Patrick Short  00:02
Welcome, everyone to the genetics podcasts. I'm really excited to be here today with Dr. Paul Nioi, the Vice President of discovery and translational research at ALNYLAM Pharmaceuticals. if you're not familiar with found ALNYLAM, they're a pharmaceutical company founded in 2002. Based on the Nobel Prize winning breakthrough, the discovery of RNA interference or RNAi, for short on ALNYLAM has had I think, at today, about five programmes approved and have another 10 Plus that are in their development pipeline focused on this RNAi technology. Paul's team is focused on early target discovery and validation, using genetics, as well as a number of other kinds of omics and next generation technologies. And I think pretty uniquely ALNYLAM has embedded genetics into their target discovery and validation process from day one. So we're going to spend a lot of time talking today about how genetics is used in the development of new medicines, some of the discoveries over the last couple of years for major population genomics programmes that Paul and his team have been involved in, and also hopefully, Paul's journey himself into genetics and drug discovery. So with that long winded intro, Paul, welcome. And thank you so much for joining me.

Paul Nioi  01:07
Hi, Patrick. Great to be here. Thanks for inviting me.

Patrick Short  01:10
I'd love to get started with how you got interested in genetics and drug discovery in the first place. Take me back to when you first decided that this is how you thought you might spend your career.

Paul Nioi  01:19
Well, my career sort of follow the development of technology, and in many ways that allows you to look at genetics. And in the first place, I was always someone that was fascinated by omics approaches. And I studied transcription factors when I was a PhD student, during my postdoc microarrays became popular. And I thought, well, this is an amazing way to be able to study transcriptomics, instead of just looking at a couple of genes, I'm going to look at whole programme of gene expression that this transcription factor that I was interested in controls. And of course, with that comes a bit of learning about how do you handle all of this data by that you make sense of it. And so I became really fascinated by, you know, the kind of big data omics type of approaches in the context of drug discovery, though, it was a really good example of something that was very hyped up and never lived up to expectations. So I was part of a group back in the early part of my career at a company called Schering plough where our goal was to use transcriptomics data to revolutionise drug discovery, we're going to discover new targets, we're going to make sure that trucks were safe. And we quickly discovered it wasn't quite that simple. And so I actually started getting interested in genetics at that point, mainly because I thought this is much simpler. Either you have the mutation, or you don't have the phenotype or you don't, there's no, you know, you don't have 20,000 transcripts going up and down you, you know, there was just a simpler problem for me to wrap my head around. And I joined Amgen in 2010. And I got really excited because we had a couple of programmes that were I think preclinical when I joined just about to go into the clinic where human genetics had played such an amazing, prominent role in the discovery of these targets in the first place. So one was PCs K nine. And everybody knows the story of you know, there are individuals that have loss of function mutations and PCs, K nine, they have very low cholesterol levels, you could also find people that have gain of function mutations, where you see the opposite phenotype. And so that was what led us to this being an interesting target at Amgen for for Lester Alenia. And similarly we had another programme, which was called Roma Susan Mab that targeted a gene called Sclerostin. And Sclerostin, is involved in maintaining the skeleton. And there are individuals that lack the Sclerostin gene, and they have big, bulky strong bones. And so again, there was a connection between what you see from genetics, and, you know, discovering a drug. And so this was a just an interesting connection. And I think also knowing that there have been so many failures in in every company's pipeline, you know, the the rate of failure in the farm industry is very high. It seems to me that it just made a lot of sense that at least if your initial discovery of a target and your understanding of disease biology came from a human, and not from a contrived animal model, or cells and culture, but actually came from studying humans, then it was much more likely to work in the clinic and of course, that has bornite. There are a number of retrospective analyses that have shown that and that was really the point where I got very, very interested in in genetics. I ended up Amgen bought DECA genetics, I ended up going to Iceland and it's very privileged to work with the call decode data for for quite some time. And you know, cardi Steffensen was it was a great mentor and taught me, taught me a lot about the ins and outs of genome wide association studies, etc. And so that was a that was a brilliant introduction to hardcore genetics for me, and then subsequently joined on island, and 2080. And our genetics efforts here five years ago. Nice. So that was my my story of how I got to where I am now.

