MYT1L deficiency impairs excitatory neuron trajectory during cortical development

Abstract

Mutations reducing the function of MYT1L, a neuron-specific transcription factor, are associated with a syndromic neurodevelopmental disorder. MYT1L is used as a pro-neural factor in fibroblast-to-neuron transdifferentiation and is hypothesized to influence neuronal specification and maturation, but it is not clear which neuron types are most impacted by MYT1L loss. In this study, we profile 412,132 nuclei from the forebrains of wild-type and MYT1L-deficient mice at three developmental stages: E14 at the peak of neurogenesis, P1 when cortical neurons have been born, and P21 when neurons are maturing, to examine the role of MYT1L levels on neuronal development. MYT1L deficiency disrupts cortical neuron proportions and gene expression, primarily affecting neuronal maturation programs. Effects are mostly cell autonomous and persistent through development. While MYT1L can both activate and repress gene expression, the repressive effects are most sensitive to haploinsufficiency, likely mediating MYT1L syndrome. These findings illuminate MYT1L's role in orchestrating gene expression during neuronal development, providing insights into the molecular underpinnings of MYT1L syndrome.

Fig. 1. Single nucleus transcriptional profiling of E14 forebrain in MYT1L animals.

A Schematic showing dissection of forebrain tissue, isolation of nuclei, massively parallel barcoding, and generation of snRNAseq libraries. Created in BioRender. Dougherty, J. (2024) BioRender.com/k40k912. B General library statistics showing mean ± SD nuclei per genotype (n = 3 biological replicates per genotype: WT, Het, and KO), median ± SD genes per nucleus, and median ± SD UMIs per nucleus. Box plots show median (center line), interquartile range (box), and whiskers extending to 1.5 times the interquartile range. C Uniform manifold approximation and projection (UMAP) showing 216,830 nuclei from the forebrain of E14 MYT1L WT (n = 3), Het (n = 3), and KO (n = 3) animals colored by cell type. D Bar plot showing the total number of nuclei per genotype across biological replicates. E Histogram showing the local inverse Simpson’s index (LISI) score with a median of 2.7, indicating that the genotypes are well mixed and integrated. F UMAP of all nuclei color-coded by cell class. G Top cluster markers for progenitors, intermediate progenitor cells (IPCs), inhibitory neurons, and excitatory neurons. H UMAP of all nuclei color-coded by cell cycle score based on expression of cell cycle genes (G2M phase in orange, S phase in yellow, and G1/G0 in blue). I UMAP feature plot showing expression of MYT1L in postmitotic excitatory and inhibitory neurons. J From left to right, plots showing: the mean ± SEM relative proportions of nuclei in each annotated cell cluster for MYT1L WT, Het, and KO genotypes (n = 3 biological replicates per genotype); Het and KO mean ± SEM proportions normalized to WT; the proportions of nuclei in the phases of the cell cycle per replicate; mean ± SEM proportions of nuclei that are in the G0/G1 phase; mean ± SEM proportions of nuclei in the S phase; and mean ± SEM proportions of nuclei in the G2M phase. “#” indicates a significant difference between WT and Het, and “*” indicates a significant difference between WT and KO (FDR adjusted p < 0.05, moderated ANOVA).

Fig. 2. Cell type-specific changes in gene expression at E14.

