MicroRNA-205-5p inhibits three-dimensional spheroid proliferation of ErbB2-overexpressing breast epithelial cells through direct targeting of CLCN3

We previously reported that microRNA-205-5p (miR-205-5p) is significantly decreased in the ErbB2-overexpressing breast epithelial cell line MCF10A-ErbB2 compared with control cells. In this study, we identified a direct target of miR-205-5p, chloride voltage-gated channel 3 (CLCN3). CLCN3 expression was induced by ErbB2 overexpression; this induced expression was then reduced to control levels by the transfection of the miR-205-5p precursor. In RNA-binding protein immunoprecipitation with Ago1/2/3 antibody, CLCN3 was significantly enriched in 293T embryonic kidney cells with miR-205-5p mimic transfection compared with negative control mimic transfection. In luciferase reporter assays using CLCN3 3′-UTR constructs, the miR-205-5p mimic significantly decreased reporter activity of both wild-type and partial mutant constructs in MCF10A-ErbB2 cells. In contrast, no inhibitory effects of the miR-205-5p mimic were detected using the complete mutant constructs. Since miR-205-5p expression in exosomes derived from MCF10A-neo cells was substantially higher than in exosomes derived from MCF10A-ErbB2 cells, we next investigated whether an exosome-mediated miR-205-5p transfer could control CLCN3 expression. To this end, exosomal miR-205-5p derived from MCF10A-neo cells was functionally transferred to MCF10A-ErbB2 cells, which served to decrease the expression of CLCN3. To assess the roles of CLCN3 in breast cancer, we next performed three-dimensional (3D) spheroid proliferation analyses using MCF10A-ErbB2 cells treated with MCF10A-neo-derived exosomes or CLCN3 shRNA stably expressing SKBR3 and MDA-MB-453 breast cancer cells. Our results showed that both treatment with MCF10A-neo-derived exosome and CLCN3 shRNA expression suppressed 3D spheroid proliferation. Collectively, these novel findings suggest that CLCN3 may be a novel direct target of miR-205-5p and this CLCN3/miR-205-5p interaction may serve a pivotal role in regulating breast cancer cellular proliferation under physiological conditions.


INTRODUCTION
MicroRNAs (miRNAs) are a class of small noncoding RNAs that regulate gene expression post-transcriptionally through binding to the 3 -untranslated regions (3 -UTRs) of target mRNAs. Via this process, miRNAs regulate various cellular activities, including cellular growth, differentiation, development, and apoptosis. Dysregulation of miRNAs is associated with various human diseases, such as cancer (Jiang et al., 2009). Therefore, miRNAs have emerged as promising prognostic and therapeutic tools for cancer management.
We previously reported that miR-205-5p is reduced by ErbB2 overexpression and that the ErbB2 tumorigenic capability to proliferate in soft agar is reduced by exogenous transfection of the miR-205-5p precursor (Adachi et al., 2011). In addition, we previously reported that ErbB2 signaling epigenetically suppresses miR-205-5p transcription via the Ras/Raf/MEK/ERK pathway in breast cancer (Hasegawa et al., 2017). Therefore in this study, we further attempted to identify an additional novel target of miR-205-5p in order to understand the comprehensive role of miR-205-5p in breast cancer. To this end, we focused on chloride voltage-gated channel 3 (CLCN3), a member of the voltage-gated chloride channel family. The volume-regulated anion channel (VRAC) contributes to cell volume regulation (Osei-Owusu et al., 2018). Dysfunction of cell volume regulation is one of the characteristics of cancer cells, leading to aberrant cell proliferation and apoptosis (Pedersen, Hoffmann & Novak, 2013). CLCN3 has been reported to play a key role in native VRAC in a variety of cancer cells (Duan, 2011;Habela, Olsen & Sontheimer, 2008;Lemonnier et al., 2004;Mao et al., 2008). Hence, CLCN3 may regulate cell proliferation and apoptosis via a VRAC-related mechanism.
Our findings in this study demonstrated that CLCN3 is a potential direct target of miR-205-5p and regulates 3D spheroid proliferation in ErbB2-overexpressing breast epithelial cells and breast cancer cells.

