Denali Therapeutics Inc. (DNLI) Earnings Call Transcript & Summary

Hello, and welcome to Denali Therapeutics Virtual R&D Day. Thank you for joining our webcast today. I am Laura Hansen, Vice President of Investor Relations. Today, we are excited to bring in our program focused on our transport vehicle platform, delivering biotherapeutics across the blood-brain barrier. Please note that the press release we issued earlier today and the slide deck for this webcast are available on the Investor Relations section of our website, denalitherapeutics.com. Before we get started, I'd like to note that the presentation is given today at the event and the responses to the questions, will contain forward-looking statements regarding Denali's future plans, business strategy, product candidates, planned preclinical studies and clinical trials, among other things. Such statements are subject to numerous important risks, uncertainties and assumptions. We, therefore, ask that you please read Slide #2 in the deck and refer to our filings with the SEC for more details. On the webcast today, I am joined by members of Denali's management team, Ryan Watts, Chief Executive Officer; Carole Ho, Chief Medical Officer; Alex Schuth, Chief Operating Officer; and Joe Lewcock, Chief Scientific Officer. We are also very pleased to welcome a special guest speaker, Dr. Simon Jones, who joins us from his U.K. location. I'd like to take a moment to review the agenda and Q&A logistics for today. We have scheduled approximately 3 hours for the webcast, including a series of presentations, 2 Q&A sessions and a 10-minute break. Ryan will begin with introductory remarks and an overview of Denali's blood-brain barrier transport vehicle platform and biotherapeutics portfolio programs. Dr. Jones will provide an overview of Hunter syndrome, and Carole will discuss our ETB is clinical program with DNL 310 in development for the treatment of Hunter syndrome. Our first Q&A session will follow, and we encourage you to make the best use of this time by asking any questions you may have with Dr. Jones during this Q&A session. We will be saying good night to him afterwards because of the late hour in the U.K. After the first Q&A session, we will take a 10-minute break, our presentations will resume at approximately 2:50 p.m. Eastern Time, with Joe, who will discuss other TV enabled programs in our portfolio. Alex will give the last presentation on unlocking the value of Denali's TV platform. We will then conduct a second Q&A session and wrap up at approximately 4:00 p.m. Eastern Time. [Operator Instructions] And with that, I'd like to turn the presentation over to our Chief Executive Officer, Ryan Watts.

Great. Thank you, Laura. Very excited for today's R&D Day. As we dive into details around our transport vehicle technology, which is designed to deliver biotherapeutics across the blood-brain barrier. Before we go into the data and the plans going forward, I would like to set the broader context and give you some background on our purpose, our principles, our approach and our recent progress. And in addition to that, some of the plans we have in the near future in the broader portfolio. So let me start first with our purpose. Denali was founded to defeat degeneration. Our goal is to develop medicines for rare neurodegenerative diseases, ALS, Parkinson's, Alzheimer's, and this has been the primary focus of the company from the beginning. We have a deep commitment to the data that we generate, and we look forward to seeing those insights translate into medicines in this huge unmet need. I want to start with our principles. So the company has founded around 3 principles. The first is genetic pathway potential; the second and the focus of today is webcast, engineering brain delivery; and the third, biomarker-driven development. I'll define these principles in subsequent slides. We also were founded on 3 business principles: the first being a broad portfolio the second parallel investments, meaning that we move multiple molecules forward for targets that we're very interested in. Examples of this is LRRK2 and RIPK1. And then finally, an immediately relevant, our strategic partnering and I'll get into details in the broader context of how partnering has played into building the organization. So let's start with the substrate for Denali. Denali is founded on the new genetic insights that have happened over the last several decades. We call these genes, degenogenes. These are genes, when mutated, that are risk factors or causative in neurodegenerative diseases. You can see here, over time, and in particular, starting in about mid- -- about 2008, a significant uptick in genetic discovery around the areas that we're focused on, Alzheimer's, Parkinson's and ALS in particular. These genes and the genetic architecture of each of these disease areas allow us to define targets and pathways that we're going after with our therapeutic approach. Our pipeline approach, our portfolio approach has 3 major principles. First, we have a diverse portfolio. We have multiple therapeutic targets, multiple modalities, as well as indications broadly in neurodegeneration. We focused highly on a differentiated approach. This includes brain delivery, but also using degenogene biology as a platform to understand the pathways that we're working on and developing biomarkers for each of these pathways. Finally, and most important for us is our data-driven mindset. We focus early on in target engagement, pathway engagement and patient phenotyping and biomarkers have been central to the success of our early programs. And as you'll see in-depth today as we tackle Hunter syndrome, we have the same commitment to identifying new biomarkers that allow us to show that our medicines are being dosed at the right level and modifying the downstream pathways. So I'm highlighting here the progress that we've made over the past 2 years. And the reason I focused on the past year, it's been roughly 2 years since our last R&D Day. It's been an extraordinary 2 years, and I'd actually say that 2020 has been incredibly productive. We have had over 10 major events. I'm going to focus in on several related to today's topic. So first, at the end of 2019, we submitted our first IND for a biotherapeutic engineered across the blood-brain barrier. That IND was accepted earlier this year. And even in the context of the challenges of this year, we were able to kick-off -- successfully kick-off and initiate a clinical trial Phase I/II in Hunter syndrome with DNL 310. In addition to that, in the middle of the year, we published 2 papers on our transport vehicle platform. We'll highlight those papers, but also additional data we've generated since the publication of those papers. And then finally, and recently, we signed a definitive agreement with Biogen, a collaboration agreement on both our LRRK2 inhibitor, but also on TV enabled biotherapeutics. Now transitioning to our portfolio, I just want to highlight that we have a broad again portfolio across multiple indications, but also multiple modalities. Today, the focus will be on large molecules, but you'll see that we also have a number of small molecule programs. I'm going to take some time now to highlight the progress on our small molecule portfolio before we focus in on the large molecule portfolio. So let's start with LRRK2. As mentioned, we recently entered a strategic collaboration with Biogen and have been defining the clinical development plan with our new partner. It's been excellent to engage with Biogen and we're looking forward to a successful launch of late-stage clinical trials with DNL151 in 2021. Our ElF2B activator program, or DNL343 is -- continues to be dosed in a healthy volunteer study. And our goal here is to enable patient study by end of the year or early 2021. This program was initially COVID delayed, but is now on track in terms of dosing. In terms of our RIP kinase program, they are both a centrally acting RIP kinase inhibitor as well as a peripherally acting inhibitor. And I'll just give some important updates here, which are new and recent updates. The IND has been submitted for DNL788. It's the new lead for RIP kinase inhibition for central nervous system diseases, including ALS, Alzheimer's and MS. The first-in-human dosing is planned by the end of this year or again, early 2021. And then our peripherally restricted compound, which is being led by Sanofi, has completed enrollment in a COVID study, and there's plans to initiate a Phase II study in cutaneous lupus in early 2021. Now I'm going to focus in on our transport vehicle technology and our lead program, ETV:IDS. Here, our goal is to establish biomarker proof-of-concept by end of year. And we'll talk in detail in Carole's presentation around the clinical design and our expectations of what we should see with this program. This data will help establish the TV platform and basically validate transferrin receptor as a viable path to get across the blood-brain barrier. We've already expanded this portfolio in partnerships with both Takeda and then recently with Biogen and specifically have added ATV:Abeta program in addition to another option program. And then our goal is to initiate IND-enabling studies for additional programs, and Joe is going to highlight new data for some of these programs as they advance to the clinic. So now let's actually talk about the invention of the transport vehicle platform and the goal of targeting the central nervous system with biotherapeutics. So to set the context, just a reminder that there's a large opportunity for biotherapeutics in neurological disease. In fact, neurological disease is the #1 cause of disability and the #2 cause of debt. There are, in fact, 4 approved biotherapeutics, and we use this definition broadly. This includes peptides, ASOs and an enzyme. And interestingly, of the 4 approved biotherapeutics, all of them are delivered directly to the brain, either intrathecally or ICV, and they act locally. So for example, motor neuron disease, those therapeutics are delivered IT. As such, there is a significant need for blood-brain barrier crossing biotherapeutics to treat broader disease. And that will be the focus of today's presentation. So let me start with some very important points around the brain and its anatomy and its biology. And these are key considerations when thinking about delivering biotherapeutics across the blood-brain barrier. So the first is that the brain is surrounded in cerebral spinal fluid and interstitial fluid. This circulate -- this is produced and circulates roughly 6 to 7 times a day in humans, and the goal of ISF and CSF is essentially to clear debris from the brain. As a result, when you inject a therapeutic into CSF, it has a very short half-life and is rapidly cleared and can be actually measured in blood. For example, an antibody injected into the brain into CSF will clear with a half-life of about one hour. As a result, directly injecting biotherapeutics in the brain faces major limitations unless you're focused on a regional area. First, the CSF flow into the brain is incomplete. As I mentioned, its primary goal is to clear. Diffusion from a point of injection relies on essentially Brownian motion, meaning that there's a limited diffusion, roughly a several millimeters. And size matters. So human brains are about 3,300x larger than a mouse brain based on weight. And as a result, direct injections are not scalable across species. Fortunately, the human brain is highly innervated with blood vessels. Every neuron is near an associated capillary. And these blood vessels are unique in that they have formed tight junctions in what is called the blood-brain barrier. As a result, ideal delivery technologies will utilize this blood-brain barrier crossing approach because of the ubiquitous nature of these vessels. Furthermore, blood-brain barrier crossing is scalable across species because we're using conserved biology. Let me illustrate with a very specific point comparing intrathecal delivery versus an IV delivery of a blood-brain barrier crossing biotherapeutics. I want to draw your attention to the left-hand diagram in which you see a lumbar injection, a single injection of a biotherapeutic. And what we and others have observed is that when this is done, you have a very high local concentration, but you're acting against a gradient. As a result, the biotherapeutic does not distribute broadly throughout the brain. However, on the right-hand side, when a BBB crossing biotherapeutic is given in the bloodstream, you see broad biodistribution, in fact, crossing the capillary beds. So the question is, how do we then approach this? How can we get biotherapeutics into the brain, utilizing some of these natural transport mechanisms. So our focus has been on the transferrin receptor? A few notes about the transferrin receptor, as I'm showing you on the left-hand side is a cross-section of a blood vessel, likely a capillary. Again, 400 miles worth of blood vessels in the human brain. Transferrin receptor actively transports iron into the brain and plays a critical role in iron homeostasis. I think importantly, it's highly expressed on brain endothelial cells as a result, it has very high capacity. In terms of targeting transferrin receptor for brain delivery, this high vascularization enables ubiquitous delivery if you are to utilize effectively the transferrin receptor. Also important, the transferrin is constitutively trancytosed. It's constantly bringing iron into the brain. And then finally, and I think very important when we think about the architecture of an effective biotherapeutic efficient transcytosis relies on the binding properties of this biotherapeutic. And in fact, affinity and valency are critical factors in designing an effective blood-brain barrier transport technology. So now turning our attention to the history of transferrin receptor. It was first described in the 1980s as being highly expressed in blood vessels. And as you can see in the middle of the diagram here, the number of publications highlighting basically blood-brain barrier crossing and blood-brain barrier properties have gone up substantially since the early '80s. For 2 decades, between 1990 and 2010, particularly, most of the effort was focused on the traditional antibodies and target [indiscernible] with relatively high affinity. And then in 2010 through 2015, there were a series of publications across all organizations showing different architectures and improved brain update. And then finally, as highlighted, we published papers in 2020 in May of this year, which I'll go into some detail with a new technology that takes advantage of this long history of transferrin receptor biology and engineering biotherapeutics to cross the blood-brain barrier. So let's start with the first publication. This publication highlights the invention of a proprietary platform known as the Transport Vehicle. I'll use the acronym TV throughout the rest of the presentation. Highlighted on the right-hand side is the actual architecture of the transport vehicle. We integrated transferrin receptor binding into one of the 2 feet of the Fc. This was done through directed evolution and identifying using the natural number of amino acid binders to the transferrin receptor. So on the right-hand side is a solved crystal structure. This is the upper right-hand corner of this interaction with transferrin receptor and what you see is it binds to the apical domain of transferrin receptor. It does not block transferrin or iron binding to the transferrin receptor. A critical proof-of-concept was now making the TV into an antibody known as the antibody transport vehicle. And in the bottom is an experiment in which we injected nonhuman primates with ATV and then looked 2 days after injection for the brain uptake across various brain regions. As I highlighted before, the goal is to have ubiquitous crossing of the blood-brain barrier. And you can see across each of these brain regions, at least 30-fold increase uptake with the ATV, again after a single injection. The image on the bottom right is actually a ceding of where the antibody goes. And you can see compared to a control antibody, very broad distribution, and this is actually in the cortex of a nonhuman primate. Next, we made a mouse in which the transferrin receptor is actually human and specifically the apical domain of the transferrin receptor in human. This is an important reagent that we've now used in many of our programs, including in the Hunter syndrome program that Carole will outline soon. What I'm showing you here is basically the transfer of antibody from blood vessels more broadly into the brain. On the left-hand side is 4 hours after a single injection. You can see at the top left is basically a control antibody. So very little staining. The bottom left is showing vascular staining for the ATV. This is, again, 4 hours after a single injection. On the right-hand side, 24 hours, you see that this staining or distribution, which is in Magenta, is now broadly throughout the brain. And in fact, it's taken up into neurons and other cell types. So there's a time-dependent transfer of the transport vehicle antibodies from blood vessels throughout the brain, and this is an important principle. The second paper published back-to-back in -- [indiscernible] highlights the utility of the transport vehicle for delivering enzymes across the blood-brain barrier. We call this the enzyme transport vehicle and the ETV was initially linked to IDS or ID Uronic 2 sulfatase, which is being developed for Hunter syndrome. One of the key pieces of data in this manuscript was showing that we could rescue enzyme activity across the various cell types in these mice. As we treat them with ETV:IDS, we see basically crossing the blood-brain barrier and an uptake into neurons, astrocytes and microglia. And in this particular experiment, we actually had associated brains, and we asked, could you reduce substrate in each of these various cell types. And you can see here a very robust reduction of substrate across cell types. The image in the center bottom of the slide is co-localization of ETV:IDS with lysosomes in neurons. So basically, injection systemically molecule crosses the bluing barrier and has taken up and rescues enzyme activity across the various cell types. Next, and this work is unpublished, we compared head-to-head differences in architecture and biodistribution. And this is a critical experiment in which we're looking at ETV:IDS using the transport vehicle technology. That has a single binding site for transferrin receptor compared to what we call IgG IDS. So this is a standard antibody fused to an enzyme. This molecule as a JCR-pharma-like molecule, has very high-affinity to transferrin receptor and 2 enzymes on the C terminus of the antibody as shown in the bottom left-hand corner. Similar to the data I showed previously, we then looked at the time course, starting at 30 minutes up to 24 hours after a single injection. These are in equivalent doses. And what you can see is that similar to the results I've shared before with ETV:IDS, over time, ETV:IDS transfers from the vascular staining to a broader staining across the brain tissue. This is shown in the panel on the right-hand side, 24 hours in high Mag, we actually see again neuronal and other cell type distribution. However, a high affinity TFR targeting antibody, the JCR like molecule remains completely trapped in blood vessels, and you can see the staining after 24 hours on the bottom right image. Now quantifying this, you can see the parenchymal staining compared to a standard antibody approach versus a ETV:IDS approach, which is a monovalent binder, gives a big difference in terms of biodistribution throughout the brain. So again, to summarize, crossing capillaries is an ideal approach to getting broad biotherapeutic distribution. However, not all approaches engineered to target the transferrin receptor are created equal and it's very important to have the right architecture and affinity to see broad biodistribution. Another benefit of the transport vehicle technology is its modularity. The ability to add this to various types of therapeutics. So I'm focused primarily on antibodies in the left-hand corner, and the goal here is to deliver antibodies in either bivalent or a bispecific format to the brain. This allows us to do it with a natural IgG structure, in other words, the amino acids that are buying to transferrin receptor are integrated into the Fc portion. And some of the examples that we'll go into detail today are ATV:TREM 2 as well as ATV:Abeta. On the right-hand side, in the focus of this and close presentation will be the delivery of enzymes across the blood-brain barrier. Here, the goal is to treat the neuronopathic symptoms of lysosomal storage diseases. And again, our lead program is IDS, and Joe will also highlight data for SGSH and Sanfilippo syndrome. On the bottom left is an example of protein transport or protein transport vehicle. And here, we can fuse again to the Fc and deliver proteins such as progranulin for FTD across the blood-brain barrier. And finally, and I think very interestingly and exciting, we can deliver antisense oligos across the blood-brain barrier, and we're going to show data today for the first time, deliver of these oligonucleotides to the brain, and basically, with the ability to knock down gene expression. So let's take a bit of time here to describe some of the other approaches targeting transferrin receptor and compare them to the transport vehicle. So on the top right are a number of molecules with different architectures. The first, which is a traditional bispecific antibody was developed by Genentech. And here what you can see is that transferrin receptor, one arm of the antibody binds the transferrin receptor where the other would bind to a target. The second molecule, which is a single-chain antibody fusion, fuses basically transferrin receptor binding domain to a standard antibody. This is known as the brain shuttle and is being developed by Roche as currently in clinical testing with Abeta antibody. The third example is a traditional antibody fused to enzymes. We call this a standard protein antibody fusion. This is the approach that JCR Pharma is taking, and the idea here is to deliver enzymes using standard antibodies. When we set out to invent a proprietary platform, the transport vehicle technology, we identified key biotherapeutic properties that were necessary in this platform. The first being architecture, the second biodistribution, and then the third is modularity. I want to highlight here some of the key characteristics around each of these 3 properties. So first is that the transport vehicle has [indiscernible] with no linkers or appendages. The second; is that it's monovalent, and it avoids transferrin receptor degradation. One of the key highlights is molecules that are high affinity or bivalent actually drive degradation of the -- of transferrin receptor, diminishing its ability to take up molecules into the brain over time. The third point around the architecture is we have the ability to silence or modify the effector function of the antibody, allowing us to dial in or out engagement with the immune system. In terms of biodistribution, which was a focus of several of the slides I presented before, it's important that we have integrated TFR binding as opposed to an appendage. Part of this is stability of the molecule, and we've seen that the transport vehicle technology is highly stable. We have optimized TFR affinity, and we've also shown cell type specific rescue, meaning that we can rescue in different cell types throughout the brain. And then finally, and I think very important is the modularity of the transport vehicle platform. I'm going to end with basically highlighting this modularity. And what our goal is in terms of going after various targets. So we see 2 large buckets of targets. The first are targets in which we can enhance therapeutic approach, and the second are targets that unlock the therapeutic approach. Let me focus in for a moment on some of the antibody approaches. So today, we will present data on ATV:Abeta, ATV:TREM2 and ATV:HER2. So one of the basic principles is that the transport vehicle can enhance biodistribution between 10- and 30-fold across all brain regions, but also very interestingly, often when we combine transferrin receptor with the target, we also have enhanced activity, and we'll show this example with HER2 and TREM2 in particular. On the right-hand side is our ability to unlock therapies. In this case, the example is that there are approved enzyme replacement therapies that do not effectively treat the neurological symptoms of disease, and our goal then is to take, for example, these enzyme replacement [indiscernible] put them on the transport vehicle and get them across the blood-brain barrier. The same is true with ASOs. If we were to inject ASO systemically, we would not see gene knockdown in brain. And basically, the goal of the TV is to unlock these targets. So with that in mind, I'm going to -- we're going to focus in now on ATV:IDS. So it's my pleasure to introduce Simon Jones. Simon Jones is a primary investigator in the U.K. at the Willink Unit Manchester Center for Genomic Medicine. Simon has worked on lysosomal storage diseases for a number of years. And as an investigator on a number of clinical trials. We'd like sign to set the context, understanding the disease that we're going after first, which is Hunter syndrome, but also more broadly, lysosomal storage diseases. And with that, I'm going to hand it over to Simon. Thank you.

