a16z Podcast - a16z Podcast: Damage-free Genome Editing -- Next in CRISPR

Episode Date: June 26, 2019

Two recent scientific journal papers show what's possible when CRISPR moves from cutting DNA tool to a full-fledged platform -- expanding its toolkit for medicine across R&D, therapeutics, and diagnostics: "Transposon-encoded CRISPR-Cas systems direct RNA-guided DNA integration" in Nature -- by Sanne Klompe, Phuc Vo, Tyler Halpin-Healy, and Samuel Sternberg (of Columbia University) "RNA-guided DNA insertion with CRISPR-associated transposases" in Science -- by Jonathan Strecker, Alim Ladha, Zachary Gardner, Jonathan Schmid-burgk, Kira Makarova, Eugene Koonin, and Feng Zhang (of the Broad Institute) What do these two papers -- both about techniques for getting rid of the need to cut the genome to edit it -- make possible going forward, given the ongoing shift of biology becoming more like engineering? Where are we in the wave of the genome engineering "developer community" building on top of CRISPR with a constantly growing suite of programmable functionalities? a16z bio general partner Jorge Conde and bio deal team partner Andy Tran chat with Hanne Tidnam about these trends -- and these two papers -- in this short internal hallway-style conversation, part of our new a16z Journal Club series. This podcast is also part of our new a16z bio newsletter, which you can sign up for at a16z.com/subscribe

Hi and welcome to the inaugural edition of the A16Z Biojournal Club. I'm Hannah. Our goal here is to take an interesting new research paper in the field and talk about why it's cool, break down a little of the science involved, and consider what the implications of this research for industry might be. So in our first take, A16Z general partner on the biofund Jorge Condé and deal team partner Andy Tran, chat with me about two papers recently published. The first, transposon encoded CRISPR-CAS System's direct RNA-guided DNA integration was published by a group under Samuel H. Sternberg at Columbia in Nature of June 2019. The second is RNA-guided DNA insertion with CRISPR-associated transposases with a team under Feng Zhang from the Broad Institute, published in science, also in June

2019. We talk about what these papers are all about in the field of CRISPR development and beyond. This mini podcast is available as part of our new A16Z Bio Newsletter. So if you like it and you want to hear more or read more, please sign up for the newsletter at A16Z.com forward slash subscribe. Let's talk about what specifically is happening here. What does what does transposon encoded CRISPRCase system direct RNA guided DNA integration? Like what does that actually mean? Can you help me understand what was the interesting science that was going on here? Yeah. So what's really interesting in the field of CRISPR lately and actually a few papers, you know, that came out in the recent time,

developed a way to basically use these transposjon machinery inside the cell and use CRISPR to direct it into a specific place in the genome and really edit the genome without cutting it open at all. So you can think of it as this scarless type of genetic modification. Essentially what a transposeon is is this sort of phenomenon that's been observed where you have sort of these genes that jump into the genome. It's kind of mysterious. What the jumping is really used for, people surmise that potentially, you know, it helps, you know, the cells and organism evolved in general. So what this paper shows is to utilize this CRISPR machinery known as Cascade, and it's formed by, you know, CAST 6, C7, CAST 8 proteins, to really insert genes into the genome.

And this Cascade protein is directed to the chromosome by using guided RNA, you know, CRISPR machinery. And it then binds to this transposes associated protein, TNIQ, and it allows them to recruit this transposable elements and then effectively integrate the gene into the genome. And this is super powerful because now we're able to, you know, add genes in the genome directly and precisely without cutting it open. It's a scarless way to modify the genome. So in some ways, this is almost like if CRISPR is about surgically inserting or editing DNA, this is almost in many ways like plastic surgery, right? It's scarless. It doesn't leave a mark and therefore it in some ways is less risky from an intervention standpoint on the genome side.

What it really boils down to is that this machinery utilizes a way to really efficiently integrate genes into the cell, into the genome specifically, without having to cut it open at all. And to the metaphor of the surgery, this is really important in the whole context of gene therapy as a whole, right? Because when we start off in gene therapy, we had these random integration of these trans genes into the cell. And, you know, this was a more stochastic process.

