NEW YORK (GenomeWeb) – Polymerases are a key component to next-generation sequencing. While DNA polymerases used in DNA sequencing applications can be extremely high fidelity — correcting errors as they add complementary bases — their counterparts used in RNA sequencing, reverse transcriptase, lack such proofreading capability.
However, a group of researchers from the University of Texas, Austin, have now engineered a reverse transcriptase enzyme, which they demonstrated in a publication in Science today, can provide such proofreading activity. The new enzyme, dubbed RTX, could help improve errors in RNA sequencing applications and could be especially useful for direct RNA sequencing on single-molecule sequencing platforms, Jared Ellefson, a postdoctoral fellow at UT Austin, told GenomeWeb.
Ellefson said the group is now in conversations with several companies about licensing RTX to include in NGS library prep kits.
Typical reverse transcriptase enzymes "introduce lots of errors" when copying cDNA, Ellefson said, "which can complicate RNA-seq." In order to design the engineered RTX enzyme, Ellefson said the team started with a high-fidelity DNA polymerase. Then, they essentially "trained it to become a reverse transcriptase," he said.
First, though, the team tracked the evolutionary history of reverse transcriptase and DNA polymerase. Reverse transcriptase enzymes belong to a single, ancient protein family. DNA polymerase evolved after reverse transcriptase, so the researchers wanted to trace the enzyme's history to see if they could identify the evolutionary processes that made DNA polymerase recognize DNA as opposed to RNA.
Ellefson said that the group identified a few point mutations that essentially act as "checkpoints" within the polymerase, enabling it to distinguish between DNA and RNA strands. "But, if you can modify those checkpoints, you can essentially get the DNA polymerase to act like a reverse transcriptase," he said.
So, the researchers started with a high-fidelity DNA polymerase and modified it at the checkpoints so that it would synthesize RNA while still keeping its proofreading feature.
"This is the first enzyme that can proofread while reverse transcribing," Ellefson said.
Next, the team wanted to see if the modified enzyme would work in an RNA-seq experiment. First, they tested the enzyme using dideoxy mismatch primers in PCR, and found that it had "robust proofreading activity on DNA template," the authors wrote.
They then looked at error rates of their engineered enzyme compared to a control retroviral reverse transcriptase and found that the error rate of RTX was between threefold and tenfold lower.
The team next tested a standard RNA-seq library prep kit from New England Biolabs, except they substituted their enzyme for the standard reverse transcriptase. Sequencing on an Illumina platform demonstrated "nearly identical coverage and expression profiles" compared to the standard enzyme, the authors wrote.
Finally, they tested its ability to directly sequence RNA using Sanger sequencing, sequencing a GATC RNA repeat. Ellefson said that the researchers are now interested in testing their RTX enzyme on Pacific Biosciences' single-molecule sequencing platform.
An added benefit of RTX, Ellefson said, is that because it was derived from a PCR enzyme, it is "extremely thermostable." Most reverse transcriptases denature at higher temperatures, he said. But being able to sequence RNA at higher temperatures could help uncover RNA that has highly structured elements, like ribosomes, he said. At higher temperatures, those structures unfold, making them more amenable to sequencing. "So we may be able to find those RNAs that you couldn't really see before because of their structure," he added.
Ellefson said that the lab is continuing to refine the RTX enzyme and plans to continue to test it in various applications. Aside from direct RNA sequencing and single-molecule sequencing, he said the team plans to use it for single-cell RNA sequencing applications.