Study population
Tissue and plasma samples from patients with CRC (colorectal cancer) or AP (adenomatous polyps) and HC (healthy control) subjects were prospectively collected at the Department of Gastroenterology and Department of Surgery, Lithuanian University of Health Sciences (Kaunas, Lithuania) during 2011–2014 years. The specimens of all investigated groups were biopsies, taken from affected or healthy colorectal tissues during routine colonoscopy or surgical tumor removal. Biopsies of the CRC group were histologically verified as being colorectal adenocarcinomas. The preneoplastic lesion group was comprised of advanced adenoma patients (conventional adenomas with high-grade dysplasia, adenomas of > 1 cm diameter or villous histology). The HC group consisted of healthy subjects, who underwent colonoscopy due to positive fecal occult blood test (FOBT), but had no history of previous malignancy and were otherwise healthy. All patients included in the study were of European descent. Clinical and phenotypic characteristics of subjects investigated in profiling and validation cohorts are presented in Table 1.
RNA isolation
Total RNA (including small RNA fraction) was extracted from snap-frozen tissue and blood plasma samples using miRNeasy Mini Kit (Qiagen, MD, USA) and miRNeasy Serum and Plasma Kit (Qiagen, USA), respectively, according to the recommendations of the manufacturer. The quality and quantity of the isolated RNA were assessed by Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific, MA, USA) or by Agilent 2100 Bioanalyzer (Agilent Technologies, CA, USA).
Small RNA library preparation and NGS
Small RNA libraries were generated with 1 µg RNA input per sample applying a standard TruSeq Small RNA Sample Preparation Kit (Illumina, CA, USA) protocol. Agilent 2100 Bioanalyzer (Agilent Technologies, USA) was used for quality and yield assessment of sequencing libraries. Small RNA libraries were randomly selected, pooled of approximately 24 samples/lane and sequenced (1 × 50 bp single-end reads) on HiSeq2500 (Illumina, USA) NGS platform.
Small RNA-Seq data analysis
The generated raw sequencing reads (fastq) were processed by cutadapt v1.9 [19] which was used to trim adapter sequences and low-quality bases (< Q20) as well as discard sequences shorter than 18 nt in length. The processed small RNA-seq reads were assigned to miRBase v21, using quantifier.pl [20] and miraligner.jar [21] for reference miRNAs and isomiRs, respectively. In order to reduce false-positives in isomiR sequence substitutions, a mismatch was considered as a real substitution when the minimum fraction rate of reads having that substitution was equal to 0.25. Additionally, only the isomiRs with uniquely mapped substitutions were kept for downstream analyses. Generated miRNA/isomiR counts were normalized using size factor normalization and variance stabilizing transformation (VST) employing the DESeq2 package [22]. The abundance differences of isomiR variations among the CRC, AP, and HC groups were examined by the Kruskal-Wallis test. Differential expression analyses of miRNAs were performed by employing negative binomial generalized linear models and Wald statistics implemented in the DESeq2 R package. The models were fitted using technical batch and age (centered) as covariates, since the average age was significantly different (P < 0.05) between CRC and HC groups. Benjamini and Hochberg correction of false discovery rate (FDR) was used for P-value adjustment. The miRNAs with PFDR < 0.01 and absolute value of log2 fold change (log2FC) > 0.5 were considered as significantly differentially expressed. Such log2FC and PFDR value threshold was chosen to detect mild changes in expression (0.5 < log2FC < 1), that are considerably consistent (PFDR < 0.01) among patients within group. Results of three-way differential expression analysis were visualized using the volcano3D R package [23]. Spearman’s rank correlation coefficient was used for correlation analysis. The raw sequencing data, as well as miRNA counts, have been deposited at the Gene Expression Omnibus (GEO) under the accession number of GSE160432 (link: www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE160432; reviewer token: upcdwguslrglfwx).
