We systematically investigate circRNA dysregulation in ASD cortex and constructed the corresponding ASD-associated circRNA–miRNA–mRNA regulatory networks. Like the previous observations for mRNAs8,78, miRNAs12 and circRNAs96 in ASD, our PCA result for circRNA expression profiles also revealed that frontal cortex and temporal cortex samples clustered together but cortex and cerebellar vermis samples grouped into separate clusters (Fig. 7). We also showed that the fold changes for the DE-circRNAs were concordant between the FC and TC, and were not biased by a small number of samples with removal of dup15q, low RIN, or high PMI (Fig. 10). While ASD and non-ASD samples can be grouped into two separate clusters based on the 60 identified DE-circRNAs by both PCA (Fig. 11) and hierarchical clustering (Fig. 12) analyses, but did not show separate clustering based on their host genes expression profiles. We thus provided a shared circRNA dysregulation signature among the majority of ASD samples.
In this study, we investigated genome-wide circRNA expression in ASD cortex samples and the corresponding ASD-associated circRNA–miRNA–mRNA axes. Regarding the identified ASD-associated circRNAs and the previously identified DE-miRNAs derived from the same cortex samples used in this study, we thus constructed 8,170 ASD-associated circRNA–miRNA–mRNA interactions (Fig. 17). Notably, within the 2,302 target genes of ASD-associated interactions, we observed significant enrichment for ASD risk genes, but not for genes implicated in monogenetic forms of other brain
disorders. Such as epilepsy, which is often a comorbidity of ASD. Regarding a recently published dataset of 102 high-confidence ASD genetic risk genes68, the 2,302 target genes also exhibited significant enrichment for the high-confidence ASD genetic risk genes (Fig. 39).
Figure 39. Enrichment of high-confidence ASD risk genes for the targets of the identified circRNA–
miRNA–mRNA interactions.
The high-confidence ASD genetic risk factors were derived from whole-exome sequencing on 35,584 ASD subjects. The P value is determined by two-tailed Fisher’s exact test.
Moreover, we observed that target genes of the ASD-associated circRNA–miRNA–
mRNA axes were significantly enriched in genes encoding inhibitory PSD proteins, but not in those encoding excitatory PSD ones (Fig. 19A). This result reflects the previous reports that there is an E/I neuronal imbalance in ASD and that inhibitory neurons are overproduced in ASD patient-derived organoids77,185,193. These results suggest that the identified ASD-associated circRNA–miRNA axes may serve as an alternative pathway to perturb key transcript levels and thereby contribute to ASD susceptibility.
On the basis of the hypothesis of circRNAs serve as microRNA sponge, we utilized differential expressed circRNA, miRNA and mRNA profiles of the ASD patients combined with experimentally validated miRNA–target interactions to reconstruct circRNA-associated network for the progression of ASD. However, the identified ASD-associated circRNA–miRNA–mRNA networks are constructed based on the correlation analysis and bioinformatics prediction, which need extensive experimental validation to investigate the interactions. As our experimental validation of circARID1A, we further characterized and functionally evaluated the circARID1A–miR-204-3p axis. Notable, we confirmed that the expressions of five SFARI genes (NLGN1, STAG1, HSD11B1, VIP and UBA6) were regulated by circARID1A via sponging miR-204-3p (Fig. 36).
The expression of both NLGN1 and STAG1 exhibited a significantly positive correlation with the circARID1A expression during ReNcell differentiation (Fig. 38). Recent evidence suggests that NLGN1 play an important role in a variety of activity-dependent response97 and memory formation98-100. Knockout of NLGN1 could cause increased repetitive behavior100, and with mutation in NLGN1 could cause abnormal social behavior in mouse models199. Alteration of NLGN1 expression in specific excitatory and inhibitory neuronal subpopulations can affect the dynamic processes of memory consolidation and strengthening98. Therefore, the identified circARID1A–miR-204-3p axis, which regulates NLGN1 expression, may provide a useful molecular mechanism of excitation and inhibition underlying long-term memory consolidation and strengthening for further developing potential therapeutic strategies to address these neuropsychiatric disorders, including ASD. Previous study has showed a higher proportion of inhibitory
neurons than excitatory neurons in ASD193. Also, ASD-associated genes were reported to be especially enriched in inhibitory neurons185. Afterward, we can differentiate ReNcell and then examine if knockdown or overexpression circARID1A alters the balance between inhibitory and excitatory neuron.
In addition, VIP and UBA6 have been demonstrated to play an essential regulatory role during rodent embryonic development101,102. VIP is known as a regulator of embryogenesis of neural tube closure; interference with VIP can result in permanent effects on adult social behavior102. It was shown that UBA6 brain-specific knockout mice exhibited social impairment and reduced vocalizations, representing a valid ASD mouse model103. As a regulator of multiple ASD-associated genes, the circARID1A–
miR-204-3p axis would be a valuable candidate for further ASD pathophysiology study.
In summary, our study provides a global view of circRNAs in ASD and non-ASD cortex. By integrating ASD candidate gene sets, miRNA and circRNA dysregulation data derived from the same ASD cortex samples, we have explored multiple lines of evidence for the functional role of ASD for circRNA dysregulation and the
corresponding circRNA–miRNA–mRNA networks. That may lead to improve ASD diagnosis and treatment in the future.