Introduction

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterized by specific social symptoms, restricted interests, stereotyped repetitive behaviors, and delayed language development [1]; its incidence is increasing worldwide [2]. Although recent genetic studies have identified high-confidence ASD-associated genetic variants [3], the molecular pathogenesis of ASD remains unclear [1].

Recent genetic studies have demonstrated that recurrent copy number variants (CNVs) are associated with a high risk for ASD [4]. The 3q29 microdeletion (3q29del) is typically an approximately 1.6 Mb deletion and contains 21 protein-coding genes [5]. Although the incidence of 3q29del is low (1 in 30,000–40,000 birth), 3q29del confers high risk for ASD both in males (odds ratio, 24.6) and females (odds ratio, 41.8), as well as for schizophrenia [6]. These exceptionally high odds ratios suggest that the cellular and molecular etiology of patients with 3q29del is an invaluable clue for elucidating the complex mechanisms of ASD. Recently, mouse models carrying a heterozygous deletion of the chromosomal region corresponding to the human 3q29 region (deficiency (Df)/+ mice) were generated and showed various neurodevelopmental and psychiatric conditions-related behavioral abnormalities, including ASD-related behavioral abnormalities [7, 8]. This highlights the relevance of Df/+ mice as an ASD model as well as various neurodevelopmental and psychiatric conditions with significant construct and face validity. Currently, the molecular pathogenesis of ASD-related phenotypes in Df/+ mice remains unclear.

Recent studies have suggested that genetic variants of the oxytocin (OXT) receptor are associated with social behavior in humans and that impaired OXT signaling is linked to ASD [9]. In mice, targeted disruption of the Oxt or OXT receptor (Oxtr) gene impairs social behavior, suggesting that the OXT system plays an important role in social behavior across species [9]. Thus, focusing on the OXT system may provide clues to shed light on the molecular mechanisms regulating social behavior in mice. Currently, limited studies have shown a link between the OXT system and behavioral phenotypes in mouse models of ASD [for example see 10, 11]. In this study, to elucidate the molecular mechanisms behind impaired social behavior in Df/+ mice, we investigated the possible involvement of OXT signaling in impaired social behavior in Df/+ mice.

Results

To investigate whether oxytocin administration ameliorates impaired social behavior in Df/+ mice, we carried out a reciprocal social interaction test with the administration of 200 µg/kg of OXT (Fig. 1a). As we and other groups previously reported [7, 8], Df/+ mice treated with saline showed decreased social interaction time compared to that of WT littermates (Fig. 1b). We revealed that intraperitoneal administration of OXT improved impaired social interaction in Df/+ mice to a level similar to that in WT littermates (Fig. 1b). However, it did not have significant effect on the social interaction time in WT littermates at the concentration used (Fig. 1b).

Fig. 1
figure 1

Administration of OXT improved impaired social interaction in Df/+ mice. a Timeline for the reciprocal social interaction test with OXT treatment. The test was carried out 30 min after administration of OXT (200 µg/kg) or saline. OXT, Oxytocin. b Social interaction time in the reciprocal social interaction test (each n = 12). The time spent sniffing, following, allo-grooming, and push-crawling of test mice toward the intruder was measured as social interaction time (two-way ANOVA for genotype, F1,44 = 8.84, P = 0.0048; for treatment, F1,44 = 4.61, P = 0.037; genotype × treatment interaction, F1,44 = 3.23, P = 0.079). WT, wild-type. c (upper) Schematic illustration of the PVN (green) and (lower) representative images of OXT-Neurophysin1 immunostaining (green) and hoechst33258 staining (blue) in the PVN of adult Df/+ mice and WT littermates. Scale bar, 100 μm. d Quantification of the number of OXT-positive cells in the PVN (each n = 4 mice). e ELISA quantification of the OXT concentration in the cerebral cortex extract (each n = 8 mice). f Quantitative real-time RT-PCR analysis of the expression levels of Oxtr and Avpr1a mRNA (each n = 4 mice). Oxtr, OXT receptor; Avpr1a, vasopressin receptor 1 A. Data are represented as the mean ± SEM. Statistical significance was analyzed by a two-way ANOVA, followed by Bonferroni Dunn post hoc tests (b) and Student’s t-test (d–f). *P < 0.05, ****P < 0.0001, n,s, not significant

Full size image

To elucidate the molecular mechanism underlying the effect of OXT in Df/+ mice, we examined the abnormalities of OXT-related signaling pathways in Df/+ mice. We assessed the number of OXT neurons in the paraventricular nucleus (PVN), where OXT neurons are densely located [9], and found that the number of OXT-positive cells was significantly lower in Df/+ mice than in WT littermates (Fig. 1c and d). We assumed that the decreased number of OXT-positive cells in Df/+ mice may result in decreased OXT concentration in the cerebral cortex, where OXT signaling plays an important role in the regulation of social behavior. To probe this possibility, we quantified levels of OXT peptide in the cerebral cortex and found decreased levels of OXT peptide in the cerebral cortex of Df/+ mice as compared to that in WT littermates (Fig. 1e). Finally, we measured the expression of the Oxtr and vasopressin receptor 1 A (Avpr1a) mRNAs, which encodes for another receptor for OXT, in the cerebral cortex by quantitative real-time RT-PCR. Results showed that the expression levels of Oxtr and Avpr1a mRNAs were not significantly changed in Df/+ mice as compared to those in WT littermates (Fig. 1f).

Discussion

Df/+ mice show social dysfunction, one of the major ASD-related behavioral phenotypes. However, the molecular pathophysiological mechanisms underlying these phenotypes remain largely unclear. In this study, we provide evidence suggesting that impaired OXT system is associated with social dysfunction in Df/+ mice. Specifically, our current results offer a new avenue for investigating the molecular mechanisms underlying the effect of 3q29del on social symptoms. Considering that the 3q29del confers an extremely high risk for ASD as well as various neurodevelopmental and psychiatric conditions [6], our study may provide important insights into the molecular pathophysiological basis of ASD as well as various neurodevelopmental and psychiatric conditions.

The molecular links between the OXT system and the 21 protein-coding genes in the 3q29 deleted region remain unknown. Among the 21 gene products, PAK2, which regulates actin cytoskeletal dynamics and dendritic spine morphology, is suggested to be mainly associated with ASD [10]. However, PAK2 mRNA expression is decreased by OXT treatment in the hippocampus [11], suggesting that PAK2 might not be involved in the effect of OXT in Df/+ mice. In addition to PAK2, a recent study suggests that four genes, including CEP19, SENP5, UBXN7, and WDR53, out of the 21 protein-coding genes in the 3q29-deleted region may be primary ASD-associated genes [12]. Unraveling the precise molecular links between impaired OXT system and dysfunction of these possible ASD-associated gene products will help elucidate the molecular pathophysiological mechanisms of social dysfunction in Df/+ mice.

Several ASD mouse models, such as POGZ-Q1038R, Nlgn3−/−, and Cntnap2−/− mice, have an impaired OXT system [13,14,15]. This suggests that, despite the etiological heterogeneity of ASD, an impaired OXT system may underpin the social symptoms in a subset of patients. Considering that the social symptoms in the patients with 3q29del may be distinct from those in idiopathic ASD patients [6] and that the social symptoms are common features of various neurodevelopmental and psychiatric conditions, further studies on the possible molecular links between an impaired OXT system and social symptoms may help advance a molecular mechanism-based stratification of groups of patient with various neurodevelopmental and psychiatric conditions as well as the development of molecular pathogenesis-based therapeutic interventions, extending well beyond OXT treatment.