Two-tailed RT-qPCR panel for quality control of circulating microRNA studies

Abstract

Circulating cell-free microRNAs are promising candidates for minimally invasive clinical biomarkers for the diagnosis, prognosis and monitoring of many human diseases. Despite substantial efforts invested in the field, the research so far has failed to deliver expected results. One of the contributing factors is general lack of agreement between various studies, partly due to the considerable technical challenges accompanying the workflow. Pre-analytical variables including sample collection, RNA isolation, and quantification are sources of bias that may hamper biological interpretation of the results. Here, we present a Two-tailed RT-qPCR panel for quality control, monitoring of technical performance, and optimization of microRNA profiling experiments from biofluid samples. The Two-tailed QC (quality control) panel is based on two sets of synthetic spike-in molecules and three endogenous microRNAs that are quantified with the highly specific Two-tailed RT-qPCR technology. The QC panel is a cost-effective way to assess quality of isolated microRNA, degree of inhibition, and erythrocyte contamination to ensure technical soundness of the obtained results. We provide assay sequences, detailed experimental protocol and guide to data interpretation. The application of the QC panel is demonstrated on the optimization of RNA isolation from biofluids with the miRNeasy Serum/Plasma Advanced Kit (Qiagen).

Figure 1

QC workflow with the Two-tailed QC panel. (A) A mix of three synthetic RNA spike-ins (cel-miR-54, spike-A, spike-B) is added prior to RNA isolation from the biofluid sample. A second mix of two spike-ins (cel-miR-76, cel-miR-2) is added before cDNA synthesis step. Optionally, a diluted isolation spike-in mix is used as a template in a “spike-only” control reaction to determine spike-in baseline signal (for details see Supplementary file section 3.2.2). Two-tailed RT-qPCR is used to quantify the spike-ins along with three endogenous microRNAs (let-7a, miR-23a and miR-451a) to evaluate the technical quality of RNA isolation, effect of inhibition and the level of haemolysis. (B) Decision chart for data interpretation and troubleshooting (see also Supplementary file section 4).

Figure 2

Optimizing input volumes of (A) human plasma, (B) human serum, and (C) rat serum for RNA isolation. Data are presented as ΔCq between Cqs obtained with the tested volume and an input volume of 200 μl (human) or 100 μl (rat). Each dot is one isolation replicate. Optimum starting serum/plasma volumes based on absolute endogenous microRNA yields are 250 μl for human plasma, 300–500 μl for human serum, and 150 μl for rat serum (blue mean profiles). Error bars on mean profiles panels indicate standard deviation (SD).

Figure 3

Effect of glycogen carrier on microRNA quantification in human plasma. Identical sample aliquots were isolated with (n = 5) or without (n = 6) addition of glycogen carrier starting from 200 μl, and quantified with the Two-tailed QC panel. Extractions with glycogen had significantly higher yields (average difference between Cq means: 1.25 cycles; paired T-test p = 0.011) and higher reproducibility (F-test, p < 0.001).

Figure 4

Assessing haemolysis in serum/plasma samples. (A) Human plasma samples with varying degree of haemolysis, corresponding A₄₁₄, ΔCq (miR-23a–miR-451a), and selected UV-Vis spectra. (B) Correlation between ΔCq (miR-23a–miR-451a) and A₄₁₄, A₅₄₀ and A₅₇₈, respectively. Exponential regression line with 95% confidence interval is shown. Dashed red line indicates A₄₁₄ = 0.25 as threshold for non-haemolysed samples. Corresponding ΔCq thresholds are ~15 (human plasma), ~11 (human serum), and ~6 (rat serum).