So think of it almost as, you know, the arrival of this paper, this paper and papers like it are showing us that we've gotten to a point where we have a fully programmable ability to integrate new genes or new DNA into a genome without having to first open up or break apart the DNA to do that. So let's go back and actually situate it into the development of the science, what it represents for where we've gotten, from where we began. Yeah. So if you want to do the really the fast forward montage version of this. Yeah. The 80s montage. Yeah, exactly. So I'm sure you'll put in the pop music behind it. Right. Yeah. The overall's in the

paintbrushes. Here we start in the 60s. And the time motion. Yeah. Let's go. But look, I mean, if you go all the way back, you know, one of the earliest sort of technologies that came to the fore was the discovery of something called the restriction enzyme. And the restriction enzyme was this ability to take this protein that could cut DNA at these predetermined sites essentially open up the DNA, and then you could introduce new genetic material, and it would eventually integrate randomly, but it would integrate into that DNA. That's what gave us the ability to do recombin DNA technology, which gives rise to the entire biotech field. One of the earliest applications of biotechnology is getting bacterial systems to integrate human insulin so that we could

coax bacteria to make that drug, or human insulin on our behalf, of course, to treat diabetics. So that starts the whole field. Now, you know, if you sort of move forward, the objective always has been, can you cure disease by repairing or replacing what's broken in DNA? Right. And so the whole field of gene therapy arises with this idea that, you know, can we introduce a corrected version of a non-functional gene in a patient and have it do the job that the non-functional gene cannot do? The first version was just put that gene into a viral vector, into almost like a delivery vehicle, introduce that into patient cells, that gets taken in by the cells, and the gene just

starts to sort of do its job. It integrates randomly in the genome, but it does the job that needs to do. Hopefully it integrates, or maybe it doesn't, but it just does the job to compensate for some mutated gene. Yeah, so it's basically almost a parallel support system that the cell has. We just had the first gene therapies approved over the course of the last couple of years. If you look at the spark therapy, gene therapy drug for this rare inherited form of blindness. And so that was one big advance forward.

Now, that is introducing a full gene and just hoping it gets taken up by the cell. Just kind of throw it in the mix. Yeah, you throw it in the mix and it has its own risk. The original discovery of CRISPRCast 9, that system, the way that system works is by making what's called a double-stranded break in the DNA molecule. So if you think of, you know, we all remember DNA from high school biology as sort of the the beautiful double helix. So imagine sort of cutting that double helix in two. Yeah, it sounds brutal. Right. And then, you know, and then putting it back together. Yeah. That's not riskless, right?

And by the way, it's very well known and documented that one of the big setbacks that the gene therapy field had a couple of decades ago was when they ran a clinical trial at the University of Pennsylvania, there was a patient named Jesse Gelsinger who died after receiving the therapy just because there's a lot of risk associated with introducing, you know, a viral capsid with a gene into a cell into human being where you could have a catastrophic result and that patient died and that actually put a big pause on how we thought about developing gene therapies for humans. But that approach will have and does have therapeutic potential and there are several companies that are pursuing developing CRISPR-Cast9-based therapeutics. As that advanced in parallel, we started to see other

gene or genome editing technologies come to the four. So the first one was one known as zinc finger nucleases, which really didn't have as much uptake as one would expect, because it's just very hard to deal with these proteins. Housing fingers work is that you actually use these proteins to bind onto DNA sequences, and every time you want to iterate to find a new target, costs tens of thousands of dollars and a few months to develop one protein. So it's enormously expensive and labor intensive. Exactly.

So it never really caught on like wildfire in the community, right? It's a very bespoke process. Right. It's, you know, one at a time type of thing. And then there were other technologies. There was Talens, which, you know, was a bit better than syncing nucleiasis. But really the big sort of shift in the field was the arrival of CRISPR. And so when scientists discovered CRISPR, you know, they noticed that it was this always constant

evolutionary warfare between bacteropages and bacteria, you know, for, you know, billions of years. Bacterophages or viruses. Yes. And so, you know, when these viruses, you know, inject their viral genome into the bacteria, that you can think of it as the bacterial immune system, whereas they're able to sense these little snippets of viral genome, they would create a vaccination card from it, and then they'll put it back into what is known as this CRISPR array.

And this would be like a vaccine card or a vaccine database, if you will. So every time it recognizes that foreign viral sequence, then it knows what to do. It would know how to snip it away, right? And then people have hijacked this machinery. So what if we can program, you know, this outside of bacteria and use it in human cells and program it to sense not viral DNA,

but specific targets in the human genome. And then then we can program and effectively target anywhere we want and sell, right? So now bring us forward to today. So is this new development, does it mean that the therapeutic potentials are essentially less risky? Or are there new possibilities that we haven't been able to do before? Or both? So basically, you know, it's really powerful

because when we talk about the first gen of gene therapy to the second gen of CRISPR, this paper really represents this third wave in this scarless genomic editing. This is basically genomic surgery at its finest, right? You know, when we think about larprescopic surgery and all these advanced surgery tools, we want to have, you know, basically a scarless methodology of doing surgery. This is a way to really cleanly integrate, you know, genomic segments into the cell without even touching, right? And then this also represents an even broader tool of, you know, the entire, you know,

