Downregulation of hsa_circ_0005654 Triggers Tumor Initiation Events in Non-Neoplastic Mammary Epithelial Cells: Relevance as a Biomarker for Early Breast Cancer Detection

Abstract

In the latest report for 2020, breast cancer (BC) had the highest incidence rate of 2.26 million cases. Ranking in the top 5 cancers for mortality, it caused 685,000 deaths. Late diagnosis as well as lack of regular screening contribute to the high death toll, especially for populations at risk. With the shortcomings of regular screening methods to detect early breast tumorigenesis and the invasiveness of biopsy screening, studies have redirected their efforts towards the utilization of body fluids for detection. CircRNAs and miRNAs are non-coding RNAs that have been shown to be involved in BC initiation and progression. CircRNAs mimic the expression of their parental gene, and may target certain miRNAs by a sponging mechanism, which produces reciprocal expression between circRNAs and their target miRNA. Using large data RNAseq (mRNA and miRNA) analysis of stage I breast cancer patient samples and multiple bioinformatic tools, our lab identified 57 gene/circRNAs/miRNA/mRNA axes that are predictive of BC. Due to their novelty, data on circRNAs was lacking and their expression in stage 1 BC remained predicted in the axes. To increase our understanding of the involvement of the proposed circRNAs in BC initiation, this study aimed to conduct in-vitro investigation of selected circRNAs from the 57 circRNAs. None of the 34 downregulated circRNAs in the BC predictive axes were previously studied for their role in BC. Literature review allowed the selection of six downregulated circRNAs that have been reported to be downregulated in other types of cancer: hsa_circ_0005654, hsa_circ_0004075, hsa_circ_0005450, hsa_circ_0087302, hsa_circ_0006539, and hsa_circ_0029405. We downregulated the expression of the six circRNAs in nontumorigenic human mammary epithelial HMT-3522 S1 cells using transient transfection of junctional-specific siRNA. Proliferation assay post transfection revealed an increased proliferation by day 3 when targeting hsa_circ_0005654. Based on this finding, and that hsa_circ_0005654, proposed within the axis PRDM5/hsa_circ_0005654/miR-183, is one of the higher confidence axes, it was selected for further studies. Downregulation of hsa_circ_0005654 in S1 cells was verified by qRT-PCR using divergent primers showing a 0.5-fold expression in si5654 (siRNA targeting hsa_circ_0005654) transfected relative to sham transfected cells. The expression of the parental gene PRDM5 of hsa_circ_0005654 was comparable between un-transfected, si5654 transfected, and sham transfected S1 cells. Downregulation of hsa_circ_0005654 expression led to enhanced migration of S1 cells (~3.2 folds in si5654 transfected compared to sham transfected cells) and disruption of lumen formation in 3D cultures (~0.5-fold decrease compared to sham transfected cells). Moreover, we showed that hsa_circ_0005654 triggers a reciprocal upregulation of its target miRNA miR-183 (~3.2 folds increase compared to sham transfected cells). To establish a model that can be used for future long-term and 3D studies, we stably downregulated hsa_circ_0005654 in S1 cells using the lentiviral system and verified its downregulation using qRT-PCR (~0.2 fold decrease compared to sham transfected cells). These in-vitro studies further reaffirmed the involvement of circRNA hsa_circ_0005654 in BC initiation as proposed in the BC predictive axes. Future studies in our lab aim to screen for the 57 circRNA and respective target miRNAs from the axes in stage I BC blood samples of patients. For this purpose, we aimed to optimize the detection of circRNAs using qRT-PCR in this study. Treatment with RNase R did not achieve earlier detection of hsa_circ_0005654 as expected and showed delayed CT values in qRT-PCR. We opted for a second approach, independent of RNase R. We found that the use of 20-fold lower concentration of total RNA and random primers in reverse transcription could possibly allow for the detection of differential circRNA trends efficiently. In conclusion, our findings provide insight into the role of hsa_circ_0005654 as tumor suppressor in the breast epithelium, whereby its downregulation is associated with tumor initiation events in non-neoplastic mammary epithelial cells possibly through sponging miR-183. We also proposed a method for the efficient detection of circRNAs using qRT-PCR to be utilized for the screening of circRNAs in blood samples.