RT-qPCR validation analysis
Quantification of hsa-miR-1246 and hsa-miR-215-5p expression in tissue samples of CRC, AP and healthy individuals were analyzed using TaqMan® MicroRNA Assays (Applied Biosystems, CA, USA) according to the producer’s recommendations in technical duplicates. MiRNA expression in plasma samples was quantified in two batches – hsa-miR-1246 and hsa-miR-215-5p expression in CRC and healthy controls plasma samples was quantified by using TaqMan® MicroRNA Assays (Applied Biosystems, USA), while expression in AP and the same healthy controls plasma samples was analyzed by using TaqMan® Advanced MicroRNA Assays (Applied Biosystems, USA). Accordingly, the results of miRNA expression in plasma of CRC and AP groups were analyzed separately. Endogenous hsa-miR-16-5p (tissue and plasma) and hsa-miR-191-5p (plasma) were used as internal normalizers in both tissue and plasma samples [24, 25]. The expression levels of hsa-miR-1246 and hsa-miR-215-5p were normalized using ΔCT method [26]. The mean value of duplicate ΔCT values of each sample was submitted to the non-parametric Mann-Whitney U test. The miRNAs with a false discovery rate (FDR) corrected P < 0.05 and an absolute value of log2 fold change > 1 were considered as significantly differentially expressed.
Target gene set enrichment analysis
Target gene set enrichment analysis for hsa-miR-1246 and hsa-miR-215 was accomplished using miTALOS v2 [27] tool. The tool was used to perform overrepresentation analysis of biological pathways with default settings using TargetScan miRNA-target collection and employing Fisher’s exact test as well as Benjamini-Hochberg multiple test correction procedure. Pathways with PFDR < 0.05 and enrichment score (E) > 1 were considered as significantly overrepresented among hsa-miR-1246 and/or hsa-miR-215 target genes.
Cell culture
Human colorectal adenocarcinoma SW620 (CCL-227™, ATCC®, VA, USA), Caco-2 (HTB-37™, ATCC®, USA), HCT116 (CCL-247™, ATCC®, USA), and gastric adenocarcinoma AGS (CRL-1739™, ATCC®, USA) cell lines were cultured in Ham's F-12K (Kaighn's) Medium (Gibco, MA, USA) supplemented with 10% fetal bovine serum (Gibco, USA) and 1% penicillin/streptomycin (100 U/mL penicillin and 100 mg/mL streptomycin, Corning, NY, USA), in a humidified incubator at 37°C with 5% CO2.
Cell transfection
For colony formation, MTT, wound healing, luciferase and protein expression assays, reverse transfection with hsa-miR-1246 inhibitor (Assay ID: MH13182, mirVana™, CA, USA) or mimic (Assay ID: MC13182, mirVana™, USA) and miR-Control (mirVana™, USA) (final concentration 100 pmol/ml) was performed using lipofectamine 3000 reagent (Invitrogen, CA, USA).
Wound healing assay
Reverse transfected HCT116 and Caco-2 cells were seeded into a 2-well silicone insert with a defined cell-free gap (Ibidi, Gräfelfing, Germany) for the wound healing assay (4 × 104 cells per well). Before the insert removal, the cells were cultured until reaching 90% confluency. The formed gap was captured at four different time points every 24 h with an inverted light microscope (Olympus IX71, Tokyo, Japan). The ratio between the remaining and the initial size of the wound area was evaluated using Image-J software (MD, USA). At least three independent experiments were performed.
MTT assay
MTT assay was used for the assessment of the metabolic activity of the cells, which reflects the quantity of viable cells. Reverse transfected SW620 and Caco-2 cells were seeded into a 96-well plate (5000 cells per well). 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma-Aldrich Co., MO, USA) (final concentration 0.5 mg/ml) was added into each well 72 h after the seeding. Cells were incubated with MTT reagent for another 2 hours and after that, the medium with reagent was removed. Formed formazan crystals were dissolved using dimethyl sulfoxide (DMSO; Carl Roth GmbH + Co., Karlsruhe, Germany) (200 Ul). Light absorbance was detected with Sunrise plate reader (Tecan, Männedorf, Switzerland) at 570 nm wavelength and a reference wavelength at 620 nm. At least three independent experiments were performed.
