The gain-of-function UBE3AQ588E variant causes Angelman-like neurodevelopmental phenotypes in mice

Abstract

Mutations in the E3 ubiquitin ligase UBE3A that cause enzymatic gain-of-function result in disease phenotypes which differ from classic Angelman syndrome. However, these phenotypes are highly heterogeneous raising questions about the mechanistic basis of such phenotypic diversity. Here, we characterize a mouse model harboring a Ube3a^Q606E gain of function variant (UBE3A^Q588E in humans). Extensive behavioral phenotyping showed that animals possessing a maternally inherited mutation (Ube3a^mQ606E) paradoxically show many behavioral deficits indicative of overall UBE3A loss-of-function. These included pronounced motor deficits, hypoactivity, and reduced stereotypic behaviors. Moreover, brain weights and MRI analysis revealed global microcephaly with a postnatal onset, consistent with phenotypes described in Angelman syndrome model mice. Additional biochemical analyses demonstrated an increased abundance of UBE3A substrates in brain tissue and immunofluorescence analyses showed that microcephaly is not caused by increased apoptotic cell death. Together, our results strongly suggest a novel mechanism by which the Ube3a^mQ606E mutation leads to enhanced self-targeted degradation of UBE3A, leading to an overall loss of enzyme activity, resulting in Angelman-like phenotypes in vivo.

Fig. 1

Ube3a^mQ606E causes motor deficits. (A) Graphical depiction of the Ube3a^mQ606E breeding scheme. (B) Graphical depiction of the behavioral assay timeline. (C–E) Latency to fall from an accelerating rotarod for all animals (C), male (D) and female (E) mice. Values represent the mean ± SE. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22, 15 male, 7 female. Repeated measures ANOVA; *p < 0.05. (F) Hang time on inverted mesh wire screen. WT n = 16: 9 male, 7 female; Ube3a^mQ606E n = 20: 13 male, 7 female. Mann–Whitney U Test, 2-tailed (exact significance). (G) Time to turn head-downward on a vertical pole. Values represent the median ± range. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22: 15 male, 7 female. Mann–Whitney U Test, 2-tailed (exact significance); *p < 0.05. (H) Time to descend from a vertical pole. Values represent the median ± range. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22: 15 male, 7 female. Mann–Whitney U Test, 2-tailed (exact significance); *p < 0.05. (I) Average startle response to a 120 dB tone above background. Values represent the mean ± SE. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 21: 14 male, 7 female. Statistical significance was tested using UNIANOVA. (J) Percentage inhibition of startle following a pre-pulse tone. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 21: 14 male, 7 female. Statistical significance was tested using repeated measures ANOVA.

Fig. 2

Ube3a^mQ606E causes hypoactivity. (A) Open Field Assay: number of beam crosses (“ambulations”) over the 60-min course of the trial. Timepoints represent the average number of ambulations ± SE binned for a 10-min period. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22: 15 male, 7 female. Repeated measures ANOVA; *p < 0.05. (B) Total time spent at rest during the open field test. Values represent the mean ± SE. UNIANOVA; *p < 0.05. (C) Time spent in the center circle of the open field. Values represent the mean ± SE. UNIANOVA. (D) Time spent at rest in the center circle of the open field. Values represent the mean ± SE. UNIANOVA; **p < 0.01. (E) Number of rearing events in the open field test. Timepoints represent the average number of rears ± SE binned for a 10-min period. Repeated measures ANOVA; *p < 0.05. (F) Swim speed measured during different phases of the Morris Water Maze. Values represent the mean ± SE. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22: 15 male, 7 female. UNIANOVA; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, #p = 0.051. (G) Path length taken to the platform during spatial acquisition trials. Values represent the mean ± SE. Statistical significance was tested using repeated measures ANOVA. (H) Total distance travelled over the course of the probe trial. Values represent the mean ± SE. UNIANOVA; ***p < 0.001. (I) Number of times the animal crossed over the platform zone in the probe trial. Values represent the mean ± SE. UNIVANOVA; **p < 0.001. (J) Total distance traveled in each quadrant during the probe trial. Values represent the mean ± SE. UNIANOVA; ***p < 0.001. (K) Path length taken to the platform during reversal spatial acquisition trials. Values represent the mean ± SE. Statistical significance was tested using repeated measures ANOVA. (L) Total distance traveled over the course of the reversal probe trial. Values represent the mean ± SE. UNIANOVA; ***p < 0.001. (M) Number of times the animal crossed over the platform zone in the reversal probe trial. Values represent the mean ± SE. UNIVANOVA; **p < 0.01. (N) Total distance traveled in each quadrant during the reversal probe trial. Values represent the mean ± SE. UNIANOVA; *p < 0.05.

