Profiling of ASD-associated circRNA–miRNA–mRNA regulatory axes

Figure 15. Module preservation Z-summary statistics of 14 modules.

Module preservation defined in ASD samples only in CTL samples (A; left), CTL samples only in ASD samples (A; right), TC samples only in FC samples (B; left), or FC samples only in TC samples (B; right).

The horizontal lines indicate each module with a Zsummary threshold for strong evidence of conservation (Zsummary > 10), week to moderate evidence of conservation (2< Zsummary <10), and no evidence conservation (Zsummary < 2).

3.4 Profiling of ASD-associated circRNA–miRNA–mRNA regulatory axes We have identified 60 DE-circRNAs and three DE-modules in ASD cortex. To further identify the corresponding ASD-associated miRNA–mRNA regulatory interactions that were potentially regulated by ASD-associated circRNAs, we extracted 58 DE-miRNAs in ASD cortex from the study of Wu et al.¹². These ASD-affected miRNAs were derived from 95 human cortex samples, of which 73 samples overlapped with the samples examined in this study. Since circRNAs often act as a miRNA sponge, we directly

assessed the role of circRNA dysregulation in alterations of the ASD-affected miRNAs.

Thus, we searched for the target sites of 41 upregulated miRNAs in 38 downregulated circRNAs and 298 circRNAs defined in two downregulated circRNA modules; and searched for the target sites of 17 downregulated miRNAs in 22 upregulated circRNAs and 21 circRNAs defined in an upregulated circRNA module.

After that, we determined the ASD-associated circRNA–miRNA–mRNA interactions by integrating the identified interactions for circRNA–miRNA and miRNA–mRNA pairs according to the common miRNA target sites of the circRNAs and mRNAs. The ASD-associated circRNA–miRNA–mRNA interactions were constructed by integrating 808 circRNA–miRNA axes with 36,512 miRNA–mRNA axes, we therefore determined 529,509 circRNA–miRNA–mRNA axes according to the common miRNA served as linker between the DE-circRNAs and DE-mRNAs. The circRNA–miRNA–mRNA axes were provided at GitHub (https://github.com/TreesLab/circRNA_ASD).

We then calculated the correlations between circRNA and miRNA expression, between miRNA and mRNA expression, and between circRNA and mRNA expression based on the same set of cortex samples (i.e., 73 samples). As shown in Figure 16, the circRNA–

miRNA–mRNA axes were simultaneously satisfied the criteria.

Figure 16. Schematic diagram representing the criteria for the identified ASD-associated circRNA–

miRNA–mRNA axes.

Therefore, a total of 8,170 ASD-associated circRNA–miRNA–mRNA axes were determined, including 356 upregulated (Fig. 17A) and 7,814 downregulated (Fig. 17B) circRNA-involved axes. The ASD-affected circRNA may provide an upstream regulation for the ASD-associated miRNA–mRNA pathways and thereby potentially contribute to ASD susceptibility.

A B

Figure 17. The 8170 ASD-associated circRNA–miRNA–mRNA interactions.

CircRNAs and miRNAs are indicated by blue circle and pink triangle, respectively. The interactions were plotted by Cytoscape software.

We provided 8,170 circRNA–miRNA–mRNA regulatory interactions that potentially contributed to ASD susceptibility for future study. The network contained 46 circRNAs, 30 miRNAs and 2,302 target genes. According to the target genes previously implicated in ASD or the experimental evidence of miRNA–mRNA binding, we further classified the identified circRNA–miRNA–mRNA axes into four categories as follows:

Category 1: The target genes have been previously reported to be ASD risk genes (i.e., SFARI or AutismKB genes).

Category 2: The target genes were reported to be DE-genes in ASD¹⁶ based on the samples overlapped with the samples examined in this study. The interactions should be either upregulated circRNA – downregulated miRNA – upregulated mRNA or downregulated circRNA – upregulated miRNA – downregulated mRNA interactions.

Category 3: The miRNA–mRNA interactions has been experimentally validated.

Category 4: Other.

The 780 axes in category 1, and in which the 188 ASD risk genes may play an important regulatory role in ASD brain. The target genes of the categories 2-4 interactions, which were identified to regulated by upstream ASD-affected circRNA–

miRNA axes, may be valuable candidates for further studies in idiopathic ASD (Fig. 18).

Figure 18. The four categories of circRNA-involved ASD-associated circRNA–miRNA–mRNA interactions.

Venn diagrams represent the overlap between the four categories of interactions, on which the numbers between brackets show the numbers of target genes. The numbers of the identified circRNA–miRNA–

mRNA interactions are shown in parentheses.

To explore the relationship between ASD and the 2,302 target genes, we first used the ToppFunn module of ToppGene Suite software¹⁸⁶ to assessed enrichment of target genes for Human Phenotype Ontology terms. We found that the target genes were significantly enriched in the phenotype ontology terms of aplasia/hypoplasia involving the central nervous system (CNS) and abnormality of forebrain, cerebrum, central mortor, and skull size, reflecting the brain morphometry differences between ASD patients and healthy individuals (Fig. 19A).

Regarding the 2,302 target genes, we further examined enrichment for ASD risk genes

67 84

databases, which have been previously implicated in ASD through genetic syndromes, candidate gene studies, common variant association, and structural variation. The target genes showed significant enrichment for both SFARI and AutismKB genes (particularly for top SFARI genes with score = 1- 3 and syndromic), but not for genes implicated in monogenetic forms of schizophrenia, Alzheimer, intellectual disability and other brain disorders (Fig. 19B). These results suggest that targets of the identified circRNA–

miRNA–mRNA interactions are enriched for genes causally connected with autism, but less so far genes connected with other brain disorders. In addition, we found that these target genes were significantly enriched in genes encoding inhibitory postsynaptic density (PSD) proteins, but not in those encoding excitatory PSD ones (Fig. 19B). This result reflects a previous observation that ASD-derived organoids exhibit overproduction of inhibitory neurons^185,193.

A B

Figure 19. Enrichment analyses of phenotype ontology (A) and 14 group of gene list (B) among the target genes of the identified ASD-associated circRNA–miRNA–mRNA interactions.

The P values are determined by two-tailed Fisher’s exact test. The red dashed lines represent the adjusted p-value (FDR) < 0.05. The enrichment odds ratios with FDR < 0.05 are provided in parentheses.