Single-cell sequencing unlocks a new level of specificity for those researching the transcriptome. But what technologies are right for a given experiment, and how should your samples inform that decision?
In this article, we review the basics of single-cell, including how it improves on bulk RNA sequencing for heterogeneous sample analysis. We then discuss limitations of the single-cell methods, as well as six questions to help researchers select the appropriate platform for their project. These questions are applied to several of the single-cell technologies offered in Psomagen labs: 10x Genomics and Parse Biosciences, as well as other tools that support single-cell research.
Single-cell sequencing makes it possible to investigate the transcriptome of individual cells in a sample. This technique can distinguish between transcript expression across different types of cells. It can also be used to identify rare transcripts, mutations, and chromatin accessibility in individual cells.
To understand why single-cell sequencing is a breakthrough technology, it’s important to understand where it improves upon existing RNA sequencing options. In bulk RNA sequencing, RNA from all cells in the sample is collected. However, bulk RNA sequencing generates an average transcriptomic readout. Data does not correspond to the transcriptome of any individual cell in the sample.
Bulk RNA sequencing isn’t the best choice for investigating heterogeneous samples. Each cell type in an organism plays a different role in biological functions, and therefore has differentially expressed RNA. Cells in different stages of their life cycle also have different RNA expression levels. None of these distinctions are visible to researchers in bulk RNA analysis.
Single-cell sequencing, by comparison, provides valuable information about each individual cell’s RNA expression. This has been invaluable in studies involving:
Oncology research, where cancerous tumors and the tumor microenvironment (TME) are heterogeneous and complex. TME cells can be important factors in tumorigenesis and growth, as well as patient prognosis, even if very few of a cell type are present. In a 2021 study out of Germany, for example, single-cell technology helped scientists classify lung adenocarcinomas into two distinct microenvironment patterns directly related to prognostic values.
Neurological disorders, where a previously under-researched diversity of neuronal and glial cells impacts the development and progression of brain disorders. In research on Alzheimer’s disease and other neurological conditions, single-cell sequencing of liquid biopsy material has revealed valuable insights into brain health. Peripheral blood and cerebrospinal fluid samples from multiple cohorts in a 2020 study underwent single-cell sequencing. This revealed enhanced T cell signaling and clonally expanded CD8+T cells in patients with Alzheimer’s disease. These discoveries would not be possible without being able to analyze heterogeneous samples.
Cardiovascular disease, where cellular changes in the heart contribute to the development of disease. Single-cell has been used to classify cells in healthy and pathological cardiovascular tissues. It has also been used to generate cell atlases and to identify new cell types implicated in cardiovascular disease.
Genetic mosaicism, where mutations lead to an organism derived from a single zygote having multiple genotypes represented in their cells. Single-cell has made it possible to detect mosaicism in more biological contexts than previously possible. In breast and ovarian cancer treatment, for example, mosaicism in the BRCA1 and 2 genes is an important factor in disease progression. Single-cell analysis was able to estimate when the BRCA2 mutation occurred in an ovarian cancer patient. This helped to show that early BRCA2 mosaicism can drive tumorigenesis.
Cellular aging, a process which varies across cell types. AgeAnno, for example, is a knowledge base of single cell annotation of human aging. Using single-cell and ATAC-seq data, researchers compiled data on aging-related genes across human tissue-cell types.
Limitations of Single-Cell Sequencing
Single-cell sequencing is not a perfect solution for every project. Although its limitations can impact the success of a single-cell sequencing project, they do not negate the insights gained from analyzing individual cells. Paying close attention to the capabilities of your technology of choice will help ensure you mitigate these limits.
Different single-cell platforms have different average cell capture rates. There can be a drastic difference between the number of cells input into a single-cell assay and the number of cells that reach the analysis stage. For example, 10x Genomics captures roughly 60 to 70% of input cells; with Parse Biosciences, the capture rate is closer to 25–30%. Understanding cell capture rates will ensure you get out of your assay what you need for impactful analysis.
The complexity of single-cell sequencing technologies and the need for specialized instruments lead to an increased cost over bulk sequencing methods. Depending on the context of your project, sample preparation and collection, optimization, library preparation and sequencing all add to the total cost. The sticker shock associated with switching to single-cell is a major roadblock to many researchers.
