Experimental validation of circARID1A interacting with miR-204-3p

As previously mentioned, circRNAs could function as miRNAs sponge. To investigate the interaction between dysregulated circRNA and miRNA in ASD, we predict DE-miRNA binding sites on DE-circRNAs that have been identified in the same ASD samples used in this study. Then, we selected a circARID1A that was back-spliced from exon 2-4 of ARID1A (chr1: 26,729,651–26,732,792) (Fig. 20A). CircARID1A was predicted to have the greatest number of target sites of DE-miRNAs (miR-204-3p) using the software RNA22¹⁷² (version 2.0) with default parameters (Fig. 20B). The circARID1A was significantly upregulated in ASD samples (Fig. 20C).

A

B C

Figure 20. CircARID1A serves as a sponge for miR-204-3p.

(A) Numbers of the predicted binding sites of the previously identified miRNAs in ASD on the DE-circRNA and DE-circRNA modules. Only the DE-circRNA–miRNA regulatory axes with a significantly negative correlation are shown. Upregulated and downregulated miRNAs/circRNAs are highlighted in red and green, respectively. The red square highlights circARID1A that was predicted to have seven target sites of miR-204-3p. (B) Schematic illustration seven predicted target sites of miR-204-3p on circARID1A. (C) Comparison of the expression of circARID1A in 134 cortex samples.

Here, the NCL junction of circARID1A has not yet been experimentally confirmed previously. The roles of both circARID1A and miR-204-3p and their regulatory interaction in CNS have not yet been investigated. We first confirmed the BSJ of circARID1A by divergent primers, which designed around the NCL junction of circARID1A. The primers are provided in Table 4. Since comparisons of different reverse transcriptase RTase products have been demonstrated to effectively detect RT-based artificially NCL junctions^30,89-92. We performed RT-PCR using MMLV- and AMV-derived RTase in parallel experiments (Fig. 21A), and PCR amplification of NCL junction validated by Sanger sequencing (Fig. 21B). Our result demonstrated that the NCL junction of circARID1A was RTase-independent, supporting that the NCL junction was unlikely to be generated by an RT-based artifact.

A B

Figure 21. Validation of the back-spliced junction of circARID1A through RT-PCR and Sanger sequencing.

(A) Comparisons of two different RTase products of circARID1A in TC/FC samples and four types of neuronal cell lines (ReN, NHA, SH-SY5Y, and U118). (B) Backsplicing site of circARID1A was confirmed by Sanger sequencing in the TC and ReNcell.

Subsequently, we treated total RNA from the examined cell lines/tissues with RNase R.

RNase R was used to confirm the existence of circRNAs. As expected, circARID1A was resistant to RNase R treatment, while linear ARID1A was significantly reduced in cell treated with RNase R, supporting the existence of the circARID1A (Fig. 22).

Figure 22. Resistance of circARID1A, ARID1A and GAPDH after the RNase R treatment.

The percentage of RNA remaining was before and after RNase R treatment in the indicated tissues/cell lines determined by qRT-PCR. GADPH was served as negative controls.

Exon2 Exon3 Exon4 ARID1A

MMLV AMV

0.0 0.3 0.6 0.9 1.2 1.5

TC FC ReN NHA SH-SY5Y U118

Relative expression (RNase R / Mock)

circARID1A ARID1A GAPDH

Intriguingly, we observed that circARID1A was commonly expressed across the brains of vertebrate species from primates to chicken, indicating the evolutionary significance of circARID1A (Fig. 23).

Figure 23. Experimental examination of the evolutionary conservation of circARID1A across vertebrate brains.

Comparison of MMLV- and AMV-derived-RTase products and the corresponding sequence chromatograms for the circARID1A event in the brains of the indicated six species.

