Novel microRNAs modulating ecto-5′-nucleotidase expression

Introduction The expression of immune checkpoint molecules (ICMs) by cancer cells is known to counteract tumor-reactive immune responses, thereby promoting tumor immune escape. For example, upregulated expression of ecto-5′-nucleotidase (NT5E), also designated as CD73, increases extracellular levels of immunosuppressive adenosine, which inhibits tumor attack by activated T cells. MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression at the post-transcriptional level. Thus, the binding of miRNAs to the 3′-untranslated region of target mRNAs either blocks translation or induces degradation of the targeted mRNA. Cancer cells often exhibit aberrant miRNA expression profiles; hence, tumor-derived miRNAs have been used as biomarkers for early tumor detection. Methods In this study, we screened a human miRNA library and identified miRNAs affecting the expression of ICMs NT5E, ENTPD1, and CD274 in the human tumor cell lines SK-Mel-28 (melanoma) and MDA-MB-231 (breast cancer). Thereby, a set of potential tumor-suppressor miRNAs that decreased ICM expression in these cell lines was defined. Notably, this study also introduces a group of potential oncogenic miRNAs that cause increased ICM expression and presents the possible underlying mechanisms. The results of high-throughput screening of miRNAs affecting NT5E expression were validated in vitro in 12 cell lines of various tumor entities. Results As result, miR-1285-5p, miR-155-5p, and miR-3134 were found to be the most potent inhibitors of NT5E expression, while miR-134-3p, miR-6859-3p, miR-6514-3p, and miR-224-3p were identified as miRNAs that strongly enhanced NT5E expression levels. Discussion The miRNAs identified might have clinical relevance as potential therapeutic agents and biomarkers or therapeutic targets, respectively.


1
Supplementary Methods

miRNA transfection
Results of the miRNA library screen were validated by individual transfections as follows: 2 x 10 5 cells were seeded in 12 well plate and cultured for 24 h reaching approx. 70 % confluency. Cells were transfected with 50 nM miRNA or siRNA using Lipofectamine RNAiMax reagent according to the manufacturer's protocol. Cells were harvested for RNA isolation and subsequent qPCR or Microarray analysis 48 h post transfection. Expression of cell surface molecules was measured by FACS 72 h post transfection.

RNA extraction and qPCR
Total RNA from frozen cell pellets was isolated using miRNeasy Mini Kit according to the manufacturer's protocol, enabling simultaneous extraction of miRNA and mRNA for expression analysis. Cell pellets were lysed using with phenol/guanidine based QIAzol lysis reagent. Subsequently, RNA was purified with silica-membrane spin columns and eluted with 30 µL nucleasefree water. In case of mRNA expression analysis, total RNA was isolated using the RNAeasy Kit according to the manufacturer's protocol. RNA samples were stored at -80 °C until further use for cDNA synthesis. RNA concentrations were measured with a NanoDrop spectrophotometer (ThermoFisher Scientific, Waltham, USA) or Qubit device using the RNA BR Assay kit (ThermoFisher Scientific, Waltham, USA). For mRNA expression analysis, 500 ng of total RNA were used for reverse transcription in a 20 µL reaction volume with oligo(dT)18 primers applying the Transcriptor First Strand cDNA Synthesis Kit (Roche Applied Science, Mannheim, Germany). To analyze miRNA expression, 40 ng of total RNA were reverse transcribed in a 15 µL reaction volume using TaqMan™ MicroRNA reverse transcription kit (Applied Biosystems, Foster City, CA, USA) according to the manufacturer's protocol including specific stem-loop primers for mature miRNA. All PCRs were run on a Veriti 96 well Thermal Cycler or QuantStudio 3 Real-Time PCR System (ThermoFisher Scientific, Waltham, USA). To perform qPCR with cDNA synthesized from mRNA, 2 µL cDNA diluted 1:5 with PCR-quality water was used in a 20 µL reaction volume using the SYBR Green PowerUP PCR Mastermix for qPCR. RPL19 and TBP encoding genes lacking putative binding site for transfected miRNAs were included as house-keeping genes for normalization. For cDNA synthesis from miRNA, 2 µL miRNA were used in a per PCR reaction performed in a total volume of 20 µL. Small nuclear RNA U6 (RNU6B) was used as endogenous control. For all samples three technical replicates were performed and relative expression was calculated according to the 2 -∆∆Ctmethod. Tables   2.1 Supplementary Tables   Table S1. Antibodies used for flow cytometry.      Table S12. For NT5E enhancing miRNAs the number of potential binding sites with 3'-UTR of selected target genes was analyzed by miRMap tool.  231 cells (B). RPL19 was used as housekeeping gene. ∆CTs were calculated for both NT5E isoforms: NT5E-1 (normal, long variant) and NT5E-2 (shorter variant). Log2 Fold changes were calculated for both isoforms normalized to corresponding mimic control-1 samples. The differences of log2FC for NT5E-1 and NT5E-2 were calculated to estimate which miRNAs asymmetrically affect the two NT5E isoforms. Conditions were compared by One-way ANOVA comparing all conditions to mimic control-1. *: p < 0.05. Figure 10: Effect of NT5E enhancing miRNAs on NT5E isoform expression. SK-Mel-28 and MaMel-05 cells were transfected with 50 nM miRNA/siRNA and 72 h after transfection total RNA was isolated and used for isoform specific qPCR. RPL19 was used as housekeeping gene. ∆CTs were calculated for both NT5E isoforms: NT5E-1 (normal, long variant) and NT5E-2 (shorter variant). Log2 Fold changes were calculated for both isoforms normalized to corresponding mimic control-1 samples. The differences of log2FC for NT5E-1 and NT5E-2 were calculated to estimate which miRNAs asymmetrically affect the two NT5E isoforms.

Supplementary
Supplementary Figure 11: Kaplan-Meier analysis of CBX6 and CNOT6L in breast cancer patients. High CBX6 expression is significantly associated with a better survival in human breast cancer patients (n = 4929). Also, high CNOT6L or NFATC3 mRNA levels are significantly associated with a better survival for breast cancer patients (n = 2032). In contrast, high NT5E/CD73 mRNA levels are significantly associated with worse prognosis for breast cancer patients. Furthermore, high miR-224 expression is linked to worse survival for breast cancer patients based on METABRIC data set (n = 1262). This indicates, that high levels of NT5E promoting miRNAs such as miR-224 or low expression of NT5E repressor such as CBX6/CNOT6L or NFACT3 are linked to progressive tumor disease reflected by shorter survival time. Plots were generated using Kaplan-Meier plotter tool 11 .
Supplementary Figure 12: Kaplan-Meier analysis of miR-134 in melanoma patients. High expression level of miR-134 levels shows a tendency to be associated with worse survival in human melanoma patients. Plot was generated using R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl). High expression level of MIR155HG gene is significantly associated with better survival in human melanoma patients. Also, in breast cancer the same tendency can be observed. Furthermore, high expression level of miRNA itself is significantly linked to better prognosis of both melanoma and breast cancer patients. Plots were generated using R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl). Noteworthy, expression of NT5E mRNA and CD274 positively correlates with MIR155HG expression. This indicates a negative feedback loop, since miR-155-5p was proven to target and inhibit both CD274 and NT5E expression.