'CRISPR is like a very accurate and efficient postal system, that can reach anywhere you want to go very precisely, but only if the ZIP code ends in a zero,' said Dr Joseph Jacobson, who led the team at the Massachusetts Institute of Technology (MIT) in Cambridge, Massachusetts. 'So it is very accurate and specific, but it limits you greatly in the number of locations you can go to.'
The current system can home in on target sites by identifying specific DNA sequences, known as protospacer adjacent motifs – or PAMs, which flank the target. The enzyme Cas9 then makes 'cuts' in the DNA to edit the sequence.
But the most commonly used form of Cas9 from the bacterium Streptococcus pyogenes, is limited to targeting about 9.9 percent of sites in the genome as it requires two G nucleotides in its PAM sequence.
In order to identify alternative enzymes with less restrictive PAM sequence requirements, the researchers developed computer algorithms to sift through potential candidates from bacterial sequences.
Once they had identified potentially useful enzymes for CRISPR, they then built and tested synthetic versions in the lab. The most successful enzyme (ScCas9) was Cas9 from a bacterium called Streptococcus canis.
'The enzyme looks almost identical to the one that was originally discovered... but it is able to target DNA sequences that the commonly used enzyme cannot,' said graduate student and co-lead author Pranam Chatterjee, who did the research with graduate student Noah Jakimo at MIT.
This new enzyme requires only one G rather than two in its PAM sequence, so opening up more locations in the genome as potential targets for genome editing.
The researchers hope to use their technique to identify other enzymes that could expand the range of the CRISPR system even further, said Dr Jacobson. 'We feel confident of being able to go after every address on the genome.'
The study was published in Science Advances.