Article Figures & Data
Figures
- Figure 1.
Loss of TrkB signaling in Cort interneurons causes gene expression changes in pathways regulating excitability. A, Schematic of control and CortCre;TrkBflox/flox mice. B, Volcano plot of bulk homogenate RNA-seq results with CortCre;TrkBflox/flox versus Control log2 fold change against −log10 p value. Orange dots represent genes that are significantly different in CortCre;TrkBflox/flox versus Control, including Npy, Syt12, Nptx2, and Chrna4. Green dots represent nonsignificant genes. See Extended Data Figure 1-1. C, GO terms in the molecular function, biological processes, and cellular component categories for genes enriched and de-enriched in cortical tissue following ablation of TrkB in Cort neurons. See Extended Data Figure 1-2. D, qPCR analysis validating genes found to be differentially expressed in bulk cortical homogenate RNA-seq of CortCre;TrkBflox/flox versus Control mice (n = 5 per genotype, Student’s unpaired t test; data are presented as the mean ± SEM: *p < 0.05 ***p < 0.001 ****p < 0.0001 vs control).
- Figure 2.
Translating ribosome affinity purification defines a unique molecular signature for cortistatin interneurons in the cortex. A, Locus of the ribosomal protein Rpl22 in the RiboTag mouse and breeding strategy used to obtain CortCre;Rpl22HA mice. Schematic of RiboTag experimental workflow. B, Validation of RiboTag allele expression in cortistatin neurons by qPCR for Cort as well as Gad, Gfap, and Bdnf exon IV (n = 3 per genotype, Student’s unpaired t test; data are presented as the mean ± SEM: **p < 0.01 ****p < 0.0001 vs control). C, Volcano plot of RNA-seq results with CortCre;Rpl22HA Input versus CortCre;Rpl22HA IP log2 fold change against −log10 p value. Orange dots represent genes that are significantly different in Input versus IP fractions, including Syt2, Nxph1, Mal, and Apoe. Green dots represent nonsignificant genes. See Extended Data Figures 2-1 and 2-2. D, GO terms in the molecular function, biological processes, and cellular component categories for genes enriched and de-enriched in Cort-expressing interneurons. See Extended Data Figure 2-3. E, qPCR analysis validating genes found to be differentially expressed in Input versus IP RNA sequencing results (n = 3 per genotype, Student’s unpaired t test; data are presented as the mean ± SEM: ****p < 0.0001 vs control).
- Figure 3.
Loss of TrkB signaling in Cort neurons alters the expression of genes important for calcium homeostasis and axon development. A, Breeding strategy used to obtain control CortCre; Rpl22HA and experimental CortCre;TrkBflox/flox;Rpl22HA mice. B, Validation of Ribotag allele expression in cortistatin cells of control and CortCre;TrkBflox/flox;Rpl22HA mice by qPCR of Cort as well as Gad, Gfap, and Bdnf exon IV (n = 6 per genotype, Student’s unpaired t test; data are presented as the mean ± SEM: *p < 0.05 **p < 0.01, ***p < 0.001, ****p < 0.0001 vs control). C, Volcano plot of RNA-seq results with CortCre;TrkBflox/flox;Rpl22HA IP versus CortCre;Rpl22HA IP log2 fold change against −log10 p value. Orange dots represent genes that are significantly different in CortCre;TrkBflox/flox;Rpl22HA IP versus CortCre; Rpl22HA IP, including Wt1, Calb1, Lgals1, Trpc6, Syt6, and Gng4. Green dots represent nonsignificant genes. See Extended Data Figure 3-1. D, GO terms in the molecular function, biological processes, and cellular component categories for genes enriched and de-enriched in Cort neurons following removal of TrkB and disruption of BDNF–TrkB signaling. See Extended Data Figures 3-2, 3-3, and 3-4.
- Figure 4.
Validation of select targets from Control versus CortCre;TrkBflox/flox RNA-seq using qPCR and single-molecule fluorescence in situ hybridization. A, qPCR analysis validating select genes (Trpc6, Calb1, Lgals1, Wt1, Syt6, Gng4) found to be differentially expressed in CortCre;TrkBflox/flox IP versus control IP RNA-seq data (n = 6 per genotype, Student’s unpaired t test; data are presented as the mean ± SEM: **p < 0.01, ***p < 0.001, ****p < 0.0001 vs control). B, Quantification of Wt1 transcripts in Cre positive cells of CortCre;TrkBflox/flox and control mice. C, D, Confocal z-projections of Cre and Wt1 transcripts in the cortex from P21 Control (C) and CortCre;TrkBflox/flox (D) mice visualized with RNAscope in situ hybridization. Wt1 transcripts (green) are more enriched in Cort neurons of CortCre;TrkBflox/flox than of Control mice. Inset depicts higher magnification of nuclei highlighted by arrows. Scale bars, C, D, 10 μm.