Patrick Short  05:38
What was it like living and working in Iceland? I had Khari. On the podcast, I don't I don't remember exactly which number episode, but it was probably about a year ago now. And he's obviously an amazing scientist, and has been at the forefront of the field for a very long time. What was that, like?

Paul Nioi  05:52
It was amazing. The place itself, I don't know if you've ever been, it's an amazingly once an amazing place. And so I was living in California time. And I've got four kids, and I only had three at the time. And so we upped sticks from sunny 1000, Oaks, California, to Reykjavik. And we had from from my personal point of view, and absolutely an amazing and amazing time, we probably could live there forever. My wife still longs to move back there someday, but we loved it. And this was 2014. And at that point, no one had data, like decode. So I think we had sequenced three times whole genome sequence three times in Icelanders, and genotype two, but half the population so 150,000 people, and then because of the genealogy exists, I think back to the year 800, or some something like that, you can do an imputation and amazing accuracy. So we had this incredible set of genetic data. And then of course, we had access to a multitude of medical records. And we were doing studies that back then, people could just imagine, you know, they couldn't do anywhere else. So it was it was incredible.

Patrick Short  07:14
That's amazing. And you mentioned this retrospective analysis that a lot of groups had done, and you've kind of seen the evolution over the past 10 years of what the genetics and just drug discovery toolkit is like, what what is it that makes genetically validated or verified, targets a lot more successful? Maybe you could talk a little bit about the major ways that genetics helps you talked a little bit about linkage to the phenotype itself, but what what else is there that's under that toolkit? Yeah, so

Paul Nioi  07:43
there's, there's kind of, I would say, two major components to it. And they're linked to the reasons that drugs fail in the clinic. So the two major reasons that you have clinical failures are a lack of efficacy, are some type of safety signal. Okay, so we unpack that. You think first about the efficacy side? You know, I would say that, in the majority of those cases, the reason that the drug sale because of a lack of efficacy is because the target the drug target is wrong in the first place that we were targeting something that doesn't actually seem to be important in the context of the human disease. And if you if you keep stepping back and say, Alright, where did this target come from? In the first place, the majority of those cases, you'll find it was from an animal model model, a hypothesis that someone had that was published in the late to sell paper or, you know, some some contrived culture system or something along those lines, not from a human. And so the the kind of, premise of it was, well, if you use genetics use human genetics as as your discovery engine, then perhaps what you're going to find is actually going to be more relevant. In in a human, of course, you can argue, well, you've had the mutation from birth, drugging someone in their 40s, or 50s, is very different, but but still, when we've looked at that question, retrospectively, and said, Okay, if we look at all of the approved drugs over the past 20 years, for example, and we say versus those that were not approved, what do we what do we find in terms of human genetic validation, and you actually see that if you have genetic validation for your target, you're somewhere between two and five times more likely to to get an approval, and that's a huge improvement from from where we are, you know, 94% of all trials or something like that. Failed historically. So that's the first piece. The other part that is often overlooked is is the safety part. And we published a study, I think four years ago now, where we essentially look at that question in a very similar way to the to the efficacy question except this time, we said, Okay, if we look at safety signals in the clinic for drugs that have failed, could we have predicted that happening from the genetics of the drug target. So this, of course, is just on target safety, not off target safety. But if you look at the genetics of the target, you may find that it has an association or there are variants that have an association with the desired phenotype, the disease that you're trying to treat. But if it's pleiotropic, and you see, let's say, three or four other phenotypes that perhaps are undesirable, can you predict those happening as well? And the answer is yes, you can. And actually, if you if you take that into account, the not just looking for evidence of on target safety, but looking at sorry, evidence of on target efficacy, but looking for evidence of on target safety, again, you have a big influence on the likelihood of being successful. So those are the two principal ways making sure that the drug that you develop from an on target point of view is safe, and stacking the odds in your favour that you're going to see efficacy in the clinic.

