Developments in Single-Cell RNA-Sequencing Technologies

Today scientists use single-cell RNA sequencing (scRNA-seq) to answer a wide range of research questions—many of which hinge on the identities or types of individual cells. Single-cell measurements can uncover crucial information about cell heterogeneity. For instance, transcriptomic profiles can identify distinct cell types or cell populations, a crucial step for studying complex microenvironments such as tumors that contain multiple cell types. The Human Cell Atlas project maps single cells in human tissues according to RNA expression and other criteria, for an invaluable reference of normal, healthy tissues. Single-cell RNA sequencing also allows the study of distinct immune cell populations during checkpoint immunotherapy treatment for cancer.

Great interest in, and fast growth of, scRNA-seq technologies over the past several years has resulted in numerous scRNA-seq protocols. Generally, they begin with single-cell isolation, and the labeling of mRNAs with nucleic acid tags. Reverse transcription and library preparation occur prior to next-generation sequencing (NGS). After sequencing, the data is demultiplexed to separate the various molecular tags, such as cell barcodes and unique molecular identifiers (UMIs) on individual transcripts, so the transcripts can be counted. The many scRNA-seq protocols available today differ in cell isolation methods, barcoding, length of transcripts sequenced, and amplification methods—but they share the goal of accurately measuring RNA transcripts in single cells.

Reducing cost and improving accuracy

Multiple manufacturers, such as 10x Genomics and NanoString Technologies, sell scRNA-seq kits for library preparation prior to sequencing on one of Illumina’s NGS sequencing platforms. Illumina’s new NovaSeq version 1.5 reagents, for their NovaSeq platform with high-throughput S4 flow cells, improve sequencing while lowering cost. “We’ve optimized the accuracy of our index and UMI reads, which have dramatically improved in the v1.5 reagents,” says Gary Schroth, VP and distinguished scientist of emerging applications department at Illumina. Illumina has also launched a new short-read flow cell that NanoString is using for whole transcriptome analysis in spatial genomics applications, which lowers the cost to $1–1.50 per million reads.

Lower sequencing costs often mean a new decision for researchers. “There comes a point where each researcher must decide whether sequencing more transcripts per cell is advantageous,” says Schroth, referencing the principle of diminishing returns due to the 10–20% of genes that are very highly expressed, and the expense of additional sequencing. “With a drop in sequencing costs comes the opportunity for Illumina users to sequence more deeply with greater sensitivity for the same price,” he says. “Deeper sequencing improves your counting statistics, your detectability, and your low end.” Of course, not everyone needs deeper sequencing—for users who are classifying cell types, for instance, sequencing the first 4,000 genes may be sufficient. Higher throughput sequencing capacities also mean longer data analysis times. Illumina has improved its informatics pipeline, DRAGEN, to speed the conversion of initial sequencing data (in the form of BCL files) into FASTQ files for analysis. This shortens the overall data analysis timeline, so researchers get the information they want sooner, in terms of number of transcripts of each gene per cell.

Multiplexing and targeting

Researchers now have greater choices for more flexibility, such as multiplexing or targeting. In March 2021, 10x Genomics released a single-cell sample multiplexing tool called CellPlex for multiplexing up to 96 samples per chip. The CellPlex assay adds oligo-tagged lipid labels that bind to cell or nuclear membranes. “Adding CellPlex tags to the 10x Genomics workflow enables the scaling of up to 30k cells per channel and 240k cells per chip,” says Michael Schnall-Levin, senior VP of research and development and founding scientist of 10x Genomics.

“CellPlex is compatible with the Single Cell Gene Expression assay, and will also integrate into our high-throughput kits that will be released together with the Chromium X instrument in the second half of 2021.” Last year 10x Genomics released their Targeted Gene Expression assay for focusing on transcripts of interest without deep sequencing. “It can reduce sequencing costs up to 90% and is compatible with the Single Cell Gene Expression, Single Cell Immune Profiling, and Spatial Gene Expression workflows,” says Schnall-Levin.

