The clustered regularly interspaced short palindromic repeats (CRISPR) system has evolved in bacteria as a form of adaptive immune system, which scientists have customized as a versatile tool for gene editing cells in vitro, and potentially in vivo to correct diseases caused by gene mutations. However, CRISPR gene editing using the bacterially derived Cas9 nuclease enzyme isn’t foolproof and the system fails about 15% of the time.
Studies by scientists at the University of Illinois (UIC) at Chicago now suggest that this failure can occur because the DNA-snipping Cas9 enzyme doesn't detach from the DNA site to which it binds, which essentially blocks the cell’s repair enzymes from accessing and repairing both of the cut strands of DNA. “We found that at sites where Cas9 was a 'dud' it stayed bound to the DNA strand and prevented the cell from initiating the repair process,” says Bradley Merrill, Ph.D., associate professor of biochemistry and molecular genetics at the UIC College of Medicine. Dr. Merrill is senior author of the team’s published paper in Molecular Cell.
The study results also showed that while the cell’s translocating RNA polymerase enzymes can knock the Cas9 off the site and allow the repair machinery to start work, they can only do so by approaching the Cas9 from one direction. This means that it is important to design the single guide RNA (sgRNA) that homes in on the correct target sequence, to bind to that sequence on the correct, template strand of the double-stranded DNA that is used by the RNA polymerase enzyme. The researchers suggest that their findings could help develop safer, more effective CRISPR-based gene therapy treatments for genetic diseases. “These data provide insights into the biology of the CRISPR system and provide a simple method of enhancing probability of successful genome editing by choosing sgRNAs that anneal to the template strand of DNA,” they conclude in their report, titled “Enhanced bacterial immunity and mammalian genome editing via RNA-polymerase-mediated dislodging of Cas9 from double-stand DNA breaks.“
The Cas9 nuclease exhibits particular biochemical and biophysical properties that allow it to be guided by sgRNAs to bind to specified sequences of DNA, and snip both strands of the molecular at that site. But Cas9 also displays other properties that distinguish it from non-RNA-guided effector nucleases of the bacterial immune system, such as restriction endonucleases, the authors write. Unlike these other endonucleases, Cas9 tends to stay bound to the target site, even after it has cut the DNA. In effect, it exhibits what the researchers call “remarkably stable enzyme-product state wherein the nuclease remains bound to the double-stranded break (DSB) it generates.” This propensity to stay bound can limit the system’s efficiency, both in its native bacterial system, and also potentially when used to edit genes in mammalian cells because the bound Cas9 isn’t free to carry out cuts at other sites.
In practice, when Cas9 is used to carry out genome editing in mammalian cells, the nuclease is typically in excess, when compared with its DNA substrates – there may only be 2-4 copies of the target sequence per cell, the authors point out. In this case, the efficiency of genome editing isn’t dependent on whether one Cas9 enzyme can act at multiple sites, but efficiency is reduced if the Cas9 doesn't detach from the DSB site, because this prevents the DNA repair enzymes from accessing the site and repairing the break. “When DNA substrates are rare, such as when Cas9 is used to edit a unique mammalian genomic sequence, persistence of Cas9-DSB could preclude repair of the DSB by the cell,” the authors write.
Working with colleagues at Rockefeller University, New York, the multidisciplinary UIC team discovered that Cas9 can be prompted to detach from the DSB site by translocating RNA polymerases. “Dislodging Cas9 from the DSB stimulates editing efficiency in cells by allowing the ends of the DSB to be accessed by DNA repair machinery,” they write. However, their studies in bacteria and in laboratory-grown mammalian cells also showed that this process of dislocation is orientation and direction specific. For the translocating RNA polymerase to dislodge Cas9 the sgRNA portion of the Cas9 complex must be attached to the DNA strand that is used as the template by the RNA polymerase.
Their studies in bacteria and in laboratory-grown mammalian cells showed that by designing the sgRNA with this in mind it is possible to effectively convert Cas9 from a single turnover nuclease, into a multi-turnover nuclease, allowing it to cut at multiple target sites in the same genome. This could be important when a bacterium needs to fight off a major phage infection, for example. “An ability of a single Cas9 molecule to digest many DNA substrates could be important when saturating levels of DNA targets need to be digested, such as when high multiplicities of infection occur during bacteriophage infection,” they comment.
When the researchers evaluated the effects of strand bias in bacterial and mammalian cell test systems, they found that the ability to dislodge the Cas9 from DSBs markedly improved the efficiency of the system and had “significant effects on genome editing and bacterial immunity by increasing mutation frequencies in mammalian cells and mediating enhanced phage interference through multi-turnover nuclease activity.”
“I was shocked that simply choosing one DNA strand over the other had such a powerful effect on genome editing,” states the paper’s lead author, Ryan Clarke, Ph.D. candidate, UIC, department of biochemistry and molecular genetics. “Uncovering the mechanism behind this phenomenon helps us better understand how Cas9 interactions with the genome can cause some editing attempts to fail and that, when designing a genome editing experiment, we can use that understanding to our benefit.”
The findings that Cas9-DSB binding represents a rate-limiting step in CRISPR-Cas9 gene editing could also help researchers design more efficient, effective and safer genome editing tools for research and potential treatments for human genetic diseases. “If we can reduce the time that Cas9 interacts with the DNA strand, which we now know how to do with an RNA polymerase, we can use less of the enzyme and limit exposure,” Dr. Merrill suggests. “This means we have more potential to limit adverse effects or side effects, which is vital for future therapies that may impact human patients.”