Imbalance of flight–freeze responses and their cellular correlates in the Nlgn3−/y rat model of autism

Abstract

Background Mutations in the postsynaptic transmembrane protein neuroligin-3 are highly correlative with autism spectrum disorders (ASDs) and intellectual disabilities (IDs). Fear learning is well studied in models of these disorders, however differences in fear response behaviours are often overlooked. We aim to examine fear behaviour and its cellular underpinnings in a rat model of ASD/ID lacking Nlgn3. Methods This study uses a range of behavioural tests to understand differences in fear response behaviour in Nlgn3−/y rats. Following this, we examined the physiological underpinnings of this in neurons of the periaqueductal grey (PAG), a midbrain area involved in flight-or-freeze responses. We used whole-cell patch-clamp recordings from ex vivo PAG slices, in addition to in vivo local-field potential recordings and electrical stimulation of the PAG in wildtype and Nlgn3−/y rats. We analysed behavioural data with two- and three-way ANOVAS and electrophysiological data with generalised linear mixed modelling (GLMM). Results We observed that, unlike the wildtype, Nlgn3−/y rats are more likely to response with flight rather than freezing in threatening situations. Electrophysiological findings were in agreement with these behavioural outcomes. We found in ex vivo slices from Nlgn3−/y rats that neurons in dorsal PAG (dPAG) showed intrinsic hyperexcitability compared to wildtype. Similarly, stimulating dPAG in vivo revealed that lower magnitudes sufficed to evoke flight behaviour in Nlgn3−/y than wildtype rats, indicating the functional impact of the increased cellular excitability. Limitations Our findings do not examine what specific cell type in the PAG is likely responsible for these phenotypes. Furthermore, we have focussed on phenotypes in young adult animals, whilst the human condition associated with NLGN3 mutations appears during the first few years of life. Conclusions We describe altered fear responses in Nlgn3−/y rats and provide evidence that this is the result of a circuit bias that predisposes flight over freeze responses. Additionally, we demonstrate the first link between PAG dysfunction and ASD/ID. This study provides new insight into potential pathophysiologies leading to anxiety disorders and changes to fear responses in individuals with ASD.

Altered motor learning and coordination in mouse models of autism spectrum disorder

“…Finally, while this review focuses on mouse models, there is a growing body of literature on genetic rat models of ASD (Berg et al, 2021;Harris et al, 2021;Dey and Chattarji, 2022), including models with mutations in ASD risk genes linked to enhanced rotarod performance discussed above, such as Fmr1 (Till et al, 2015;D'Elia et al, 2022;Schiavi et al, 2023) and Nlgn3 (Thomas et al, 2017;Anstey et al, 2022). However, as genetic accessibility of rat models is a recent development, assessment of rotarod performance in these models remains to be performed.…”

Cerebellar contributions to fear-based emotional processing: relevance to understanding the neural circuits involved in autism

“…Intriguingly, this activity pattern is reliant upon input from the cerebellum, routed through the medio-dorsal thalamus (Frontera et al, 2023). Thus, disruption of cerebellar output could potentially lead to fear response imbalance (as recently observed in Katsanevaki et al, 2020;Anstey et al, 2022) via modulation of 4 Hz oscillations in the prefrontal cortex. The studies described above indicate that cerebellar interactions with a multitude of brain regions, including those not classically considered involved in motor control, are important in the aetiology of autism (and other neurodevelopmental disorders).…”

“…Fearbehaviour execution centres, such as the PAG are also crucial for appropriate fear-response performance (Tovote et al, 2015). Recently, Anstey et al (2022) demonstrated hyperactivity within dorsal PAG and an imbalance of fear-related defensive responses in a rat model of autism. Such an imbalance in fear-related behavioural response may in turn provide unexpected feedback to prediction error systems, leading to cerebellar internal model maladaptation.…”

“…Impaired short-and long-term fear memory (Banerjee et al, 2016) Homer1 overexpression in BLA (Banerjee et al, 2016) ARID1B Successful fear memory acquisition and normal fear recall (Ellegood et al, 2021) Smaller cerebellum and larger hippocampus (Ellegood et al, 2021) PCDH10 Reduced freezing levels (Ferri et al, 2021) Decreased gamma oscillation in BLA (Ferri et al, 2021) NLGN3 Aberrant fear behavioural expression. Preference towards flight responses (Anstey et al, 2022) Hyperexcitability in dorsal PAG (Anstey et al, 2022) SYNGAP1 Impaired fear extinction (Katsanevaki et al, 2020) Decreased cortico-cortico coherence during non-rapid eye movement sleep (Buller-Peralta et al, 2022) FMR1 Impaired fear recall (Fernandes et al, 2021) Amygdala dysfunction (Fernandes et al, 2021). Impaired Purkinje cell long term depression and startle response (Koekkoek et al, 2005).…”

Dynamic Foraging Behavior Performance Is Not Affected byScn2aHaploinsufficiency

“…Towards this end, other animal models may be more appropriate for different traits. For example, rats can learn and perform more complex behavioral tasks than mice, potentially allowing for study of aspects of learning more closely related to those engaged by children (Wong et al, 2020;Fernandes et al, 2021;Harris et al, 2021;Anstey et al, 2022). Furthermore, other rodent species, like prairie voles, form highly complex social networks and enduring relationships that persist throughout life, allowing for study of social interaction and attachment phenotypes that are not innate to mouse (Young et al, 2002;McGraw and Young, 2010).…”