Colony formation assay
Both reverse transfected Caco-2 and SW620 cell lines were seeded into a 6-well plate (250 cells per well). After 2 weeks of incubation with complete medium, the cells were washed with PBS, fixed with 10% formaldehyde (Sigma Aldrich, Germany) for 20 min and stained with 1% crystal violet (Alfa Aesar by Thermo Fisher Scientific, Kandel, Germany) for 15 min at room temperature. The number of colony-forming units was calculated using ImageJ software (USA) in three independent experiments.
Target prediction
Putative target genes, having 8-mer binding sites for hsa-miR-1246 were retrieved from the TargetScan v7.2 [28] database. Since the hsa-miR-1246 showed increased expression levels (oncogenic) in CRC, the targets were selected based on their function in carcinogenesis, i.e., targets having tumor-suppressive function were chosen for further investigation.
Protein extraction and Western Blot
Reverse transfected SW620 and Caco-2 cells were seeded in 6-well plates (1 × 106 cells per well) for protein expression experiments. 72 h after transfection, cells were lysed using 1× radioimmunoprecipitation assay (RIPA) buffer (Abcam, Cambridge, UK) containing phosphatase and protease inhibitor cocktail (Sigma Aldrich, USA). Total protein concentration was evaluated using Pierce BCA Protein Assay Kit (Thermo Fisher Scientific, USA). The extracted protein was fractionated by SDS-PAGE using 4–12% Bis-Tris Plus Mini Gels and transferred to 0.45 µm PVDF membrane. PVDF membranes were blocked using WesternBreeze™ Blocker/Diluent (Part A and Part B) (Thermo Fisher Scientific, USA) at room temperature for 1 h. Antibodies directed against AXIN2 (1:1000 dilution; Cat. No. ab109307; Abcam, UK), CFTR (1:500 dilution; Cat. No. ab2784, Abcam, UK) and GAPDH (0.4 µg/ml concentration; Cat. No. AM4300; Ambion by Thermo Fisher Scientific, CA, USA) were used. Protein signals were detected and visualized using ChemiDocTM XRS + System (Bio-Rad, CA, USA) and ImageLab Software v5.2.1 (Bio-Rad, USA). GAPDH protein was used as an endogenous control. At least three independent experiments were performed.
Dual light luciferase assay
Wild-type (WT) and mutant (MT) seed region sites of hsa-miR-1246 containing 3’UTR sequence of target genes were constructed and cloned into pMIR-REPORT-Luciferase vector (Invitrogen, USA) (Fig. 3B). Constructed insert sequences were verified by Sanger sequencing using BigDye Terminator v3.1 kit (Applied Biosystems, USA) and 3500 Series Genetic Analyzer (Applied Biosystems, USA). Previously well-established workflow for dual luciferase assay using AGS cells was applied for this experiment [29, 30]. Cells were co-transfected with hsa-miR-1246 mimic and miR-Control, pMIR-REPORT-Luciferase-WT vector, pMIR-REPORT-Luciferase-MT vector, and pMIR-REPORT-ß-galactosidase control vector (Invitrogen, USA). Luciferase activity was detected after 48 h of incubation by the Dual-Light™ Luciferase & β-Galactosidase Reporter Gene Assay System (Invitrogen, USA) and normalized with ß-galactosidase activity using Tecan Genios Pro (Tecan, Switzerland). At least three independent experiments were performed.
Statistical analysis
All data from the functional experiments in CRC cell lines are given as means ± standard deviation (SD) of at least three independent experiments. Data between groups were compared using a two-tailed Student’s t-test for the normally distributed data or the Mann-Whitney U test for the not normally (shown by Shapiro-Wilk test) distributed data. All statistical calculations were performed with R Studio software version v3.6.0 (MA, USA). False discovery rate (FDR) corrected P < 0.05 was considered statistically significant.