Fig. 3

Ube3a^mQ606E causes decreased stereotypic behavior, but no change in social or depressive phenotypes. (A) Number of marbles buried. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22: 15 male, 7 female. Values represent the mean ± range. Mann–Whitney U Test, 2-tailed (exact significance); ****p < 0.0001. (B) Percent of time spent floating (inactive, only minimal movements to keep head above water) in the forced swim test. Values represent the mean ± range. WT n = 27: 15 male, 12 female; Ube3a^mQ606E n = 26: 10 male, 16 female. Mann–Whitney U Test, 2-tailed (exact significance). (C) Tube Test: percentage of “dominant” bouts; i.e. in which an animal caused its opponent to back out of the tube. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 17: 10 male, 7 female. Pearson’s Chi-Square test, 2-sided (exact significance). (D–E) Social preference indexes in the 3-chamber social approach assay. (E) depicts preference for a novel, conspecific mouse over an empty cup. Values represent the mean ± range. (F) depicts preference for a novel conspecific over a familiar conspecific. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22: 15 male, 7 female. UNIANOVA. (F) Number of total entries into either chamber over the course of the trial period in the 3-chamber task. Values represent the mean ± SE. UNIANOVA; *p < 0.05.

Fig. 4

Ube3a^mQ606E mice display microcephaly with a postnatal onset. (A) Representative image of 3-month WT (left) and Ube3a^mQ606E (right) brains post gross dissection. Brain weights were 0.44 and 0.43 g, respectively. (B) Cross-sectional analysis of littermate WT and Ube3a^mQ606E postmortem brain weights. Values represent the mean ± SE in grams (g). P1: WT n = 24, mUbe3a^Q606E n = 13; P5: WT n = 11, mUbe3a^Q606E n = 10; P10: WT n = 11, Ube3a^mQ606E n = 12; P30: WT n = 27, mUbe3a^Q606E n = 28; 3 month: WT n = 13, mUbe3a^Q606E n = 9; 8 month: WT n = 12, Ube3a^mQ606E n = 14. Genotype effect from 2-way ANOVA with Tukey’s multiple post-hoc analysis for within-subject comparisons; *p < 0.05, ***p < 0.001, ****p < 0.0001. No significant sex-dependent effect was observed. (C) Body weights of animals used in behavioral studies. Values represent the mean ± SE. WT n = 17: 10 male, 7 female; Ube3a^mQ606E n = 22: 15 male, 7 female. Statistical significance was tested using an unpaired t-test. (D) Representative MR image of adult whole brain (left) and coronal segmentation. A – anterior, P –posterior, D – dorsal, V – ventral. (E) Quantification of whole brain volume in postnatal day 12 mice. WT n = 8; 5 male, 3 female; Ube3a^mQ606E n = 8; 3 male, 5 female; 2-way ANOVA; *p < 0.05. (F) Quantification of whole brain volume in 8.5 month mice. WT n = 10; 5 male, 5 female; Ube3a^mQ606E n = 10; 5 male, 5 female. 2-way ANOVA; ****p < 0.0001. (G) Representative image of brain segmentation as viewed from a coronal section (left) and example segmentation of cerebral cortex (right). (H–L) Quantification of absolute volumes (left) and volumes normalized to whole brain volume (right) of regions of interest in 8.5 month old mice: cerebral cortex (E), cerebellum (F), midbrain (G), amygdala (H), and lateral ventricles (I). *p < 0.05, 2-way ANOVA; **p < 0.01, ****p < 0.0001.

Fig. 5

Proteomic alterations and apoptosis in Ube3a^mQ606E mutants are lacking. (A) Schematic showing tissue collection and processing for proteomics. (B) Volcano plot depicting differentially regulated proteins for pooled samples. Significantly differentially regulated proteins are indicated in blue. P-values were adjusted with Benjamini–Hochberg multiple comparison correction (FDR = 0.05). All proteins with p-values, log2FC, and FDR are provided in Table S1. (C) Representative image of cleaved caspase 3 immunohistochemistry. Arrowheads point to CC3-positive neurons. (D) Total number of cleaved caspase 3 cells in whole brain as estimated by stereological results. Values represent the mean ± SE. 2-way ANOVA with Sidak’s multiple testing correction. (E) Total number of cleaved caspase 3 cells in whole brain normalized to whole brain volume as estimated by stereological results. Values represent the mean ± SE. 2-way ANOVA with Sidak’s multiple testing correction. (F) Lipid Peroxidation (MDA) Assay for amount of lipid peroxidation in adult (3.5 month) cortical lysate. Values represent the mean ± SE. 2-way ANOVA with Sidak’s multiple testing correction.

Fig. 6

The Ube3a^mQ606E mutations causes an overall loss of UBE3A activity in vivo. (A–D) Western blot analysis and quantification of proteins extracted from the cortex (A,B) and hippocampus (C,D) of age P21 WT and Ube3a^mQ606E animals. WT n = 4; Ube3a^mQ606E n = 5; *p = 0.03, **p = 0.01, ***p < 0.01, Unpaired t-test. (E–H) Western blot analysis and quantification of proteins extracted from the cortex (E,F) and hippocampus (G,H) of age P21 WT and Ube3a^m-/p+ animals. WT n = 3; Ube3a^m-/p+ n = 4; *p = 0.03, **p = 0.01, ***p < 0.01, Unpaired t-test.