However, advancements in available technology are decreasing the cost of single-cell. Parse Biosciences, for example, offers a highly multiplexed, instrument-free protocol that eliminates the need for specialized instrumentation. This option can significantly decrease the cost of sample analysis compared to other methods.
Multiplexing options allow researchers to label and pool samples prior to encapsulation in a 10x assay. Known as cell hashing, this method, which uses tagged antibodies from BioLegend, makes it possible to run multiple samples in one lane. This significantly reduces the cost of library preparation.
Thousands of unique genes can be expressed in a sample at any given time. In bulk sequencing, all of these genes will be represented in your sequencing output — however, all of those genes are not present in any individual cell. With single-cell, you can pinpoint which cells are expressing which genes.
However, limitations to single-cell sequencing mean you will not see 100% of expressed genes in your data. Many factors impact the efficacy of gene capture, including:
Primers used for transcript capture. Encapsulation methods like 10x Genomics utilize an oligo dT primer covered bead to capture mRNAs. Limited physical space and interaction can restrict gene capture.
Sequencing depth. If the samples are not sequenced at a sufficient depth of coverage, some less-expressed transcripts will not be evident in sequencing data above background.
Like the rest of our challenges surrounding single-cell sequencing, technological advancements are addressing this issue. Changes in 10x Genomics protocols between the HT and GEM-X technologies, for example, allow for improved gene capture of fragile cells, rare transcripts, and cells with low RNA content. GEM-X boasts up to 80% gene capture, compared to 65% on Next GEM.
In some instances, Parse Biosciences gene capture is an improvement on other processes. Parse Biosciences uses both 3’ polyA capture and random priming to capture both mRNAs and some noncoding RNA. Depending on your sample type, one of these newer technologies may be a good choice to ensure sufficient gene capture.
Maintaining sample viability and the biological accuracy of a sample’s transcriptome is the key component ensuring successful single-cell sequencing. Many factors impact the success of this process, including:
Sample preparation and collection steps, which can cause cell loss, damage, or unwanted transcriptional changes.
Presence of dead cells and debris which can confound downstream data.
Choosing the right cell capture methods for your sample type.
Proper project design to overcome noise.
Defining the scope of your investigation is a key first step in choosing the correct single-cell platform. Whether you are interested in RNA, DNA, protein, or some combination of the three, that will impact your choice of platform. Variables in this decision will impact what happens downstream during your project.
Another part of this question is: how are you collecting your cells? We’ve already discussed how cell capture rate decreases the total number of sequenceable cells. If you are working with a rare or precious sample, this will impact which technology is the best fit.
Similarly, preservation and shipment of your samples will decrease cell count. If cryopreservation is necessary, that process and the washing steps that follow will decrease your cell count by 50%.
Some workflows, like Parse, don’t require this cryopreservation step. Immediate fixing of cells allows you to head straight into the assay without performing these washes.
The organism you’re studying plus the focus of your study will also determine which technology is most feasible. Many assays support only human or mouse, with fewer being applicable to any species.
Dead and dying cells provide an uninformative transcriptional profile, but these cells cannot be filtered out of a library prior to sequencing. In order to minimize wasted sequencing cost, it is important to remove these cells from your sample. Additionally, debris from these cells can be encapsulated with a live cell, captured, and sequenced, confounding your overall data set. While analysis packages are available to aid in removal of this background, utilizing bioinformatics to “save” your data isn’t the best approach. Putting a high quality sample into the prep is the optimal way to yield high quality, reproducible results.
With the LevitasBio LeviCell system, it’s possible to separate dead and dying cells and debris from viable samples. Using gentle levitation, live, buoyant cells with intact membranes are able to float, while permeabilized membranes of dying cells sink to the bottom.
Batching processes will impact the cost of multiple samples run on different platforms. In some cases, you can run multiple samples at the same time, reducing overall cost.
With all of these questions in mind, will you be able to do everything you need to do in order to ensure a successful run? If a combination of the sample enrichment and assay you need fall within your budget — great! If not, you may find that you need to sacrifice sequencing depth or total sample number.
Making the jump from bulk sequencing to single-cell can be intimidating — but it doesn’t have to be! With these considerations in mind, you can select a platform that best explores your sample, species, and research question. Still not sure what the best option might be for you? Our single-cell experts can help you find a combination of services that fit within the scope and budget of your project.