Myt1l haploinsufficiency leads to obesity and multifaceted behavioral alterations in mice

Abstract

Background: The zinc finger domain containing transcription factor Myt1l is tightly associated with neuronal identity and is the only transcription factor known that is both neuron-specific and expressed in all neuronal subtypes. We identified Myt1l as a powerful reprogramming factor that, in combination with the proneural bHLH factor Ascl1, could induce neuronal fate in fibroblasts. Molecularly, we found it to repress many non-neuronal gene programs, explaining its supportive role to induce and safeguard neuronal identity in combination with proneural bHLH transcriptional activators. Moreover, human genetics studies found MYT1L mutations to cause intellectual disability and autism spectrum disorder often coupled with obesity.

Methods: Here, we generated and characterized Myt1l-deficient mice. A comprehensive, longitudinal behavioral phenotyping approach was applied.

Results: Myt1l was necessary for survival beyond 24 h but not for overall histological brain organization. Myt1l heterozygous mice became increasingly overweight and exhibited multifaceted behavioral alterations. In mouse pups, Myt1l haploinsufficiency caused mild alterations in early socio-affective communication through ultrasonic vocalizations. In adulthood, Myt1l heterozygous mice displayed hyperactivity due to impaired habituation learning. Motor performance was reduced in Myt1l heterozygous mice despite intact motor learning, possibly due to muscular hypotonia. While anxiety-related behavior was reduced, acoustic startle reactivity was enhanced, in line with higher sensitivity to loud sound. Finally, Myt1l haploinsufficiency had a negative impact on contextual fear memory retrieval, while cued fear memory retrieval appeared to be intact.

Limitations: In future studies, additional phenotypes might be identified and a detailed characterization of direct reciprocal social interaction behavior might help to reveal effects of Myt1l haploinsufficiency on social behavior in juvenile and adult mice.

Conclusions: Behavioral alterations in Myt1l haploinsufficient mice recapitulate several clinical phenotypes observed in humans carrying heterozygous MYT1L mutations and thus serve as an informative model of the human MYT1L syndrome.

Keywords: Autism; Obesity; Social behavior; Transcription factor; Ultrasonic vocalization.

Fig. 1

Generation of Myt1l mutant mice by CRISPR/Cas9 gene editing. A Schematic of the gene editing strategy. The guide RNA sequence (blue text) targets seven base pairs downstream of the translation start site in exon VI of the mouse Myt1l locus and generates a frameshift deletion. The two forward and one reverse primers (red arrows) used for genotyping are noted. The forward primers, F1 and F2, incorporate the presence and absence of the seven base pairs, respectively. B PCR genotyping of Myt1l mutant mice. As expected, the primer pair F1 and R1 (left lanes) shows a PCR amplification of only wild-type alleles, whereas the primer pair F2 and R1 (right lanes) amplifies only the deletion allele, allowing unequivocal genotyping of wild-type Myt1l^+/+, heterozygous Myt1l^+/-, and homozygous Myt1l^−/− mice. (C) Immunoblotting of Myt1l in WT, HET, and KO E18.5 brains. Whole brain lysates were subjected to Western blotting using two different Myt1l antibodies with similar results. MAP2 was used as a loading control. Note with both antibodies an extra band(*) appeared in mice carrying the mutant allele corresponding to a Myt1l-related protein of ~ 158 kDa

Fig. 2

Myt1l is essential for postnatal survival. A Genotype distribution of Myt1l offspring at various ages. The dashed bars represent the expected percentage of mice for a genotype, assuming Mendelian inheritance from two heterozygous parents. The solid bars show the observed distribution based on 10–13 litters. Approximately 25% of mice were WT, 50% HET, and 25% KO at ages E13.5, E15.5, and E18.5. In contrast, only 12% of KO were present at PND 0. Numbers in bars represent number of mice per genotype/total number of mice. The chi-square test indicated a significant difference between the expected and observed distribution of KO pups at PND 0 (p = 0.011). n.s. = not significant; *p < 0.05. B Kaplan–Meier survival curve comparing the three genotypes after birth. WT and HET survive past PND 7, while all Myt1l KO die by PND 1. A logrank Mantel–Cox test showed a significant trend for the survival probability of KO pups (p = 0.0019)