Thank you, Ryan, and thank you all for listening. Good afternoon, good morning, whatever it happens to be. So as Ryan said, I'm a pediatrician from Manchester in the U.K., our large clinical center service for children with lysosomal storage diseases. And we've been involved in clinical trials and developing clinical therapies for the MPS II group for about 15 years. Starting off with the Phase III Elaprase trials and really my experience there as a younger man, was to really -- was really the frustration of whilst having a treatment for the somatic aspects of the disease, really not addressing the full unmet need in that patient population. And so we've been involved in quite a bit of work since then trying to address that. So I'll give some slides talking about Hunter syndrome in particular, but also particularly thinking about the CNS as a target. So my disclosures, I work with almost everybody involved or many of those involved in developing therapies for Hunter syndrome and across the LSD portfolio, our aim is to improve outcomes for patients. And no matter sort of what the modality, so we're involved in a number of different modalities of therapy. Okay. So the MPS diseases, I know from these sorts of meetings before that many of you are very -- fairly familiar with this group of disorders. But I hope that this initial slide, although quite to essentially illustrates the key pathway that we're talking about. We're talking about rare genetic disorders, in which the mutations in that gene mean there is a deficiency of an enzyme. All of these enzymes are in the catabolic pathway for glycosaminoglycans or MPS chemicals. These are ubiquitous, really important chemicals with many different biological activities. But we can be -- we can generalize slightly about these and the generalizations that we can make, I think, point towards the phenotype we get with these diseases, but these are all truly multi-systemic disorders with defects in every cell in the body in many cases. So if we think of the heparan sulfate, so HS is heparan sulfate, there's a storage substrate primarily in this group of diseases is caused by a series of enzymes, primarily only involved with the heparin sulfite degradation. And these disorders give rise to a single clinical phenotype with occasional variabilities. No -- and this clinical phenotype has noticed Sanfilippo, divided into the different genes or enzymes involved. And important, and I know Denali will mention this later on. But my main -- just my main topic here will be Hunter syndrome. So defects or diseases in the MPS group, where there is primarily dermatan sulfate storage give rise primarily to like an MPS VI, as we can see here, a somatic visceral skeletal cardiac disease. And so with very little CNS. So these young people do not have a progressive or significant CNS disease, although to be fair, very few of them have an entirely normal brain scan, but there's not a predominance of CNS disease. Whereas those MPS disorders, where we have both heparan and dermatan sulfate storage, for example, MPS1 , 2 and 7 are the classic MPS diseases, especially MPS1 and 2 and that because we have heparan and dermatan sulfate storage, we see both peripherals and sematic visceral skeletal disease as well as, in many cases, progressive central nervous system disease. And that's what makes these particular disorders so challenging to treat. So let's talk about MPS II now. These are all individuals I've looked after with MPS II. And what I'm trying to illustrate with this is the true multi-systemic nature of the disorder. So while some of these individuals will have more CNS disease and others more somatic disease, they all have an element of both. But also the real spectrum here. So the young man, who is in the top central photograph has a complete gene deletion so a very severe form of MPS II. And that's pretty severe somatic and CNS disease. However, the young man on the bottom right has much less in the way of somatic disease than the man on the bottom left, and the young man on the bottom left of this picture, I've looked after both of those young men. Whereas the CNS disease is really very different. The young man on the right, he -- Ryan has had really -- fairly rapidly progressive CNS disease, whereas the young man on left, the bottom left, was still walking and talking in his late teens. So a bit of a disparity in the CNS and the peripheral storage manifestations, which can lead to complications and difficulty with accurate phenotyping and prediction. And that, of course, makes clinical trials and interpretation of clinical trial results challenging. This is truly, however, multisystemic disease with cardiac involvement, significant respiratory involvement, and we know that the primary cause of death for individuals with MPS II is either airway and respiratory disease and/or cardiac valve abnormalities. What we see from the photograph of the hand here that actually everything is involved and everything can lead to disease and disability. So it's a challenging disorder to treat. And we are just in the last few years, really starting to identify and understand that there's also a tail. And it's not just a pediatric disease, and this tail is the very, very slowly progressive, very attenuated adult onset disorders. And the not so young man on the top right, diagnosed at the age of 54, only in the last couple of years, and he's been stopped and true Northern Englishman believed that he just had a hard life, and that's why things were hard and his hands didn't work very well and everything hurt, and he didn't hear very well. So a remarkable story actually. So this is a much more diverse disease than we thought, but a disease with significant ongoing unmet need. Let's try and focus on neuronopathic MPS II here in the next 3 slides. Because I think it's most relevant to the talks we'll hear later on from Denali. So the overall incidence of MPS II is between the literature between 1 in 100,000 and 1 in 150, 000, and that's relatively stable throughout countries and different ethnic backgrounds. And I think that's fairly real. It's unlike all of the other MPS disease as an X-linked disorder. And again, unlike other X-linked, lysosomal disorders like fiber disease, female involvement of MPS II in females is incredibly rare with only a handful of reported female cases published. Also unlike in February, there are very, very few females, and this is predominantly a disease of males. So if you read the literature in MPS II, they will describe a severe classical disorder, known as classical or severe neuronopathic MPS II, which affects about 2/3 of the boys in which there is a rapidly progressive CNS disease. The rest of the boys have no CNS disease. And in fact, have a purely sort of somatic visceral disorders. Now that's a growth over simplification, as you might expect, for agenetic disease. And of course, there is a true spectrum here. And I think that's important to realize and that there is a small group within that 2/3 of neuronopathic involvement, children who have indeed got some learning challenges, but they appear to be at least over 10 to 15 years nonprogressive. And so that also complicates recruitment of trials and complicates prediction of phenotype. As there's no current specific treatment as a standard of care for neuronopathic MPS II, occasionally, boys born and identified and diagnosed very, very early in life and they usually in the first year are sometimes offered hemopoetic stem cell transplantation or bone marrow transplantation, we've done that here. Usually, for the very rare cases, where, again, you have a boy diagnosed shortly after birth, usually because of a family history of CNS Hunters, neuronopathic Hunters. And we would, in that case, sometimes offer hematopoietic stem cell transplantation, where there's a suitable donor. The outcomes are a bit more uncertain than in MPS I, where early PMT is standard of care. It is not standard of care at this point in MPS II, and the outcomes are not as clear and not as good. So there remains a significant unmet need, as I've said. Intravenous Elaprase is used in the CNS affected group. They were excluded from all the clinical trials because they simply were not of an age and ability that they could be involved and performed the clinical endpoints adequately. We could talk about that later on. But IV Elaprase has been used in many centers in many countries in this neuronopathic group since the drug was licensed. And undoubtedly, we see a peripheral effect. Undoubtedly, there's an immediate or a temporary benefit in terms of quality of life. But the evidence base remains really very poor and practice does vary significantly from a reimbursement point of view from country to country. And again, as I've said, we are going to [indiscernible] the disease as the [indiscernible] unmet need still in MPS II, in my view, as a pediatrician. So as I said, I think around about 50% of cases of MPS II have this classical progressive CNS disease with initial developmental delay, more noticeable after the second year, a plateau of development, so not gaining new skills, but not losing skills and then a regression or loss of skills or, if you like, a true dementia. And long-term survival of boys with this form of the disease is often in the early teens because of the aggressiveness of the CNS disease. Somewhere between 20% to 30% of boys will have some neurological involvement. So a cognitive impairment and initial developmental delay, but that doesn't appear to be progressive. And so that's how we end up with the overall feeling that 2/3 of boys have some CNS involvement. The boys with classical and rapidly progressive disease, and we don't just have this dementia, but they often have a significant and really damaging behavioral phenotype with significant sleep disturbance and the impact on the overall family and family functioning is huge and cannot be underestimated. So this is not simply a loss of IQ points, which is often the way that we measure it. But that's not what the families will complain of and tell us about in clinic. It's highly nursed that we get through the next day, how do we go to the sick market, how do we survive as a family. And the support of measures that I have to offer these guys at the present time are all relatively per. So I will use IV Elaprase. I sometimes have to use psychotropic medications. We use medicines to help with sleep, medicines to help with pain, and we offer as good and as bespoke palliative care as we can, but ultimately, all of the children and young men that I look after with neuronopathic disease will die despite all of the current interventions we have to offer. And as I said before, and I think important and challenging once again in Hunter's and different to MPS I is that there's a less clear relationship between the severity of the Somatic disease and of the CNS disease. So lots of aspects of this would make it challenging, but a lot here to try and treat and really to illustrate how important it is to treat this. So I've mentioned that we do use IV Elaprase in neuronopathic MPS too. Here's a couple of siblings who I've treated. They've got a severe, rapidly progressive CNS phenotype with a number of other affected older individuals in their family, uncles, et cetera. We diagnosed the old and young man, both these boys have Hunters, the older man -- the older boy I diagnosed when he was a boy, 18 months old, and we started him with IV [indiscernible] at that time. And we can see the benefits of somatic treatment to him and that he initially grew really well. He had very little somatic features. But as time is going on, he's now about 12 years old, has developed really significant neuronopathic disease and is increasingly unable to walk, not because of terrible skeletal disease or bad heart and lungs, but because of progressive neurodegeneration. Interestingly, the younger sibling was treated shortly after he was born from a few weeks of aging, and in fact, is now significantly taller than his older brother, just showing the effect that actually these enzymes can have on the skeleton and often called a hard-to-treat part of the body in MPS diseases. But if we get in early enough. So again, he's now starting to deteriorate from a neurodevelopmental point of view. But the effect of very early treatment in this family has been dramatic in many ways. Yes, we can show benefit early on in neuronopathic MPS with just peripheral treatment. But ultimately, this is not enough. I think I've talked about this. The growth, yes, and we've proven that the growth improves in both neuronopathic and attenuated patients. And there's a lot of stuff in these patients that we really can't -- a lot of very hard to measure elements. There's a parental talk about general well-being, parental talk about the lack of ear, nose and throat infections and upper respiratory tract infections and overall less pain and stiffness. But they often will describe this primarily as my child is happier. And that's a pretty difficult thing to quantify as in most of these families will not identify their child as miserable before you start treatment, but recognize how much better they are. But we've yet been able to demonstrate, as we would expect any impact on overall cognition or prevention of the neurological decline with intravenous Elaprase. We have also seen a probable positive effect on survival with intravenous Elaprase, both for the attenuated boys and the neuronopathic boys. And this would be expected given an improvement in the somatic disease. However, the impact on survival is significantly less in the neuronopathic affected patients. So thinking about the unmet needs. We've gone on about it, and I've said it multiple times and I'll still say it again, the central nervous system remains the greatest unmet need for the majority of young men with MPS II. However, there are still evidence of significant residual disease in ART-treated patients currently with the -- in terms of the skeletal system, we don't have completely normal bones and joints. There's concern that the heart valves may still be affected, although very slowly progressive in most of the affected boys. And then there's potentially some airway to kill disease. And later on, that still needs addressing. I think it's important to say that we also see a significant heterogeneity and partial response to other prayers. So when we look at individual boys as well as a whole population, we're not completely addressing the storage, and there were differences in response to treatment, some of which are based on the age at onset of treatment other than the 2 boys that I have shown, but also there's differences between individuals that we don't fully understand, and we currently use 0.5 milligrams per kilogram weekly of Elaprase. And I think the current sense is that like in many of these enzyme replacement therapies, we probably are under dosing these patients. And I think if we knew then 15 years ago, what we knew -- know about these disorders and about the biomarkers, I think we may approach it differently. So perhaps because of the dose, perhaps because we're not individualized dosing -- individualized the dosing to these patients, we have this real heterogeneity, which is a huge problem when you think about clinical trial development. If we've got small numbers of patients who are themselves heterogeneous, we need to make every patient count in a clinical trial to get the most thorough and complete response we can in each patient in order to show benefit. But that's not just important for clinical trials. My job is a pediatrician, my responsibility is primarily to my patients, and I want to be able to treat individuals and know that I'm getting as good a response in each individual as I possibly can. So we've got some way to go on that. There are some problems still with prediction free to tap, and we may discuss that later on. It's going to be a lot of discussion, I think, in Carole's presentation about the role of biomarkers. And I just want to sort of talk briefly about how biomarkers have been used so far. So the obvious biomarker in MPS diseases, especially when it comes to somatic disease, where the majority of studies have been focused so far is in the store glycosaminoglycans themselves. And you'll see multiple publications looking at Elaprase from development -- early development through to licensing and beyond, looking at reduction in urine total glycosaminoglycans traditionally done through a relatively nonspecific method, which just measured all of the glycosaminoglycans, normalized to creatinine in urine. And whilst nonspecific, no very sensitive alters with time in any individual as well as disease individuals. Whilst the really crude biomarker and probably reflecting renal tubular storage rather than anything else in other tissues. Still, actually, this was a biomarker that was useful. That was accepted by the regulators in terms of dose-finding and for helping the registration of the various enzyme therapies we have for MPS diseases so far. So yes, a really crude biomarker in many ways and nonspecific biomarker, but it has served a degree of its purpose. And so you see here pre and post Elaprase urinary total GAGs in patients under the age of 6 and over the age of 6 with MPS II in one of the clinical trials. And so you see a dramatic reduction, and that's wonderful. And I think in the early days of enzyme replacement therapy development in the MPSs, we were just glad to see a reduction, and that's wonderful. And we focused not so much on this. I hope you can see the significant variability in response. So yes, there's a dramatic reduction in the mean. But actually, are we getting all of the individuals to have a really good reduction? No, we're not. And if we look at the red line is essentially the normal range of total urinary Glycosaminoglycans group and these individuals, you'll see that actually, we're not getting anywhere near normal in any of these patients, which may be fine for some organs and may not -- may not be in others. And so the residual disease that we see in many of our patients may be due to an incomplete reduction. And I think the individual patient variability that we see here and we see in our patients to this day is something that bothers me significantly. So total urine -- total GAGs in urine lack sensitivity, but it also critically lacks an ability to reflect the storage burden within CNS. So of course, we can measure and know there are a number of different measures of direct GAG, so not total GAG but Dermatan sulfate or Heparan sulfate in blood. We can measure those in urine as well. But that's still not the brain. And so Denali, amongst a number of other groups, have worked on direct measurement of Dermatan sulfate and Heparan sulfate in brain and, importantly, in CSF. Of course, it's important to measure this in the brain because that's what we want to treat. But we need to know the relationship between the brain measurement and the CSF measurement as in terms of my patients, I'm only going to get brain. So I'm only going to get cerebrospinal fluid, if I'm lucky. I'm not going to get brain. And so understanding the relationship between these is important. And actually, one thing I'll say here without trying to get into a whole biomarker discussion, which could take me all night, is different assays are measuring different things. There are multiple different sulfation states and ways and approaches and methods of measuring Dermatan and Heparan sulfate. And so the more validation work you can do on this, the better in terms of knowing what is actually useful and important to measure. Interpretation of the levels of these GAGs in CSF is also really dependent on -- in terms of when we're thinking about and intervention totally dependent on the nature of that intervention. So -- and I'm an investigator in the intrathecal iduronate 2-sulfatase studies. I've also performed bone marrow transplants on patients with MPS II. And we've looked at gene therapies for MPS II and a number of other approaches. And so we will be interpreting the CSF guides and change in CSF guides differently, depending on how we have delivered our therapy. If I've delivered the enzyme directly into the cerebral spinal fluid space itself through an intrathecal or [indiscernible] injection, then the CSF that I remove pre injection a few weeks or a month later on will reflect the CSF reduction in GAGs in the -- perhaps the tissue or the fluid in which we've directly dosed. So it's likely to in terms of what we're trying to dose, i.e., the brain, is likely to overestimate clearance from the brain. Whereas in a disease like -- sorry, with an enzyme delivered like a blood-brain barrier crossing enzyme that's delivered to the bloodstream. And we would -- possibly, it would get into the brain and into neurons itself. One would only imagine that the cerebral spinal fluid level of guides would underestimate the brain clearance. So I think it's important when we think about the relevance of CSF GAGs to think about what we're measuring, where and how we're measuring it, and what's the nature of the therapy. I think that's all very important. Lastly, I'll just think a little bit about clinical endpoints. So just when we thought this was a hard disease and have to develop trials for and to treat, I think it's worth thinking about the challenges of the clinical endpoints here. So if we are primarily aiming at the CNS disease, then it would seem that some of the most critical clinical endpoints will be neurocognitive endpoints. And so thinking about that developmental trajectory that I've talked about, initial global development and delay in development, the plateauing and the subsequent deterioration loss of skills, et cetera. And this, we can do through cognitive testing. There's a wide range of competitive tests. And so we've had a number of international consensus confronted and papers written to try and improve the approach to this in all of the different trials and therapeutic approaches. So that there's some uniformity, and we can compare and all being using best practice. And we've been, as a center, here are involved in those because we are involved with so many different MPS trials. I think this is really important. But moving beyond cognition, there's all of the really important family stuff that we talked about. The really significant behavioral disease. The patient-reported outcome measures and quality of life and what is quality of life in a family who have one or more boys with neuronopathic MPS II. And this, I think, we're still really scratching the surface of the attempt to try and develop behavioral tools but certainly, my experience from a number of trials in this area is that simply using generic tools off the shelf really struggle to get at what is important for these families. And so that's really challenging. And so what have been used beyond the urine biomarkers before. Well, the Elaprase trials use combinations of pulmonary function tests, they're much beloved in MPS disease, 6-minute walk test, essentially as composite endpoints to look at multiple disease systems and what the effect of peripheral enzyme has on cardiorespiratory function as well as well as musculoskeletal disease. And yes, they're absolutely useful. But you need a cooperative child who's at least 5 or 6 years old to do these tests. And so that's why they traditionally are never done or have never been done in young man with neuronopathic MPS II. Okay. I've talked enough, I think, so far. And I'll hand back to my colleagues. Thank you.