A Summary plot showing the numbers of nuclei and genes detected in each cluster (left). Bar plot (middle) displays the number of differentially expressed genes (DEGs) upregulated in WT (red) or KO (blue) conditions (n = 3 biological replicates per condition). DEG burden analysis (right) shows the distribution of DEGs per cell type, normalized by the number of nuclei in each cluster. Statistical significance was assessed using a two-sided Mann-Whitney U test with Benjamini-Hochberg correction for multiple corrections (*FDR adjusted p < 0.05). Box plots show median (center line), interquartile range (box), and whiskers extending to 1.5 times the interquartile range. B MYT1L gene dose-dependent expression patterns of DEGs in Im ExN_2 and L5-6 ExN_1 clusters, categorized as MYT1L-activated (decreased expression in KO) and MYT1L-represssed (increased expression in KO). Box plots show median, interquartile range, and whiskers extending to 1.5 times the interquartile range. Dot plots showing enriched GO biological processes terms in MYT1L-repressed (C) and MYT1L-activated (D) DEGs. Overrepresentation of genes within GO terms was determined using a one-sided hypergeometric test, with p values adjusted for multiple testing using the Benjamini-Hochberg method. E Plot showing the frequency distribution of annotated protein classes among the DEGs. F Heatmap displaying identified regulons (columns) for each WT cluster (rows), colored by the Regulon Specificity Score (RSS). RSS measures the specificity of each regulon’s activity for each cluster. Cluster classes are annotated on the right, and differentially expressed regulons are annotated in each column. G Representative plots showing the ranked Regulon Specificity Score plots for Im L5-6 ExN_1, L5-6 ExN_1, and Sst InhN clusters. Differentially expressed regulons are highlighted in red and labeled on the plot.

Fig. 3. Loss of MYT1L disrupts excitatory neuron maturation at E14.

A UMAP of all E14 nuclei colored by pseudotime. B Box plots showing distributions of nuclei from excitatory neurons along pseudotime per genotype (n = 3 biological replicates per genotype: WT, Het, and KO). “#” indicates a significant difference between WT and Het distributions, and “#” indicates a significant difference between WT and KO distributions (FDR adjusted p < 0.05, two-sided Kolmogorov-Smirnov test). Box plots show median (center line), interquartile range (box), and whiskers extend to 1.5 times the interquartile range. C Representative plots showing the relative differences in distributions of MYT1L Het and KO nuclei compared to the WT distribution in the RG_2 and Im ExN_3 clusters. D Box plots showing the distributions of inhibitory neuron nuclei along pseudotime per genotype (n = 3 biological replicates per genotype: WT, Het, and KO). “#” indicates a significant difference between WT and Het distributions, and “*” indicates a significant difference between WT and KO distributions (FDR adjusted p < 0.05, two-sided Kolmogorov-Smirnov test). Box plots show median, interquartile range, and whiskers extending to 1.5 times the interquartile range. E Diagram displaying the breakdown of direct MYT1L targets identified by CUT&RUN that were transcription factors and showed gene expression changes across pseudotime in the excitatory neuron trajectory. F Heatmaps showing scaled expression of WT (left), Het (middle), and KO (right) excitatory neuron pseudotemporal genes. Each row represents a gene, with rows sorted according to their expression peak in pseudotime. Black tick marks on the right indicate rows where genes are MYT1L direct targets as determined by CUT&RUN in (E). G Scatterplot showing the Kullback-Liebler divergence metric used to identify differential pseudotemporal expression profiles in KOs compared to WT. H Representative traces of the differential pseudotemporal gene expression profiles for Tet2 and Hbp1 across genotypes.

Fig. 4. Loss of MYT1L disrupts proportions of excitatory neurons at P1.