Fig. 3

No overt anatomical defects in the developing mouse brain of Myt1l mutant mice. A, B Sox2 staining (cyan) in E15.5 Myt1l (N = 3/genotype) shows no gross difference in VZ compared with littermate controls. C The number of Sox2-positive neurons was quantified as the percentage of Sox2-positive area in a 100 × 100 µm area in the ventricular zone. D Coronal sections stained for Tbr2 (green), Ctip2 (red), and counterstained with the nuclear stain DAPI (blue). E The size of layers expressing Tbr2 and Ctip2 remains similar between all three genotypes (N = 3/genotype). F The density of Ctip2 positive neurons was assessed by counting in a field of 50 × 100 µm in the central CP. G Nissl-stained whole brains from E18.5 embryos. H Myt1l mutants show similar gross brain anatomy, including the rostral and caudal cortex (motor cortex region M1, sensory cortex barrel region S1), corpus callosum, striatum, septum, hippocampus, thalamus, and hypothalamus, compared to WT and HET. N = 3/genotype. (Left and right hemispheres of one animal are indicated as arrowheads leading in the same direction; MZ marginal zone, CP cortical plate, IZ intermediate zone, (d)SVZ (distal) subventricular zone, MAZ multipolar cell accumulation zone, (p)VZ (proximal) ventricular zone, V ventricle)

Fig. 4

Effects of Myt1l haploinsufficiency on maternal odor preference in the homing test. A Maternal odor preference is intact in Myt1l^+/- mouse pups. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor preference (P). P: F_1,52 = 214.668; p < .001; all other p > 0.050. B Area crossings are unchanged in Myt1l^+/- mouse pups. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050 C Body weight is unchanged in Myt1l^+/- mouse pups. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. D Body temperature is not altered in Myt1l^+/- mouse pups. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. All data are means ± SEM, combined across males and females; * p < 0.05; **p < 0.01; ***p < 0.001

Fig. 5

Effects of Myt1l haploinsufficiency on the emission of isolation-induced ultrasonic vocalizations in the homing test. A, B Emission of isolation-induced ultrasonic vocalizations is altered in Myt1l^+/- mouse pups. Representative spectrograms showing sequences of isolation-induced ultrasonic vocalizations (A: Myt1l^+/+ littermate control, B: Myt1l^+/- mouse pup). Please note the higher peak frequency that is characteristic for many isolation-induced ultrasonic vocalizations emitted by Myt1l^+/- mouse pups. C Total calling time is unchanged in Myt1l^+/- mouse pups. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. D Call number is not altered in Myt1l^+/- mouse pups. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. E Number of calls in the low-frequency cluster is not affected by Myt1l haploinsufficiency. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. F Number of calls in the high-frequency cluster is increased in Myt1l^+/- mouse pups. ANOVA with the between-subject factors genotype (G) and sex (S): G: F_1,52 = 5.760; p = .020; all other p > 0.050. G–L Density plots depicting the distribution of individual isolation-induced ultrasonic vocalizations in Myt1l^+/+ littermate controls (G-I; ~ 20,000 calls) and Myt1l^+/- mouse pups (J-L; ~ 30,000 calls). Color coding reflects frequencies as percentages. All data are means ± SEM, combined across males and females; *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 6

Effects of Myt1l haploinsufficiency on body weight gain and body temperature after weaning. A Body weight gain is increased in male and female Myt1l^+/- mice. Representative picture of two females from the same litter (left: Myt1l^+/+ littermate control, right: Myt1l^+/- mouse). B Body weight gain after weaning measured at PND 378 (relative to PND 25). ANOVA with the between-subject factors genotype (G) and sex (S), followed by unpaired t tests when appropriate. G: F_1,50 = 26.381; p < .001; S: F_1,50 = 5.532; p = .023; GxS: F_1,50 = 4.054; p = .049. C Body weight gain trajectories after weaning. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor age (A), followed by individual ANOVAs when appropriate. A: F_16,800 = 596.784; p < .001; AxG: F_16,800 = 13.965; p < .001; AxS: F_16,800 = 6.314; p < .001; AxGxS: F_16,800 = 3.621; p < .001, G: F_1,50 = 13.428; p = .001; S: F_1,50 = 37.161; p < .001; all other p > 0.050. D Body temperature is not altered in male and female Myt1l^+/- mice. ANOVA with the between-subject factors genotype (G) and sex (S); all p values > 0.050. Of note, N = 2 male Myt1l^+/+ littermate control mice died a natural death between months 11 and 13 without evidence of injuries from fighting and were excluded from the body weight gain and body temperature analyses. All data are means ± SEM, combined across males and females if not otherwise indicated; *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 7

