Analysis: human immunity to Cas9 bigger #CRISPR therapeutic hurdle than off-targets?

Some in CRISPR-Cas9-land who are focused on potential future clinical applications are kind of rejoicing or at least sighing a breath of relief.

This upbeat swing in the atmosphere (from investors especially) was sparked by retraction of that paper, the one initially reporting tons of supposed off-target CRISPR-Cas9 activity in mice, which turned out to be a “nothing burger” according to one investment site. Off-target activity definitely still needs to be on people’s radar screens, but it’s a problem that’s not nearly so widespread as that paper incorrectly argued it was.

All’s well now on the translational CRISPR front, right?

Well, kinda wrong.

If you are all excited about the idea of putting CRISPR-Cas9 into patients because maybe it’ll help them via gene therapy or maybe it’ll help you as an investor make money, sit down.

I have one sobering word for you. Immunity. Ring a bell?

It was only a few months ago, but it seems like an eternity in what I call “CRISPR years” (kind of like dog years, moving much faster than normal human years), that we learned from a preprint on Biorxiv from a team led by Matthew Porteus that many of us humans seem to have antibodies against the nuclease Cas9 and there was some T cell reactivity. In other words these data were suggesting that if you put CRISPR-Cas9 into patients for an attempt at gene therapy things may go haywire as the human immune system reacts.

On this blog after that preprint came out, I advised people to keep calm and CRISPR on as solutions could likely be found, but now there’s yet another new Biorxiv preprint out bolstering the notion that humans may be widely immunoreactive to Cas9.

CRISPR expert Gaetan Burgio’s reaction to this preprint on Twitter included a “Wow, just wow!” comment because of how big a deal this appears to be.

The new preprint is entitled “High prevalence of S. pyogenes Cas9-specific T cell sensitization within the adult human population – A balanced effector/regulatory T cell response” and comes from a group with senior author Michael Schmueck-Henneresse.

It reports that human T cells clearly react to Cas9 in various assays and likely will trigger strong immune responses in humans who receive CRISPR-Cas9 machinery. This kind of T cell situation was suggested by the earlier Porteus-led preprint too, which also reported Cas9 antibodies in many humans, but the new data concretely show the earlier results were not fluky. Human immunity Cas9 is a real, reproducible phenomenon. See Fig. 1-B above from the new paper showing in just a few of the many experience done that one can see clear T cell reactivity to S.Pyogenes Cas9.

The real gut-punch for those counting on CRISPR-Cas9 being a therapy is the phrase in the preprint “immunosuppressive treatment must be considered” for deploying CRISPR-Cas9 gene editing for patients.

Now, those in the stem cell field and broader cell therapy field are familiar with immunosuppressive therapy being needed for some things like allogeneic cellular approaches, but I can’t imagine the CRISPR-Cas9 translational field wanting to go down that path. Also, the authors of the new preprint point out that some immunosuppressive drugs probably wouldn’t help with certain kinds of T cell responses that are activated by Cas9.

Some may still say, “Well, we don’t have to use Cas9”, which is true as there are other nucleases that may be suitable, but other microbial nucleases may have similar immune-related problems or other issues.  Switching bacterial strains is an overlapping strategy as the above one to try to work around the immunity issue as well, focusing on strains that maybe don’t infect humans very often or at all, but the human immune system could still react more than one might predict to bacterial nuclease proteins as human cells expressing them cleave the proteins and present antigens on their surfaces and/or die so end up releasing the proteins systemically. We’ll see as other approaches continue to be tried.

Transient use of Cas9 RNPs is another potential way to go to help avoid this immunity situation as RNPs are present only transiently, but they may yield lower efficiency too than other approaches such as AAV transduction. Also, Zinc-finger and TALEN approaches seem unlikely to raise immunity problems (in part depending on their delivery mechanisms) based on initial findings with them so far so that’s encouraging and the cool base editor approach is out there too.

I believe there are going to be ways to deal with this immunity thing, but it’s a real, substantial hurdle.

The larger take home message here is that it’s going to be a winding trek from bench to beside for CRISPR technology. That’s just the reality of much of translational science anyway so perhaps some folks just thought naively it’d be a simple, linear CRISPR path to patients…and keep in mind that we cannot forget potential CRISPR off-targets either in the therapeutic pipeline even with that mouse paper debacle.