An important application for single-cell transcriptomics is the profiling and typing of immune cells. 10x Genomics recently released a new version of their Single Cell Immune Profiling assay for their Chromium platform. “It substantially improves gene sensitivity of up to 60%, leading to novel cell discoveries and lower spend per reaction,” Schnall-Levin says. A research group headed by Diether Lambrechts, at the Katholieke University Leuven, recently examined immune responses of mild versus critical COVID patients using the Single Cell Immune Profiling assay. They found evidence of impaired T cell effector functions and macrophage differentiation that may shed light on the development of severe COVID.

Droplets and nanowells

Many scRNA-seq applications are well-served by high-throughput, droplet-based systems such as the commercial 10x Genomics platform or open-source systems like inDrop and Drop-seq. In these systems, individual droplets enclose single cells, serving as closed vessels for the reverse transcription reaction. Other applications, however, can benefit from a different configuration.

Takara Bio’s ICELL8 Single Cell System uses nanowells instead of droplets to isolate single cells for sequencing reactions. Despite having lower throughput compared to droplet-based platforms, it offers more flexibility in chemistry and applications while sequencing hundreds of cells in parallel. “Only one reaction can occur in a droplet, but you can do multiple stacking reactions in a nanowell because it’s an open container,” says Suvarna Gandlur, associate director for NGS marketing at Takara Bio. “Nanowells are best when you want to do complex reactions that add multiple reagents as time progresses.” Reactions are performed on the ICELL8 chip, which prepares a library ready for an Illumina sequencer.

The ability to perform multiple chemical reactions allows you to sequence full-length transcripts from single cells. “Making a library out of every fragment of a long transcript is only possible in a well-based method,” says Gandlur. Unlike droplet methods, which capture only one end of each transcript, “well-based methods can reveal the full length of transcripts, which give you information on gene fusions and isoforms.” Using nanowells is also particularly valuable for RNA sequencing of single cells that are too large for droplet-based methods, such as cardiomyocytes. “Even small, specialized groups of cells, such as organoids, can be isolated in nanowells and sequenced,” she adds.

Spatial transcriptomics

Spatial transcriptomics, where transcripts are detected while retaining their spatial information within a tissue under study, has seen a quick rise in interest. Nature Methods chose spatially resolved transcriptomics as “Method of the Year” for 2020, due to its ability to reveal heterogeneous cell types in situ. These quickly evolving technologies are supported by platforms such as NanoString’s GeoMx Digital Spatial Profiler and 10x Genomics’s Visium Spatial Gene Expression. Augmenting spatial transcriptomics with co-localization of DNA and/or protein expression opens further opportunities for applications.

The resolution of spatial transcriptomics is improving beyond single cells. For example, Resolve Biosciences’s new Molecular Cartography platform uses microscopy and 3D imaging with proprietary chemistry and bioinformatics capabilities to increase resolution for subcellular transcriptomics. “Our customers are now able to understand where molecular processes occur in a tissue sample,” says Jason Gammack, co-founder and CEO of Resolve Biosciences. “The dataset generated from Molecular Cartography is rich in information with the precise 3D coordinates of millions of transcript copies in a given tissue sample.” Madeline Lancaster’s research team, at the MRC Laboratory of Molecular Biology in Cambridge, used the platform to quantify the spatiotemporal expression of 100 transcripts at subcellular resolution in brain organoids.

Schroth notes a recent change in the users of single-cell and spatial transcriptomics. “Sometimes these are cell biologists coming into genomics for the first time,” he says. “10x Genomics and and NanoString have put together really good end-to-end solutions that help customers with the choices you need to make, so you don't have to be a technology savvy person to do single-cell genomics or transcriptomics anymore.” Indeed, with manufacturers making these tools accessible to researchers without expertise in sequencing, the applications of scRNA-seq have only just begun.