Single-cell RNA sequencing (RNA-seq) is a powerful tool to reveal cellular heterogeneity, discover new cell types and characterize tumor microevolution. However, losses in cDNA synthesis and bias in cDNA amplification lead to severe quantitative errors. We show that molecular labels--random sequences that label individual molecules--can nearly eliminate amplification noise, and that microfluidic sample preparation and optimized reagents produce a fivefold improvement in mRNA capture efficiency.

Our understanding of the development and maintenance of tissues has been greatly aided by large-scale gene expression analysis. However, tissues are invariably complex, and expression analysis of a tissue confounds the true expression patterns of its constituent cell types. Here we describe a novel strategy to access such complex samples. Single-cell RNA-seq expression profiles were generated, and clustered to form a two-dimensional cell map onto which expression data were projected. The resulting cell map integrates three levels of organization: the whole population of cells, the functionally distinct subpopulations it contains, and the single cells themselves—all without need for known markers to classify cell types. The feasibility of the strategy was demonstrated by analyzing the transcriptomes of 85 single cells of two distinct types. We believe this strategy will enable the unbiased discovery and analysis of naturally occurring cell types during development, adult physiology, and disease.

Single-cell analysis of gene expression is increasingly important for the analysis of complex tissues, including cancer, developing organs and adult stem cell niches. Here we present a detailed protocol for quantitative gene expression analysis in single cells, by the sequencing of mRNA 5' ends. In all, 96 cells are lysed, and their mRNA is converted to cDNA. By using a template-switching mechanism, a bar code and an upstream primer-binding sequence are introduced simultaneously with reverse transcription. All cDNA is pooled and then prepared for 5' end sequencing, including fragmentation, adapter ligation and PCR amplification. The chief advantage of this approach is the great reduction in cost and time, afforded by the early bar-coding strategy. Compared with previous methods, it is more suitable for large-scale quantitative analysis, as well as for the characterization of transcription start sites, but it is unsuitable for the detection of alternatively spliced transcripts. Sample preparation takes 3 d, and two sets of 96 cells can be prepared in parallel. Finally, the sequencing and data analysis can take an additional 4 d altogether.

Reverse transcriptases derived from Moloney Murine Leukemia Virus (MMLV) have an intrinsic terminal transferase activity, which causes the addition of a few non-templated nucleotides at the 3´ end of cDNA, with a preference for cytosine. This mechanism can be exploited to make the reverse transcriptase switch template from the RNA molecule to a secondary oligonucleotide during first-strand cDNA synthesis, and thereby to introduce arbitrary barcode or adaptor sequences in the cDNA. Because the mechanism is relatively efficient and occurs in a single reaction, it has recently found use in several protocols for single-cell RNA sequencing. However, the base preference of the terminal transferase activity is not known in detail, which may lead to inefficiencies in template switching when starting from tiny amounts of mRNA. Here, we used fully degenerate oligos to determine the exact base preference at the template switching site up to a distance of ten nucleotides. We found a strong preference for guanosine at the first non-templated nucleotide, with a greatly reduced bias at progressively more distant positions. Based on this result, and a number of careful optimizations, we report conditions for efficient template switching for cDNA amplification from single cells.

No abstract