Patrick Short  11:20
Maybe we can talk about the safety side a little bit more, because we've talked to a lot of guests on the efficacy side. And I think you explained both really clearly. On the safety side, it seems like there might be two, two major pieces. One is what you described, which is sort of like a fee was right, where you're taking the gene, looking at all the all the possible phenotypes that it relates to not just the one maybe the disease you're interested in, but other off target ones. But then I think there's another maybe more de risking side of it, which is looking for human knockouts,  who have the gene of interest completely knocked out and are otherwise healthy, righ t? Maybe you can talk a little bit about where the where the data points leads you to potential safety concerns to be on the lookout for and where maybe it tells you that knocking this gene out with a small molecule or gene therapy or otherwise is actually very likely to be a okay, because they're humans walking around with this gene knocked out and yeah, in the real world.

Paul Nioi  12:15
Yeah, it's it's a really, it's a really good example, actually, of where you can use genetics for that question. So let me give you an example to illustrate this. So we have a drug on island called Blue Mascherano. It's approved in the US to treat a disease called primary hyperoxaluria, or Pah, and this is a defect in the liver, but that causes an inability to convert or to metabolise something called glycolate. And when you lack this particular enzyme, you produce an something soluble called oxalate, which ends up in the blood, it precipitates, and it lodges in the kidney and often presents its kidney failure in kids. So we had an idea hypothesis that well, we could stop back conversion. So we knew what the genetic defect was. It's a mutation in a gene called EGT in the liver, and we knew that biochemistry is really well worked out. And we knew that the the enzyme that catalyses the conversion of glycolate to oxalate is called h2o One or go one likely oxidase. And if we block that we're not done with SI RNA, we were confident that we could stop oxalate from being produced, but we didn't know if it was going to be safe. Right. So that that became the question. Well, we know that we know the biochemistry, we certainly know what's going on. It's a good idea, but I treat this, but is this going to be okay, what happens if you've got a tonne of likely building up that isn't going to be a bad thing? So we did a study with the barn and Bradford cohort that we actually found one individual who was no for ha a one org likely oxygen?

Patrick Short  13:57
I remember reading this paper a while back. Yeah, go on. But I do remember reading. Yeah, again, how amazing. So

Paul Nioi  14:02
we were, we were excited. So there's a South Asian cohort, as you probably already know, and there's a high rate of consanguinity within within that population. And so this person had long stretches of artha zygosity. And we found that, you know, in one of the strategies that she was on both alleles had the same loss of function mutation in HL one. So we were able to collaborate with the PIs there and actually have her volunteer to come back into the clinic and do essentially a type of deep phenotyping study to look around. So she was in her 50s She had had kids, she she was otherwise very healthy to some blood, their glycolate was through the roof, like 12 times the upper limit of normal. Wow. And and, you know, we sequenced excetera and made sure that she was a carrier that mutation but She was living, you know, otherwise healthy life. And that was the piece of evidence that gave us confidence that, okay, here's someone that's had this from birth by chemically, they've definitely got it, we can see that they've got it. And then yet, you know, they're there, they're actually healthy. And so that was really instructive for us to say, Okay, this is probably going to be well tolerated when we we take it to into the population, of course, we're, in some cases with the master, and we're talking about treating children. So you know, it even adds that extra level of of concern that we want to make absolutely certain that this is this is going to be saved. So that's like the most beautiful example if you can find people that are know for for your target, and they have and you see a benefit, and you don't see any downside to it. That's the most wonderful example. Most of the time, you don't have that most of the time, you're dealing with maybe more common variants with lower effect sizes, missense mutations where you don't necessarily know the upregulating time, right, what is it doing to the function and so but you can still perform a few losses you as you pointed out, you can look at the the associations that come from that. And then often there may be some lab work that you need to do in combination just to characterise the variants a little bit just to understand what they're doing. And I'm taking so that can be a little bit more complicated, but still, you know, very instructive when it comes to the biology of the target, which is essentially what you're trying to understand. Yeah,

Patrick Short  16:35
no, that's a that's a beautiful example of it. The since the days of 140 150,000 genotypes and 3000 whole genomes in isolated you've now got collaborators like the Born in Bradford study who I greatly admire, I think they've done a tremendous amount of science. And they've also engaged with the community in a really deep level that I don't think we see a lot of Population Genomics programmes do. We've got the UK Biobank, we've got many others around the world, what's changed from the Iceland and decode days when that really was the the largest and most comprehensive programme to where we are today?