A UMAP projection showing 98,797 nuclei in 36 clusters from the forebrain of P1 MYT1L WT (n = 8) and Het (n = 4) animals. B From left to right: summary plot showing the numbers of nuclei and genes detected in each cluster; bar plot displaying the mean ± SEM relative proportions of nuclei in each annotated cell cluster for MYT1L WT and Het genotypes; mean ± SEM proportions of Het normalized to WT; and number of differentially expressed genes (DEGs) per cell type that are upregulated in WT (light blue; n = 8 biological replicates) and upregulated in Het (medium blue, n = 4 biological replicates) (*FDR adjusted p < 0.05, moderated t-test). Dot plots showing enriched GO biological processes terms in (C) MYT1L-activated (decreased expression in Het) and (D) MYT1L-repressed (increased expression in Het) DEGs. Overrepresentation of genes within GO terms was determined using a one-sided hypergeometric test, with p values adjusted for multiple testing using the Benjamini-Hochberg method. E UMAP visualization of excitatory neuron clusters colored by pseudotime. The arrow indicates the inferred developmental trajectory from early (blue) to late (yellow) pseudotime. F Distribution of excitatory neuron subtypes along the pseudotime axis for WT (light blue; n = 8 biological replicates) and Het (blue; n = 4 biological replicates). Box plots show median (center line), interquartile range (box), and whiskers extending to 1.5 times the interquartile range. Asterisks indicate statistically significant differences between WT and Het distributions (*FDR adjusted p < 0.05, two-sided Kolmogorov-Smirnov test). G Percentage change in cell proportion along pseudotime in Het samples normalized to WT for two excitatory neuron subtypes: Im L2-4 ExN (top) and Im ExN_1 (bottom).

Fig. 5. Sensitivity of excitatory neurons persist throughout neurodevelopment to P21.

A UMAP projection showing 96,505 nuclei in 39 clusters from the cortex of P21 MYT1L WT (n = 6) and Het (n = 6) animals. B From left to right: summary plot showing the numbers of nuclei and genes detected in each cluster; bar plot displaying the mean ± SEM relative proportions of nuclei in each annotated cluster for MYT1L WT and Het genotypes; mean ± SEM proportions of Het normalized to WT; and number of differentially expressed genes (DEGs) per cell type upregulated in WT (light blue; n = 6 biological replicates) and upregulated in Het (blue; n = 6 biological replicates). Dot plots showing enriched GO biological processes terms in (C) MYT1L-activated (decreased expression in Het) and (D) MYT1L-repressed (increased expression in Het) DEGs. Overrepresentation of genes within GO terms was determined using a one-sided hypergeometric test, with p values adjusted for multiple testing using the Benjamini-Hochberg method. E, F Representative immunofluorescence images of brain sections from WT and MYT1L Het mice from a medial region (R2) and posterior (R3) region of the cortex. Upper panels show whole-brain sections stained for DARPP32 (green), NeuN (magenta), and DAPI (blue). Lower panels show higher magnification of the boxed areas in the upper panels, showing DARPP32 and NeuN staining separately. G Box plots showing quantification of DARPP32 and NeuN double-positive cell counts in three cortical regions (R1, R2, and R3) of WT and MYT1L Het mice. Each dot represents a counted section and is color-coded by animal. Box plots show median (center line), interquartile range (box), and whiskers extending to 1.5 times the interquartile range. Statistical significance was assessed using a linear mixed model comparison to test the effect of genotype while accounting for section differences and individual mouse variability. Significance is indicated by asterisks (***FDR adjusted p < 0.001, ANOVA).

Fig. 6. Integrated analysis of MYT1L deficiency across neurodevelopment.

A UMAP projection showing integrated data of 412,132 nuclei from E14, P1, and P21 datasets. Left panel: nuclei colored by cell classes. Middle panel: nuclei colored by 54 distinct clusters. Right panel: nuclei colored by developmental stage. B Mean ± SEM proportions of major cell classes across developmental stages in WT (top; n = 17 animals) and Het (bottom; n = 13 animals). C Mean ± SEM proportions of excitatory neuron subtypes across developmental stages in WT (top; n = 17 animals) and Het (bottom; n = 13 animals). UL ExN: upper layer excitatory neurons, DL ExN: deep layer excitatory neurons, Im: immature. D Venn diagram showing the number of differentially expressed genes identified by pseudobulk analysis in deep layer excitatory neurons (DL ExN) associated with age, genotype, and age:genotype interaction. E Expression patterns of genes showing significant age:genotype interaction across developmental stages in WT (blue) and Het (red) animals. Each line represents a gene, with box plots showing the distribution of scaled gene expression values. Box plots show median (center line), interquartile range (box), and whiskers extending to 1.5 times the interquartile range.