To study a sample’s transcripts—the pieces of RNA—scientists turn to next-generation sequencing. In 2009, a team of scientists from Yale University wrote that studies using RNA sequencing (RNA-Seq) “have already altered our view of the extent and complexity of eukaryotic transcriptomes.”1 In addition, they claimed that “RNA-Seq also provides a far more precise measurement of levels of transcripts and their isoforms than other methods.”

Approaching a decade later, molecular biologists know that RNA-Seq provides the most sensitive method of exploring the RNA being transcribed from genes embedded in DNA. It reveals that the genes get turned on or off in response to chemical and physical changes to the organism due to environmental stressors, disease, genetic disposition, or general development.

RNA-Seq provides the most sensitive method of exploring the RNA being transcribed from genes embedded in DNA.

Before RNA-Seq, scientists used microarrays to analyze transcripts. “These methods suffer from background and cross-hybridization issues, and they do not allow the detection of RNA transcripts from repeated sequences,” explained Daniela Munafo, field applications scientist for next-generation sequencing at New England Biolabs (NEB). “Also, they enable only measurement of the relative abundance of the RNA transcripts included in the array design, consequently offering a limited dynamic range, and are unable to detect very subtle changes in gene expression levels, which is critical in understanding any biological response.”

RNA-Seq, by contrast, can determine many features: expression levels of specific genes, including allele-specific expression; differential splicing; RNA editing; and transcript fusion events, Munafo noted.

With RNA-Seq, other features also help. “The ability to ‘read’ the transcriptome sequences without prior knowledge enables identification of novel events and the ability to study organisms that have not been investigated before,” said Keerthana Krishnan, development scientist for applications and product development at NEB. “The single-base resolution of RNA-Seq allows identification of mutations and allele-specific expression.”

Sequencing single cells

Instead of applying RNA-Seq to a bulk sample, advancements in cell-isolation technology allows scientists to focus RNA-Seq on individual cells. “Sensitive and high-resolution sequencing of single cells or subpopulations of cells can reveal the complexity of samples,” Krishnan said. “In the case of disease models of progression, this would help identify cells of mutational origin.”

When asked about the key applications of single-cell RNA-Seq, Dan Norton, senior product manager for next-generation sequencing applications at Bio-Rad Laboratories, noted that “While the application space is incredibly broad, immunology has been a huge driver. This is primarily due to tractability—it is easy to get and analyze immunological cells versus primary cells derived from solid tissue. From autoimmune diseases to immuno-oncology, researchers are examining conditions that impinge on the immune repertoire and studying how to harness the immune system to fight disease.”

Developmental biology is another major application area. Norton pointed out that applications range “from using stem cells to recapitulate normal body processes, disease states and organs to understanding key developmental stages of cellular differentiation.”

Tissue and tumor characterization is also highly prominent. “By studying cellular taxonomy within a sample, researchers can answer questions about what populations exist and how they change in response to specific conditions and disease, and about how to identify general cell populations,” Norton said. “Tissue characterization is across the board, from pancreatic islets to liver, heart and tumor biopsies.”

By looking at the RNA in single cells, “you can see what transcripts co-exist in certain cell types within a tissue, the different types of cells within a given population, and how they change in response to temporal and/or conditional stimuli,” Norton explained. “At its most basic level, single cells provide important resolution versus bulk sample RNA.”

With current technology, standard taxonomy studies are a great fit, but to gain a more advanced functional understanding involving pathway analysis, RNA-Seq still poses a challenge. “At the moment, we really don’t know how much of the single-cell transcriptome we’re capturing,” Norton said. “With current single-cell methods, is there 25% more we can capture, or 250% more we can capture?”

In some cases, such as cancer, scientists want to identify very rare cell populations due to their potential importance in clonal evolution, but achieving this detection level is not always practical. “To detect extremely rare cell populations, it is simply not economical to sequence the required number of cells to the necessary depth,” Norton added.

Last, the method of obtaining the cells also creates a stumbling block for the moment. There is “no standard protocol to dissociate tissue without often sacrificing viability or causing transcriptome changes due to cell stress,” Norton explained. So, despite the great promise of single-cell RNA-Seq, some improvements will take this technology even farther.

Measuring more

Just as scientists want to sequence long strings of DNA to get a better look at the complete picture without the need to put back together a perplexing puzzle, the same can be said of RNA-Seq. Some methods of doing that exist.

According to Emily Hatas, director of agbio applications at Pacific Biosciences, “The full-length transcript isoform sequencing method developed by PacBio, the Iso-Seq method, employs our long-read Single Molecule, Real-Time—SMRT—Sequencing technology to sequence transcripts in their entirety within a single long read.” Methods that sequence shorter reads break up the transcripts, sequence them, and then use bioinformatics to put the data together. “This reassembly process is prone to errors, especially when attempting to determine transcriptional start and termination sites or exon chaining in alternative transcription,” Hatas said.

Biocompare’s RNA Search Tool
Find, compare and review RNA products
from different suppliers Search

Using the Iso-Seq method, one group of scientists working with PacBio showed that a chicken’s transcriptome rivals the complexity of a human’s.2 In previous attempts to compare organisms, the researchers noted the “difficulty in transcript identification from short read RNA-Seq data.” With longer reads, they pointed out: “The relative proportion of alternative transcription events revealed striking similarities between the chicken and human transcriptomes while also providing explanations for previously observed genomic differences.”

Digging deeper in data

Whatever advances scientists make in RNA-Seq, data analysis will remain a crucial step. One method involves clustering, literally grouping data from similar things, such as transcripts. For example, transcript data from people with the same type of cancer could be grouped.

One group of scientists applied clustering of samples from RNA-Seq to cancer data.3 “Our evaluation considers strategies regarding expression estimates, number of genes after non-specific filtering, and data transformations.” They concluded: “Results support that clustering cancer samples based on a gene quantification should be preferred.”

For scientists interested in using clustering with data from RNA-Seq, many methods exist in common languages, including R.

The best results from RNA-Seq depend on sample acquisition and preparation, sequencing methods and platforms, and computational methods of analyzing the data. The choices in all of these pieces keep growing, and some work better in answering some questions than others. Nonetheless, molecular biologists know that the applications of RNA-Seq promise to solve many more of nature’s riddles, because this technique is just getting started.

References

1. Wang Z, et al. “RNA-Seq: a revolutionary tool for transcriptomics,” Nat. Rev. Genet. 10:57–63, 2009 [PMID: 19015660].

2. Kuo Ri, et al. “Normalized long read RNA sequencing in chicken reveals transcriptome complexity similar to human,” BMC Genomics 18:323, 2017. [PMID: 28438136].

3. Jaskowiak PA, et al. “Clustering of RNA-seq samples: comparison study on cancer data,” Methods. 2017 [epub ahead of print; PMID: 28778489].

Image: RNA sequencing reveals the complex patterns of transcription in cells, and that advances basic and applied research. Image courtesy of Bio-Rad Laboratories.