Terrific. Well, thank you, Simon, for that very insightful overview of the Hunter's syndrome MPS II. I want to segue now to our history of involving ourselves in the development of Hunter's syndrome and what brought us to this disease. 3.5 years ago, when we recognized the potential of our transport vehicle to deliver enzymes to the brain, we identified Hunter's syndrome is our flagship therapeutic area to validate this platform. There were 3 considerations that we had in selecting the disease. The first and most important was the high unmet medical need and the ability to potentially deliver a transformative therapy. The second was having disease biology that was tractable and understandable and also have an availability of an animal model, where we could pursue biomarker-driven drug development. And the third, an area that we are passionate about is the ability to conduct a biomarker-driven clinical development plan. We have been grateful and humbled by the engagement of physicians like Dr. Jones, clinicians as well as scientists, passionate advocacy groups such as the MPS society and patients and families. We have launched an engaged Hunter website, which is directed towards the patient community for us to share our approach to addressing this disorder and also provide updates on our progress. And this is a website view of our Engage Hunter website. The therapy that we are developing is called ETV:IDS or enzyme transport vehicle, IDS, which stands for iduronate 2-sulfatase, the enzyme that is missing or deficient in Hunter's syndrome. This is also called DNL 310. For the purposes of this presentation, I will refer to this as ETV:IDS. Brain delivery, as noted, remains a critical unmet medical need for Hunter's syndrome. DNL 310 ETV:IDS aims to treat the cognitive aspects of Hunter's syndrome while maintaining physical benefit in the entire body with IV administration. The goal of this therapy is to treat Hunter's syndrome with a single weekly administered IV therapeutic. Understanding the biology of the disease is critical for us to apply a biomarker-driven drug development strategy. The biology of the Hunter's syndrome is actually quite simple, given that it is due to the loss of activity of IDS, which is a single gene mutation. This is the approach that we are taking to look at target engagement of looking at the substrate that accumulates due to this deficiency in this enzyme, which are GAGs or as Dr. Jones noted, Heparan sulfate and Dermatan sulfate. These accumulate in cells, including neuronal cells in the brain. However, in addition to this accumulation, secondary lysosomal dysfunction ensues, where additional lipids accumulate in the lysosome and cause dysfunction. These things include gangliosides and BMP, substrates that accumulate not because of deficiency in enzyme that cleaves them, but because of the overall lysosomal dysfunction that occurs due to the accumulation of these GAGs that are metabolized by IDS. This is a critical area we want to understand in our animal models and then also in our patient studies, the effect that we are having not only on GAG reduction, but also these pathway biomarkers that represent lipid accumulation in the lysosome. Our understanding then is the accumulation of these lipid proteins in the lysosome leads to lysosomal dysfunction and cellular dysfunction and this leads to the cognitive impairment that results in the pathology of neurodegeneration, which can be measured by biomarkers such as neurofilament. Current enzyme replacement therapy does not effectively cross the blood-brain barrier, and therefore, is unable to address these pathway and pathology manifestations of the accumulated GAGs in the brain. So now how do we translate this biology into a biomarker-driven development strategy. This is a paradigm that we have used across all of our programs at Denali, where we systematically and rigorously address target engagement, pathway engagement, and eventually, this is going to help us find the dose to run a clinical endpoint study to demonstrate patient impact. Target engagement in this context is looking at reduction of Glycosaminoglycans. And specifically, for us in this paradigm is to look at in the CSF. In patients, we cannot access brain, so it is critically important in our preclinical development strategy to correlate these reductions of GAGs in the CSF with brain. This helps us define our early dose finding. The next step is to increase our confidence in the dose selection by monitoring biomarkers of pathway engagement. These biomarkers of pathway engagement include biomarkers of lysosomal function. And specifically, we will look at several lipids that accumulate in the disease that we can reduce in animal models with treatment of ETV:IDS. Finally, we also want to look at neuronal axonal injury with a biomarker neurofilament light. In our biomarker-driven development strategy, we accomplished these first 2 steps in the Phase I/II study that is currently enrolling. These steps will enable us to select the optimal dose to take forward into a patient endpoint study. So how does the strategy convert to our specific studies that we're looking at. So this is a road map of the studies that we'll review today. So to start in our preclinical strategy, we will present published as well as new data on preclinical efficacy studies in a Hunter mouse model. And juxtapose that with fluid biomarker analysis in Hunter patients that will demonstrate the relevance of these clinical biomarkers that we will look at. We'll share that data today. In addition, we'll provide an overview of a clinical observational biomarker study that's ongoing that will extend our understanding of these biomarkers in Hunter's patients and correlate these with clinical endpoints, which will be important to support future registration. I'll also provide an update of our Phase I/II clinical studies in Hunter's patients, where we have significant data readouts in the next year, including at the end of this year, we are on track to provide data on target engagement, GAG reduction, which will ungate our transport vehicle platform and provide proof of concept. The second data readout will be in mid-2021, which includes our pathway engagement data and neurofilament data, which will ungate Phase II/III startup for our registrational study with clinical cognitive endpoints. We are using this approach to inform a robust and systematic strategy to identify the bright dose to take into the Phase II/III study to maximize our probability of success and patient impact. This slide provides a detailed overview of the data that we'll review today. So starting with our preclinical efficacy studies in the Hunter mouse model, which are combined with understanding of patient biomarkers that are relevant, we will look at target engagement, GAG reduction, improvement in lysosomal function as evidenced by lipid biomarkers that accumulate in the disease and neurofilament light reduction. In addition, in a chronic dosing paradigm, we'll show you data using ETV:IDS in this mouse model that demonstrates correction of skeletal abnormalities, improvement in motor skills and rescue of cognition in a mouse model of this disease. The totality of this data provides translatable fluid biomarkers of lysosomal function and neuronal axonal injury that we will then apply to our Phase I/II study, which is currently enrolling. As noted, at the end of 2020, we expect to have data on target engagement and GAG reduction as well as safety that will ungate this TV platform and the potential of this platform across many modalities in our pipeline that Ryan discussed earlier. By mid-2021, we expect to have the data on pathway engagement that will ungate the start of activities of the Phase II/III study. And finally, the Phase II/III study will look at cognitive and functional endpoints in this disease. The stepwise derisking of the development program we feel is very important. And as noted by Dr. Jones, understanding the dose really matters and adequate dosing to address the wide range of clinical endpoints is going to be critical to successfully transform the treatment for this disease. So now let's jump into the data. I'll start with this paper that was published earlier this year, and we presented some of this data in the past. This paper demonstrated our ability to get ETV:IDS across the blood-brain barrier distribute widely and localized to multiple different cell types in the brain and reduce both Glycosaminoglycan accumulation as well as secondary biomarkers of lipid accumulation in the lysosome. This study and all of the studies that I'll present today using this preclinical mouse model utilizes this IDS knockout model, which is a cross between an IDS knockout animal, so this animal does not have the enzyme that is deficient or absent in the disease. And this is cross to a chimeric mouse that has the human apical portion of the transparent receptor. This allows us to dose mouse with the clinical candidate that is the candidate used in humans and allows that to bind to the blood-brain barrier in the mouse. As you can see in the figure here, the vehicle, which treated normal animals, so in most of the slides that I'll show, this will be in black text. This is a non-disease control. You don't see any accumulation of these gangliosides. However, with the vehicle control -- vehicle-treated disease animal, which is in gray, you see the accumulation of gangliosides and the deep structures of the brain. With IDS current standard-of-care therapy, there is minimal reduction in this accumulation of gangliosides in these deep structures of the brain. However, with ETV:IDS, you can see a near-complete resolution of these accumulated gangliosides in the brain. This data and as well as data that I'll present today demonstrates that we can achieve high concentrations and broad distribution of enzyme in the brain. And while IDS does not reduce significantly this lysosomal lipid accumulation in the brain, ETV:IDS is able to do this. So I'm now going to present additional data on our preclinical mouse model. But in order to put this into context, I wanted to share the paradigm for the patient samples that we looked at to demonstrate the relevance of the biomarkers that we'll look at in the animal model of the disease. This is a paper that was published in July of 2020, which is a paper that characterizes fluid biomarkers in neuronopathic MPS II patients. It is notable that we have non-MPS controls in this study as well as MPS II patients that are on enzyme replacement therapy as well as 2 individuals that received hematopoietic stem cell transplant. This is the first paper to our knowledge in Hunter patients that correlates abnormalities and GAGs with biomarkers of secondary lysosomal dysfunction, which includes gangliosides and BMP as well as looking at neurofilament and inflammation. So now we'll look at the data in the mouse model of the disease juxtaposed with the clinical significance based on the data that we've collected in the patients in the paper that was just referenced. So I will use this format for several of the slides presenting data, where on the left side, we have patient data from these patient samples from the published study. And on the right side, we have the mouse model data. So what you can see on the left side of this panel is in Hunter's patients, the primary storage substrate GAGs is elevated in serum as well as CSF. However, in serum, this difference is not statistically significant, and there's only about a twofold increase in MPS patients compared to non-MPS controls. This reflects the fact that standard-of-care therapy can reduce GAGs in the serum. However, you can see in the CSF, there's a different picture here where there is a statistically significant difference in the level of GAGs, and this is approximately elevenfold and is statistically significant. On the right, what you can see is in CSF in the normal control animal in black, there is no accumulation of GAGs. In the gray disease animal, there was accumulation of GAGs, but this can be reduced with clinically relevant doses of 1 mg per kg and 3 mg per kg after 13 weeks of treatment. This reduction, very importantly, is also correlated with reduction in GAGs in the brain that can be seen on the right panel. This type of experiment allows us to measure a translatable fluid biomarker in CSF and predict levels of GAGs in brain. And what you can see here is at these clinically relevant doses, 50% reduction in GAGs is associated with downstream pathway effects, which I'll present next. So on this slide, we look at gangliosides. Gangliosides accumulate in the lysosome is a result of secondary defects to the lysosome due to the accumulation of GAGs. This is clinically relevant in that ganglioside accumulation in other rare diseases on its own can cause neurocognitive deficits. And in Hunter syndrome, it has been shown in pathologic samples that gangliosides accumulate in the brain. Similar to the last slide that I showed you, as you can see in serum, there is not much of a difference between MPS II and non-MPS controls because standard-of-care therapy can address the accumulation of gangliosides in the serum. However, in the CSF, there is a statistically significant difference that's fourfold. And on the right, you can see that at 1 mg per kg and 3 mg per kg, both clinically relevant doses, there is a more than 50% reduction of gangliosides in the CSF that is correlated with even greater reduction of gangliosides in the brain. This data suggests that we can demonstrate modulation of pathway activity and evidence of restored lysosomal function with treatment of ETV:IDS in this mouse model of disease. On the next slide, we have additional lysosomal lipids that also accumulate in the disease in the CSF, which is glucocerebrosidase as well as BMP. In the left side, in the patient samples, you can see that both of these increases are statistically significant in patients with MPS II that are either treated with enzyme peripheral enzyme replacement therapy or have had hematopoietic stem cell transplant. On the right, you can see in the same paradigm in the normal control animals in black, there is not an accumulation. In the gray dots, you can see the accumulation of these substrates in CSF and brain. And with treatment with clinically relevant doses of ETV:IDS of 1 mg and 3 mg per kg, there is a reduction in CSF, which is mirrored by the reduction in brain. This data, like the data on gangliosides demonstrates an effect on lysosomal function and improvement of this lysosomal function and processing of these lipid substrates in the brain of IDS knockout-treated animals. So now I'm going to segue from looking at lysosomal function to neuroaxonal injury. And to provide an overview of this, we'll first look at the natural history of neurofilament light both in patients as well as in the animal model to provide context to the data they'll present for the animal model. So as you can see here in patients, there is an elevation of neurofilament light, which reflects neuroaxonal injury in the brain in both serum and CSF. Standard-of-care Elaprase therapy is not able to reduce these neurofilament level elevations in either the serum or the CSF and the increase in patients of this