A MYT1L syndrome mouse model recapitulates patient phenotypes and reveals altered brain development due to disrupted neuronal maturation

Abstract

Human genetics have defined a new neurodevelopmental syndrome caused by loss-of-function mutations in MYT1L, a transcription factor known for enabling fibroblast-to-neuron conversions. However, how MYT1L mutation causes intellectual disability, autism, ADHD, obesity, and brain anomalies is unknown. Here, we developed a Myt1l haploinsufficient mouse model that develops obesity, white-matter thinning, and microcephaly, mimicking common clinical phenotypes. During brain development we discovered disrupted gene expression, mediated in part by loss of Myt1l gene-target activation, and identified precocious neuronal differentiation as the mechanism for microcephaly. In contrast, in adults we discovered that mutation results in failure of transcriptional and chromatin maturation, echoed in disruptions in baseline physiological properties of neurons. Myt1l haploinsufficiency also results in behavioral anomalies, including hyperactivity, muscle weakness, and social alterations, with more severe phenotypes in males. Overall, our findings provide insight into the mechanistic underpinnings of this disorder and enable future preclinical studies.

Keywords: ADHD; ASD; Autism; Chromatin Accessibility; Hyperactivity; ID; Neuronal Differentiation; Social Motivation; Transcription.

Figure 1.. MYT1L frameshift mutation results in protein haploinsufficiency, physical anomalies, and obesity

(A) qRT-PCR revealed the trajectory of MYT1L mRNA expression across mouse brain development (n = 3). (B) Western blot showed a parallel trajectory of protein levels. (C) IF of MYT1L protein (red) revealed expression in MAP2+ (green) cortical plate (CP), intermediate zone (IZ), and a few in SOX2+ (white) progenitors in the ventricular zone (VZ)/subventricular zone (SVZ). Scale bar, 50 μm. (D) Quantification of MYT1L+ fraction within different cell types (n = 3). (E) Schematics for MYT1L-KO mouse line generation. (F) Sanger sequencing of c.3035dupG mutation on MYT1L mutant allele. (G and H) Western blot on P1 whole brain lysates confirmed MYT1L protein reduction in Het mice. (I–K) On physical examination, a subset of Het mice displayed (I) fifth finger clinodactyly and (J and K) abnormal hindlimb posture. (L) Het mice weighed significantly more than WT mice as adults, which was more pronounced in females. Data are represented as mean ± SEM. See also Figure S1 and Table S5 for statistical test details.

Figure 2.. MYT1L haploinsufficiency causes microcephaly and white-matter thinning in corpus callosum

(A) Sectioning strategy for Nissl staining. Scale bar, 3 mm. (B) Diagram of different brain structures examined. (C and D) Adult Het mice had decreased brain weight (C) and (D) decreased cortical volume. (E) Coronal images acquired from DTI. (F) Fractional anisotropy (FA) map for visualization of white-matter tracts. Scale bar, 0.5 cm. (G) Three-dimensional reconstruction of different white-matter tracts via FA maps, including corpus callosum (CC; green), cerebral peduncle (CP; red), internal capsule (IC; blue), fimbria (yellow), and cortex (blue). (H) DTI recapitulated smaller brain phenotype in Het mice. (I) Histogram showed adult Het mice had decreased corpus callosum volume. Data were normalized to brain volume. Data are represented as mean ± SEM. See also Figure S2 and Table S5 for statistical test details.

Figure 3.. Chromatin accessibility and RNA-seq analysis define molecular consequences of MYT1L loss in the developing brain

(A and B) Less accessible (A) and (B) more accessible regions in MYT1L mutant E14 mouse cortex identified by ATAC-seq (FDR < 0.1). (C) Homer motif analysis of less accessible DARs over more accessible DARs. (D) GO analysis of less accessible DARs associated genes showed the disruption of neurodevelopmental programming in mutants. (E) Heatmap for differential gene expression in mutants (FDR < 0.1). (F and G) GO analysis of DEGs revealed an upregulation of early neuronal differentiation pathways (F) and (G) a downregulation of cell proliferation programs. (H and I) GSEA analysis revealed iN signature genes increased expression (H), while (I) MEF genes decreased expression in mutants cortex. See also Figure S3 and Table S5 for statistical test details.

Figure 4.. MYT1L loss disrupts progenitor proliferation by precocious cell-cycle exit

(A) IF for nuclei (DAPI; blue), apical progenitors (SOX2; green), intermediate progenitors (TBR2; gray), and postmitotic neurons (TBR1; red) in the E14 mouse cortex. (B–E) KO mouse cortex had significantly less cellular density (B) and (C) fewer apical progenitors, with normal (D) intermediate progenitors and (E) postmitotic neurons. (F) Myt1l mutants have significantly more early cell stage populations but less later cell stage population. (G–I) KO mice have fewer proliferating cells compared with Het and WT littermates. (J and K) EdU labeling for a 1.5 h window revealed decreased cell proliferation rate in mutant mouse cortex compared with WT. (L and M) Co-staining for Ki-67 and EdU (20 h after labeling) experiments (L) found (M) a larger Q fraction value in KO but not in Het mouse cortex. White dashed lines in (L) indicate the border where proliferating cells started to exit the cell cycle and differentiate. Data are represented as mean ± SEM. Scale bars, 25 μm (A), 50 μm (G), and 100 μm (J and L). See also Figure S4 and Table S5 for statistical test details.