Tables
Figure Data structure Type of test Type I error control Notes 1B Gene counts for differential expression analysis Moderated t tests with linear regression (empirical Bayes) FDR < 0.1 limma Bioconductor package: voom approach 1C Gene set enrichment analysis Hypergeometric test FDR < 0.05 clusterProfiler Bioconductor package: compareClusters approach 1D Normalized qPCR data Student's t test p < 0.05 2B Normalized qPCR data Student's t test p < 0.05 2C Gene counts for differential expression analysis Moderated t tests with linear mixed effects modeling (empirical Bayes) Bonferroni < 0.05 limma Bioconductor package: voom approach 2D Gene set enrichment analysis Hypergeometric test FDR < 0.05 clusterProfiler Bioconductor package: compareClusters approach 2E Normalized qPCR data Student's t test p < 0.05 3B Normalized qPCR data Student's t test p < 0.05 3C Differential expression analysis Moderated t tests with linear regression (empirical Bayes) FDR < 0.05 limma Bioconductor package: voom approach 3D GO enrichment analysis Hypergeometric test FDR < 0.05 clusterProfiler Bioconductor package: compareClusters approach 4A Normalized qPCR data Student's t test p < 0.05 4B Normally distributed Student's t test p < 0.05
Figure 1-1
Differential gene expression analysis of CortCre vs CortCre;TrkBflox/flox bulk cortex for all expressed genes. Columns represent Symbol (mouse gene symbol), logFC (log2 fold change comparing experimental to control animals; positive values indicate higher expression in experimental animals), t [moderated t statistic (with empirical Bayes)], P.Value (corresponding p value from t statistic), adj.P.Val (Benjamini–Hochberg-adjusted p value to control the FDR), B (log odds of differential expression signal), gene_type (gencode class of gene), EntrezID (Entrez Gene ID), AveExpr [Average expression on the log2(counts per million + 0.5) scale], Length (coding gene length), and ensemblID (Ensembl gene ID). Download Figure 1-1, CSV file.
Figure 1-2
Gene ontology analysis of differentially expressed genes between CortCre and CortCre;TrkB flox/flox bulk cortex. Columns represent Cluster (label for set of differentially expressed genes), ONTOLOGY (gene ontology type: CC, cell compartment; BP, biological process; MF, molecular function), ID (gene ontology ID), Description (gene ontology set description), GeneRatio (fraction of differentially expressed genes were in the GO set), BgRatio (fraction of differentially expressed genes that were not in the GO set), Pvalue (p value resulting from hypergeometric test), p.adjust [Benjamini–Hochberg-adjusted p value (FDR)], qvalue (Storey-adjusted p value), geneID [gene symbols corresponding to the differentially expressed genes in the GO set (i.e., from GeneRatio above)], and Count [number of differentially expressed genes in the GO set (numerator of GeneRatio, to avoid forced Excel conversion to dates from some fraction)]. Download Figure 1-2, CSV file.
Figure 2-1
Differential gene expression analysis of Cort Cre Input vs IP for all expressed genes. Columns represent Symbol (mouse gene symbol), logFC (log2 fold change comparing IP to Input samples; positive values indicate higher expression in IP samples), t [moderated t statistic (with empirical Bayes)], P.Value (corresponding p value from t statistic), adj.P.Val (Benjamini–Hochberg-adjusted p value to control the FDR), B (log odds of differential expression signal), gene_type (gencode class of gene), EntrezID (Entrez Gene ID), AveExpr [average expression on the log2(counts per million + 0.5) scale], Length (coding gene length), and ensemblID (Ensembl gene ID). Download Figure 2-1, CSV file.
Figure 2-2
CSEA of IP-enriched genes in Cort neurons. CSEA of IP-enriched genes identifies Cort interneurons. Bullseye plot of the output of CSEA reveals a substantial over-representation of Cort-positive neuron cell transcripts at multiple pSI levels among those transcripts (n = 100) found to be enriched in our IP samples from Cort neurons. Box highlights Cort-positive neurons. Download Figure 2-2, TIF file.