Paul Nioi  17:11
Well, in a lot of ways, the big changes be the ability to affordably sequence large numbers of people. And you gave UK Biobank as an example. It's a perfect example of this type of approach where it started in 2006. So this was before I even arrived in Iceland, it started in 2006. But they recruited half a million people between 2006 and 2010. And then in 2018, on ALNYLAM got together with Regeneron, and several other big pharma companies, to sequence the exomes of all 500,000 people in that cohort. And because we were looking at exome ism, because the cost of sequencing had dropped precipitously from, you know, the 10 years prior to that, it was something that, you know, coming together in a free and competitive way with with a group of other companies where essentially, you're just collaborating on generating data than ever wanting to ask whatever question they want about the data, but then, you know, you're splitting the cost of that by whatever it was 810 companies are involved, it becomes realistic to be able to do something like this. So and then, you know, you have a data set is just, I couldn't even conceive this something like this 10 years ago, so it's incredible.

Patrick Short  18:43
When you look at a programme like the UK Biobank, if we just take the UK Biobank as an example, I'm interested to hear from you what the most useful directions for the future be. Is it about increasing the scale, say, from 500 to 5 million? Is it about the same people but more diverse types of data? So we know the UK Biobank is doing imaging and taking all kinds of useful samples to do more, maybe deeper phenotyping on that same population. There's also an important piece around enrolling a more representative population, not just in the UK, but of the world. And then also this recontact ability to the longitudinal aspect of these programmes. So how important is it to have a snapshot in time versus a sample every year, for example, to light understand how a disease might progress, I'd love to hear from you if you could pick one or two of those that are on your wish list where the focus would be.

Paul Nioi  19:39
So we recently are maybe a year or so ago now joined a new initiative in the UK called our future health, which you can think of as as the big brother to UK Biobank and they're not quite the same, but so I'm actually the chair of the founders board or for our future health for this next year. So our goal, again is based in the UK or our goal is to recruit 5 million people in in the UK. And for that cohort to represent the population of the UK as a whole in terms of socio economic status, but also importantly, race and making sure that we have groups that are historically understudied actually represented the levels that that these populations are finding within within the UK, and they have access to medical records, we're going to generate genetic information on these individuals and volunteers. So that kind of answers your question in a lot of ways, because there's a couple of things. One is, you know, as you sequence more and more people, you it becomes, you know, the the really interesting variants are rarer, and rarer and rarer. And you need bigger and bigger cohorts to actually have enough carriers of each of these interesting rare variants to be able to make sense of what they're doing. So the scale is really important. And then, of course, the other way to look at that question as well. If you have different ethnicities represented in the data, then you're going to find interesting variants that are specific to each of those populations, and that maybe are not finding the bound in the white European population, but are actually very prevalent in you know, another, another group, so that these are important ways that we can, you know, explore the role of each gene and disease. So that part's really important. The other thing that's key with our future health is ability to recontact. Participants, we have a really interesting example that we found in UK Biobank, whereas as I'm sure you know, you cannot be current type participants were so we have an for a condition called hereditary a TTR amyloidosis. And this is a condition caused by mutation in a gene called Transthyretin missense mutation, number of different missense mutations, the most common of which is a mutation called V 122. I, and it's found in dominantly, almost exclusively in people of West African origin. And then UK Biobank, you find it in individuals on black Caribbean descent, mostly, when we look in UK Biobank at carriers of this mutation, we find around 400 carriers, three homozygous carriers, a percentage of those people had signs in their medical records that were consistent with them having the disease. So they had carpal tunnel syndrome, they had autonomic dysfunction, evidence of neuropathy, a number of the hallmark symptoms, and so we would love to have had an ability to say, hey, you should go and see the doctor or have them contacted by a physician, and worked out to see if indeed, they did they, they did have a condition. And by the way, even though, if you look in the medical records, none of them are diagnosed, none of them zero. Right. So that is, you know, it's part and parcel of UK Biobank, but it's something that we're hoping with our future health, it's going to be a way of pioneering, you know, the ability to detect genetic disease earlier and actually intervene in a way that is going to help help those patients. So that's the other really, really key part of that for us. The UK

