T-cell
© The AtlanticScanning electron micrograph of a T cell, colored blue
A human T cell as seen through a scanning electron microscope
2018 is supposed to be the year of CRISPR in humans. The first U.S. and European clinical trials that test the gene-editing tool's ability to treat diseases-such as sickle-cell anemia, beta thalassemia, and a type of inherited blindness-are slated to begin this year.

But the year has begun on a cautionary note. On Friday, Stanford researchers posted a preprint (which has not been peer reviewed) to the website biorXiv highlighting a potential obstacle to using CRISPR in humans: Many of us may already be immune to it. That's because CRISPR actually comes from bacteria that often live on or infect humans, and we have built up immunity to the proteins from these bacteria over our lives.

It's the first time this concern has been aired so publicly, and the preprint kicked off something of a firestorm. "We had no anticipation it would be picked up so broadly on social media. I don't even have a Twitter account. I just heard this from others," says Matthew Porteus, a pediatrician and stem-cell researcher at Stanford who led the study and is working on a clinical trial for sickle-cell anemia.

Not all CRISPR therapies in humans will be doomed. "We don't think this is the end of the story. This is the start of the story," says Porteus. There are likely ways around the problem of immunity to CRISPR proteins, and many of the early clinical trials appear to be designed around this problem.

Porteus and his colleagues focused on two versions of Cas9, the bacterial protein mostly commonly used in CRISPR gene editing. One comes from Staphylococcus aureus, which often harmlessly lives on skin but can sometimes causes staph infections, and another from Streptococcus pyogenes, which causes strep throat but can also become "flesh-eating bacteria" when it spreads to other parts of the body. So yeah, you want your immune system to be on guard against these bacteria.

The human immune system has a couple different ways of recognizing foreign proteins, and the team tested for both. First, they looked to see if people have molecules in their blood called antibodies that can specifically bind to Cas9. Among 34 people they tested, 79 percent had antibodies against the staph Cas9 and 65 percent against the strep Cas9.

Then, they looked to see if a particular type of immune cells called T cells can recognize the Cas9 proteins. This time they studied T cells from 13 healthy adults. Six of them-or 46 percent-reacted to the staph Cas9. None of them did against the strep Cas9.

The Stanford team only tested for preexisting immunity against Cas9, but anytime you inject a large bacterial protein into the human body, it can provoke an immune response. After all, that's how the immune system learns to fight off bacteria it's never seen before. (Preexisting immunity can make the response faster and more robust, though.)

In statements to The Atlantic, three of the leading companies in CRISPR human therapy-Editas Medicine, CRISPR Therapeutics, and Intellia Therapeutics-all downplayed the new findings, citing various ways their therapies could get around the immune system. (Porteus is a scientific founder of CRISPR Therapeutics, though this study was performed independently.) A September 2017 presentation from a scientist at Editas Medicine also detailed some of the ways to test for immune reactions to CRISPR, anticipating a potential problem.

Here are some possible strategies to get around the immune system that are being discussed and tested:
  • Only use CRISPR outside of the body: Instead of delivering CRISPR/Cas9 into the body, you take cells out of the body, use CRISPR to edit their genes in a lab, and return Cas9-free cells. This is the strategy pursued by CRISPR Therapeutics for the inherited blood disorder thalassemia and in various trials using CRISPR to modify immune cells to attack cancer.
  • Only use CRISPR in places the immune system cannot reach: Some sites of the body are immunoprivileged, meaning the immune system can't really attack invaders there. The eye, which Editas Medicine is targeting for inherited blindness, is one of those sites.
  • Modify Cas9 or use a different CRISPR protein altogether: It may be possible to redesign Cas9 to hide it from the immune system or to find other bacterial proteins that can do the job of Cas9 without provoking the immune response. Many different bacteria have CRISPR systems. "We already have lots of Cas enzymes and could get many more," George Church, a geneticist at Harvard and a founding scientific advisor of Editas, wrote in an email.
  • Express Cas9 only transiently: Once Cas9 has made its edit, it doesn't need to stick around. A spokesperson for Intellia noted that it's still unclear how the immune system responds to continuous versus transient expression of Cas9. The company says its lipid-nanoparticle delivery system can get cells to make Cas9 only transiently, but enough for the editing to happen in rodents and non-human primates.
The one type of therapy where immune response may be most dangerous and unavoidable is when Cas9 is produced for a prolonged period of time in a non-immunoprivileged site. In the liver, for instance, the immune system could end up attacking the Cas9-making liver cells.

The danger of the immune system turning on a patient's body hangs over a lot of research into correcting genes. In the late 1990s and 2000s, research into gene therapy was derailed by the death of 18-year-old Jesse Gelsinger, who died from an immune reaction to the virus used to deliver the corrected gene. This is the worst-case scenario that the CRISPR world hopes to avoid.