Replicable in vivo physiological and behavioral phenotypes of the Shank3B null mutant mouse model of autism

Abstract

Background: Autism spectrum disorder (ASD) is a clinically and biologically heterogeneous condition characterized by social, repetitive, and sensory behavioral abnormalities. No treatments are approved for the core diagnostic symptoms of ASD. To enable the earliest stages of therapeutic discovery and development for ASD, robust and reproducible behavioral phenotypes and biological markers are essential to establish in preclinical animal models. The goal of this study was to identify electroencephalographic (EEG) and behavioral phenotypes that are replicable between independent cohorts in a mouse model of ASD. The larger goal of our strategy is to empower the preclinical biomedical ASD research field by generating robust and reproducible behavioral and physiological phenotypes in animal models of ASD, for the characterization of mechanistic underpinnings of ASD-relevant phenotypes, and to ensure reliability for the discovery of novel therapeutics. Genetic disruption of the SHANK3 gene, a scaffolding protein involved in the stability of the postsynaptic density in excitatory synapses, is thought to be responsible for a relatively large number of cases of ASD. Therefore, we have thoroughly characterized the robustness of ASD-relevant behavioral phenotypes in two cohorts, and for the first time quantified translational EEG activity in Shank3B null mutant mice.

Methods: In vivo physiology and behavioral assays were conducted in two independently bred and tested full cohorts of Shank3B null mutant (Shank3B KO) and wildtype littermate control (WT) mice. EEG was recorded via wireless implanted telemeters for 7 days of baseline followed by 20 min of recording following pentylenetetrazol (PTZ) challenge. Behaviors relevant to the diagnostic and associated symptoms of ASD were tested on a battery of established behavioral tests. Assays were designed to reproduce and expand on the original behavioral characterization of Shank3B KO mice. Two or more corroborative tests were conducted within each behavioral domain, including social, repetitive, cognitive, anxiety-related, sensory, and motor categories of assays.

Results: Relative to WT mice, Shank3B KO mice displayed a dramatic resistance to PTZ seizure induction and an enhancement of gamma band oscillatory EEG activity indicative of enhanced inhibitory tone. These findings replicated in two separate cohorts. Behaviorally, Shank3B KO mice exhibited repetitive grooming, deficits in aspects of reciprocal social interactions and vocalizations, and reduced open field activity, as well as variable deficits in sensory responses, anxiety-related behaviors, learning and memory.

Conclusions: Robust animal models and quantitative, replicable biomarkers of neural dysfunction are needed to decrease risk and enable successful drug discovery and development for ASD and other neurodevelopmental disorders. Complementary to the replicated behavioral phenotypes of the Shank3B mutant mouse is the new identification of a robust, translational in vivo neurophysiological phenotype. Our findings provide strong evidence for robustness and replicability of key translational phenotypes in Shank3B mutant mice and support the usefulness of this mouse model of ASD for therapeutic discovery.

Keywords: Anxiety; Autism; Gamma oscillations; Pentylenetetrazol; Repetitive behavior; Shank3B; Social behavior.

Fig. 1

Representative EEG after PTZ injection. a Representative 10-min EEG from WT mouse shows three clinical epileptic stages after PTZ administration progressing from a healthy baseline (i), to developing epileptic spike trains (ii), and a run of frequent myoclonic seizures as indicated by markers (iii). b Representative 10-min EEG from Shank3B knockout exhibits epileptic spike-trains but has noticeably fewer myoclonic seizures than wildtypes

Fig. 2

Analysis of PTZ-induced seizures in cohort 1 (a-c) and in cohort 2 (d-f). a and d Incidence and latency to first myoclonus after PTZ injection. Kaplan–Meier survival curve is used to display percentage incidence of first myoclonus (y-axis) and its latency (x-axis), after PTZ injection (40 mg/kg). Curve comparison in first cohort shows that the incidence of myoclonus rate of (100%) in WT group was significantly higher than in the Shank3B KO group (67%) in a, which was replicable in the second cohort with only 50% knockouts experiencing a myoclonus in comparison to 100% of WT mice in d. Moreover, the Shank3B KO mice had a significantly longer latency to the first myoclonic seizure relative to WT controls. b and e Myoclonic seizure count. The frequency of PTZ-induced myoclonic seizures per 20 min of recording was significantly reduced in Shank3B KO mice relative to WT as replicated in two separate cohorts (b for cohort 1 and e for cohort 2) . c and f Epileptic spike count. The number of epileptic spikes on the EEG was also significantly lower in Shank3B KO mice relative to WT mice in both cohorts (c for cohort 1 and f for cohort 2). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Fig. 3

EEG gamma power. Spectral analysis of the 1-h pre-PTZ EEG shows higher power in the gamma frequency band (30–80 Hz). These results were replicable in both cohorts **p < 0.01. (a for cohort 1 and b for cohort 2)

Fig. 4

Locomotor activity at baseline. The total locomotor activity measured from 1 week baseline of male Shank3B knockout (KO) mice was significantly lower than wildtypes. The motor hypoactivity in KO as compared to WT was significant in both, cohort 1, *p < 0.05, and cohort 2, ***p < 0.001 (a for cohort 1 and b for cohort 2)

Fig. 5

Self-grooming. Shank3B null mutant mice (KO) spent more time in self-grooming during a 10-min session in a clean empty cage, as compared to wildtype littermates (WT), replicated in two independent cohorts. a Males, cohort 1, p < .001. b Females, cohort 1, *p < .02. c Combined males and females, cohort 1, *p < .001. d Males, cohort 2, p = .09. e Females, cohort 2, *p < .05, f Combined males and females, cohort 2, *p < .01

Fig. 6

Social interactions. Shank3B male null mutant mice (KO) interacting with an estrus WT female displayed deficits on some parameters of reciprocal social interactions, as compared to wildtype littermate males (WT), replicated in two independent cohorts. Time spent in nose-to-anogenital sniffing over the 5-min test session was less in Shank3B KO than WT (a cohort 1, *p < .05; e cohort 2, *p < .05). Time spent in nose-to-nose sniffing was not significant in cohort 1 (b), but reached significance in cohort 2 (*p < .01). Ultrasonic vocalizations emitted during the 5-min male-female interaction sessions were lower in KO than WT in both cohorts on total number of calls (c cohort 1, *p < .02; g cohort 2, *p < .02) and on calls per minute (d cohort 1, *p < .01; h cohort 2, *p < .02)