Patrick Short  23:34
Biobank isn't alone in having this recontact challenge. What what do you think it is that has changed in the last couple of years that has made everyone a lot more focused on this and made it difficult before?

Paul Nioi  23:46
I think that there's a couple of things and by the way, I would say that it's I'm not sure that it has fundamentally changed, there's still a raging debate about whether this is actually doable or not, partly because of concerns over overwhelming the health care system, because not everything is 100% penetrant. And you can be a carrier, but you might be you know, you might go your whole life and you know, you never have whatever the condition is by you might feel like you should go and see your GP and so we have to we have that needs to be worked out. We have to figure out how do you people data in a responsible way, and you balance the burden that's going to be placed on the healthcare system if you were to just dump everything on to participants, so and that hasn't been worked out, frankly. So that part I would say is a work in progress and has hasn't changed. But I do think that there is a there is a growing desire to shift the focus of medicine from treating chronically ill patients to to one of prevention and of course, this would be one tool to be able to do that if you you know, use Someone had a high polygenic risk of type two diabetes, could you? Could you intervene early and actually prevent them from developing full blown TTD? Or, you know, could you find patients that are predisposed to hypertension or have hypertension and treat them before they develop cardiovascular disease? And, and by doing so maybe then you can create a more sustainable health care system. So, so that's sort of the long term goal of efforts like our future health and, you know, so one step along the way as I did you engage the participants and give them information and give them the power to act on your health?

Patrick Short  25:44
Yeah, I completely agree. It's, I think it's an amazing an amazing initiative. And a huge part of the reason I moved to the UK initially to do my PhD was how forward thinking the government has been here and backing us. First was 100,000 genomes genomics England well, even before that was UK Biobank, and many of the early cohort studies like children of the 90s. And and I just think it's such a lead. In this regard. I wanted to ask one follow up on the topic of Population Genomics programmes and target discovery before we then get into a little bit more about RNAi. And the work you're doing at El nyalam. You mentioned earlier about how genetics can be used as a kind of discovery engine. And we talked for you through a couple of examples around really significant monogenic disease genes, I wanted to hear your thoughts around how you use genome wide association studies, polygenic scores, cases where you may not have a single gene that has an enormous effect on a phenotype that you can focus on. But actually, you may have hundreds that have small to medium effects. How do you transition into this world where the genetic story isn't quite so clean as as the monogenic world?

