Unlocking Applications for RNA Sequencing

RNA sequencing is a genomics technology that gives scientists a transcriptome-wide view of gene expression and messenger RNA (mRNA) splicing. The tool is already providing insights in cancer research, from identifying clinical biomarkers to helping better characterize disease subtypes, but experts say they are only beginning to unlock its potential.

The technique, which leverages next-generation sequencing, first emerged in the mid-2000s, as researchers realized they could use the same sequencing technology on RNA that had been used successfully with DNA. The technology has evolved over the last 15 years, offering the ability to examine RNA both in bulk and at the single-cell level. Going forward, newer techniques such as long-read RNA sequencing, which can determine the nucleotide sequence of long sequences of DNA between 10,000 and 100,000 base pairs at a time, could further expand the understanding of RNA biology.1

“We are still in the early days of the clinical application of RNA sequencing,” Jason D. Merker, MD, PhD, associate professor of pathology and genetics at University of North Carolina, Chapel Hill, told ASH Clinical News. “We’re beginning to see incorporation of targeted RNA sequencing to complement DNA-based sequencing assays, but I think there are many more potential future applications of RNA sequencing in addition to current use for fusion detection, including identifying splice isoforms and studying gene expression profiling.”

How It Works

To perform RNA sequencing, RNA is initially extracted from tissues and fragmented. Then, through reverse transcription, the RNA fragments are converted into complementary DNA and amplified through polymerase chain reaction for sequencing. Several bioinformatics tools are available to assist in quality control, read alignment, transcript assembly, and expression analysis. The sequencing data can then be aligned to a reference genome or reference transcripts and used to create a genome-scale map showing the transcriptional structure and expression level of each gene.2,3

“There’s an advantage of RNA sequencing over the other methods we use,” said Lukasz Gondek, MD, PhD, assistant professor of oncology at Johns Hopkins University in Baltimore. “RNA-sequencing is more reliable, accurate, and provides better reproducibility than the older, microarray-based methods so that we can more precisely assess the level of gene expression.”

With prior techniques, such as microarrays, researchers could only glean the signal intensity of mRNA in the cell. With newer sequencing technology, scientists can see the sequence of nucleotides in each RNA molecule, as well as the post-transcription processing.

For instance, Dr. Gondek offered, it is known that one gene may result in different RNA isoforms through alternative splicing, or skipping of certain exons, in the RNA. “Because we can sequence the RNA molecule, we can precisely measure the different isoforms that are being transcribed from the same DNA template,” he explained. “The same DNA sequence may produce different RNA isoforms that are cell- or tissue-specific and may result in the production of proteins with unique functions.”

Genes responsible for RNA splicing are frequently mutated in human cancers, Dr.
Gondek added. “With RNA sequencing, we can measure the impact of these mutations on RNA processing.”

Bulk Versus Single-Cell Analysis

RNA sequencing is primarily done in bulk, on samples consisting of heterogeneous mixtures of many cells, rather than at the single-cell level. In bulk-level sequencing, gene expression is analyzed for the entire sample without drilling down to the different cell types within the sample.

Around 2009, researchers began developing single-cell RNA sequencing techniques to look at genetic material at the single-cell level. In this case, scientists break the sample into individual cells, allowing them to identify cell types within the sample and measure the gene expression profile for each cell.

“Let’s say we take 10,000 cells, then we’ll be able to look at the RNA expression in individual cells,” Dr. Gondek said. “That’s how we find new cells or cell states that we didn’t know existed before. With single-cell RNA sequencing, we can better define rare cell populations within the tissue. We can actually assess qualitative and quantitative characteristics of the transcriptome at the single-cell level.”

In contrast, bulk sequencing may miss a signal from a small group of cells since it could be washed out in a larger sample. “We cannot truly separate the cells and the cell function based on bulk RNA data because we are lumping together all of this RNA that is coming from different cells,” Dr. Gondek said.

How to Use It

While single-cell sequencing is considered a significant advance in sequencing technology, experts say one technique is not better than the other. It all comes down to how you use it.

“It’s meant to answer different questions,” said Romanos Sklavenitis Pistofidis, MD, an instructor at the Dana-Farber Cancer Institute and the Broad Institute of MIT and Harvard.