We examined the expression profiles of circARID1A and its corresponding colinear mRNA counterpart in various human tissues. The result showed these two isoforms exhibited different expression patterns. CircARID1A was particularly enriched in brain, whereas its corresponding mRNA counterpart was not (Fig. 24A). Importantly, regarding the relative expression of these two isoforms in the brain, circARID1A was significantly more abundant than its colinear form; in contrast, circARID1A was expressed at a relatively low level as compared to its colinear form in the non-brain tissues (Fig. 24B)

A B

Figure 24. The relative expression of circARID1A and its corresponding co-linear mRNA counterpart in 10 normal human tissues.

(A) The expression levels of brain are used to normalize the relative expression values of the other tissues.

(B) The relative expression of circARID1A and ARID1A in 10 normal human tissues.

We also demonstrated that circARID1A was widely expressed in 24 human brain regions by Human Brain cDNA Array (OriGene). These results thus suggested the biological importance of circARID1A in human brain (Fig. 25).

Figure 25. RT-PCR analysis of circARID1A expression in 24 human brain regions.

To function as a miRNA sponge, circRNAs not only need to harbor miRNA-binding sites but also be expressed at high levels in the cytosol. To test if circARID1A plays a

0.0 Relative expression circARID1A / ARDI1A)

*

* *

** *

*** *** **

regulatory role of miR-204-3p activities, we confirmed that circARID1A was indeed predominantly expressed in the cytoplasm in both ReN and NHA cells (Fig. 26).

Figure 26. Subcellular localization of circARID1A and ARID1A.

The expression levels in nucleus are used to normalize the relative expression values in cytoplasm.

GAPDH and U6 snRNA are examined as a cytosol marker and a nucleus marker, respectively.

To investigate the downstream effect of circARID1A, knockdown and overexpression of circARID1A in different types of neuronal cell lines were established. We found that circARID1A knockdown and overexpression did not significantly affect the expression of its corresponding colinear mRNA counterpart (Fig. 27).

Figure 27. qRT-PCR analyses of the expression of circARID1A and ARID1A after circARID1A

knockdown or overexpression in various neuronal cell lines.

NC, negative control. KD, knockdown. OE, overexpression.

0

Notably, miR-204-3p expression significantly increased and decreased after knockdown and overexpression of circARID1A, respectively (Fig. 28). We hypothesized that circARID1A may serve as a sponge for miR-204-3p.

Figure 28. qRT-PCR analyses of the correlations between the expression of circARID1A and miR-204-3p after circARID1A knockdown or overexpression in various human neuronal cell lines.

To verify our prediction, luciferase reporter assay was performed to confirm the relationship between circARID1A and miR-204-3p. Various human neuronal cell lines were co-transfected the Gluc-circARID1A reporter, together with either miR-204-3p mimic or scramble mimic. Dual-luciferase activity of GLuc-circARID1A showed that miR-204-3p significantly reduced the luciferase activity as compared with scramble control (Fig. 29).

Figure 29. Luciferase reporter assay for the interaction between circARID1A and miR-204-3p.

To further examine the specificity of the binding between circARID1A and miR-204-3p, We also mutated miR-204-3p binding sites from the luciferase reporter (Fig. 30A) and found that transfection of miR-204-3p had no significant effect on the luciferase activity of the GLuc-circARID1A mutated reporter as compared to the scrambled negative control (Fig. 30B). These results thus confirmed the regulatory role of circARID1A as a miR-204-3p sponge.

A B

Figure 30. Site-directed mutagenesis of the potential binding sites of miR-204-3p in GLuc-circARID1A reporter construct.

(A) Schematic illustration of mutated putative miR-204-3p binding sites in luciferase reporter vectors labeled by red. (B) Luciferase assays of the NHA cells co-transfected with mutated reporter mimics.

In addition, we also showed that miR-204-3p overexpression did not significantly affect circARID1A expression (Fig. 31). These results thus confirmed the negative regulation between circARID1A and miR-204-3p and the regulatory role of circARID1A in the miR-204-3p sponge.

Figure 31. The expression level of circARID1A after overexpression of miR-204-3p in ReN and NHA cells.

P values are determined using two-tailed t-test.

0.0 Relative Luciferase activity ^{P = 0.257}

0.0

3.6 Regulation of ASD risk genes via the identified circRNA–miRNA