age is about 5 to sixfold. When we characterize neurofilament increase in the IDS knockout model, you can see that by 2 months of age, neurofilament and therefore, neuroaxonal injury is already beginning. By 5 months of age, this is starting to increase. And by 9 months of age, the increase in neurofilament light is in the order of magnitude similar to what we see in the patient samples. We decided to look at the effect of ETV:IDS on both of these paradigms of time course in the knockout animal paradigm 1 and paradigm 2. Paradigm 1 is earlier treatment and paradigm 2 is later treatment that may be more typical of what we will see at the age that we treat patients in the clinic. So here is the data treating IDS knockout animals with ETV:IDS. In the middle panel, you can see paradigm 1, which is the earlier treatment. And here, you can see in the disease animals, there is an increase in neurofilament, which is reduced by almost 50% in the 1 mg per kg and the 3 mg per kg dose that is administered IV. Similarly, in the second paradigm where we treat later, there is also a reduction that is statistically significant in neurofilament light. This difference is less pronounced than you can see in the earlier paradigm. But what I will show you is that with chronic dosing, this is associated with downstream benefits on -- in the mouse model on motor as well as cognitive endpoints. So based on this data, at clinically relevant doses of 3 mg per kg at earlier and later stages of disease, ETV:IDS completely in the earlier paradigm and substantially in the later paradigm reduces neurofilament light levels, which is reflection of improved neuroaxonal function. So now I'm going to share some of the new behavioral data that we have generated using this Hunter IDS knockout model. On the left, a reminder of the gross motor impairment that occurs in Hunter patients, where you can see that patients after the age of 3 or 4 that have the neurocognitive manifestations of this disease fall off the normal development age curve. In the ETV -- in the IDS knockout animal treated with ETV:IDS in orange, you can see that in a treadmill and pole test, they recover function and look very similar to the non-disease controls in black. On the left is a treadmill test and with increased treadmill speed, the animals in orange that are treated with ETV:IDS maintain their ability to complete the task. Similarly in the pole test, which is a measure of the animal's ability to descend a pole with shorter times being better than longer times, you can see in orange a restoration of the ability to descend the pole as compared to the gray bar -- the gray dots, which are demonstrating the animals that are disease that are treated with vehicle. So this data shows that at 3 mg per kg, which is, again, a clinically relevant dose, ETV:IDS improves locomotive performance and agility in the Hunter IDS knockout mouse. So moving to the next set of experiments. This looked at a cognitive test in a Hunter knockout model, which is both a learning and memory test. On the left, again, we see the disease in Hunter's patients where the -- for patients that have the neurocognitive manifestations of disease. At ages greater than 3, patients start to fall off the normal development curve. In the IDS knockout model, we looked at an active place avoidance test. This is a learning and memory test that requires that the animal first learn to avoid the orange area of a rotating platform, where noxious stimuli, a shock is provided. They are provided with multiple training tests on day 2 and 4 where they avoid this noxious stimuli. And then subsequently, on day 8, there is a memory learning task where they have to remember now what they learned from the previous training tests and avoid the area where the noxious stimuli is applied. On the left, we can look at the training trials, where you can see the gray disease animal that's treated with vehicle requires many more entrants to learn this cognitive task. However, in the orange dots, the IDS-treated animal -- the ETV:IDS-treated animal has a similar profile to the non-disease animal in black. In the reinstatement trial, which is a trial where they have to remember what they've learned in the previous study and avoid that noxious stimuli. You can see that in the ETV:IDS-treated animals on the far right bar, they have a normal distribution of good and top performers, unlike the middle bar where animals that have the disease and were treated with vehicle have very few good and no top performers. This data tells us that at 3 mg per kg in this animal model, ETV:IDS can normalize the spatial learning and memory deficits that are seen in the Hunter IDS knockout model. In addition to these behavior and cognitive tests, we also looked at other areas where peripheral enzyme therapy does not fully address the manifestations of the disease. And as Dr. Jones noted, skeletal disease is a continued area where current standard of care does not fully stress the skeletal manifestations of the disease, which include, as you can see on the left, changes in posture, scoliosis, claw hand and also brittle bones that can result in fracture and in the x-ray that you can see here, osteonecrosis of the hip. We looked at the skeletal manifestations of the disease in the knockout animal. And what you can see is that in the IDS knockout animal in the middle, there is increased trabecular bone density as well as thickness of cortical bone density in the femur. With ETV:IDS, this normalizes these findings and looks very similar to the black picture, which is the animals that do not have disease. As you can see, this is quantified on the right in the bar graph. And this data in totality suggests to us that, that at 3 mg per kg ETV:IDS can correct abnormal increase trabecular and cortical bone mass in the femur of Hunter IDS knockout animals. Now to summarize this data and put it into context of the competitive landscape. On the top bar of this graph, I've summarized the data we've seen here. Based on the totality of the mouse data that we've shared with you today and that's been published, where we have dosed animals with short-term dosing as short as 4 weeks to as long as about 13 to 17 weeks, we can demonstrate a consistent at least 50% reduction in GAG with clinically relevant doses. This is now being tested in the clinic in our Phase I/II study, and we expect to have this target engagement data at the end of this year. We have also demonstrated an animal model that we can demonstrate lysosomal function improvement by multiple lipid biomarkers that accumulate in the lysosome, including gangliosides, glucocerebrosidase and BMP. We've also demonstrated with 2 paradigms of treating both at an earlier stage and a later stage of disease in the IDS knockout model that we can preserve neuroaxonal degeneration as measured by neurofilament. In addition, these studies were all done with IV-administered therapy, which would enable us in these animal models to dose with only 1 therapy, and this we expect to translate into humans where this IV therapy would replace both standard of care as well as address the CNS and manifestations of the disease. We've also shown in these animal models very clear correlation between CSF effects as well as brain effects. Other therapies that are currently in development have not shown published data demonstrating the pathway effects on lysosomal dysfunction and neurodegeneration that has been published. In addition, some of these therapeutics require intrathecal or intra-cisterna magna administration, which would require that these therapies also be delivered on top of standard of care, which would begin to give an IV. Based on the totality of our data, we are looking at a proof of concept about 50% GAG reduction after 8 weeks of dosing in patients, which we anticipate will be associated with subsequent lysosomal function improvement and neurodegeneration biomarker improvement. And this reduction in GAGs is competitive with other programs that have dosed in patients and looked at shorter-term dosing. So now let's segue to the clinical trials that we are currently enrolling for Hunter syndrome. In October of last year, we initiated our observational treatment study -- our observational study of treatment responsive biomarkers in Hunter syndrome. This is an observational study that will continue to build on the data that I presented today for us to understand fluid biomarkers in patients and now correlate these to clinical outcomes. Given the challenges that Dr. Jones mentioned previously in clinical studies and clinical endpoints, we feel that this data is very critical to put context into our effects on biomarkers and how those correlate with cognitive and clinical endpoints in patients. In addition, in August of this year, we initiated our Phase I/II interventional study of DNL310 in patients with Hunter syndrome. This study, as noted previously, has 2 data readouts in the next year, which include at the end of the year, proof-of-concept readout based on target engagement and then in mid-2021 data on pathway engagement and neurofilament. I'll now spend a few minutes discussing the detailed design of the Phase I/II study. So this study is in neuronopathic as well as non-neuronopathic Hunter's patients. Patients can be either treatment naïve or on Elaprase for greater than 4 months. For those patients that are on Elaprase, they switch on day 1 of the study to DNL310 and are no longer on Elaprase during the study. Part 1 of the study is a dose escalation for 24 weeks, and part 2 will include a safety extension of 18 months, which will also be used to support the safety database at filing. Cohort A, which is a cohort that has been enrolling is in neuronopathic patients aged 5 to 10. Given the importance of treating earlier in disease to have the maximal impact on the progression of this disease, we are looking to also study this in earlier patients as young as 2 years old to enable us to enroll younger patients in our registrational study. So cohort B will enroll a broader range of patients from 2 to 18 years of age. The primary endpoint is safety in this study, looking at adverse events, infusion-related reactions, and very importantly, urine GAGs to demonstrate that the effects of ETV:IDS are similar to Elaprase in the periphery. In addition, our secondary endpoints will get CSF GAGs for target engagement in PK, and our key pharmacodynamic and exploratory endpoints are the endpoints that we looked at in the IDS knockout mouse of biomarkers of lysosomal dysfunction as well as neuroaxonal injury. On the right, we have a schema of our design of this Phase I/II study, which is notably an intrapatient dose escalation study. So patients come into the study, the starting dose is 3 mg per kg, which we've shown in the previous animal model data is a very relevant dose for seeing effects both in the CSF as well as in brain. Patients in the study will then dose escalate based on tolerability to -- up to a higher dose, dose D and will continue on that dose weekly. Cohort B in the younger patients will start at a dose that is to be determined based on the data generated from cohort A. As noted, the starting dose of 3 mg per kg is anticipated to reduce CSF GAGs and notably is 3x the enzyme activity compared to Elaprase. This is because the molecular weight is 2x higher with the ETV:IDS construct. And as you can see, the dose is about sixfold higher than what is currently administered with Elaprase. We believe that this dose will be effective both in the periphery as well as demonstrate evidence of target engagement and pathway engagement. Cohort A is on track for biomarker proof of concept at 8 weeks of dosing by the end of the year. In addition, as noted by mid-2021, we will have data demonstrating the effects on lysosomal endpoints as well as neuronal injury with neurofilament in both the CSF and the blood. So in summary, for the data that we've presented today, we have shared preclinical efficacy model studies in the Hunter knockout mouse along with fluid biomarker data in Hunter's patients that demonstrates the translatability of the biomarkers that we're looking at in the clinic to understanding target and pathway engagement in the brain. We've also shared information on our clinical studies, an observational study that is enrolling to help us understand further the correlation between these biomarkers and clinical endpoints as well as an interventional study to demonstrate an effect on the biomarkers that we shared in the IDS knockout mouse and demonstrate similar effects in humans. This data will both ungate the potential of this platform for a number of different modalities and programs in our pipeline at the end of 2020. And by mid-2021, the data from the Phase I/II study will ungate the start-up activities for the Phase II/III study. This very rigorous and systematic approach of understanding our dose before going into Phase II/III we believe is critical for success in this area and will maximize the probability of having an impact on patients in effect on the neurocognitive endpoints that currently remain a significant unmet medical need. This program and our proof of concept is successful, also ungates the platform that is enabled by the number of programs that are enabled by this transport vehicle technology. And this extends to multiple other platforms, including our antibody transport vehicle, our protein transport vehicle as well as oligonucleotide transport vehicle. And notably, as far as our enzyme transport vehicle platform, almost 2/3 of lysosomal storage diseases have a neuronopathic component, and this opens the doors to treatment of this high unmet medical need across a number of rare diseases. So in conclusion, DNL310 has the potential to treat the neurologic and physical manifestations of Hunter syndrome with a single IV therapeutic replacing standard of care. Novel biomarkers that we have demonstrated are correlated in patients as well as in the hunter knockout animal demonstrate that in the clinic, we can use these stock biomarkers to look at the GAG reduction as well as pathway engagement in terms of lipid reduction in lysosomal function rescue, reduce neurodegeneration and preservation of neuroaxonal injury as well as improvement in motor and cognitive function. We've shown that the brain and CSF GAGs are highly correlated, and these are also correlated with these downstream effects on lysosomal function and neuronal function. In terms of the data that will be coming out over the years on our development program, step 1, which will be available at the end of the year is demonstrating our effect on target engagement. This ungates the platform of our transport vehicle-enabled programs as well as triggers Phase II/III planning and the start of engagement of Phase III global regulatory meetings. Step 2 is looking at pathway engagement of lysosomal function and reduction in neurofilament, and this will enable specific dose selection and a go decision for Phase II/III in mid-2021. And the final step will be at the end of our Phase II/III study, which will look at neurocognitive endpoints, and this will enable registration and the ability to bring therapy to patients. Okay. With that, I will segue back to Ryan for the Q&A.