Figure 2-3
Gene ontology analysis of differentially expressed genes between CortCre Input vs CortCre IP. Columns represent Cluster (label for set of differentially expressed genes), ONTOLOGY (gene ontology type: CC, cell compartment; BP, biological process; MF, molecular function), ID (gene ontology ID), Description (gene ontology set description), GeneRatio (fraction of differentially expressed genes were in the GO set), BgRatio (fraction of differentially expressed genes that were not in the GO set), Pvalue (p value resulting from hypergeometric test), p.adjust [Benjamini–Hochberg-adjusted p value (FDR)], qvalue (Storey-adjusted p value), geneID [gene symbols corresponding to the differentially expressed genes in the GO set [i.e., from GeneRatio above)], Count (number of differentially expressed genes in the GO set (numerator of GeneRatio, to avoid forced Excel conversion to dates from some fraction)]. Download Figure 2-3, CSV file.
Figure 3-1
Differential gene expression analysis of CortCre IP vs CortCre;TrkBflox/flox IP for all expressed genes. Columns represent Symbol (mouse gene symbol), logFC (log2 fold change comparing experimental to control animals; positive values indicate higher expression in experimental samples), t [moderated t statistic [with empirical Bayes)], P.Value (corresponding p value from t statistic), adj.P.Val (Benjamini–Hochberg-adjusted p value to control the FDR), B (log odds of differential expression signal), gene_type (gencode class of gene), EntrezID (Entrez Gene ID), AveExpr [average expression on the log2(counts per million + 0.5) scale], Length (coding gene length), and ensemblID (Ensembl gene ID). Download Figure 3-1, CSV file.
Figure 3-2
Gene ontology analysis of differentially expressed genes between CortCre IP vs CortCre;TrkBflox/flox IP. Columns represent Direction (+1 is upregulated in experimental compared with control, −1 is downregulated in experimental compared with control), Cluster (label for set of differentially expressed genes), ONTOLOGY (gene ontology type: CC, cell compartment; BP, biological process, MF, molecular function), ID (gene ontology ID), Description (gene ontology set description), GeneRatio (fraction of differentially expressed genes were in the GO set), BgRatio (fraction of differentially expressed genes that were not in the GO set), Pvalue (p value resulting from hypergeometric test), p.adjust [Benjamini–Hochberg-adjusted p value (FDR)], qvalue (Storey-adjusted p value), geneID (gene symbols corresponding to the differentially expressed genes in the GO set [i.e., from GeneRatio above]), and Count [number of differentially expressed genes in the GO set (numerator of GeneRatio, to avoid forced Excel conversion to dates from some fraction)]. Download Figure 3-2, CSV file.
Figure 3-3
Cort-enriched and TrkB-dependent genes in SFARI. Rows indicate SFARI genes (from either the human or mouse model databases, as described in the text) that were differentially expressed in at least one dataset (Cort Cre vs Cort Cre;TrkB flox/flox bulk cortex; Cort Cre Input vs IP; or Cort Cre IP vs Cort Cre;TrkB flox/flox IP). TRUE indicates that gene was significant in that particular χ2 enrichment test for that dataset and SFARI gene set. The first column indicates the Gencode ID. Download Figure 3-3, CSV file.
Figure 3-4
Cort-enriched and TrkB-dependent genes in Harmonizome database. Each Excel tab indicates the enrichment analyses from each disease gene set in the Harmonizome database. For “Bulk” and “IP genotype” tabs, columns represent Harmonizome disease set description, OR (odds ratio of being differentially expressed and in the disease set compared with being differentially expressed and not in the disease set), p.value (p value from χ2 test), adj.P.Val (Benjamini–Hochberg-adjusted p value), setSize (number of genes in the disease gene set), numSig (number of significantly differentially expressed genes in the gene set), ID (Harmonizome ID), and sigGenes [genes significantly differentially expressed and in the disease set (i.e. those genes driving the enrichment)]. For the “IP vs Input” tab, columns represent Harmonizome disease set description, Enrich_OR (odds ratios from genes significantly more highly expressed in Cort neurons than in Input), Enrich_Pval (corresponding p value from χ2 test), Deplete_OR (odds ratios from genes significantly more highly expressed in Input vs Cort neurons), Deplete_Pval (corresponding p value from χ2 test), and setSize (number of genes in the disease gene set). Download Figure 3-4, XLSX file.
In this issue
