Highly multiplexed single-cell quantitative PCR

We present a microfluidic device for rapid gene expression profiling in single cells using multiplexed quantitative polymerase chain reaction (qPCR). This device integrates all processing steps, including cell isolation and lysis, complementary DNA synthesis, pre-amplification, sample splitting, and measurement in twenty separate qPCR reactions. Each of these steps is performed in parallel on up to 200 single cells per run. Experiments performed on dilutions of purified RNA establish assay linearity over a dynamic range of at least 104, a qPCR precision of 15 %, and detection sensitivity down to a single cDNA molecule. We demonstrate the application of our device for rapid profiling of microRNA expression in single cells. Measurements performed on a panel of twenty miRNA in two types of cells revealed clear cell-to-cell heterogeneity, with evidence of spontaneous differentiation manifest as distinct expression signatures. Highly multiplexed microfluidic RT-qPCR fills a gap in current capabilities for single-cell analysis, providing a rapid and cost-effective approach for profiling panels of marker genes, thereby complementing single-cell genomics methods that are best suited for global analysis and discovery. We expect this approach to enable new studies requiring fast, cost-effective, and precise measurements across hundreds of single cells. We analyzed 1454 samples in order to validate the performance of our device for performing single-cell RT-qPCR, and 362 samples in order to demonstrate it's efficacy on single-cells. For the validation experiments, 208 samples were in order to determine the lower sensitivity; 208, 207, and 208 samples were to determine amplification uniformity and mRNA-assay linearity, from each of three technical replicates; and 208, 208, and 207 samples were to determine amplification uniformity and microRNA-assay linearity, from each of three technical replicates. For the single-cell mRNA expression experiments, 175 single cells and 4 no cell controls (NCC) were measured using an assay panel designed to interrogate genes that spanned a wide range of abundance, and included i) three genes typically used as endogenous controls in bulk gene expression experiments (human GAPDH, RPPH1, and TBP), ii) a fusion transcript associated with CML (BCR-ABL), iii) a heterogeneously expressed gene (GGCX), and iv) a negative control (murine Gapdh). For the single-cell microRNA expression experiments, we measured 95 single K562 cells, 81 BaF3 cells, and 7 no cell controls against a panel of miRNA.