Effects of Myt1l haploinsufficiency on locomotor activity and habituation learning in the activity box. A–A′ Locomotor activity is increased in Myt1l^+/- mice. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor time (T). T: F_29,1508 = 3.848; p < .001; TxG: F_29,1508 = 1.819; p = .005; TxS: F_29,1508 = 1.634; p = .018; G: F_1,52 = 8.682; p = .005; S: F_1,52 = 11.948; p = .001; all other p > 0.050. B–B′ Rearing behavior is increased in Myt1l^+/- mice. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor time (T). T: F_29,1508 = 2.760; p < .001; G: F_1,52 = 17.237; p < .001; GxS: F_1,52 = 5.395; p = .024; all other p > 0.050. C–C′ Locomotor activity remains high in Myt1l^+/- mice during the second exposure to the activity box. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor time (T). TxS: F_29,1508 = 1.853; p = .004; G: F_1,52 = 15.266; p < .001; S: F_1,52 = 22.731; p < .001; GxS: F_1,52 = 9.918; p = .003; all other p > 0.050. D–D′ Rearing behavior remains high in Myt1l^+/- mice during the second exposure to the activity box. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor time (T). T: F_29,1508 = 4.570; p < .001; TxS: F_29,1508 = 2.086; p = .001; G: F_1,52 = 11.525; p < .001; all other p > 0.050. All data are means ± SEM, combined across males and females; *p < 0.05; **p < 0.01; ***p < 0.001; ^#p < 0.05 versus day 1

Fig. 8

Effects of Myt1l haploinsufficiency on anxiety-related behavior, social behavior, interaction-induced ultrasonic vocalizations, self-grooming behavior, spatial working memory, and nest building. A Distance traveled in the open field. ANOVA with the between-subject factors genotype (G) and sex (S). S: F_1,52 = 8.198; p = .006; all other p > 0.050. B Time spent in the center of the open field. ANOVA with the between-subject factors genotype (G) and sex (S). S: F_1,52 = 18.534; p < .001; all other p > 0.050. C Time spent in the open arms of the elevated plus maze. ANOVA with the between-subject factors genotype (G) and sex (S). G: F_1,52 = 8.747; p = .005; S: F_1,52 = 7.494; p = .008; all other p > 0.050. D Entries into open and closed arms of the elevated plus maze. ANOVA with the between-subject factors genotype (G) and sex (S). S: F_1,52 = 5.382; p = .024; all other p > 0.050. E, F Emission of female-induced ultrasonic vocalizations is not altered in Myt1l^+/- mice. Representative spectrograms showing sequences of female-induced ultrasonic vocalizations (A: Myt1l^+/+ littermate control, B: Myt1l^+/- mouse). Please note the presence of low-frequency noise due to locomotor activity. G Social approach behavior is intact in Myt1l^+/- mice. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor preference (P). P: F_1,52 = 45.199; p < .001; all other p > 0.050. H Call number is not altered during male–female social interaction in Myt1l^+/- mice. ANOVA with the between-subject factor genotype (G); all p > 0.050. I Call number is not altered during female–female social interaction in Myt1l^+/- mice. ANOVA with the between-subject factor genotype (G); all p > 0.050. J Self-grooming. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. K Alternation quotient Y-maze. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. L Nest width. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor time (T), followed by unpaired t tests when appropriate. T: F_1,260 = 25.511; p < .001; G: F_1,52 = 5.591; p = .022; all other p > 0.050. M Nest quality. ANOVA with the between-subject factors genotype (G) and sex (S). G: F_1,52 = 9.644; p = .003; all other p > 0.050. All data are means ± SEM, combined across males and females if not otherwise indicated; *p < 0.05; **p < 0.01; ***p < 0.001. ^#p < 0.05 versus 50% chance level

Fig. 9

Effects of Myt1l haploinsufficiency on motor performance and learning on the accelerating rotarod. A Motor performance in the standard task with an accelerating rotarod (4–40 rpm) is reduced in Myt1l^+/- mice and reflected in a lower latency to fall, despite evidence for intact motor learning. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor trial (T), followed by unpaired t tests when appropriate. T: F_1,260 = 51.953; p < .001; G: F_1,52 = 7.309; p = .009; S: F_1,52 = 11.197; p = .002; all other p > 0.050. B Motor performance in the challenge task with an accelerating rotarod (8–80 rpm) is reduced in Myt1l^+/- mice and reflected in a lower latency to fall. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor trial (T), followed by unpaired t tests when appropriate. T: F_1,260 = 6.701; p < .001; TxS: F_5,260 = 2.512; p = .030; G: F_1,52 = 8.933; p = .004; S: F_1,52 = 21.467; p < .001; all other p > 0.050. All data are means ± SEM, combined across males and females; *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 10