Figure 5.. Long-term MYT1L deficiency results in arrested maturation of neuronal chromatin and expression patterns

(A and B) Less accessible (A) and (B) more accessible regions in adult Het mouse PFC identified by ATAC-seq (FDR < 0.1). (C) Homer motif analysis of less accessible DARs over more accessible DARs. (D) GO analysis of DAR-associated genes showed the dysregulation of neurodevelopmental programming in adult Het mouse PFC. (E) Heatmap for differential gene expression in adult Het mouse PFC (FDR < 0.1). (F and G) GO analysis of DEGs revealed an upregulation of early neurodevelopmental pathways (F) and (G) a downregulation of neuron maturation and functions. (H and I) GSEA analysis revealed that “early fetal” genes increased their expression (H), while (I) “mid-fetal” genes remained unchanged in adult Het mouse PFC compared. (J) Repressed genes upon MYT1L loss in PFC significantly overlapped with induced neuron and neuronal signature genes. (K) MYT1L regulated genes were implicated in other ID/ASD mouse models and human genetic datasets. See alsoFigure S5 and Table S5 for statistical test details.

Figure 6.. MYT1L haploinsufficiency disrupts baseline neuronal properties and dendritic spine maturity but not neuronal morphology

(A–D) MYT1L loss led to less negative membrane potential (A), (B) reduced membrane resistance, (C) decreased membrane capacitance, and (D) smaller membrane time constant in cortical pyramidal neurons. (E) Neuronal soma and dendrites tracing in Neurolucida. (F) Sholl analysis revealed no dendrite complexity change across genotypes. (G) Het neurons showed increased mEPSC amplitude distribution compared with WT neurons. (H and I) Analysis of individual events of mEPSC and mIPSC revealed that the charges of Het neurons’ mEPSCs are slightly larger (H), (I) while mIPSCs are slightly smaller. (J) Representative images of spine tracing and subtypes identification using Neurolucida. (K and L) Het neurons had more apical spines (K) with (L) general increase in different spine subtypes. (M) Het neurons had a higher percentage of immature spines (stubby, thin) but less mature spines (mushroom) compared with WT. Data are represented as mean ± SEM. See also Figure S6 and Table S5 for statistical test details.

Figure 7.. Myt1l haploinsufficiency results in heightened USV production and muscle weakness and fatigue

(A and B) Timelines for (A) developmental assessments and (B) post-weaning assays. (C) Comparable early postnatal weight trajectories. (D–H) Hets produced fewer USVs than WT (D), which did not differ from WT calls on (E–G) temporal (call duration, pause duration, sound pressure level) or (H) spectral (mean frequency) features. (I and J) MYT1L loss was not associated with ambulation scores at P8 (I) or (J) grasping reflex. (K) Hets exhibited latency to righting reflex similar to WT mice. (L) Latency to exhibit negative geotaxis was comparable, but MYT1L loss was associated with increased falls from the inclined apparatus. (M) Hets were unable to remain suspended by fore or hindlimbs as long as WT mice. (N) Hets fell from the grip strength mesh screen at a narrower angle than WT mice. (O) As adults, Hets hung on an inverted screen for a shorter latency than WT mice. (P) Hets exhibited a longer latency than WT mice to climb to the top of a 90° screen. (Q–S) Times to balance on an elevated platform (Q), (R) latencies to down a pole, and (S) latencies to climb up a 60° wire mesh screen were comparable. (T) Hets initiated movement at a similar latency to WT mice. (U) Percentage inhibition of startle following a prepulse was similar in Het and WT mice. For (C), (D), and (U), grouped data are presented as mean ± SEM. For (E)–(H), (K), and (L) (left) and (M)–(T), grouped data are presented as boxplots with thick horizontal lines denoting group medians, boxes 25th to 75th percentiles, and whiskers 1.5 × interquartile range (IQR). Individual data points are open circles. See Table S5 for statistical details.

Figure 8.. Myt1l haploinsufficiency altered social behaviors

(A) MYT1L loss was associated with losses in the social dominance assay. (B) Social approach test schematic. Investigation zones are demarcated by the dotted red lines. (C) In the sociability trial, Hets spent less time investigating the social stimulus than WT mice and failed to show an increase in time spent in the social versus empty investigation zones. No difference was observed in social novelty. (D) Sociability and social novelty preference scores were comparable. (E) Hets spent more time in the center chamber during both trials compared with WT mice. (F) In the sociability trial, Hets entered the zone surrounding the social stimulus fewer times and failed to show an increase in entries into the social cup zone versus the empty cup zone. In the social novelty trial, Hets entered the zone surrounding the novel mouse less than WT mice. (G and H) Social operant assay and timeline schematics. (I) C57BL/6J mice show consistency in the maximum level of effort they will exert for access to social interaction reward, demonstrating that performance in the social operant test is reproducible across test days. (J) This maximum effort is driven by the social aspect of the reward, as demonstrated by the difference in performance between mice that received the social interaction reward and mice that did not. (K) The time series of task acquisition demonstrates that Myt1l WT and Het mice learn to discriminate between correct and incorrect holes for access to a social interaction reward during FR1 training. (L) All mice that meet learning criteria are motivated to work harder for the social interaction reward when more effort is required in FR3 testing. (M and N) Day to reach criteria during social operant training (M) and (N) breakpoint reached during PR3 testing were not different between Het and WT mice. (O) Het males achieved fewer social rewards compared with WT males. (P) Het males and females exhibited a comparable number of correct nose pokes as WT littermates. (Q) During a reward, Het males trended toward less total time in the social interaction zone compared with WT males. Regardless of genotype, males spent more time in the social interaction zone compared with females. (R) Het males spent less total time in the social interaction zone than WT males. Regardless of genotype, males spent more time in the social interaction zone compared with females. (S) Female and male Hets traveled farther distances during 1 h social operant trials compared with WT mice. Overall, females traveled farther distances than males during social operant trials. For (C)–(F), (I)–(L), and (N)–(S), grouped data are mean ± SEM. Individual data points are open circles. See also Figures S7 and S8 and Table S5 for statistical details.