genomic toolkit landscape. The real promise on the near term side is that we get to a point where you can make scarless integrations of genetic material into the DNA. So it's less traumatic in that regard. So if you can do that more precisely, there's less scar tissue that hopefully is a better intervention altogether. So that holds great promise from a potential for future therapeutics based on this type of technology. I think the other thing that's worth noting is that in a relatively of the short period of time, the programmability of these kinds of systems has improved, has improved dramatically. So the kinds of things that we could do, that we can do now with, you know, with CRISPR, based on these kinds of advances, gives us a very broad repertoire and

toolkit to work with, whether it's for therapeutic applications or for diagnostic applications or for any other number of things. When most people think about CRISPR, developing CRISPR for human health, they're thinking about therapeutic applications. But the reality is there's also a lot of potential for diagnostic applications. So going back to the original discovery of the CRISPR-Cast system, this was essentially the memory bank or immune system for bacteria to remember what viruses had attacked it before so they could protect themselves going forward. And the way it does that is by essentially cutting that viral genetic material so it's ineffective, essentially basically, you know, cutting it off at its Achilles heel, so to speak.

And so, as you can imagine, if you're hijacking that capability for therapeutic purposes, you could also hijack that capability for diagnostic purposes. And a way that that could potentially work, for example, is if you know what bacteria or what virus or even what mutation in DNA you're looking for, in a, say, a human sample, blood sample, or urine sample or anything like that, if it's present, it will get cut by the right-cast system. You program the CRISPR-Cast system to say these are the sequences of genetic material that I am looking for in the sample. This is basically the search engine. Like you're doing essentially a Google search.

And you basically just look for the kill switch to have been activated. You just look for the identifier. So you basically say, if I want to look in this patient sample, let's say I want to look for a specific bacteria or a specific virus or specific genetic mutation associated with disease, you can say if you find the presence of any one of these sequence, Those are almost like the search terms, cut them. And when you cut them, you can essentially engineer the system to send out some sort of a reporter, a reporter. Usually it's a visual marker. And so basically if it lights up,

it's because the CRISPRCAS system cut the DNA you told it to look for. And that has a potential diagnostic application. And you could run that diagnostic essentially without a lap, right? Because it just happens with the biology. So a whole other kind of new tool, essentially. Yeah. And I think the diagnostic applications for these kinds of technologies of technologies are actually pretty intriguing because most of the way we do diagnostics

is based on developing some very, you know, specific biological or chemical assays. So look for something and if so have a reaction take place and one that I can visualize and quantify or quantitate in some way. But here you're just basically letting the biology do the work for you. In the CRISPR toolkit, you know, a lot of the initial, you know, applications was using, you know, CRISPR CAS-9, this one nuclei. and actually the diagnostic application that Jorge was talking about was actually using these other nuclei known as, you know, CAS 13 and CAS 12.

And so all these different CRISPR proteins, CAS 9, 12, 13, X, Y that we're continuing to discover has all different fundamental applications, right? Even fundamentally changing what these CRISPR nucleases even do, it doesn't even cut anymore. It can do, you know, scarless editing. And then we can even add different applications on it. You know, there's new applications adding, you know, deamination. is just to do base editing. So we can really do single base pair resolution editing.

There's really the final frontier of precision in terms of genomic modification. And I think what's really important is that we've really seen this shift from when it became this random bespoke science and really turning into full-fledged modified engineering tool. And this paper that we talk about here is not only a great representation of this engineering biology thesis, but also a pretty huge potential step change. for the field as a whole. You can program this to turn genes up and down as opposed to just editing them.

There's even work that's ongoing to use this technology to image DNA directly, which is a pretty remarkable thing because since this is acting locally on DNA, you can add all kinds of agents to make it imagable, so therefore you can observe,

you know, chromatin or genome structure correctly. So there's a lot that can be done with this technology, and you hear the old adage about the pickaxes for the gold. And with these tool kits, I mean, we are quite, quite literally panning for gold here.

I mean, where these things get found, they're found in soil and in, you know, ocean vents. New York City Subway. The New York City Subway. So these people are quite literally looking in nature because nature has ingenious ways to do a lot of the things that we're trying to do from an engineering or engineering biology

or programmable biology standpoint. And so I think it's a remarkable moment to take pause and see how far this technology has come in a relatively short period of time. This generations were competent DNA or restriction enzymes that really gave rise to the biotechnology industry. I think this toolkit, this CRISPR toolkit, as we're describing it and discussing it,

represent sort of the next frontier for what will happen in biology. So an incredible development of precision in what the tools can do and at the same time a huge expansion of what that will enable us to do going forward. Thank you guys so much for joining us. Should I say on the A16Z Journal Club?

No, no, please don't say that. Yeah, okay.