Bulk sequencing has some advantages, too, Dr. Pistofidis said. Since the technique has been around longer, scientists have had more time to establish various protocols to adjust the technique based on the type of RNA to be assessed or the quality of the tissue sample available. This type of flexibility becomes even more important when dealing with clinical samples, which tend to be of poorer quality, Dr. Pistofidis said.

Bulk sequencing is also more comprehensive than single-cell sequencing in describing the transcriptome, he added. “Not all cells that express a single gene will appear to express it by single-cell RNA sequencing.”

Bulk analysis has assisted in the development of biomarkers, understanding drug resistance, and discerning how investigational drugs affect disease. But single-cell RNA sequencing “really shines” when analyzing the subclones in a single tumor, Dr. Pistofidis said.

“Imagine that you have a patient whose tumor has 100 cells that are resistant to treatment x, and then 1 million cells that are not resistant to treatment x,” he explained. “If we do bulk RNA sequencing, 100 out of 1 million is not such a large proportion. Depending on how deeply we sequence, we may not see the signal. But, if you do single-cell RNA sequencing, that group of treatment-resistant cells may pop up as a cluster that is defined by different characteristics.”

Additional RNA sequencing approaches are emerging that offer long reads of the RNA – up to tens to thousands of kilobases in length – and the ability to capture the full gene transcript. However, these approaches are lower throughput and still expensive, according to Francois Aguet, PhD, a computational biologist at the Broad Institute, who is a member of the Genotype-Tissue Expression (GTEx) Project Consortium, which built a comprehensive atlas of tissue-specific gene expression and regulation.

These newer approaches aim to bridge some of the gaps that currently exist with bulk and single-cell sequencing, such as limited information on alternative splicing, Dr. Aguet explained.

“We’re still not capturing alternative splicing well enough because we are fragmenting the RNA with current high-throughput sequencing approaches,” he said. “For genes that have many known transcripts, it is difficult to reconstruct which transcripts are being expressed. In that case, long-read technologies will be necessary, and I expect those to become cheaper and scale up over the coming years.”

Disease Insights

RNA sequencing has already yielded insights into characterizing acute myeloid leukemia (AML) and multiple myeloma.

In a 2019 study published in Cell, Peter van Galen, PhD, of Harvard Medical School, and colleagues adapted a single-cell RNA sequencing technology and incorporated long-read nanopore sequencing to help characterize AML hierarchies that are relevant to disease progression and immunity.4 Researchers profiled more than 30,000 cells from 16 patients with AML and more than 7,500 cells from five healthy individuals, and obtained genotyping information for nearly 4,000 cells.

After using a machine learning classifier to distinguish malignant from normal cells, the researchers discovered six malignant AML cell types that shared features with normal hematopoietic cells. They also identified genetic subclones in FLT3-mutated AML, including an FLT3-ITD subclone that primarily contained primitive cells and an FLT3-TKD subclone that primarily contained differentiated cells, even though they were found in the same tumor. In addition, the researchers showed that FLT3-ITD expression suppressed differentiation of AML cells in vitro. “These inter-tumoral, intra-tumoral and in vitro findings suggest that FLT3 variants differentially affect AML differentiation, and may explain the association of FLT3-ITD with poor patient outcomes,” Dr. van Galen and colleagues concluded.

In a 2015 Nature Medicine study, Ravindra Majeti, MD, PhD, of Stanford University School of Medicine, and researchers used RNA sequencing as part of an analysis evaluating why patients with AML and an IDH mutation have a distinctly favorable response to BCL2 inhibition.5 The findings from the experiment indicated that IDH1/2 mutation status is a factor affecting sensitivity to the BCL2 inhibitor ABT-199. The researchers asserted that it could be useful in identifying patients who would be likely responders to pharmacologic BCL2 inhibition.

In myeloma, researchers led by Irene M. Ghobrial, MD, of Dana-Farber Cancer Institute, published a study in Nature Cancer that focused on characterizing precursor states of multiple myeloma and the native tumor microenvironment.6 After performing single-cell RNA sequencing of bone marrow cells from healthy donors and donors with multiple myeloma in various stages, they found that immune changes occur early in the course of disease, even before patients develop symptoms. “These results provide a comprehensive map of immune changes at play over the evolution of premalignant [myeloma], which will help develop strategies for immune-based patient stratification,” the authors concluded.

