CRISPR’s Ability to Knock Down RNA May Be Artificially Inflated

The gene editing tool CRISPR has many uses for cell biologists. One of the most popular is the disruption of nucleic acids, like DNA and RNA. While CRISPR DNA knockouts targeting particular genes permanently change the genome, CRISPR RNA knockouts disrupt gene expression instead, changing how cells read their DNA without directly modifying it.

But in a new study, researchers led by Johns Hopkins University RNA biologist Bin Wu suggested that previous estimates of how well CRISPR knocks out these RNA molecules have been inflated by an issue with the technology used to measure knockdown success. The new study, in which the team also proposed a way to fix the problem, was published in Nature Biotechnology.¹

When a Dead Enzyme Does the Job

The researchers found the anomaly when comparing the performance of different CRISPR systems to knock down RNA. One of these included a CRISPR system that uses Csm proteins (Csm), rather than Cas proteins, to mediate the knockdown because the Csm system doesn’t damage off-target RNA molecules. When testing Csm, they noted that the CRISPR system was effective at knocking down RNAs, but they also saw that an RNase-dead version of the enzyme that binds but cannot cut target RNA registered the same level of knockdown effectiveness.

This finding made no sense.

The authors began an exhaustive troubleshooting process, checking whether, for example, Csm’s binding action was enough to damage the target RNA. But they found an explanation only when they tested a quantitative technique used to measure knockdown called reverse transcription polymerase chain reaction (RT–qPCR).

This technique measures nucleic acid content by massively multiplying amplicons, specific sections of DNA or RNA. Choosing the right amplicon to measure as a proxy for knockdown success needs to be done carefully. Picking the wrong amplicon might incorrectly amplify chopped-up nucleic acid fragments. To avoid this, chosen amplicons usually cross over the site where the guide RNA binds and the CRISPR system cuts. This is called a spanning amplicon. Wu’s team found, however, that the knockdown measured using these amplicons with the inactive RNase-dead version of Csm was 90 percent—the same as an active version of the enzyme.

Identifying an RT–qPCR Problem and Finding an Enzymatic Solution

When the team moved their amplicon target to a different site further from the guide RNA binding site, normal measurements returned. The team then tested other RNA-targeting CRISPR systems and found that all of them shared the same inaccuracy when using RT–qPCR around spanning amplicons.

The researchers explored the mechanisms of this misleading measurement. They found the issue was in the guide RNA CRISPR systems use to identify targets. This guide RNA had remained bound to its target, blocking the reverse transcriptase enzyme in the RT–qPCR assay from reading the target sequence. This produced the misleading assessment that the target wasn’t present and that knockdown had been successful.

Luckily, the authors concluded that using alternative reverse transcriptase enzymes called ultraMarathonRT, which knocks the blocking guide RNAs out of the way, produced an accurate reading once again. The findings show that assessing CRISPR function requires care, but that modern biology has a suite of tools to circumvent even stubborn problems.