*2.5. Highly Expressed Exo-miRNA Targets Are Efficiently Downregulated and Endo-miRNA Targets with Low Expression Levels Are Efficiently Upregulated by the Transfection of the Exo-miRNA Duplex*

Among exo-miRNA targets, the genes with high expression levels in the normal conditions were more efficiently downregulated by the transfection of exo-miRNAs compared to those with low expression levels (Figure 6A). The results showed good agreement with the previous study [31]. In contrast, the endo-miRNA targets with low expression levels in the normal conditions were uncovered to be intensively upregulated by the exo-miRNA transfection, and those with high expression levels were not upregulated efficiently (Figure 6B). So, we investigated the number of endo-miRNA target sites in gene with different expression levels. As a result, a large number of target sites were found in the mRNAs with low expression levels, but a small number of the sites were detected in the mRNAs with high expression levels (Figure 6C), suggesting that many genes with low expression levels might be downregulated by endo-miRNAs, but those with high expression levels are not repressed by endo-miRNAs in the normal condition. Thus, the changes of expression levels of the exo-miRNA target genes might be also suitably regulated by the endo-miRNAs pre-situated in the RISCs.

**Figure 6.** Mean fold changes of target genes of exo-miRNAs and endo-miRNAs, according to the expression levels. Differential fold changes (log2) of target gene expression levels of exo-miRNAs (**A**) and endo-miRNAs (**B**) and the number of the top 20 endo-miRNA target sites (**C**), according to the expression levels of exo-miRNA targets. The averaged numbers of exo-miRNA target sites are 130, 217, 237 and 215 (**A**), and those of endo-miRNA target sites are 92, 164, 176 and 202 (**B**) in the 3' UTRs of genes with differential fold changes of 0~7, 7~9, 9~11 and 11~.

#### **3. Experimental Section**

#### *3.1. Cell Culture and miRNA Synthesis*

Human HeLa cells were cultured and used in reporter assays and microarray analyses. Cells were cultured at 37 °C in Dulbecco's modified Eagle's Medium (Invitrogen, Carlsbad, NM, USA), supplemented with 10% heat-inactivated fetal bovine serum (Sigma, St. Louis, MO, USA). They were plated on 24-well culture plates (1 × 105 cells/mL/well) 24 h prior to transfection. Transfection was carried out using Lipofectamine 2000 (Invitrogen, Carlsbad, NM, USA). Each RNA strand of the miRNA duplex was chemically synthesized (Sigma, St. Louis, MO, USA) and annealed to form the duplex structures the same as those shown in miRBase [28]. The sequences of the synthetic miRNAs (let-7b, miR-1, miR-21, miR-22, miR-28, miR-30c-1, miR-186, miR-199b, miR-200b, miR-330, miR-335, miR-346, miR-466, miR-574 and miR-3126) are listed in Table S1.

#### *3.2. Construction of Luciferase Reporters*

All of the reporter plasmids constructed were derivatives of psiCHECK-1 (Promega, Fitchburg, WI, USA). Oligonucleotides with target sequences completely matched to each miRNA strand (cm-target) were chemically synthesized with cohesive *Xho*I/*Eco*RI ends (Table S2). They were then inserted into the corresponding restriction sites of psiCHECK-1 to generate miRNA cm-targets (miR-200b-3p target, miR-7b-5p target, miR-21-5p target and miR-330-5p target). Each of the inserted targets was expressed as part of the 3' UTR region of *Renilla* luciferase mRNA in transfected cells.

HeLa cells growing in 24-well plates were transfected simultaneously with miRNA target (100 ng), pGL3-Control (Promega, 0.5 µg) and miRNA (50 nM). The cells were harvested 24 h post-transfection and the relative luciferase activity (*Renilla* luc activity/firefly luc activity) was determined using a Dual-Luciferase Reporter Assay System (Promega, Fitchburg, WI, USA). The pGL3-Control encoding firefly luciferase served as a control for the calculation of relative luciferase activity for miRNAs.

#### *3.3. Microarray Analysis*

HeLa cells (1 × 105 cells/mL) were transfected with 50 nM of each of 15 miRNA duplexes. At 24 h post-transfection, total RNA was purified using an RNeasy Kit (Qiagen, Hilden, Germany). The steps were repeated four times and RNA quality assessed using a NanoDrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, USA) and a Bioanalyzer (Agilent, Santa Clara, CA, USA). RNAs recovered independently were mixed equally for cDNA synthesis using an Agilent One Color Spike Mix Kit (Agilent, Santa Clara, CA, USA). Cy3-labeled cRNA was synthesized using a Quick Amp Labeling Kit (Agilent, Santa Clara, CA, USA) and was hybridized to an Agilent Whole Human Genome Microarray (4 × 44 K multi-pack format), according to the manufacturer's protocol. RNA from mock-transfected cells treated with transfection reagent in the absence of miRNA was used as a control. Transcript expression values were calculated using Microarray Suite 5.0 (MAS5: Affymetrix, Santa Clara, CA, USA) [32] with quantile normalization [33]. To identify transcripts whose expression was upregulated or downregulated, the cumulative distribution of expression changes for transcripts containing the site was compared with that for transcripts with no canonical site. NCBI's Reference Sequence (RefSeq) was used to identify mRNAs with sequences complementary to the seed regions of the transfected miRNAs. Data are presented as an MA plot (M = intensity ratio, A = average intensity) and a cumulative frequency distribution. Changes in expression are shown as fold changes (log2).