These types of studies show the utility of RNA sequencing today as a tool that can tease out explanations for disease activity, said Mark J. Levis, MD, PhD, professor of oncology and director of the Adult Leukemia Program at Johns Hopkins University. “That’s where it works beautifully, as pure basic science to explain a clinical phenomenon,” he said. “It’s a laboratory tool to help explain what we see in the clinic.”

Long Road to the Clinic

While RNA sequencing in its various forms has contributed to greater understanding of hematologic diseases, it is far from being integrated into clinical use on any large-scale basis, experts noted.

Dr. Merker said the primary, but limited, clinical application in hematology is for bulk sequencing in the identification of gene fusions, which are important for establishing diagnosis and prognosis of a malignancy, as well as for therapy selection for many hematologic malignancies.7

“One of the current major uses of RNA sequencing in the clinical realm is detection of gene fusion events. Some of these events can be difficult to detect with DNA sequencing technologies, and so increasingly, these clinical assays are being supplemented with targeted RNA sequencing for fusion detection,” Dr. Merker said. “That’s an area where it is being used clinically for a variety of audiences, regardless of clinical practice setting.”

Another clinical use for RNA sequencing is to prognosticate patient outcomes for certain hematologic malignancies based on the patient’s RNA profile, Dr. Gondek said. “In acute leukemias, myelodysplastic syndromes, and other diseases, we can precisely measure the fusion partner by using RNA sequencing,” he said. “Some of the laboratories, including our molecular lab, use RNA sequencing to detect the known fusion genes in certain myeloid malignancies.”

Scientists are also exploring the use of RNA sequencing to assess response to therapy through measurement of measurable residual disease, Dr. Gondek added. “We can see very precisely and with a very sensitive test whether the malignant cells are still detectable or indetectable using this technique,” he said. “Even though it can be done at the DNA level, I know a lot of investigators are looking into it at the RNA level.”

Dr. Levis, who has been using RNA sequencing as part of his research, advised physicians to maintain a healthy level of skepticism about its use in the clinic. “Scientists are looking at genes and using RNA sequencing to see what happens to gene expression patterns, but from a direct clinical standpoint, RNA sequencing itself has not amounted to much,” he said.

When it comes to single-cell RNA sequencing, which is not as far along as bulk technologies, there is also a risk for technical errors, due to the small amount of material that must be amplified. That process can be uneven and result in skewed gene expression ratios, Dr. Levis warned, so it is important to view those results through a critical lens.

Other barriers to bringing RNA sequencing into the clinic relate to throughput speed, cost (ranging from hundreds of dollars per bulk sequencing experiment to thousands for single-cell analysis, plus expensive equipment), and the need for complex bioinformatics, he said. Still, the biggest barrier may just be scientists’ lack of complete knowledge of the genome.

“The problem with RNA sequencing right now, still, is that we can’t see the pattern. It’s a universe of subtle ups and downs and connections that we don’t fully understand,” Dr. Levis said. “I have no doubt that 10 years from now, when we have yet a greater understanding of the functions of the cell, bioinformatics will presumably be that much better at picking out patterns.” —By Mary Ellen Schneider

References

  1. Stark R, Grzelak M, Hadfield J. RNA sequencing: the teenage years. Nat Rev Genet. 2019;20(11):631-656.
  2. Hong M, Tao S, Zhang L, et al. RNA sequencing: new technologies and applications in cancer research. J Hematol Oncol. 2020;13(166).
  3. Wang Z, Gerstein M, Snyder M. RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet. 2009;10(1):57-63.
  4. van Galen P, Hovestadt V, Wadsworth Ii MH, et al. Single-cell RNA-seq reveals AML hierarchies relevant to disease progression and immunity. Cell. 2019;176(6):1265-1281.
  5. Chan S, Thomas D, Corces-Zimmerman M, et al. Isocitrate dehydrogenase 1 and 2 mutations induce BCL-2 dependence in acute myeloid leukemia. Nat Med. 2015;21(2):178-184.
  6. Zavidij O, Haradhvala NJ, Mouhieddine TH, et al. Single-cell RNA sequencing reveals compromised immune microenvironment in precursor stages of multiple myeloma. Nat Cancer. 2020(1):493-506.
  7. Byron SA, Van Keuren-Jensen KR, Engelthaler DM, et al. Translating RNA sequencing into clinical diagnostics: opportunities and challenges. Nat Rev Genet. 2016;17(5):257-271.