Interest in single-cell analysis is growing rapidly, and for good reasons. Not every cell—even in the same tissue—is alike, and the differences among them can reveal heretofore unknown biology. Researchers, once content with (or having settled for) bulk measurements are asking which genes or proteins are being expressed together in any given cell, for example, or what genetic changes have taken place as a cancer progresses.
Single-cell analysis is not new, of course. Antibody-based methods such as flow cytometry and immunohistochemistry (IHC), in-situ hybridization (ISH) methods to look at nucleic acids, and other imaging techniques, have allowed a view into the workings of individual cells for some time now. Yet the thirst to explore more parameters, often at greater throughput, than these allow—along with the advances in sequencing and other technologies and the accompanying bioinformatics—has led to platforms and insights only imagined a short time ago. In this article, we will review some of the methods by which single cells are being analyzed, including multiomic combinations.
Why single cells?
Many questions cannot be easily answered by averaging the measurements from a population of cells. The most obvious of these, perhaps, are questions about the diversity of the population itself—how many different sub-populations does it comprise? Are these distinct or continuous, and is one a transition to another? Needles in haystacks—rare cell types such as cancer stem cells or circulating tumor cells—can be identified, and their phenotypes examined. For example, single-cell RNA sequencing (scRNA-seq) has allowed researchers to find sub-populations associated with drug resistance or self-renewal.1 Cell lineages can be tracked though development and disease.
There are many different ways to analyze single cells. “Going from the bottom up: genomics, epigenomics (chromatin state, methylation state of the cell), through to transcriptomics, and even at the protein level,” says Ronald Lebofsky, senior scientist at Bio-Rad Laboratories. “And then you start to see combinations of those unitary analyses.”
The key to single-cell analysis is keeping all of the pieces of information from a cell together.
The key to single-cell analysis is keeping all of the pieces of information from a cell together. If the cells are to be mixed together then they either need to remain intact (as is the case with multiple fluorescently tagged cells seen under a microscope or run through a flow cytometer, for example), or prior to their mixing those pieces of information must be identified as having come from the same cell. The latter is the principle behind barcoding, which adds a unique nucleic acid index sequence to all the pieces of the cell’s genome or RNA transcripts. For plate-based protocols where single-cell reactions take place in microwells, this becomes an issue when the readout is to be pooled, as is generally the case with next-generation sequencing (NGS). The principle is the same for droplet-based techniques, in which the reactions (including barcoding) take place on a single cell in an isolated droplet before the contents are allowed to mix.
scRNA-seq
Of all the single-cell techniques, scRNA-seq “—doing gene expression profiling by transcriptomics at the single-cell level—is the most useful for people,” opines Mike Lucero, product fellow at 10X Genomics. 10X’s droplet-based Chromium™ Single Cell 3′ Solution allows hundreds to millions of transcriptomes to be profiled, with NGS library preparation in under a workday. “That’s why we did that first—it’s the most mature, and it also has the most utility.”
Amplification of a single cell’s-worth of RNA is a relatively high-throughput and straightforward process compared with that of DNA. And there are currently no comparable methods to amplify other analytes such as proteins, lipids, or metabolites, which may be present in quantities below the typical affinity range of standard detection reagents (and while mass spectrometry may allow their detection it falls short in terms of throughput).
The Illumina® Bio-Rad® Single Cell Sequencing Solution works with the BaseSpace Sequence Hub (pictured above) which has completely integrated sequencing and data analysis. Image courtesy of Bio-Rad.
ddSEQ is Bio-Rad and Illumina’s collaborative “turnkey kind of solution” to droplet-based scRNA-seq, backed by “several months to years of product development involving key robustness studies,” says Lebofsky.
There are also academically developed droplet scRNA-seq approaches. For example, “Drop-seq is open-source, all the reagents and tools that you need to process the data and also acquire the data are freely available or available for purchase—it’s kind of a DIY, and much cheaper, alternative,” says one of its inventors Evan Macosko, now an assistant professor of psychiatry at the Broad Institute of MIT and Harvard.2 Protocols and an equipment list can be found at www.dropseq.org. “It takes maybe a couple of weeks to learn how to process and do the actual run. But our protocol has been downloaded 50,000 times, so somebody’s doing it and figuring it out, and it’s not just microfluidics experts.”
Depending on the chemistry involved, scRNA-seq can look at just the 3’ ends of transcripts or full-length (longer read) sequencing. The former is best for quantitation of transcripts, while the later lends itself to identification of splicing isoform variations.
Biocompare’s Flow Cytometry Search Tool
Find, compare and review flow
cytometers from different suppliers Search
Rather than querying the entire transcriptome, BD’s new Rhapsody™ Single-Cell Analysis System “attempts to reduce overall experimental cost using a targeted amplification approach, whereby pre-selected genes may be used to amplify genes …that are thought to most contribute to understanding cellular differences,” says product manager Kieren Patel. Oligo targets are attached to magnetic beads that capture the transcripts of interest and serve as barcoded primers for the library preparation. Targeting greatly reduces the cost of sequencing, as well as the complexity of analysis.
There are well-based commercial, contract, and academic platforms that process tens to hundreds of cells. These tend to be far less expensive on a per-run basis (but considerably higher per-cell). Yet “for a lot of applications, numbers really matter” for Macosko. “When you get enough statistical power you really start to see the patterns.”
Other and multiple parameters
Of course, many researchers are interested in more than just how many of which transcripts a given cell expresses. Whereas the genome can be assumed identical across normal somatic cells from a given individual, for example, this is not the case in oncology and immunology, and single-cell genomics can give answers that cannot be garnered from bulk measurements.
Techniques such as flow cytometry and mass cytometry (a sort of fusion of flow cytometry with mass spectrometry) can simultaneously examine a considerable number of distinct protein epitopes for which antibodies are available—including some post-translational modifications and proteolysis products—in millions of single cells.
And there are techniques to combine some of these single cell measurements as well. For example, RNA and DNA were simultaneously measured back in 2005. And just this summer two groups reported techniques termed REAP-seq and CITE-seq in which antibodies conjugated to oligos were used to look at both protein and RNA in single cells.3
For now, sequencing is destructive of the cell, so “the best you can do is integrate with surface markers,” remarks Macosko. He predicts that soon both upfront and downstream selection—like gating on a flow cytometer—will complement the current state-of-the-art single cell analysis techniques.
References
1. Yuan GC, et al., “Challenges and emerging directions in single-cell analysis,” Genome Biol, 18(1):84, May 8 2017. [PMID: 28482897]
2. Macosko EZ, et al., “Highly Parallel Genome-wide Expression Profiling of Individual CellsUsing Nanoliter Droplets,” Cell, 161(5):1202-1214, May 21 2015. [PMID: 26000488]
3. Stoeckius M, et al., “Simultaneous epitope and transcriptome measurement in single cells,” Nat Methods, 14(9):865-868, Sept 2017. [PMID: 28759029]
4. Peterson VM, et al., “Multiplexed quantification of proteins and transcripts in single cells,” Nat Biotechnol, Aug 30, 2017, [Epub ahead of print]. [PMID: 28854175]
Image: Shutterstock Images