Effects of Myt1l haploinsufficiency on spatial learning and reversal learning. A Spatial learning and reversal learning in the Barnes maze is intact in Myt1l^+/- mice. B Intact spatial learning and reversal learning is reflected in a decrease in the latency needed to reach the target hole and the reversal target hole, respectively. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor trial (T). For spatial learning, T: F_15,780 = 41.186; p < .001; S: F_1,52 = 6.936; p = .011; all other p > 0.050. For reversal learning, T: F_15,780 = 28.583; p < .001; S: F_1,52 = 4.138; p = .047; all other p > 0.050. C Affinity for the target hole during spatial learning is unchanged in Myt1l^+/- mice. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. D Intact spatial learning in Myt1l^+/- mice is also reflected in a clear preference for visiting the target hole during the last spatial learning day. Gray highlighting indicates target. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor preference (P). P: F_8,416 = 70.310; p < .001; all other p > 0.050. E Primary errors during spatial learning are unchanged in Myt1l^+/- mice. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. F Affinity for the reversal target hole is slightly reduced in Myt1l^+/- mice. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. Paired t test for comparing reversal target hole (triangle) and initial target hole (circle); Myt1l^+/- mice: t₂₉ = 1.615; p = .117; Myt1l^+/+ mice: t₂₅ = 2.222; p = .036. G, H Despite a clear preference for the initial target hole, intact reversal learning in Myt1l^+/- mice is also reflected in the rapid development of a secondary preference for visiting the reversal target hole during the first reversal learning day. Gray highlighting indicates target. Repeated-measures ANOVA with the between-subject factors genotype (G) and sex (S) and the within-subject factor preference (P). For the initial target hole, P: F_8,416 = 72.597; p < .001; all other p > 0.050. For the reversal target hole, P: F_8,416 = 9.625; p < .001; PxS: F_8,416 = 3.657; p < .001; all other p > 0.050. I Primary errors during reversal learning are unchanged in Myt1l^+/- mice. ANOVA with the between-subject factors genotype (G) and sex (S); all p > 0.050. All data are means ± SEM, combined across males and females; *p < 0.05; **p < 0.01; ***p < 0.001. ^#p < 0.05 versus 4.5s chance level or adjacent holes to target

Fig. 11

Effects of Myt1l haploinsufficiency on acoustic startle reactivity and pre-pulse inhibition of acoustic startle. A–C Acoustic startle reactivity is increased in male but not female Myt1l^+/- mice. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor startle sound intensity (I), followed by unpaired t tests when appropriate. I: F_4,208 = 84.359; p < .001; IxGxS: F_4,208 = 3.648; p = .007, S: F_1,52 = 4.662; p = .035; GxS: F_1,52 = 4.574; p = .037; all other p > 0.050. D–F Pre-pulse inhibition of acoustic startle is unchanged in male and female Myt1l^+/- mice. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor pre-pulse sound intensity (I), followed by unpaired t tests when appropriate. I: F_3,156 = 95.448; p < .001; IxS: F_3,156 = 9.664; p < .001; IxGxS: F_3,156 = 3.071; p = .030; S: F_1,52 = 8.443; p = .005; all other p > 0.050. D′–F′ Pre-pulse inhibition of acoustic startle is unchanged in male and female Myt1l^+/- mice also after correcting for baseline differences in startle reactivity. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor pre-pulse sound intensity (I), followed by unpaired t tests when appropriate. I: F_2,104 = 80.919; p < .001; IxS: F_2,104 = 5.880; p = .004; S: F_1,52 = 6.643; p = .013; all other p > 0.050. All data are means ± SEM, combined across males and females if not otherwise indicated; *p < 0.05; **p < 0.01; ***p < 0.001

Fig. 12

Effects of Myt1l haploinsufficiency on fear conditioning. A–C Fear-related freezing behavior is enhanced in male and female Myt1l^+/- mice during acquisition on the training day. Gray highlighting indicates tone presentations (1 min inter-stimulus intervals). Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor time (T), followed by unpaired t tests when appropriate. T: F_12,624 = 190.370; p < .001; TxG: F_12,624 = 3.535; p < .001; TxS: F_12,624 = 2.355; p = .006; TxGxS: F_12,624 = 2.015; p = .021; G: F_1,52 = 4.803; p = .033; S: F_1,52 = 25.623; p < .001; all other p > 0.050. D–F Fear-related freezing behavior is enhanced in male but not female Myt1l^+/- mice during contextual recall. Repeated-measures ANOVAs with the between-subject factors genotype (G) and sex (S) and the within-subject factor time (T), followed by unpaired t tests when appropriate. T: F_9,468 = 23.286; p < .001; TxG: F_9,468 = 2.660; p = .005; G: F_1,52 = 6.682; p = .013; S: F_1,52 = 16.678; p < .001; all other p > 0.050. All data are means ± SEM, combined across males and females if not otherwise indicated; *p < 0.05; **p < 0.01; ***p < 0.001