#### *3.4. Quantitative RT-PCR*

Total RNA was reverse-transcribed using a Transcriptor High Fidelity cDNA Synthesis kit (Roche, Basel, Switzerland). The resultant cDNA samples were incubated with FastStart Universal SYBR Green Master (Roche, Basel, Switzerland) at 95 °C for 10 min, followed by PCR amplification. PCR product levels were monitored using an ABI PRISM 7000 sequence detection system and analyzed with ABI PRISM 7000 SDS software (Applied Biosystems). The expression of each target gene was first normalized to that of β-actin and then to the mock-transfection control. The primer sets used are listed in Table S3.

#### **4. Conclusions**

In this study, we quantified the changes in expression levels of endo-miRNA target genes resulting from the transfection of exo-miRNA duplexes. The expression levels of endo-miRNA target genes with seed-complementary sequences were increased by the transfection of exo-miRNA duplexes, while exo-miRNA target gene expression was reduced (Figures 1–3 and Figures S1–S3). These results suggest that exo-miRNA duplex transfected into the cells may compete with endo-miRNAs for the RISC, which may be saturated with endo-miRNAs under normal conditions in HeLa cells.

In miRNA-mediated gene silencing, the structure of the RNA-protein complex is known to be altered [23]. The miRNA duplex in RISC loading complex is unwound, yielding single-stranded RNA, which is loaded onto the RISC and recognizes target mRNAs through base-pairing in the seed region [4–6]. Our results indicate that the expression of the target genes of a given endo-miRNA differed according to the exo-miRNA duplex that was transfected (Figures 2 and 7A), whereas with a given exo-miRNA duplex, the fold changes in the target gene expression of the endo-miRNAs examined were mostly equivalent, except for a limited types of exo-miRNAs (miR-1, miR-28, miR-199b and miR-335), despite differences in the structures and sequences of the endo-miRNA duplexes (Figure 3 and Figure 7B). One of the possible explanations of these results is that the RISC exchange reaction might be occurred associated with single-stranded endo-miRNAs and not endo-miRNA duplexes or target-paired endo-miRNAs (Figure 7), because double-stranded endo-miRNAs in the RISC might not be replaced with exo-miRNAs at similar levels, due to their different structures and sequences. Most miRNA-RISCs might be in the form of single-stranded miRNAs in RISCs, so as to be readily replaced by double-stranded miRNAs.

**Figure 7.** Predicted model for the RNA-induced silencing complexes (RISCs) replacement. Different types of exo-miRNAs transfected into cells may be replaced with single-stranded endo-miRNAs loaded on the RISC with different efficiencies (**A**); however, different types of endo-miRNAs may be replaced with a given exo-miRNAs with similar efficiencies (**B**).

Furthermore, it was apparently revealed that endo-miRNAs constantly repress the expression of endogenous mRNAs with endo-miRNA target sites, and their repression is probably relieved by the replacement of endo-miRNAs on the RISC by the exo-miRNAs transfected (Figures 5 and 6). Thus, global gene expression by endogenous miRNAs might be fluctuated by the transfection of exo-miRNAs or the increase of expression levels of endo-miRNAs. Competition similar to that shown here between exo-miRNA duplexes and endo-miRNAs may also occur among newly transcribed endo-miRNAs and may provide a mechanism for orchestrating cellular programs.

#### **Acknowledgments**

We thank Yuki Naito for helpful discussions about the bioinformatics analysis and Kenji Nishi for technical advises on microarray experiments. This work was supported by a Grant-in-Aid for Scientific Research (grant numbers 21115004, 21310123), the Cell Innovation Project from the Ministry of Education, Culture, Sports, Science and Technology of Japan and the Core Research Project for Private University matching fund subsidy to K.U.-T.

#### **Conflict of Interest**

The authors declare no conflict of interest.

#### **References**



Reprinted from *IJMS*. Cite as: Tomaselli, S.; Bonamassa, B.; Alisi, A.; Nobili, V.; Locatelli, F.; Gallo, A. ADAR Enzyme and miRNA Story: A Nucleotide that Can Make the Difference. *Int. J. Mol. Sci.* **2013**, *14*, 22796-22816.