Great. Excellent, Carole, thank you. We've got Simon with us and Carole and actually a number of fantastic questions. We have a large number here. I'm going to focus on the question specifically for you, Simon, for Dr. Jones here, and we'll go through them as quickly as we can and cover as many as we can. We will take a short break and start back up at 2:50 Eastern Time. So I just want to make sure we get to that point. So at least a 5-minute break, but we've got some great questions here. So let's begin. So -- and I'll read them verbatim. So for Dr. Jones, how would he interpret different levels of change in CSF GAG for predicting clinical benefit? And actually, I'll add to that, says in terms of clinical endpoints, what are the most important domains of neurodevelopmental measures? And how do these differ across age cohorts? How much change? And in what time frame would he see as clinically meaningful? So basically, the question is around CSF GAGs and also neurodevelopmental measures or clinical endpoints. So Simon, I'm going to hand it to you.

Thanks, Ryan. An excellent question. And I'll discuss this over the next 2 hours. And I think it's difficult. So I think as I said earlier, I think it's important when we think about interpretation of CSF GAGs in terms of what that might mean, it's important to think about how your therapeutic agent was delivered. Do we have a very clearly proven effective treatment for training the brain disease in MPS II today that we can learn from gold standard treatment? No. And so it's very, therefore, challenging to know any human, what level of change in CSF GAG actually will predict clinical benefit. And I think the fact that we have here and what gives me encouragement in moving forward and doing a trial like this is that we are evident in the most model of CSF GAG and brain GAG and the correlation between the 2 of those in response to this therapeutic agent. And I think that's important. We've not had that before from small animal studies. And so we've had with previous therapeutics. In this area, we've had brain GAG data from mouse models, and we've had to then make a jump into humans already get to look at CSF, which, as we said, for most of the intrathecal administrations, you're measuring that exactly where you administer the therapeutic. And you've got to make that real leap of faith that this is meaningful. It implies something that's happening in the brain, but you don't know in the human. And the aspect around scaling from the animal model to the human experience here is really critical. And I think it's evident, not just from the presentation that we've heard already, but just from a general, it just makes sense and makes biological sense that scaling when we're thinking about direct to CSF injections is going to be really challenging when we think of the differences between the size of the brains between the small animal, the large animal models and humans compared with the therapy delivered through the blood-brain barrier, where the distribution and scaling as long as we can get that therapeutic across the blood-brain barrier is dealt with. And I think the reason that we failed before or struggled before to deliver things directly has been a, distribution; b, dose; c, patient selection. So a long-winded answer but that's my fish.