Paul Nioi  26:54
Yeah, it's, it's a good question. It's almost like, you know, going back to good old fashioned genotyping arrays and trying to scratch in your head and wondering what to make of all these common signal. So we've gotten a bit more sophisticated. But so I would, I would say a couple of things. One is, I do think that monogenic diseases, obviously, you can go to OMIM, and read all about them. And, you know, a lot of those targets and ideas were, where we started, but on island that was the invaluable source of data that we had. And what's what's different about g y, especially in this year, we take the UK Biobank data, for example, is that when you when you sequence people on on that sort of scale, so you have X ohms genomes on half a million people, of course, you'll find carriers of pathogenic variants, and you'll see evidence of disease, but what's not in all moments because often it's there's no reason for it to be there is if you carry rare loss of function variants, but it's benefiting you in some way. Now, you might find something like, you know, PCS canine and very low cholesterol, but they're there, we've seen numerous examples of rare loss of function variants in genes, that, that confer a beneficial effect on individuals are not found in, you know, the classic genetic databases. For example, we published a paper last year on finding where we identified loss of function carriers in a gene called inhibin, e r i n HPE. What we did is we did we aggregate the variants, so we did a burden test. So we took all of the loss of function variants, different flavours across the whole 500k cohort. And we lump them together and did a gene based analysis with some with some filtering, but just focusing on on each gene and we found that individuals that had the loss of function in inhibit II had more beneficial distribution of fat in their body, so less fat around the middle, and more on the hips, as measured by by waist to hip ratio and, and that fat in the middle of your body, and I suffer from it myself, and it's as bad for you, you know, it has particular qualities that increase your risk of diseases like Type two diabetes, cardiovascular disease, nach, I blood. And of course, what we see in this in these carriers is not only as fat distributed differently, but they have lower risk of the diseases that I that I just mentioned. So that actually became a new target for us. And but it was, you know, it was just a straight up. It was one gene. Yeah, it was it was loss of function. And it was easy to interpret from that point of view, and we've seen a number of things like that. So those are the easiest things to act on, because you know, for sure what the gene is. And you know, What the What the variant is doing to the gene. And so therefore, you can, you know, you can come up with a pretty good idea, but what your drugs should, the more complicated thing is, you know, honestly, I, we haven't really wrap their heads around specific use cases, for polygenic scores at this point in time, it is going to be part of our future health, actually, they're going to develop, and they're going to come up with polygenic scores, I think, for the whole cohort over the course of time for a variety of conditions. And so we'll get a really good insight from that and a population level as to how powerful these these things can be. I mean, we know what are some examples, right, but you do have similar effect sizes for for a polygenic score in terms of our risk of disease, as you do for, you know, one of the bonnet gigantic version of the of the same disease, but I'm picking what, what's the important part? Yeah, amongst however many, you know, variants are making up that score is, is a tricky thing. What's the what's the critical node to go after?

Patrick Short  31:12
Yeah, I'm interested to see it feels like that's a problem that a lot of people are figuring out how to crack because I don't I don't know the answer. We had Peter Donnelly on a previous episode, and probably he knows the answer, I'll get him back on what he thinks. But it seems like the kind of thing that it's either going to align on a number of pathways and like you said, you can figure out the critical note, or it's not, and it's going to be complete chaos. And then we just use it for screening and early detection rely on other kinds of interventions. But probably, it's a mix of both. I wanted to shift gears a little bit to RNA interference. This is the cornerstone technology that I mentioned in the intro, Nobel Prize winning that was behind Alnylam initial entrance into precision medicine and the market. Maybe you could talk a little bit about what RNA RNAi is, how does it work, and also just compare and contrast to some of the other precision medicine type of techniques like CRISPR, cast nine gene editing that people may if they're not too steeped in the field, be unclear on what the difference is between RNAi gene editing and some of the other concepts.

