Guiding CRISPR/Cas13 Activity Using Chemically Modified Guides

 

In a recent episode of the NEB podcast, Lessons from Lab and Life, I had the privilege of interviewing Dr. Neville Sanjana, a core faculty member at the New York Genome Center. During this discussion, he explained his recent collaboration and publication detailing the use of chemically modified guide RNAs to enhance CRISPR-Cas13 knockdown in human cells. Details from this discussion are included below. With gene and cell therapies already in the clinic and massive development underway for new therapies and diagnostics, this method that fine-tunes T cell transcriptomes is provocative.

Tell us about your recent publication.

This work, which was a collaboration with Synthego^® and New England Biolabs^® and some other folks in all of our groups and labs, explored chemically modified guide RNAs for use in an RNA targeting the CRISPR enzyme called Cas13.

 

What is the current standard for gene knockdown and what are the advantages of using chemically modified guide RNAs for gene knock down? 

I think the natural comparison for RNA targeting CRISPR enzymes or really any RNA targeting system is going to be RNA interference, which has been a huge discovery over the last 20, 30 years. And I think the most exciting thing about RNAi, RNA interference, is that it's already entering the clinic. But there are some drawbacks.

It's well known in the literature that RNAi is prone to off target effects. And there's now advanced algorithms that can help you design different kinds of RNAi, like shRNA that are virally delivered, or siRNA that are chemically synthesized. You can use these algorithms to predict what is a better RNAi target. But these off target issues, they're kind of always a concern with these approaches. 

The other issue is that RNAi doesn't use a CRISPR enzyme, but uses an endogenous nuclease, these proteins like Dicer and the RISC complex and these proteins have a particular localization within the cell that they work in. So they're found mostly in the cytoplasm. For RNA targets in the nucleus like non-coding RNAs, which is something that my lab studies a lot – non-coding genome regulation and non-coding RNAs – they're really not possible to target using RNAi methods. 

Now there's a lot of RNAs that are not in the cytoplasm. Those are much easier to target. So, when we say that this is now entering the clinic, which is truly amazing for RNAi whatever 20 plus years from the initial discovery to be... There are FDA approved therapies that use these RNAi reagents. I think the first one is Alnylam drug used to treat a very rare form of amyloidosis and there's many of these different hereditary mutations. And so, I mean, it's rare. It's a debilitating progressive disease, fatal, and it's just caused by buildup over time of this amyloid in brain, heart, all sorts of organs. And for this RNAi is great. It targets, regardless of the mutation you might have in this gene, TTR, the RNAi can target the gene and degrade it. And that transcript is available in the cytoplasm. So it’s an mRNA, a messenger RNA that can be degraded. 

I think we're just seeing kind of the beginning of these RNAi therapeutics, and I think there's a lot they can do. There are some things that they can't do. And I think in both categories.

This is a little bit of a tangent, but with Cas9, or with DNA targeting CRISPRs, you see so many different approaches for say sickle cell anemia now, right? That's a disease where there's many different gene editing approaches. So I think it's the same thing with these kinds of drugs that we're seeing entering the clinic that use RNAi, that there actually might be many therapeutic opportunities that are RNAi based, and that are Cas13 based. But Cas13 offers some very unique opportunities, like targeting in the nucleus.

 

 

What was your motivation to want to do this work in Human T cells?

That's a great question. So, in our lab we also have a lot of interest in understanding immunotherapy and other breakthrough technologies of the last decade. And understanding both in our earlier work, why does it fail? In many cases, it doesn't succeed for the patient. What kinds of tumor mutations enable tumors to evade immunotherapies? 

And we're also very interested, with very recent work in the lab, that hopefully I come back to tell you about next time, in how do we engineer better T cells? How do we supercharge T cells so we can avoid that? Can we make the failure rate drop down? 

Chimeric antigen receptor T cells, they're basically approved for very specific kinds of cancer, mostly liquid tumors, blood cancers. But there's a whole lot of cancer that isn't that, right? Solid tumors. We do a lot of work on melanoma, and some work on pancreatic cancer. You don't really see CAR T therapies there. And so, it'd be great if we could engineer CAR T therapies that are effective there. So our motivation would is working with primary cells like human T cells , so that we can do things like this ¬– that we can engineer better CAR T, better TCRT, better kinds of these T cell immunotherapies.

And one thing that is, I think, foremost in the minds of synthetic biologists or genome engineers that are doing this kind of work, is that CAR Ts, even in cases where they work, you commonly have very severe side effects during treatment like CRS or cytokine release syndrome.

And that might be that we're just pushing these T cells to be very aggressive, very active, and maybe that we need just a little bit more finesse in the approach. So, one thing you might imagine doing is on the initial infusion of these, engineer T cells back into the patient. Imagine supercharging them for a week or two by targeting specific proteins' transcripts using Cas13 knockdown. And having that be on for a focus period of a week, two weeks and then go away. And so that's where I think Cas13 can be perhaps better than permanent genome modification where you say, knock out the immune checkpoint PD-1 or something like that in these T cells. This is something that we've been trying to do since the work that we did with you guys.

But you could imagine flipping that around also, that you could temporarily modify tumor cells and go to the other side of that immune or that checkpoint synapse where they have all these things that they display, these immunoglobins on their cell surface that actually push the T cells away. So, you could imagine transiently knocking those down.

