Brain and Spinal Cord Cell Profiling with Barcoding

 

Science  13 Apr 2018: Vol. 360, Issue 6385, pp. 176-182

Single-cell profiling of the developing mouse brain and spinal cord with split-pool barcoding

Alexander B. Rosenberg, et.al.

Department of Electrical Engineering, University of Washington, Seattle, WA, USA.

Department of Bioengineering, University of Washington, Seattle, WA, USA.

Allen Institute for Brain Science, Seattle, WA, USA.

Molecular Engineering and Sciences Institute, University of Washington, Seattle, WA, USA.

Institute for Stem Cell and Regenerative Medicine, Seattle, WA, USA.

Paul G. Allen School of Computer Science and Engineering, University of Washington, Seattle, WA, USA.

[paraphrase]

To facilitate scalable profiling of single cells, we developed split-pool ligation-based transcriptome sequencing (SPLiT-seq), a single-cell RNA-seq (scRNA-seq) method that labels the cellular origin of RNA through combinatorial barcoding. SPLiT-seq is compatible with fixed cells or nuclei, allows efficient sample multiplexing, and requires no customized equipment. We used SPLiT-seq to analyze 156,049 single-nucleus transcriptomes from postnatal day 2 and 11 mouse brains and spinal cords. More than 100 cell types were identified, with gene expression patterns corresponding to cellular function,   regional specificity,   and stage of differentiation. Pseudotime analysis revealed transcriptional programs driving four developmental lineages, providing a snapshot of early postnatal development in the murine central nervous system. SPLiT-seq provides a path toward comprehensive single-cell transcriptomic analysis of other similarly complex multicellular systems.

More than 300 years have passed since van Leeuwenhoek first described living cells, yet we still do not have a complete catalog of cell types or their functions. Recently, transcriptomic profiling of individual cells has emerged as an essential tool for characterizing cellular diversity. Single-cell RNA-sequencing (scRNA-seq) methods have profiled tens of thousands of individual cells, revealing new insights about cell types within both healthy and diseased tissues. Unfortunately, since these methods require cell sorters, custom microfluidics, or microwells, throughput is still limited and experiments are costly. We introduce split-pool ligation-based transcriptome sequencing (SPLiT-seq), a low-cost, scRNA-seq method that enables transcriptional profiling of hundreds of thousands of fixed cells or nuclei in a single experiment. SPLiT-seq does not require partitioning single cells into individual compartments (droplets, microwells, or wells) but relies on the cells themselves as compartments. The entire workflow before sequencing consists just of pipetting steps, and no complex instruments are needed.

In SPLiT-seq, individual transcriptomes are uniquely labeled by passing a suspension of formaldehyde-fixed cells or nuclei through four rounds of combinatorial barcoding. In the first round of barcoding, cells are distributed into a 96-well plate, and cDNA is generated with an in-cell reverse transcription (RT) reaction using well-specific barcoded primers. Each well can contain a different biological sample, thereby enabling multiplexing of up to 96 samples in a single experiment. After this step, cells from all wells are pooled and redistributed into a new 96-well plate, where an in-cell ligation reaction appends a second well-specific barcode to the cDNA. The third-round barcode, which also contains a unique molecular identifier (UMI), is then appended with another round of pooling, splitting, and ligation. After three rounds of barcoding, the cells are pooled and split into sublibraries, and sequencing barcodes are introduced by polymerase chain reaction (PCR). This final step provides a fourth barcode, while also making it possible to sequence different numbers of cells in each sublibrary. After sequencing, each transcriptome is assembled by combining reads containing the same four-barcode combination.

Four rounds of combinatorial barcoding can yield 21,233,664 barcode combinations (three rounds of barcoding in 96-well plates followed by a fourth round with 24 PCR reactions), enough to uniquely label over 1 million cells. Even larger numbers of barcode combinations can be achieved by performing experiments in 384-well plates or through additional rounds of barcoding. In addition, by performing the first step in a 384-well plate, up to 384 different biological samples could be combined in a single experiment.

In this work, we profiled hundreds of thousands of cells using only basic laboratory equipment with a library preparation cost of ~$0.01 per cell. In our analysis of more than 150,000 single-nucleus transcriptomes from two early postnatal stages, we identified 69 types of cells in the brain and 44 types in the spinal cord. We defined many new molecular markers for specific cell types and explored gene expression in four different developmental lineages.

SPLiT-seq’s compatibility with fixed cells and fixed nuclei overcomes challenges faced by other scRNA-seq methods. Fixation can reduce perturbations to endogenous gene expression during cell handling and makes it possible to store cells for future experiments. Moreover, the use of nuclei bypasses the need to obtain intact single cells, which can be challenging for many complex tissues. SPLiT-seq’s compatibility with formaldehyde-fixed nuclei suggests that it may be used to profile single nuclei from formalin-fixed, paraffin-embedded tissue.

SPLiT-seq enables flexible and scalable cell and sample multiplexing. The use of the first-round barcode as a sample identifier makes it possible to profile a large number and variety of samples in parallel, thus minimizing batch effects. As the number of unique barcodes grows exponentially with the number of barcoding rounds, larger numbers of cells than presented here could be processed by adding a fifth barcoding round or by switching to a 384-well plate format. Although for such large cell numbers, sequencing cost may currently be forbidding, it is easy to imagine applications, such as targeted sequencing of gene panels, which would even now benefit from very large cell numbers and only require shallow sequencing depth.

Our hope is that the increased scale and accessibility provided by the low cost and minimal equipment requirements of SPLiT-seq will further accelerate the widespread adoption of scRNA-seq.

[end of paraphrase]

 

Return to — Embryonic Development of Brain