No problem, Simon. I'm just going to add a little bit to that, and Carole, you as well. And I think there were at least 4 questions around this topic. So I think it's worth it to spend a little bit of time. One of the questions, and those of you that can see the slides, I've moved back to Slide 15 was the differences between intrathecal GAG measurements in the intrathecal trial versus what you would expect from a BBB crossing biotherapeutic. And I just want to highlight here again in the bottom left-hand corner that when you deliver a therapy directly, you're relying on basically diffusion from that point. And the CSF measures will, as you -- I think I'm going to use verbatim, Simon, your words, can overestimate the amount of GAG reduction through the rest of the brain, whereas -- and I think you may have said this, I don't know if it's entirely true, but you may underestimate when you measure CSF, let's say, a brain distributed biotherapeutic. In our hands, we actually do see, on average, better brain reduction than CSF. And Carole presented many of those data and you can go back and look at them. So in general, it's almost a 1:1 correlation, but usually superior in brain. I think really important is we -- if we go to another slide here, I'm now going to go to Slide 21, which is comparing ETV:IDS versus IgG IDS. We actually -- and again, this is answering one of the questions that came up. We actually don't see as strong a correlation with GAG reduction. And so we've done a number of experiments here, and I just will say, stay tuned. We want to publish these head-to-head comparisons. But the challenge has been, when you have a vascular localized therapy versus a broad brain distribution, the CSF to brain relationship isn't as strong as what Carole has presented. So let's turn to another question here. And I think, Carole, I'm going to direct this towards you, and then, Simon, you can add to it. So on Hunter syndrome, the goal ultimately would be to show CNS benefit. What do you think you could see in a Phase I after 3 to 6 months? And what would you look at? And what would FDA be okay with in terms of CNS endpoint for a Phase III? And how do you power this? So Carole, I'm going to start with you.

Yes. So those are great questions. And maybe I'll start with what the FDA is expecting to see to our understanding and then address the questions about what we can see in the Phase I/II study. So in the end, for registration, we will need to demonstrate an effect on cognitive behavioral endpoints. And I think that's very important because we want to be able to demonstrate that what these biomarker changes show are correlated with clinical outcomes. At this time, looking at solely biomarker endpoints is not going to be sufficient for approval. And actually, in many ways, that makes sense, if you look at what we just discussed in terms of effects on biomarkers in the CSF. It really is context dependent in the sense that where you collect that CSF relative to where you've administered the therapy is very important interpretation of that. Because our therapy is delivered IV when we are measuring biomarkers in the CSF, as shown with the animal model studies, we are really reflecting what is being seen in the brain. And as noted, what we tend to see is actually a greater reduction in the brain than we're even seeing in the CSF because we're actually distant from the brain in terms of collecting those biomarkers. In terms of the Phase I/II study, our goals are to really see effects on these biomarkers that, again, we've correlated in the animal model, looking in the brain and also have demonstrated the relevance of these biomarkers as they are abnormal in patient samples. In terms of whether we will see clinical endpoints, I think the estimate is that for this size of study and particularly with the age range that we're studying in the first cohort, it would be unlikely to be able to see neurocognitive outcomes. But certainly, we will be monitoring this and enduring the safety extension phase with longer-term treatment, the benefit could be seen in that study.

Great. Thanks, Carole. So Simon, a question for you. What proportion of your patients would be candidates for a treatment such as DNL310?

Yes. It's a very good question. So we've thought long and hard about this in our own center with our own patient groups, thinking ahead about all of the different therapeutics that there could be or that are in development or in Phase II. And given that prediction of phenotype is not always clear at the age in which patients are clinically diagnosed. And that's between 2 and 4 years old in the main. And then at that age, it sometimes is clear that there is CNS disease or sometimes it's because of genotype, we can accurately predict the phenotype. But it's not always clear at that age. And so I think if I had a 2- to 4-year-old who had a little bit of developmental delay, I think I would be very keen to be starting that individual on a treatment that treated the potential CNS involvement as well as this semantic involvement rather than a therapy that treated just their somatic involvement. Partly on a percentage basis, there's a 2/3 chance that we can -- that patient will absolutely need a CNS treatment. But ultimately, we just can't predict that. And you can't go back in time. So I think if it's a therapy -- this is a therapy that has a safety tolerability profile very similar to Elaprase, for example, then you would use this treatment for all of those patients that you see in that 2 to 4 age group. The question gets way more complex when we think about newborn screening, but everything about newborn screening gets more complex. So I think that's -- but we'll deal with that when we have to.

Okay. Thanks, Simon. I mean I just want to comment. There are an enormous amount of awesome questions here, which is really great to see the participation. So I wish we could spend 3 hours answering these questions. I'll go to the next one. And I'll take the lead here, and then, Carole, I'll have you add to it. So what is the relative ratio of CNS to peripheral distribution of ETV:IDS? What does the overall biodistribution look like outside the brain and have you looked at Elaprase in the bone density experiment, what would it look like at a clinically relevant dose in that model? So let me start with the beginning. So -- and I think this is really important. Our goal is to replace Elaprase. And I think one of the important pieces of data that we shared today is that our starting dose is 3 mg per kg. So on an activity level or an enzyme level, our dose is actually 3x that of Elaprase. Elaprase is given at 0.5 mg per kg once a week. And we're giving 3 mg per kg once a week, but remember, the DNL310 is about 2x the molecular weight. In our hands, when given at the same enzyme level, so basically, the same amount of Elaprase and the same amount of DNL310, we have an equivalent and potentially superior peripheral knockdown of Glycosaminoglycans, so GAG. So it's actually -- the peripheral biodistribution is actually is as good or better than Elaprase. The question is, have you looked at Elaprase in the bone density experiment? We have not looked at that head-to-head. And getting back to what we think is a clinically relevant dose in our models, again, 3 mg per kg translating from a PK perspective and an activity perspective, it's likely that in larger animals, DNL310, given that 3 mg per kg is at least equal to the mouse 3 mg per kg, but probably superior again with better PK. Carole, would you like to add anything to that?

