Evaluation of two commercial global miRNA expression profiling platforms for detection of less abundant miRNAs

Abstract

Background: microRNAs (miRNA) are short, endogenous transcripts that negatively regulate the expression of specific mRNA targets. miRNAs are found both in tissues and body fluids such as plasma. A major perspective for the use of miRNAs in the clinical setting is as diagnostic plasma markers for neoplasia. While miRNAs are abundant in tissues, they are often scarce in plasma. For quantification of miRNA in plasma it is therefore of importance to use a platform with high sensitivity and linear performance in the low concentration range. This motivated us to evaluate the performance of three commonly used commercial miRNA quantification platforms: GeneChip miRNA 2.0 Array, miRCURY Ready-to-Use PCR, Human panel I+II V1.M, and TaqMan Human MicroRNA Array v3.0.

Results: Using synthetic miRNA samples and plasma RNA samples spiked with different ratios of 174 synthetic miRNAs we assessed the performance characteristics reproducibility, recovery, specificity, sensitivity and linearity. It was found that while the qRT-PCR based platforms were sufficiently sensitive to reproducibly detect miRNAs at the abundance levels found in human plasma, the array based platform was not. At high miRNA levels both qRT-PCR based platforms performed well in terms of specificity, reproducibility and recovery. At low miRNA levels, as in plasma, the miRCURY platform showed better sensitivity and linearity than the TaqMan platform.

Conclusion: For profiling clinical samples with low miRNA abundance, such as plasma samples, the miRCURY platform with its better sensitivity and linearity would probably be superior.

Figure 1

Assessment of the assay specificities of the miRCURY and TaqMan platforms. Utilizing the two synthetic samples #1 and #2 with known miRNA contents enabled us to investigate the false positive rate of the individual platforms at given Ct detection thresholds. The rates for the individual platforms are provided in (A) miRCURY and (B) TaqMan. Clearly, the number of false positives increased with increasing Ct detection threshold indicating that carefully chosen Ct detection thresholds could potentially reduce the number of false positives without affecting the true positives. In order to determine the thresholds for the two platforms the raw Ct's of the synthetic samples were compared to no-RT controls. As illustrated by plotting the raw Ct's for the duplicate measurements of the synthetic sample #1 and the no-RT-control for miRCURY (C) and TaqMan (D) thresholds set at Ct = 38 and Ct = 30, respectively would dramatically reduce the number of false positives. #FP, number of false positives.

Figure 2

Assessment of the platforms ability to discriminate miRNAs with close sequence homology. Using the synthetic samples, with known miRNA content, it was investigated how many of the false positives detections (miRNAs not in the sample) at a given Ct detection threshold could be related to sequence homology to a miRNA in the sample. Plotted are the number of false positives detected in the synthetic samples at given Ct detection thresholds and it is indicated how many of these can be related to sequence homology. (A, B) TaqMan analysis of synthetic samples #1 and #2. (C, D) miRCURY analysis of synthetic samples #1 and #2. A Poisson randomness test was used to evaluate if the fraction of homology related false positives at a given Ct detection threshold was significantly higher than expected. The expected fraction was defined as the number of potential false positives (the number of assays on a given platform targeting miRNAs not present in the investigated sample) that have sequence homology to miRNAs in the sample. As the synthetic samples #1 and #2 contain the same miRNAs (but in different concentrations) the expected fraction is the same for the two samples. For the TaqMan platform the expected fraction is 93 out of 565 (16%) miRNAs and for the miRCURY platform it is 102 out of 552 (18%). **p < 0.0001, *p < 0.05.

Figure 3

Assessment of the capability of the platforms to recover known miRNA copy number differences. The ability of the platforms to recover known four-fold miRNA copy number differences between samples was assessed by calculating ΔCt's for 125 synthetic miRNAs common to both the TaqMan and miRCURY platforms. Shown are boxplots of the ΔCt's obtained for TaqMan (A, C) and miRCURY (B, D). Recovery was assessed both in the synthetic samples i.e. consisting solely of synthetic miRNAs (A, B) and the plasma RNA samples spiked with synthetic miRNAs (C, D). Separate plots were made for miRNAs that were four fold higher (ΔCt = 2) and lower (ΔCt = -2) in sample 1 compared to 2.

Figure 4

Evaluation of assay linearity at low miRNA input levels for 125 spike-in miRNAs common to the TaqMan and miRCURY platforms. Linearity was assessed based on a four point dilution series, ranging four orders of magnitude, of the spiked plasma RNA sample #1. For ease of platform comparison the data presented was restricted to the 125 spiked miRNAs queried by both platforms. The template input per PCR reaction of these 125 miRNAs ranged from 5 - 5,000 template copies (n = 62) and 20-20,000 (n = 63). Each dilution point was measured in duplicate and the linearity of each of the 125 assays was estimated by calculating the squared Pearson correlation coefficient, r² of these measurements. Plotted are the obtained r² values for (A) the TaqMan platform and (B) the miRCURY platform. Dashed lines correspond to number of assays with r² ≥ 0.9.