This week on the Podovirus Podcast, we explore advances in microbiome editing with Jesus Fernandez-Rodriguez, VP of Technology at Eligo Bioscience. Eligo, based in Paris, is using modified phages to target and edit bacterial genes in situ. They recently published a paper in Nature on base editing of E. coli and Klebsiella in the mouse gut without using CRISPR’s typical cut-and-repair method, which avoids killing the bacteria—crucial for maintaining microbiome balance.

Jesus breaks down how Eligo shifted from antimicrobial phage approaches to gene editing within the microbiome and where this technology could take us next. He shares how Eligo’s technology goes beyond traditional phage therapy, using phages not to kill but to modify bacteria in the body and how this could potentially be used to tackle diseases we hadn’t linked to bacteria before.

Listen to the full conversation on the Podovirus Podcast!

Or watch it on YouTube!

Jessica Sacher: I feel like Eligo has been a company I’ve had my eye on ever since we started Phage Directory in 2017. I don’t remember when exactly you started, but it was in the early days of me getting into this space and Xavier, he was always at the phage conferences and I’ve always been rooting for you guys and meeting your scientists and it seemed like you had a really cool mission. And all of us in the phage space, I feel, are hoping for successes so the whole field can be lifted. And definitely looking at using phages to edit the microbiome. I guess I want to know if that was always your focus or was it more antimicrobial, but also how it led to where you are now.

Jesus Fernandez: So you’re very right. We started Eligo more than eight years ago now as an antimicrobial company. So it was David Bikard, one of the co-founders of Eligo, that was working with Xavier. And then he got this paper out where he could use CRISPR, deliver it through a phage that would kill bacteria through a sequence-specific killing, right? So that was the start of Eligo. So we had a lot of success doing that, actually, had Eligo using modified phages to deliver a payload that contains CRISPR.

It’s actually, it’s important to know that what we do at Eligo is not really a phage. It’s based on the classical Cosmid technology, 2.0, let’s say. So we apply all the synthetic biology techniques that we know today to build particles that are really good at delivering, and with circuits that have been engineered to do exactly what we want to do.

We started using CRISPR sequence-specific antimicrobials. We had several programs with that. And then we realized with all the CRISPR technology booming in the last years, we realized that actually our technology was great to do not only that, but a way of modifying the microbiome, not only with having specific killing in mind without disturbing the rest of the bacteria that live, for example, in the gut or in the skin, but also to make them express things so we can modify the phagemids and the Eligobiotics actually to contain payloads that are going to be expressed by the target bacteria, or as we show in the paper, to modify genes in situ without the need to make really large disruptions in microbiome like antibiotic treatments; directly in the complex environment of the natural microbiome.

Jessica Sacher: Okay, awesome. So you edit, and then you either kill or you modify. Okay, I guess what I want to dive into, and I’ve always wondered this about microbiome therapeutics, is to what extent is it known for diseases — how much is it a bottleneck deciding what to target versus targeting it?

Jesus Fernandez: Yeah. So that’s a really good question. So whenever I talk about what we do at Eligo, CRISPR killing or editing, base editing, you know, people really have human genes in mind. People know we know a lot about mutations that cause diseases in humans. And the thing is that this is kind of a paradigm shift. It’s not just having a human with a patient with a mutation in the gene that you can edit. We have realized over the last year that the genes in the microbiome are very, very important for human diseases. It’s exactly what Eligo was trying to do, bring this technology to patients by touching upon the genes that exist in the microbiome. So there’s not a lot of specific targets known yet for bacterial genes. So we know bacteria cause diseases. It could be degenerative diseases, could be cancer. They could affect, for example, a treatment a patient is following, like an anti-cancer treatment. Bacteria can modify the molecule and make it inactive. They can sequester it.

So they can really do a lot of things that affect the health of patients, but it’s hard to pinpoint exactly which bacteria are doing that and what’s the gene that is causing that. And I think — and there’s a lot of people working on this right now —I think the technology that we show in the paper provides a really great way of interrogating specific gene functions in a natural environment.

So for example, you say, we think this bacteria might have a gene that does X, and has an impact in the patient. So we can really go there, modify the gene, inactivate it without killing the bug, and see if that has an effect on the disease, you know? So I think it’s a really great way of interrogating, having this kind of correlation between a gene and an impact in disease. So I think there’s going to be a shift in the way we do things. And I think that’s exactly what we need, finding these specific targets.

Jessica Sacher: Yeah, that’s so interesting. Do we not know targets because there haven’t been the technologies to probe them and to interrogate them?

Jesus Fernandez: Exactly. That’s one of the reasons. You can do things in vitro and you can say, I put this bacterium together with these cells and something happens. That’s never going to be a real disease setup. For example, in an animal model, when you have a disease model, some bacteria cannot be cultivated in vitro. So you might not be able to really do that directly because they need to be part of a consortium or it’s really just not possible to cultivate them. So at the end, having this technology that allows it to do something in the natural environment, to me is the key to really find specific targets that we can go after.

Jessica Sacher: Yeah. And touching on what you said about uncultured organisms and all the bugs that we barely know anything about because they don’t play well in the lab or in animal models. Can you use your technology equivalently on those kinds of targets, or is it going to be the same problem? How do you think about that?

