A Novel Multi-Gene Detection Platform for the Analysis of miRNA Expression

Abstract

The study of miRNAs and their roles as non-invasive biomarkers has been intensely conducted in cancer diseases over the past decade. Various platforms, ranging from conventional qPCRs to Next Generation Sequencers (NGS), have been widely used to analyze miRNA expression. Here we introduced a novel platform, PanelChip™ Analysis System, which provides a sensitive solution for the analysis of miRNA levels in blood. After conducting miRQC analysis, the system's analytical performance compared favorably against similar nanoscale qPCR-based array technologies. Because PanelChip™ requires only a minimal amount of miRNA for analysis, we used it to screen for potential diagnostic biomarkers in the plasma of patients with oral cavity squamous cell carcinoma (OSCC). Combining the platform with a machine learning algorithm, we were able to discover miRNA expression patterns capable of separating healthy subjects from patients with OSCC.

Figure 1

PCR efficiency of 9 representative miRNA assays on miRSCan™ PanCancer Chips. Real-time qPCR was carried out for each miRNA cluster using 3-fold serially diluted cDNA template synthesized from Universal Human miRNA Reference RNA (UHmiRR; miRQC A). The resulting Cq values were plotted against the respective miRNA concentrations to derive the PCR efficiency for each assay. All 9 assays fell within acceptable PCR efficiency of 90–110%.

Figure 2

Correlation between miRQC sample replicates. Real-time qPCR was carried out on duplicate miRQC samples to evaluate the reproducibility between miRSCan™ PanCancer chips. Correlation analysis was performed using the resulting two qPCR datasets of duplicate samples (samples 1, 3, 5, 7 for replicate 1 and samples 2, 4, 6, 8 for replicate 2). Filled circles represented data from replicate 1, and open circles represented data from replicate 2. Correlation of 0.91 indicates that miRSCan™ PanCancer chips have high reproducibility.

Figure 3

Expression distribution of detectable miRNAs from double positive replicates. Real-time qPCR was carried out on duplicate miRQC samples. Double positives are miRNAs detected in both duplicates. 131 double positives out of 164 total miRNA candidates were detected after applying the detection cut-off (Cq < 34). Expression range of the double positives is 18.64 log2-units, showing the system’s ability to detect a wide range of template concentrations.

Figure 4

Dynamic range of known concentration of spike-in synthetic has-miR-10a-5p. A set of 10-fold serial dilutions of synthetic hsa-miR-10a-5p miRNA were spiked into the same amount of Universal Human miRNA reference RNA (samples 16–22, miRQC A) to generate cDNA for qPCR analysis. 20 ng of the RNA input was used for the spike-in test. qPCR results illustrated a dynamic range of at least 7 orders of magnitude ranging from 80 to 8 × 10⁷ copies per nanowell.