoshkhoo S.
        • Vogt D.
        • Sohal V.S.
        Dynamic, cell-type-specific roles for GABAergic interneurons in a mouse model of optogenetically inducible seizures.
        Neuron. 2017; 93: 291-298
        • Ayzenshtat I.
        • Karnani M.M.
        • Jackson J.
        • Yuste R.
        Cortical control of spatial resolution by VIP + interneurons.
        J Neurosci. 2016; 36: 11498-11509
        • Lee A.T.
        • Cunniff M.M.
        • See J.Z.
        • Wilke S.A.
        • Luongo F.J.
        • Ellwood I.T.
        • et al.
        VIP interneurons contribute to avoidance behavior by regulating information flow across hippocampal-prefrontal networks.
        Neuron. 2019; 102: 1223-1234.e4
        • Stanco A.
        • Pla R.
        • Vogt D.
        • Chen Y.
        • Mandal S.
        • Walker J.
        • et al.
        NPAS1 represses the generation of specific subtypes of cortical interneurons.
        Neuron. 2014; 84: 940-953
        • Goff K.M.
        • Goldberg E.M.
        Vasoactive intestinal peptide-expressing interneurons are impaired in a mouse model of Dravet syndrome.
        eLife. 2019; 8
        • Stephens S.H.
        • Franks A.
        • Berger R.
        • Palionyte M.
        • Fingerlin T.E.
        • Wagner B.
        • et al.
        Multiple genes in the 15q13-q14 chromosomal region are associated with schizophrenia.
        Psychiatr Genet. 2012; 22: 1-14
        • Koga A.T.
        • Strauss J.
        • Zai C.
        • Remington G.
        • De Luca V.
        Genome-wide association analysis to predict optimal antipsychotic dosage in schizophrenia: A pilot study.
        J Neural Transm (Vienna). 2016; 123: 329-338
        • Montano C.
        • Taub M.A.
        • Jaffe A.
        • Briem E.
        • Feinberg J.I.
        • Trygvadottir R.
        • et al.
        Association of DNA methylation differences with schizophrenia in an epigenome-wide association study.
        JAMA Psychiatry. 2016; 73: 506-514
        • Bekenstein U.
        • Mishra N.
        • Milikovsky D.Z.
        • Hanin G.
        • Zelig D.
        • Sheintuch L.
        • et al.
        Dynamic changes in murine forebrain miR-211 expression associate with cholinergic imbalances and epileptiform activity.
        Proc Natl Acad Sci U S A. 2017; 114: E4996-E5005

      Linked Article