Paul Nioi  32:14
Yeah, happy to do that. So in a nutshell, you know, RNAi, in its natural form, is a mechanism that every cell in your body has to defend itself from viral infection, RNA viruses, in particular, that's why it exists. So within each cell, there is a protein complex that's called risk, which is RNA induced silencing complex is that is what the term stands for. and its job is to recognise double stranded RNA floating around in the cell, which shouldn't be there, of course, because RNA is single stranded, not double stranded. So if it sees double stranded RNA, binds to it, and unwinds the two strands takes one of the strands and it goes searching for things that that match that sequence, ie, you imagine that you'd be infected with an RNA virus. And it's grabbed a bit of that RNA genome, and now it's hunting for the intact genome, when it finds it. It chops up into into debts and silences expression of viral genes. And that's its role. But as with with all exciting discoveries, we figured out ways of hijacking that system to do other things. So the the original data that started on island was actually exactly that it was the experiment that showed if you if you make synthetically double stranded RNA, and you make it complimentary to a human gene, and you introduce it into cells and culture, well, that will then get incorporated into risk. And risk will go hunting for what it thinks is a foreign invader, but actually tricked into to go and looking for, you know, your gene of interest and a human cell. It'll find it and it will, it will chop the transcript and stop it from being expressed. So it's called gene silencing is a term that's often used. So that that was the idea. And so ALNYLAM started because we, you know, we thought, well, if you can do that, in vitro, imagine if you could turn this into a Madison, imagine if you could silence any gene in any cell. The then you can drug anything, you could silence anything you could you don't have this issue of drug ability anymore. You're able to go after any gene encoding anything and it could be transformative and that has been instilled is the mission of our islands. So we started in 2002, but double stranded RNA does not look like us. molecule drug doesn't look like an antibody floating around, we had to figure out ways of of protecting the RNA and stopping it from being degraded, getting it into cells, cells that you know of interest, and then do what it needed to do silencing that silencing the gene. And so really the first decade plus of our Lyrans existence with solving some of these fundamental problems, we figured out how to stabilise the RNA, we figured out how to incorporate it into lipid nanoparticles or to add conjugates that target the drug to the liver. And now we have an amazing platform where we can target any gene any transcript and we have this amazing profile of our of our drugs where you know, we we take increase around for example, which is a PCS canine silencing si RNA that was discovered here on island, you can give that drug once every six months, when you see this, this clamped, knockdown of of the PCS canine transcript in the liver. So imagine that sort of a profile, you know, you only need to go you take a pill every day you don't see your doctor once every six months, get your cholesterol check, get your VISTA of of increased Saran and off the goes. So it's it's an amazing, amazing platform. It truly is. It's a whole new class of medicines.

Patrick Short  36:24
And what is the major barrier to moving from the liver to every other organ in the body? What What's the biggest set of challenges than I imagined? It's not genetics and target discovery, but maybe some other things?

Paul Nioi  36:36
You're absolutely right. As soon as you get into an issue, there are obvious targets, to go after your back to home and you're like, oh, there's all this stuff that's been known for ages, you know, we don't need to do anything to create the we just need to get into, into this, this tissue delivery is the big challenge. So with the liver, ultimately, what we settled on was was we add a sugar on to the end of the ssrna. It's called galnac. And it's recognised by a receptor on the surface of liver cells hepatocytes called the esala glycoprotein receptor as as as GPR. And that's its whole job. The reason it exists on the surface of liver cells is to grab things from the blood and internalise them and its posts to target them for degradation. But we can get around that. And so our S RNA gets internalised into the sides, there's loads of as GPR in the surface of a beach hepatocyte, it turns over with high frequency. So it's a beautiful system. The challenge is, well, what's the equivalent system in different tissues? And how do we how do we exploit that and achieve delivery there. So we are now in the CNS with our with our first or a there. So we published the paper last year, I think in nature biotech, showing that we have novel conjugates. So not the sugar that I mentioned, but something different that we can put onto the end of the SI RNA. And if you administer it, administer that, at least in preclinical species interest equally. So into the into the CSF into the spine, that you get broad distribution and silencing targets in the brain. So we're still pending clinical data, we don't have the human clinical data yet. But that could potentially be, you know, another tissue that we've been able to target. And we continue to work on this and you know, in terms of like, Where can we go next? And I will add one thing, which is genetics is a key part of deciding where we go next, because we are looking at what are the targets? Let's say we could target lung or we could target muscle? Or we could target, you know, blood, you know, immune cells? What do we know about genetic conditions and what we might be able to do there, and it helps us just to kind of organise our thoughts when we're balancing work to our efforts,

Patrick Short  38:58
it makes total sense. And you've got, you've got only a limited amount of resources that you can focus on different places, you've got to pick your you've got to pick your next route very carefully. Yes, exactly. A little bit of a maybe left field question here. But how do you you you can do it, you can learn a lot from naturally occurring human data. And we've talked a lot about that today. But how do you know that there's not something that you're missing, that's just not naturally occurring? That the what you were just telling me about the very expensive toolbox that you have of knocking out any gene in the liver, it may just be that there, there may be a gene or a set of genes that you could selectively knock down, that does something spectacular, but it's just as an observed in the human data. So you don't know about it? Are you thinking about how to wrap your hands around that kind of question?

