Heterozygous ambra1 deficiency in mice: a genetic trait with autism-like behavior restricted to the female gender

Abstract

Autism-spectrum disorders (ASD) are heterogeneous, highly heritable neurodevelopmental conditions affecting around 0.5% of the population across cultures, with a male/female ratio of approximately 4:1. Phenotypically, ASD are characterized by social interaction and communication deficits, restricted interests, repetitive behaviors, and reduced cognitive flexibility. Identified causes converge at the level of the synapse, ranging from mutation of synaptic genes to quantitative alterations in synaptic protein expression, e.g., through compromised transcriptional or translational control. We wondered whether reduced turnover and degradation of synapses, due to deregulated autophagy, would lead to similar phenotypical consequences. Ambra1, strongly expressed in cortex, hippocampus, and striatum, is a positive regulator of Beclin1, a principal player in autophagosome formation. While homozygosity of the Ambra1 null mutation causes embryonic lethality, heterozygous mice with reduced Ambra1 expression are viable, reproduce normally, and lack any immediately obvious phenotype. Surprisingly, comprehensive behavioral characterization of these mice revealed an autism-like phenotype in Ambra1 (+/-) females only, including compromised communication and social interactions, a tendency of enhanced stereotypies/repetitive behaviors, and impaired cognitive flexibility. Reduced ultrasound communication was found in adults as well as pups, which achieved otherwise normal neurodevelopmental milestones. These features were all absent in male Ambra1 (+/-) mice. As a first hint explaining this gender difference, we found a much stronger reduction of Ambra1 protein in the cortex of Ambra1 (+/-) females compared to males. To conclude, Ambra1 deficiency can induce an autism-like phenotype. The restriction to the female gender of autism-generation by a defined genetic trait is unique thus far and warrants further investigation.

Keywords: Ambra1; autism composite score; autophagy; cognitive rigidity; heterozygous null mutant mice; repetitive behavior; social interaction; ultrasound communication.

Figure 1

Normal sensory function is found in male and female Ambra1^+/− versus Ambra1^+/+ mice. The upper row presents results for female, the lower row for male mice. (A,B) Visual cliff test; (C,D) hearing curve; (E,F) buried food finding; and (G,H) hot plate test. Mean ± SEM presented; respective sample sizes are indicated in the figures.

Figure 2

Spontaneous and novelty-induced activity of Ambra1^+/− versus Ambra1^+/+ mice are widely comparable. The upper row presents results for female, the lower row for male mice. (A–F) LABORAS standard readouts; (G–L) open field; and (M,N) hole board. Female Ambra1^+/−appear slightly more active. Mean ± SEM presented; respective sample sizes are indicated in the figures.

Figure 3

Motor performance and sensorimotor gating of male and female Ambra1^+/− versus Ambra1^+/+ mice are normal. The upper row presents results for female, the lower row for male mice. (A,B) Body weight; (C,D) motor performance and motor learning; (E,F) startle response to a 120 db stimulus; and (G,H) pre-pulse inhibition. Mean ± SEM presented; respective sample sizes are indicated in the figures.

Figure 4

Anxiety- and depression-relevant tests in Ambra1^+/− versus Ambra1^+/+ mice show comparable results. The upper row presents results for female, the lower row for male mice. (A–H) Elevated plus-maze readouts; (I–L) sucrose preference and forced swim test as readouts of depression. Mean ± SEM presented; respective sample sizes are indicated in the figures.

Figure 5

Cognitive testing of male and female Ambra1^+/− versus Ambra1^+/+ mice reveals impaired cognitive flexibility in female Ambra1^+/−mice. The upper row presents results for female, the lower row for male mice. (A,B) Novel object recognition, no-delay task. (C,D) Spatial alternation in Y-maze. (E,F) Spatial learning in the Morris water-maze; hidden platform acquisition curve presented. Figure inserts represent time spent searching for the platform in the target quadrant during the probe trial. (G,H) Reversal learning in the Morris water-maze. Mean ± SEM presented; respective sample sizes are indicated in the figures.

Figure 6

Comparison of autism-readouts in Ambra1^+/− versus Ambra1^+/+ mice of both genders discloses an autism-like phenotype in female Ambra1^+/−mice. The upper row presents results for female, the lower row for male mice. (A,B) Communication: ultrasound vocalization; (C,D) social competence: nesting score; (E,F) social interaction in pairs; (G,H) social preference tested in the tripartite chamber; (I,J) social memory tested in the tripartite chamber; (K,L) stereotypies/repetitive behaviors: circling, measured in LABORAS; and (M,N) stereotypies/repetitive compulsive behaviors: marble burying. Mean ± SEM presented; respective sample sizes are indicated in the figures.

Figure 7

Composition and intercorrelations of the autism severity score in Ambra1^+/− versus Ambra1^+/+ mice of both genders are presented. The upper row presents results for female, the lower row for male mice. (A,B) Presentation of z-standardized autism-relevant readouts to be integrated into the composite score; mean ± SEM presented; respective sample sizes are indicated in the figures. (C,D) Intercorrelation pattern of the autism-relevant readouts including Cronbach’s alpha as a quality measure for internal consistency of the autism severity composite score.

Figure 8

Autism composite score results for Ambra1^+/− versus Ambra1^+/+ mice of both genders underline the female autism-like phenotype in Ambra1^+/−. The upper row presents results for female, the lower row for male mice. (A,B) Highly significant genotype-dependent score difference in female but not male mice; mean ± SEM presented. (C,D) Frequency distribution of autism composite score bins dependent on genotype; (E,F) Composite score presentation of all individual animals reveals a clear discrimination between genotypes in female but not male mice.

Figure 9

Neonatal development and pup vocalization in Ambra1^+/− versus Ambra1^+/+ mice uncover early ultrasound communication deficits in females. The upper row presents results for female, the lower row for male mice. (A,B) Body weight over time is comparable between genotypes; (C,D) no genotype differences in physical development, neurodevelopment (PR, placing response; SRR, surface righting reflex; CA, cliff avoidance; NGR, negative geotaxis reflex; TS, tactile startle; ET, ear twitch; ARR, air righting reflex), and neuromotor coordination (OF, open field traversal; WS, wire suspension). Overall test results are expressed as days to reach criterion. (E–H) Ultrasound vocalization of pups briefly separated from their mother on postnatal day 8/9. Mean ± SEM presented; respective sample sizes are indicated in the figures.

Figure 10

Quantification of Ambra1 mRNA and protein in cortex of male and female Ambra1^+/− versus Ambra1^+/+ mice delivers a first hint to potentially explain the behavioral gender difference. (A) mRNA; (B) sample western blot; (C) western blot quantification; (D) note the significantly higher relative reduction (delta) from the respective wildtype control (Ambra1^+/+) of Ambra1 protein in cortex of female as compared to male Ambra1^+/− mice. M, male; F, female; Mean ± SEM presented; respective sample sizes are indicated in the figures.