Cells
MCF10A-ErbB2 and MCF10A-neo cells were previously generated in our laboratory (Adachi et al., 2011) and cultured in DMEM/F12 with the addition of 5% horse serum, 20 ng/mL EGF, 10 µg/mL insulin, and 500 ng/mL hydrocortisone. The human breast cancer cell lines MDA-MB-453 and SKBR3 cells, as well as the human embryonic kidney cell line 293T, were cultured in DMEM with the addition of 10% fetal bovine serum. DMEM/F12, DMEM, and fetal bovine serum were purchased from Thermo Fisher Scientific (Waltham, MA, USA). EGF, insulin, and hydrocortisone were purchased from Sigma (St. Louis, MO, USA).

miRNA precursor transfection
The transfection of miRNA Precursors (Pre-miR TM hsa-miR-205-5p miRNA Precursor or Pre-miR TM miRNA Precursor-Negative Control #1, both purchased from Thermo Fisher Scientific) were performed using the RNAiMAX reagent (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer's instruction. Briefly, the cells were seeded in 12-well plates 1 day before transfection, and 15 pmol/well miRNA Precursor was transfected using Lipofectamine RNAiMAX into 30%-50% confluent cells. At 48 h post-transfection, the cells were harvested for RNA extraction. At 72 h post-transfection, the cells were harvested for protein extraction.

RNA isolation and real-time RT-PCR of CLCN3
Total RNA was purified by RNAiso Plus (Takara Bio, Kusatsu, Shiga, Japan) according to the manufacturer's instructions and then treated with RNase-free DNase I (Takara Bio). Subsequently, the RNA was cleaned up using an RNeasy Mini kit (Qiagen). Briefly, 2 µg of total RNA was reverse transcribed into cDNA using the High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific). Quantitative real-time PCR was then carried out on an MJ-Mini thermal cycler in conjunction with a MiniOpticon Real-Time PCR system (Bio-Rad, Hercules, CA, USA) under the following conditions: an initial denaturation step at 95 • C for 10 s, followed by 40 cycles at 95 • C for 10 s and 60 • C for 30 s. Dissociation curve analysis was performed for each reaction to guarantee the specificity of amplification. The final concentrations of the PCR reaction components were as follows: 1X SYBR Premix Ex Taq II (Perfect Real Time) (Takara Bio), 0.4 µM forward and reverse primers and 5 µL template cDNA for 20 µL reaction. The primer sequences were as follows: CLCN3 (forward: 5 -ACATGCACCACAACAAAGGC-3 ; reverse: 5 -TTTCGGTTTTGAGCCACACG -3 ), ZEB2 (forward: 5 -TGTTTCTGCAAGTGCCATCC-3 ; reverse: 5 -ACACTGAAGCTGGTGCAAAG-3 ) and β-actin (forward: 5 -ATTGCCGACAGGATGCAGA-3 ; reverse: 5 -GAGTAC TTGCGCTCAGGAGGA-3 ). Expression level of CLCN3 was normalized to β-actin using a standard curve method.

Western blotting
Western blotting was performed as previously described (Adachi et al., 2011). Whole cell lysates were subjected to SDS-PAGE, and separated proteins were transferred to a 0.2-µm PVDF membrane. Blocking was performed with 5% dry milk in 0.05% PBST. The membrane was then blotted with the specific primary antibody. After washing in 0.05% PBST, the membrane was probed with the corresponding secondary antibody conjugated with horseradish peroxidase. After washing in 0.05% PBST, the membrane was visualized by the SuperSignal R West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific) and analyzed using the ChemiDoc XRS-J image analysis system (Bio-Rad). The antibodies used in this study are shown in the Supplementary Material and Methods.

RNA-binding protein immunoprecipitation
RNA-binding protein immunoprecipitation (RIP) was performed using the miRNA Target IP kit (Active Motif, Carlsbad, CA, USA) following the manufacturer's instructions. Briefly, 293T cells (2.5 × 10 6 ) were seeded in a 100-mm dish and transfected with 750 pmol of miR-205-5p mimic or negative control mimic using Lipofectamine RNAiMAX. After 24 h, cells were lysed and RIP assay was performed using anti-Ago1/2/3 antibody or negative control IgG. The immunoprecipitated RNA was purified and subjected to real-time RT-PCR analysis. The levels of CLCN3, ZEB2 or β-actin were detected and normalized to the input levels.