Paul Nioi  39:46
Yeah, that's a really good question. There's actually I think about this as there's three classes of genes based on what we know from genetics. There's those are brilliant drug targets because you know, The people that have lost a function have got all kinds of benefits and they each other do well, otherwise healthy, but you've got some tree that is really beneficial.

Patrick Short  40:11
The genetic superheroes, so to speak a genetics,

Paul Nioi  40:14
yes. And then you have genes where you know, it'd be a really bad idea to target because you see that they may be the cause loss of function may be the cause of a genetic disease, for example. And you know, very clearly, that's not something you would want to recreate the drug. And then you have genes where you don't know anything, because there's so constrained evolutionarily, you just don't find variants that you can anchor on to understand what is the importance of that gene, presumably, they have some role in development. But that doesn't mean that they're bad drug targets, it doesn't mean that if you target them, you know, postnatally, right, die, they're actually going to do something bad. And I know, a really interesting example of this is a gene called angiotensinogen, or AGT. So angiotensin is an important part of the way in which your body controls blood pressure. But it's a highly constrained gene. And we know from animal studies that if you knock it out, it has a big impact on the development of the kidney, the kidney doesn't develop properly, and it's lethal. And that's probably what why you don't find it in humans, why it's depleted from loss function. But it's the target of a number of classes of drugs, angiotensin converting enzyme inhibitors or angiotensin receptor blockers. And it treats blood pressure, we have an siRNA that is called se suranne. That is in clinical trials at the moment, that knocks down EGT transcripts. And we've shown in phase one, that we could more blood pressure in patients volunteers. But you wouldn't tell that necessarily from I mean, there are common variants that sort of give you a hint that there's, you know, has a role. But it's not a strong signal. So, so there, there has to be some scrutiny place, which I think is where you were going on that subset of genes that are constrained and could be, right could be really good targets. So I think there's some functional piece to that I know, I spent a lot of time at the beginning, telling you how bad an idea was to look in a mouse and sales and culture. But I think if done in combination with genetics, in large functional genomic screens, for example, do have a place and they do help you to unpick some some the role of some of those genes where maybe all you have is a common variant that's got a very small effect, the genes constrained but you can learn something from some lab data that you could generate that's gonna help you to figure out well, maybe this is actually really important. So

Patrick Short  43:05
yeah, it's really interesting. And there are a lot of those genes. Thank you took that question in a direction I wasn't expecting. And I'm it's gonna be really thinking we're, we're very close to running out of time here. I want to ask you one more question, which is just what is it that you're most excited about right now could be a technology, new piece of science, anything that isn't on everybody's radar yet that you think would be useful to share?

Paul Nioi  43:28
Not on everyone's radar? Yes.

Patrick Short  43:31
It can be on some people's radar, you get extra points, depending on how little it's on everyone's radar.

Paul Nioi  43:37
My I actually I don't think our future health is on everybody's right. I agree with that. Yeah. And that is what I'm I am most excited about, because I think not only is it going to be an amazing cohort, because of the scale of it, but I think the downstream potential that it has to change healthcare is so exciting that I am really all in on it. Yeah,

Patrick Short  44:01
I couldn't agree more. It's a good choice. I thought you might pick that one. It's, I don't think it's I think it will be in the next year or two, much more, especially at first data releases start to come out. Well, Paul, thank you. I really enjoyed this. I learned a tonne. It was amazing listening to the stories that you've were able to tell about genetics and drug discovery. And it seems like you've had a front row seat to some of the most interesting programmes on Earth. So thank you. I really appreciate you taking the time. Great. Thanks, Patrick. And thanks everyone for listening. As always, we appreciate any feedback you have in the episode please feel free to share it with a friend. If you'd like to leave a review on your favourite podcast player. We're on Apple Spotify, everywhere in between. Thanks again and we'll see you next time.


Transcribed by https://otter.ai