Jesus Fernandez: So actually we talk about that in the paper. And while we focus on Enterobacteria here, I think one of the most important aspects of what we do is that you don’t need to have a specific vehicle, specific Eligobiotic particle. You can program to target just E. coli. We show you, we can jump the species barrier as long as they’re relatively close. For example, we jump from E. coli to Klebsiella. We can edit Klebsiella using a Lambda-based particle by changing some of the receptor binding determinants. So yes, different targets will most probably need different vehicles to be developed. And then it’s telling what we need to bind to make them deliver well. But I think what we show here is the first step in that race and that will lead us to really bring this technology to the clinic. It would be great to have somebody treated with this kind of technology and cure a disease. This is what we are after.

Jessica Sacher: Yeah, exactly. I want to talk about how you distinguish, for someone who’s not in the CRISPR space, maybe someone who is more of a phage biologist, base editing versus CRISPR. Like how do we think about the different technologies and the types of ways you’re going to edit something?

Jesus Fernandez: Yeah. So that’s a good question. And I go back to what I said before about the image that people have about CRISPR in eukaryotic cells, human cells and bacteria. I think the main difference here is when you express CRISPR in a human cell, CRISPR will cleave the genome, and the repair mechanism of the cells, the human cells in this case, will repair that; the cell would not die, it will stay alive and will have a modification. Well, the vast majority of bacteria don’t have this kind of repair mechanism. So if you cleave the genome of a bacterium, it will start degrading its own genome and die. That’s the basic basic difference.

Okay, so that’s why people use CRISPR Cas9 as an editing tool for human cells, but you don’t really edit bacteria this way — you can, but the efficiency is very low because most of them will die. So that’s why some specific mutations for example, when you specifically want to change this A to a G, you know, this is what we show in the paper. You need more specific tools, and that’s exactly what we’re applying to bacteria, just because the nuclease will kill them while the base editor will modify the gene without killing them.

Jessica Sacher: That’s so fascinating. I never realized that distinction. Do people use base editing in human editing projects or do they need to if they have the repair mechanism?

Jesus Fernandez: Yeah, you can use it, it depends on what your target is and what you need to do. For some specific targets, diseases, modifications that you need in the lab, you might just want to have a one base mutation or something more or less complex. For more complicated things or more specific things where you really want to have something specific inserted or changed, then you go for these base editing tools that we use in the paper, because that allows you to have a deeper control on what you want to do and exactly what kind of modifications you want to insert.

Jessica Sacher: How do you do things like add a whole gene or delete a whole gene?

Jesus Fernandez: So again, this really depends on the mutations. So for example, in the paper we introduce, we break the start codon of a gene. But we don’t really insert anything there. There’s no start codon, so there’s no protein being made. But that only requires the change of one base. So there’s really not any insertion going on there. To do something, like you were saying, inserting a gene, and that’s where I’m going to more complex or let’s say fancier gene editing, then you rely on new technologies or technologies that are different to the base editors that we use here. Like a prime editor, which will let you insert short fragments of DNA into the target site of the genome. You have transposon-based methods to allow you to insert stuff into the genome. So this technology exists. It’s new. People are developing that now. And I have no doubts that these kinds of tools will become really important for therapy in the coming years.

Jessica Sacher: Okay. So there’s a bunch of tools. Some of them play well with bacteria where they’re not going to cleave the DNA and cause the whole cell to die. Some of them don’t, but you guys have been focusing on this base editing to make pinpoint changes. I noticed in your paper that it’s very high effectiveness, like close to 100% effective on the population. Is that mostly attributable to the lack of the death of the cells because of the base editing, or more than that?

Jesus Fernandez: Yeah, so there’s more than that there. And I think we’ll also talk about this in the paper. I think to me, one of the key aspects of what we do at Eligo, let’s say that there are two ways of doing this. Whenever you want to target a bacterium, you have a new bug, and then you’re going to use a phage to deliver something. I’m going to develop, in this case, a cosmid phage; an Eligobiotic. You can go for really fancy phages that do a lot of things. Or you can go for a phage that you know a lot about. In this case we use Lambda, which is really a tame phage. You cannot really infect a lot of wild type E. coli strings with Lambda. It will just not do much. But the good thing about this is that you really understand what Lambda does because there’s so much literature on this phage. There’s been so many things done with this phage that you really understand what you need to touch. The first thing we saw was like, people use Lambda, it works fine with lab E. coli strains, but it uses a receptor that is down-regulated or mutated in the gut. So the first thing we did is, okay, let’s try to find a receptor. Let’s try to make Lambda find something that the cells cannot lose in the gut. In this case, we modified it to recognize OmpC, for example.

And also the famous tail fibers or the tail spikes that people talk about in phages, Lambda also has them, the wild type Lambda at least. So we modify them to be very, very strong binders to the target strain. And by changing these two receptor binding domains in the phage capsid, we could really have binding curves that were really, really accurate. We could select variants of these Eligobiotics that would bind strongly to one thing and not to the other. We could make Lambda degrade capsules of Klebsiella, or of other pathogenic E. coli, just by changing and understanding what we want to target. So going back to your question, I think it’s really important to really understand what you have in your hands so you can engineer this to make it a really good binder. So for me, reaching that 100% was really a lot about engineering the phage particle to make a really good Eligobiotics that would really recognize a strain with high affinity, even in the crowded or complex conditions of the mouse gut.

[…]

For the rest of the interview, check out Spotify or Youtube (head to the 17-min mark to pick up where we’ve left off here). Or email Jessica if you want the full transcript!

Want to learn more?

📕 Read:
Eligo’s In Situ Base Editing paper, published July 2024 in Nature: https://www.nature.com/articles/s41586-024-07681-w

💻 Check out Eligo Bioscience’s website, especially their open roles!
https://eligo.bio/