Three-dimensional (3D) spheroid proliferation assay
The 3D spheroid proliferation assay was performed using the Cultrex R 3D Spheroid Colorimetric Proliferation/Viability Assay (Trevigen, Gaithersburg, MD) following the manufacturer's instructions. Briefly, 3,000 cells were plated in 50 µL medium containing Spheroid Formation ECM in a 3D Culture Qualified 96-well Spheroid Formation plate and cultured for 72 h. In an experiment using CLCN3 shRNA stable cells, 50 µL medium was added to each well and cells were cultured for additional 72 h. In an exosome treatment experiment, 50 µL medium plus 10 µL PBS or 10 µL exosomes derived from MCF10A-neo cells were added to each well, and cells were cultured for an additional 72 h. Cellular proliferation was assessed by MTT analysis, and absorbance was measured on a Biotrak II Plate Reader (GE Healthcare, Chicago, IL) at a wavelength of 562 nm, with background subtracted at 690 nm.

shRNA expression plasmid construction
The

Stable cell generation
Retroviral infection was performed as previously described (Adachi et al., 2011;Hasegawa et al., 2017). shRNA-expressing retroviruses were prepared by transient co-transfection with pSINsi-DK II-CLCN3 shRNA or pSINsi-DK II-control shRNA and the amphotropic helper virus pSV-A-MLV into 293T cells by using calcium phosphate precipitation. SKBR3 and MDA-MB-453 cells were cultured with fresh retroviral supernatants in the presence of polybrene for 48 h and then subjected to selection by 1.5 mg/mL G418 (Sigma) for SKBR3 and 1 mg/mL G418 for MDA-MB-453.

Exosome isolation and exosomal RNA purification
Exosomes were isolated using Total Exosome Isolation (from cell culture media) (Thermo Fisher Scientific) following the manufacturer's instruction. Briefly, 1× 10 6 cells were seeded in a 10 cm dish and cultured in serum-containing medium for 24 h. After washing cells with serum-free medium, the cells were cultured in serum-free medium for 48 h. Culture medium was then harvested and centrifuged at 2,000× g for 30 min. The supernatant was incubated with the Total Exosome Isolation (from cell culture media) reagent at 4 • C overnight and then centrifuged at 10,000× g for 1 h at 4 • C. The supernatant was then removed, and the exosome-containing pellet was resuspended in 100 µL PBS. Exosomal RNA was purified using the Total Exosome RNA & Protein Isolation Kit (Thermo Fisher Scientific) following the manufacturer's instructions. Confirmation of exosome isolation was checked by evaluating exosomal marker protein expression (Fig. S1).

Exosome treatment
Cells (4 × 10 5 ) were seeded in a 6-well plate and cultured in serum-free medium with 60 µL exosome suspension in PBS or 60 µL PBS for 24 h. Cells were harvested and applied to Real-time RT-PCR analysis for miR-205-5p and CLCN3 and 3D spheroid proliferation assays.

MiR-205-5p inhibits expression of CLCN3 in breast epithelial cells
We previously established breast epithelial cells that stably overexpress ErbB2 (MCF10A-ErbB2) and the associated control cells (MCF10A-neo). In this previous study, we reported that the overexpression of ErbB2 inhibits the expression of miR-205-5p (Adachi et al., 2011). We next searched for potential target genes of miR-205-5p using in silico analysis (miRBLAST-B, Cosmo Bio, Tokyo, Japan) and narrowed down candidate genes by literature search and real-time RT-PCR analysis. Then we selected CLCN3 as one of the candidates.
To determine whether miR-205-5p expression correlates with CLCN3 expression in breast epithelial cells, we further examined CLCN3 expression in MCF10A cells, MCF10A-neo cells, MCF10A-ErbB2 cells, negative control precursor-transfected, and miR-205-5p precursor-transfected MCF10A-ErbB2 cells by western blotting. Our results revealed that the expression of CLCN3 increased in MCF10A-ErbB2 cells compared with MCF10A and MCF10A-neo cells and that the elevated CLCN3 expression level was reduced by transfection with the Pre-miR-205-5p precursor (Fig. 1).