And that sounds kind of crazy. Oh, you're just going to deliver this to a tumor and transiently reshape the tumor microenvironment? But we already have approved therapies that are in a similar vein. Like for melanoma, we have oncolytic viruses that are approved, where you just locally inject at the site of the melanoma. You inject the virus. And so, you could imagine engineering a virus with Cas13 that, so this is maybe getting a little away from chemical modification, but you can imagine engineering virus Cas13 that does something similar. Remodels that tumor microenvironment, makes it easier for your CAR T to come in.

So I think this was what we had in mind when we started working with you guys. This was the long-term mission that we're making slow process toward.

 

 

What was your motivation for wanting to protect the guide RNAs?

There was a pretty simple rationale for them, but I think it's something that probably everybody can understand now, now that much of the world has been exposed to modified RNA through the COVID vaccines that have just been truly amongst the most amazing science we've seen this last year.

So by chemically modifying the guide RNAs, what we do is we extend their usually short half-lives. Even water is dangerous for these short RNAs. So, we extend their usually very short half-lives. And can make Cas13 work – instead of just working for hours, it can now work for days or weeks.

But what we didn't know before we started working with all of you was how do we go about doing this? This hadn't been done with really any Cas13 before. And so, the question was, do we just maximally modify every single base? Are there certain places, certain modifications or certain placement of the modifications that are better?

And that's really, I think the heart of the paper that we put together was saying, what modifications are good and where do we put them? And I think it was worthwhile to do this study because I don't think we could have just looked at Cas9 and guessed it right off the bat that we just do the same things.

I mean, first of all, the guide RNAs, this is perhaps a bit technical discussion, but the guide RNAs for Cas13 are a bit different than for Cas9. They have a quite a different structure. And so, I think it's really great that we did this very systematic study and I'm very grateful to for Synthego for entertaining a whole bunch of requests during that process.

 

What do you see as the advantages of using the protein to introduce the nuclease and editing reagents?

That's a great question. And this is something that we very explicitly tested in the paper. During the revision, I think this got buried a little bit maybe in the supplement, but we did do some work where we actually looked at the timing of the delivery of Cas13 protein alongside the chemically modified guide RNAs.

And something that we saw in many different forms of delivery, I think we looked at protein mRNA and having plasmid or integrated trans gene, we saw that if the Cas13 protein is not available when you put in those chemically modified guides, things are just a lot less efficient. They still work. You still see knockdown, but it's a lot less efficient.

We have this experiment with mRNA with the Cas13 delivered as mRNA. And there, you can see that if you co-deliver, you really don't get efficient knockdown. But if you deliver the mRNA in advance such that the Cas13 protein is already made and waiting around in the cell, then when you put in those chemically modified guides, things work really well.

And so I think the really nice thing that we were able to take away from that is that we can actually do this ex vivo, we can make these RNPs, these ribonucleoprotein complexes, where we get this really nice Cas13 protein that came from your group. And then complex it just in a test tube, basically, just in like a salt buffer and then deliver it with electroporation or nucleofection directly into these primary cell types. So that was super exciting work that we were able to do and was totally motivated by this idea that chemically modified guide RNAs, if they're just waiting around – you’re going to lose some effects, some impact if the protein is not there.

 

What impacts on future therapeutics and/or diagnostics do you see this work is having?

That's a very open ended question. I think the diagnostics front is a little bit easier to address, right? Because we've already seen, I mean, less so with Cas13d, which is what our three groups here worked on together, but more with some of the other Cas13 orthologs like Cas13a and b.

And I know that your group, Brett at NEB, you guys have been very involved with labs all over the place with the Cas13 diagnostics making Cas13a and b enzymes. And I think that's very cool because people have now shown you can freeze dry it, you can deliver this stuff in the field and it can work basically anywhere.

It can be quickly reprogrammed. So, if you have to update for Omicron variant or something like that, and you have to change out the guide RNA, it's really simple. So I think Cas13 is undeniably with techniques like Sherlock from say from Feng Zhang's lab already had very impressive impact on the field of diagnostics.

So I think about diagnostics. A lot of what my lab does are these high throughput functional genomic screens, right? We do this, like using Cas9 to target every gene in the genome, understand what genes are important for cancer drug resistance or immunotherapy resistance.

And again, thinking about what are the unique properties of Cas13? I mean, this is a little bit different than what we've been talking about with the protein and chemically modified guides, but I think we're very excited to use Cas13 also to screen non-coding RNAs. Again, these nuclear localized RNAs, and you might say, well, why is that relevant to diagnostics?

We don't know a whole lot about some of these classes of RNA, like these long non-coding RNAs, or circular RNAs, or enhancer RNAs, or a million other RNAs that are in the nucleus. And I think because we can, as your group did, you guys stuck this NLS tag, this nuclear localization sequence, on Cas13, we can just direct it wherever we want.

In the bacteria, there's no nucleus, right? So, we can just sculpt the protein to go wherever we want it to go. It doesn't require other proteins. It just needs Cas13 to kind of do its thing. And so for me, I think that's a very exciting thing that we can potentially discover, say, diagnostic, prognostic markers in that non-coding RNA space that might be useful for different indications, different types of cancer, things like this.