No. I think that was perfect, Ryan. And I think that this data is important because with the starting dose of our Phase I/II study, we are starting at an enzyme activity, as noted, that's higher than Elaprase activity. And so that enables us to potentially have greater effects on some of the peripheral endpoints of disease as well as addressing the central manifestations of disease.

I'm going to focus on one or 2 more questions for you, Simon, in part because we're going to have a larger Q&A, and we have many more questions that we can address as they're focused more on us. So for -- a question for Dr. Jones, how much follow-up do you think would be needed to establish a benefit on neurocognitive measures in a pivotal trial?

Yes. Again, an excellent question. I always get slightly stressed at the questions that these investor meetings are better than the questions [indiscernible] meetings, which makes me worried about where the smart people are. I mean this depends very much on what's your patient population. If you have a patient population who are clearly symptomatically neurologically involved and actively declining, then you are going to be able to establish new alteration to their neurocognitive trajectory, if you like, fairly rapidly within 6 to 12 months. And if you're taking a -- and that's if you have a fairly homogeneous population. And if you have a patient population that you're trying to select and detect and start treatment on before there's very clear neurological disease or neurocognitive disease, they are not more challenging. You may have a greater opportunity for benefit, but you clearly have to wait longer to see the separation between what you hope will be your effectively treated nondeclining patients and a control population or a historical control population, which will be those patients naturally declining. And that may take more time like 2, 3 years.

I think with that, we're going to take a break. We're now well into the break. We'll gather again at 2:50 Eastern Time, where we'll just -- we'll kick off the discovery portfolio in some of our new programs. And then we will focus in on broader sort of value realization for the TV, and then we'll get back to all of these excellent questions. So with that, Simon, thank you so much. Enjoy your evening. If you want to stay with us, you're welcome to. And with that, let's take a break. Take care.

Thanks. [Break]

All right. Welcome back, everyone. Thanks for taking a short break. We now look forward to diving into our discovery portfolio. I'll hand it over to Joe Lewcock, our Chief Scientific Officer, who will lead the way here. Joe, go forward.