MiR-205-5p directly targets CLCN3 3 -UTR in breast epithelial cells
An Argonaute protein (Ago) plays a crucial role in the maturation process of miRNAs as a component of the RNA-induced silencing complex. We next performed RIP assay with anti-Ago1/2/3 antibody to validate the interaction between miR-205-5p and CLCN3. RIP assay revealed that the relative enrichment of CLCN3 in Ago immunoprecipitation complex was significantly increased in 293T cells transfected with miR-205-5p mimic compared with negative control mimic group ( Fig. 2A). MiR-205-5p mimic transfection resulted in the similar enrichment of ZEB2, a known target of miR-205-5p, whereas didn't change the enrichment level of β-actin. We further evaluated whether CLCN3 is a direct target of miR-205-5p. We predicted a putative miR-205-5p binding site in the CLCN3 3 -UTR (Fig. 2B) and constructed luciferase reporter plasmids containing wild-type CLCN3 3 -UTR or the three different mutations at the putative miR-205-5p binding site (Fig. 2C).

Functional exosomal miRNA is transferred into ErbB2-overexpressing breast epithelial cells
Since the expression of miR-205-5p was significantly reduced in MCF10A-ErbB2 cells compared with MCF10A-neo cells as previously described (Adachi et al., 2011), we next determined the expression of miR-205-5p in the exosomes from these cells. We found that miR-205-5p expression in exosomes isolated from MCF10-neo cells was much higher than in exosomes isolated from MCF10-ErbB2 cells (Fig. S2). It has been reported that exosomal miRNAs can be transferred between cells and mediate target gene repression and physiological function (Bovy et al., 2015;Mittelbrunn et al., 2011;Santos et al., 2016). Therefore, we treated MCF10A-ErbB2 cells with exosomes derived from MCF10A-neo cells to determine whether functional miR-205-5p could be transferred. Our results showed that miR-205-5p expression in MCF10A-ErbB2 cells treated with MCF10A-neo-derived exosomes was increased by about 7-fold compared with vehicle (PBS) treatment (Fig. 4A). Moreover, CLCN3 expression in MCF10A-ErbB2 cells treated with MCF10A-neo-derived exosomes was significantly decreased compared with vehicle treatment (Fig. 4B).

CLCN3 mediates 3D spheroid proliferation in ErbB2-overexpressing breast epithelial cells and breast cancer cells
Since our data indicated that CLCN3 is one of the potential targets of miR-205-5p, we investigated the possible biological function of CLCN3 in ErbB2-overexpressing breast epithelial cells and breast cancer cells. We analyzed 3D spheroid proliferation of MCF10A-ErbB2 cells treated with exosomes derived from MCF10A-neo cells because we previously found that miR-205-5p inhibited 3D colony formation in soft agar using MCF10A-ErbB2 cells. Our results showed that the treatment of MCF10A-neo-derived exosomes decreased 3D spheroid proliferation by about 40% compared with vehicle ( Fig. 5A). In addition, we established CLCN3 shRNA or control shRNA stably expressing cells using the ErbB2overexpressing breast cancer cell lines SKBR3 and MDA-MB-453 (Fig. S3) and analyzed the 3D spheroid proliferation of these stable cells. Inhibition of CLCN3 expression in 3D spheroids was confirmed by real-time RT-PCR (Fig. S4). Our results showed that CLCN3 shRNA stable cells have significantly decreased 3D spheroid proliferation compared with control shRNA stable cells in both SKBR3 and MDA-MB-453 cells (Fig. 5B, Fig. S5).

