Editors' ChoiceGenomics

Single-cell nuclei rise from the dead

See allHide authors and affiliations

Science Translational Medicine  06 Sep 2017:
Vol. 9, Issue 406, eaao6120
DOI: 10.1126/scitranslmed.aao6120

Abstract

High throughput, single-cell transcriptomic profiling can rapidly profile frozen mouse and human tissues.

Single-cell transcriptomic profiling is rapidly advancing in throughput while its cost declines thanks to innovative single cell manipulation technologies, clever barcoding, and next generation sequencing. Microfluidics recently expanded throughput by enabling isolation, lysis, and RNA capture in continuous trains of aqueous nanoliter droplets, making it possible to profile >10,000 single cells in one experiment. Unfortunately, a focus on fresh cells left behind the vast collection of potentially informative frozen tissue samples. Nuclei from archived samples are teeming with RNA transcripts that have been successfully sequenced with low throughput techniques. Now, Habib and colleagues leverage the scalability of droplet-based microfluidics to perform high-throughput single-nucleus RNA-Seq on frozen archived mouse and human samples in a method they call DroNc-Seq.

They began by optimizing microfluidic channel dimensions to maximize the number of transcripts detected and minimize the occurrence of more than one nucleus in a single droplet. On average, nuclei yielded fewer genes than intact cells did, but throughput and capture efficiency were comparable. Expression of most genes was highly correlated between single nuclei and cells, but lncRNA and mitochondrial transcripts were strongly enriched in nuclei and cells respectively, as expected. Approximately half of single nuclei reads mapped to introns, which the authors attribute to nascent transcripts and did not explore in detail. Using shallow sequencing, the authors focused on exon-mapped reads of nuclei from murine frozen brain tissue. They also profiled single nuclei from postmortem human samples that had been frozen for up to 6 years. Despite the variable quality of some samples, they were able to measure gene expression in nuclei of neuronal and nonneuronal cells from both organisms. In both humans and mice, clustering of single nuclei transcriptomes mapped to known cell subsets in anatomically distinct brain regions and compared well with previously reported results from low-throughput single nuclei sequencing.

Single-cell sequencing technologies were initially limited to fresh samples processed at the time and location of collection. Now, by enabling single-cell profiling of frozen samples, techniques such as DroNc-Seq allow archived tissue samples to rise from the dead and teach us about human health, disease, and the effects of our therapies.

Highlighted Article

View Abstract

Navigate This Article