DISCUSSION
To understand the comprehensive role of miR-205-5p in breast cancer, we attempted to identify a novel target of miR-205-5p that may be involved in breast cancer progression and performed analyses focusing on CLCN3. On the basis of our observation that CLCN3 expression was increased in the ErbB2-overexpressing breast epithelial cells MCF10A-ErbB2, and ectopic transfection of the miR-205-5p precursor reduced the elevated CLCN3 expression levels, CLCN3 may prove to be a miR-205-5p target. Interestingly, we additionally found that miR-205-5p expression was significantly reduced in the ErbB2overexpressing breast cancer cell lines stably expressing CLCN3 shRNA (Fig. S6). This may occur by negative feedback loop mechanisms between miRNA and its target genes (Herranz & Cohen, 2010;Liu, Duan & Duan, 2018). Interaction between CLCN3 and miR-205-5p was also verified by the significantly high enrichment of CLCN3 in miR-205-5p mimic transfected 293T cells via RIP analysis. These observations were further confirmed by reporter assays revealing that the miR-205-5p mimic significantly decreased the reporter activity of CLCN3-3 -UTR-wt. Additional observations that the miR-205-5p mimic did not show significant effects on the reporter activity of the complete mutant, CLCN3-3 -UTR-mut1, strongly support that CLCN3 is likely a novel potential target of miR-205-5p. The seed sequence, nucleotides 2 to 8 of the miRNA, has been recognized as a critical determinant of canonical miRNA-target interaction. Although there are imperfect seed matches between CLCN3 and miR-205-5p, recent reports revealed that imperfect seed matches could be compensated for by extensive pairing with the seed-distal 3 end of the miRNA (Brancati & Grosshans, 2018;Broughton et al., 2016). According to our data showing that the miR-205-5p mimic also significantly decreased the reporter activity of the partial mutants, CLCN3-3 -UTR-mut2 and CLCN3-3 -UTR-mut3, it is suggested that the miR-205-5p and CLCN3 interaction needs both seed and seed-distal pairing. CLCN3 is a member of the voltage-gated chloride channel family and functions as a Cl − /H + transporter in intracellular membranes (Duran et al., 2010;Guzman et al., 2013). In addition, several studies have shown that CLCN3 is involved in cell proliferation, apoptosis, drug resistance, and invasion in many cancers (Lui et al., 2010;Su et al., 2013;Xu et al., 2010;Zhang et al., 2013). In cellular proliferation, CLCN3 plays an important role by controlling cell cycle progression (Wang et al., 2002;Xu et al., 2010). Knockdown of CLCN3 by siRNA reduces cells in S phase, while increasing those in G 0 /G 1 phase, in rat basilar arterial smooth muscle cells and inhibits cellular proliferation by downregulating the expression of cyclin D1 and cyclin E in mouse mesenchymal stem cells (Tang et al., 2008;Tao et al., 2008). CLCN3 also accelerates the G 0 /G 1 to S phase transition in the cell cycle by enhancing the phosphorylation of ERK1/2 and upregulating cyclin E and cyclin D1 in multiple myeloma cells (Du et al., 2018). Thus, there are several reports on the functions of CLCN3 in cancer. However, the roles of CLCN3 in breast cancer cell proliferation remain unclear. Our findings in this study indicate that CLCN3 promotes 3D spheroid proliferation in ErbB2-overexpressing breast epithelial and cancer cells. These findings should improve our understanding of the significance of CLCN3 in breast cancer cellular proliferation. In addition, we showed that miR-205-5p expression in exosomes isolated from MCF10-neo cells was higher than in exosomes isolated from MCF10-ErbB2 cells. Our data of exosome treatment experiments further indicate that exosomal miR-205-5p may be functionally transferred between breast epithelial cells and inhibit 3D spheroid proliferation by downregulating CLCN3. Although the inhibitory effect of exosomal miR-205-5p on CLCN3 expression was marginal, it was likely due to the amount of exosomal miR-205-5p transfer being considerably small in comparison with that of miR-205-5p precursor or mimic transfection. The important thing was that miR-205-5p could be successfully and functionally transferred by exosome treatment. We also found that expression of miR-200 family members in exosomes from MCF10A-neo cells was significantly higher than in exosomes from MCF10-ErbB2 cells, but the transfer of miR-200 family members didn't