Thanks, Ryan. Really excited to talk to everyone today about our TV discovery portfolio. And this is going to include data on a number of programs that we haven't previously disclosed before. So really excited to take everyone through some new data today. And I'm just going to start by giving a brief overview of where we stand with our TV portfolio overall. Our portfolio has actually grown to a total of 12 programs, all using our TV technology. This includes ATVs, ETVs, PTVs and OTVs across a range of indications. Of these 12 programs, 9 of these programs are in discovery, where we have 2 programs at the IND-enabling stage, namely our PTV:Progranulin program that's partnered with Takeda as well as our ATV:TREM2 program for the treatment of Alzheimer's disease. And then, of course, you heard in the last section about our lead program, our ETV:IDS program, where we hope that the clinical data that we're going to be receiving at the end of this year is going to have broad readthrough into our TV portfolio overall and really increase our potential for success with a range of programs. So you saw this slide before from both Ryan and Carole. And I just wanted to come back to it quickly to remind folks that the TV platform is highly modular. And because we're engineering the TV sequence into the Fc domain, what we can do then is couple it to a standard antibody, which is shown on the top left, whether it could be a bivalent or a bispecific antibody while retaining normal binding to its targets. But we can also then conjugate it to things such as enzymes, as you saw with our ETV:IDS molecule, or proteins, which I'll talk about a bit with our PTV molecule. And more recently, we've even extended on to conjugating these molecules to oligonucleotides to deliver to other types of therapeutics to the brain. And so when we think about how we use our TV platform, we've really broken up our TV targets into 2 categories. And the first is, and Ryan mentioned this a bit earlier around the ability of the TV platform to enhance targets. And this is really most applicable for our ATV platform, where we can take therapeutic antibodies and we can enhance their activity. And that comes in really 2 ways. The first way is through increasing the concentration of antibody present in brain, and we can increase that by between 10 and thirtyfold to really get much better target engagement with our ATV coupled molecules than a standard antibody. But the second, which is a bit more unexpected, and I'll show you a couple of examples of it today, is that TfR binding actually seems to enhance the activity of some of these antibodies. And that's through the antibodies binding to their desired target as well as TfR on the cell surface. And this seems to result in certain situations in greater potency, which is something we're really excited about. The second set of targets are ones where the TV really unlocks these targets. It wouldn't be possible to deliver these types of biotherapeutics to brain through systemic administration without a BBB crossing platform. And these are ones we're really excited about where -- because it turns a non-druggable target into a druggable target. And so during the course of my presentation today, I'm going to be talking about every one of the molecules on this slide, except for IDS. So I'll go over quickly our ATV:Abeta, ATV:TREM2 and at ATV:HER2 programs, and then I'll shift gears to talk a little bit about our ETV:SGSH program, our OTV program and our PTV Progranulin program. Now given that, that's going to be a lot to cover in 30 minutes, I'm not going to be able to dive too deeply into any of these programs, but rather I'm going to focus on how these programs are enabled by our TV technology. And I'll start by going over our ATV:Abeta program. So this is one of the programs that was part of our recently signed Biogen collaboration. And what I'm showing you here on the slide is on the left-hand side of the slide is really just a schematic of our ATV:Abeta molecule. So it's a standard antibody format. It retains bivalent binding to Abeta, but it has a TV sequence engineered in as well as what we call a cisLALA sequence that allows these antibodies to retain a factor function and do so safely without having any TfR-mediated [indiscernible] effects. And so the data I'm showing you here is from a relatively simple single-dose study, either in our TfR knock-in mice that expressed the apical domain of human transparent receptor, or those mice crossed to a 5XFAD, which is a mouse model of Alzheimer's disease that develops amyloid plaques. And as you could see, both in the gray shaded section on the right hand of this slide, which is just our TfR knock-in, as well as on the green shaded section on the right-hand side of the slide, across 3 different therapeutically relevant dose levels, you see a very, very nice increase in the amount of antibody getting into the brain of these animals 24 hours after a single dose with the ATV sequence compared to a standard antibody. So similar to what we've shown you before for ETV:IDS is clear in this case that the ATV sequence dramatically increases the amount of antibody in brain. But what's perhaps more impressive is now if we look at the engagement of these antibodies with their target as measured by the immunodecoration of amyloid plaques by antibody. And so that's what you're looking at in the images on the left-hand side of the slide, where amyloid plaques that are present in the brains of these mice are labeled in green, where human antibody, which is -- which detects our ATV:Abeta antibody or standard antibody is detected in purple. And what you could see is that, again, this is 24 hours after a single dose, that there's really not a lot of colocalization of the purple and the green, meaning that those antibodies are engaging with plaques with a standard antibody. However, that's really significantly increased with treatment of our ATV:Abeta molecule. And note that this is at a 10 mg per kg dose, which is highly clinically relevant dose. This data is quantified then on the right-hand side of the slide, where you can see, again, across dose levels at 1, 3 and 10 mg per kg that you get significantly increased colocalization of antibody with amyloid plaques with our ATV molecule compared to Abeta alone. In fact, our colocalization at 1 mg per kg with the ATV molecule is equivalent, if not higher, than 10 mg per kg with the standard antibody. So it does seem like that with this molecule, we'll be able to get significantly higher target engagement with the same level of dose as compared to a standard antibody. So now I'll shift gears to our ATV:TREM2 molecule. And just a reminder that the therapeutic hypothesis that underlies this program is that enhancing TREM2 function through an agonist antibody has the opportunity to improve microglia function. And so for a bit of background, I'll tell you that TREM2 is a cell surface receptor that is expressed within the brain pretty exclusively on microglia. And they gained a lot of interest in the Alzheimer's field when human genetics identified a number of different mutations, coding mutations in TREM2, that increase the risk of Alzheimer's disease. And a table summarizing those mutations is on the bottom left of the slide. And through the work of many labs, I think the consensus in the field is that these TREM2 mutations result in a loss of normal TREM2 function, which results in poorly functional microglia. So it brings forward the hypothesis that improving TREM2 function could increase microglial function in the brain and potentially have a benefit for Alzheimer's disease. So here at Denali, we've really spent a long time diving deeply into this field and trying to gain a better understanding of TREM2 function in order to enable our programs. And that's highlighted by the 3 papers shown on the right side of this slide, which have been published by Denali scientists. The first was around understanding TREM2 function within microglia, where we showed that TREM2 is not only important for recognizing and phagocyte hosting, lipids and other debris, but it was also important for the microglial response to be able to digest those lipids and cargo effectively. In the second paper, we focused on another genetic risk factor for AD, a gene called PLCG2. And that's also shown on the table on the bottom left of the slide, where we found that actually at mutations in PLCG2 rather than causing increased Alzheimer's risk were actually protective for AD. And we and other groups went on to show that these mutations in PLCG2 resulted in an increased function of PLCG2 and that this acted downstream of TREM2. So putting this together, it provides a direct genetic evidence that increasing the signaling through the TREM2 signaling pathway can be protective in Alzheimer's disease. And so consistent with that, we recently published a third paper in collaboration with Christian Haass in Munich, demonstrating that agonist antibodies for -- against targeting TREM2 were able to have beneficial effects in vivo in mouse models of Alzheimer's disease. But so rather than going through any of that data in detail, I'll shift to what happened when we coupled our ATV:TREM2 molecule to our TV platform. And that's what I'm showing here. And again, similar to the data I showed you for ATV:Abeta. Now we're looking at a single dose of ATV:TREM2 -- at ATV:TREM2 dosed into normal mice, not an Alzheimer's disease model. And what we saw here was first, consistent with what we saw before, we had increased levels of antibody get into the brain with our ATV:TREM2 molecule. But interestingly, when we then looked at the target of these antibodies, the microglia within the brain, and we did that by staining with the microglia marker IBA1, which is shown in the images on the right side of the slide. That compared to a control antibody, a normal TREM2 antibody did have some impact on these microglia and brain, which was encouraging. However, the ATV:TREM2 molecule had a far more dramatic impact on microglia biology, indicating that either through increased brain penetration or increased potency of the molecule or both, that we're having a much more pronounced effect on microglia with this molecule. So we then went on to looking to this in a bit more detail using a multi-dose study. Now in this study, we tried to dissect the brain uptake component from the efficacy on microglial component, and we did that through dose normalizing by dosing 10x more normal antibody -- normal TREM2 antibody, then we did ATV antibody. And what you're looking at on the graph in the bottom left of the slide is that when we took this 10x differential in our dosing strategy, the actual concentration of antibody in brain, which is shown on the right 2 bars, was roughly equivalent between our normal TREM2 antibody and ATV version of our TREM2 antibody. However, when we went in and looked at the effect on microglia in this dosing paradigm, what we found was that there was again a dramatically increased impact on microglia as measured by IDA1 with the ATV:TREM2 molecule rather than the TREM2 molecule alone. And so this was really encouraging to see this positive effect on potency. And although I don't have time to get into it today through this data as well as a number of cell-based experiments, we believe that you actually get an increase in potency of these molecules through coordinate binding of TREM2 2 as well as TFR on the surface of microglia. So this is something that we're really excited about and pursuing further in with in-house experiments. So then I'll shift gears to talking about a potential application for the ATV platform outside of neurodegeneration. Moving to neuro oncology with our ATV:HER2 molecule. So ATV:HER2 molecule combines a very well-defined strategy of targeting the growth factor HER2, which is over-expressed in many cancers, and conjugating it to our TFR platform. And so what we first wanted to do here was to take a look at how -- similar to what -- with our past experiments, whether the TV sequence was again able to increase brand concentrations and it was able to do that very effectively. And that's what's shown in the graph in the middle part of this slide, showing that whether we connect -- or when we include the ATV sequence on trastuzumab or pertuzumab, in both cases, we get quite a significant increase of antibody concentration in brain. But what was perhaps more interesting was, again, looking on the right side of the slide, now this is a cell-based experiment and again, very akin to what I told you about for TREM2. But what you're looking at here is comparing an ATV version of trastuzumab versus trastuzumab alone. And in this particular cell line, and we saw there's across a few different cell lines, is that there is dramatically increased activity from the ATV:HER2 molecule -- pardon me, as compared to HER2 alone. And again, through a series of experiments, I won't have time to tell you about, we were able to demonstrate that this was, again, because of coordinate binding between TREM2 -- bet TFR and HER2, leading to increased TFR and HER2 internalization in these cells. So I think this is really exciting because keep in mind, this is a cellular experiment, and so there is no blood-brain barrier uptake component to this. And even in the absence of that, we're seeing increased activity with this ATV:HER2 molecule. So this cell-based data, we then went on to confirm whether this might also be the case in vivo. And to do that, we used a standard xenograft model using HER2-positive cells. So again, remember that the blood-brain barrier, there is not a blood-brain barrier component to this model. It's really just looking at the anti-tumor activity of our ATV molecules as compared to the normal molecules. And what you could see here is across a range of dose levels, again very similar to what we saw with our in vivo studies that in these xenograft models that the ATV, the combination of ATV [indiscernible] and ATV: HER per was significantly better than those molecules alone or the standard antibodies in the absence of the ATV sequence. So very encouraging data from these programs, and we're excited to keep them going. But now I'm going to shift gears to talk a little bit more about our ETV portfolio. So Carole and Simon did a great job talking about ETV:IDS, and I wanted to talk to you a little bit about a second enzyme in our portfolio, our ETV:SGSH molecule. And a diagram of what that molecule looks like is shown on the left part of the slide. And you might note that ETV:SGSH is a -- it's a biozyme architecture rather than a monozyme as with IDS, and the reason for that is to really mimic normal physiology as SGSH normally functions as a dimer. This biozyme sequence we found to have much greater activity than comparable monozyme. So we were able to generate this biozyme molecule conjugated to our TFR platform, and what I want to show you is some of our initial in vivo data using these molecules, and why we're encouraged that the data that we saw with IDS will also hold true for other enzymes. And so what you're looking at here is, again, using our TFR knock in animal. And in this experiment, crossed to animals that have a mutation in the SGSH gene, leading it to be nonfunctional. And what you can see is in the brain of these animals, if you compare the first 2 bars on the left-hand plot, you see that I -- there is a dramatic accumulation of heparan sulfate in these animals as compared to their wild-type littermates. And what we were able to show is a single treatment with our ETV:SGSH molecule was able to result in a 50% reduction in the amount of -- and the extent of heparan sulfate accumulation in these knockout animals. And this was significantly improved compared to FC-SGSH control. So the exact same molecule absent our TV sequence. And so this was really exciting to us because first, we were able to show -- again, we were able to reduce the accumulation of these GAGs within brain. Again, much like Carole showed you with IDS, that reduction in brain heparan sulfate correlated very nicely to a reduction in CSF heparan sulfate, and that's shown on the graph on the right-hand side of the slide. And lastly, this data that I'm showing here is quite comparable to our initial single-dose data with our ETV:IDS enzyme, leading us to believe that we feel like we have a very good chance of read-through of what we've seen with our IDS molecule being able to translate to additional enzymes in the future. And based on that, we have started work on our SGSH molecule. We've also started work on an additional ETV molecule for the treatment of lysosomal storage disease with the neuronopathic component. And so we now have 3 programs ongoing within our portfolio, and we're really poised that if our ETV:IDS data reads out positively that we can push these programs forward using the know-how that we've developed very quickly into subsequent clinical studies. And also to note that, although each one of these diseases is relatively rare as a whole lysosomal storage diseases at a relatively large patient population, and so there is potential for further expansion in this area as well. Okay. Now I'm going to pivot a bit from our ETV program to our PTV program, our Protein Transport Vehicle program to talk about our Progranulin and PTV program. And what I'm going to talk about is how excited we are about this particular program, and what we've learned about how similar biology exists between progranulin and some of our enzyme-link programs. So in order to do that, I'm going to first take a step back and go through a little bit more about progranulin biology, and how it relates to neurodegenerative disease. So it starts with mutations in progranulin are one of the most common causes of frontal temporal dementia. And there is a number of different mutations that have been identified in progranulin and relatively uniformly, these mutations result in a decrease in progranulin levels. And this decrease in progranulin levels can be measured in both the CSF and plasma. And I'm showing you on the bottom left part of this slide, an example of some data we generated at Denali, which is very consistent with the literature, showing that patients with Progranulin-driven FTD, do, in fact, have a much lower level of progranulin present in their CSF as compared to healthy controls. And -- but what we started to learn when we looked into progranulin more ethanol was that even though progranulin is often measured in extracellular fluids, it seems that the real functional role for progranulin is within the lysosome. And that's what's shown in the 2 microscopy images in the middle of this slide, where you're looking at progranulin in red; co-localized with a marker of lysosomes, LAMP-1, shown in green. And these are in cultured human microglia IPS-derived microglia. And so what we've really found through a number of experiments is that progranulin within lysosomes is really the functional component of progranulin. And when it's absent, as you can see in the progranulin knockout cells on the right-hand side that you get lysosomal dysfunction, and this can be seen visually by microscopy through these in large lysosomes precedent these cells. So of course, the end result of progranulin haplo insufficiency in these patients is neurodegeneration and neuroinflammation. So progranulin FTD patients are known to have a very high level of microglial activation, and then they, of course, have a neurodegeneration that can be measured through increases in neurofilament light and CSF. And, that's what's shown on the bottom right of the slide. But in many -- I think our data suggests, in many ways, this is very similar to other lysosomal storage diseases that result in neurodegeneration for very similar mechanistic reasons. And so we believe that then the storing functional progranulin within the lysosome is going to be key to a therapeutic effect. I think that really distinguishes us from some of the competitors in this space that are really trying to increase the amount of progranulin in that can be measured in fluids. Sometimes that comes at the expense actually delivering progranulin effectively to the lysosome. So I'm going to go into this in a little bit more detail and talk to you about how we measure progranulin. How we measure lysosomal dysfunction due to progranulin deficiency. And so the way we started to do that here was to do an unbiased analysis of lipid species that are present in the brains of mice lacking progranulin versus wild-type brains. And that's what's shown in this volcano plot, on the left-hand part of the slide, and what we saw here was that there was a differences in a number of lipid species, and in particular, there was differences in the lipid -- various species of the lipid BMP, which I'll go into in a little bit more detail on the next slide as well as another lysosome. And then when we look at these changes in BMP levels, they're not age dependent. They start at very young ages and really argue that this is a core component of the pathology the onset of the pathology within these animals, and that's what's shown on the right-hand side of the slide on the top. On the bottom part of the slide, this other lysosomal lipid that I pointed out in the Volcano plot, this -- the accumulation of this lysosomal lipid further increases with age, congesting that this lysosomal lipid is then a secondary consequence of the initial lysosomal dysfunction. And so to dive into BMP in a bit more detail, and you've heard about BMP before. Carole mentioned it in her earlier talk around our ETV:IDS molecule. We Spoken about it before with respect to our LRRK2 compound. So I wanted to take a little bit of time to dive into a bit more detail on what BMP is, and what does it reflect in terms of cell biology and lysosomal biology, in particular. So BMP is a really specialized lipid. And the picture of one type of BMP is shown on the top right of the slide. And the reason I call it a specialized lipid is it's a phospholipid that is pretty exclusively found within the late endosome and lysosome, and more specifically, it's found on these structures known as intraluminal vesicles within lysosomes. And so these are actually small vesicles within lysosomes, and the function of these vesicles is to aid in the activity of normal lysosomal enzymes. And that's why I let people believe that when lysosomes are not functioning properly and these lysosomal enzymes are not functioning properly that you get changes in BMP as a compensation to that effect. So what's interesting is that these intraluminal vestibules can also be exacytos by cells in the form of exosomes, and this is why BMP can be found in extra cellular fluids such as CSF and plasma. So it provides a convenient way in fluids to actually have a proxy of what's happening within the lysosomes of cells. And so because BMP is so specific to lysosome, this is why we feel like it's an excellent marker for monitoring lysosomal function for both our PTV:Progranulin program as well as other programs. And so I'm going to show you a bit of data on how we use that here. And so now this is a single-dose study with our PTV:Progranulin molecule and the architecture can be seen there on the right, where the progranulin molecule, which is actually a number of different granular repeats is conjugated to our TV sequence. And we dosed again these animals, we dose this construct to animals lacking normal progranulin expression. And what you can see, again, in the first 2 bars on the left-hand side of the -- or the left-hand chart is that BMP levels are dramatically reduced in progranulin knockout animals as compared to their wild-type littermates. And so what we did then was to dose these animals with PTV:Progranulin, and what we found was that a single dose of PTV:Progranulin was able to rescue these BMP levels up back up to comparable levels to wild type. And this will only happen with our PTV:Progranulin molecule and a control molecule that have the FC domain to match exposure, but not the TV sequence was unable to elicit this rescue. And so this is in contrast to the liver, where also the -- you could see these BMP changes, and that's shown on the right-hand side of the slide. And much like Carole showed you with our ETV:IDS molecule, in the periphery, both the FC Progranulin and the PTV:Progranulin molecule were able to rescue BMP levels. But in the brain, only the delivery of PTV through transfer end media transcytosis was able to elicit or to generate sufficient levels to elicit this effect. So we then wanted to dive into this into in a bit more detail. And so what I'm showing you here is now looking at a multi-day study looking at the impact of PTV:Progranulin in progranulin knock-out animals. And so this is looking at 1, 3, 7 and 10 days after PTV treatment. And what you can see on the left-hand part of the slide is that our PTV sequence is -- reaches higher levels within brain as compared to a non-TV sequence, but it's cleared relatively quickly from the brain. However, when we look at BMP over this dosing window, with the first time point being at one day, you're already getting very nice correction. So the gray line at the bottom is roughly the level -- the lower level of BMP in the granular knockout versus the black line is the granular level in the wild type. And by one day, you get a rescue, but this rescue is sustained throughout the 10-day experiment, even though all of the PTV -- the full-length PTV has been cleared from the brain. And again, this provides very nice data to suggest that actual active progranulin is acting in the lysosome, and that stays within the lysosome even after the IgG portion is cleared. So this rescue of BMP, importantly, also correlates with a rescue of microglial activation as measured by TREM2 levels. So TREM2 levels are increased in activated microglia, and as you could see by 3 days after PTV administration, we're able to rescue those TREM2 levels back down to wild hype levels. And this is also sustained throughout the course of the experiment. So we then decided to push it even a bit further and now using a lower dose. So a 10x lower dose of PTV:Progranulin compared to our initial experiment and looking at 2, 3 and 6 weeks, instead of days, looking at weeks after initial dosing. And what you can see here is -- and this data is part of what got me really excited is that even 2 weeks after a 10x lower dose, you're still able to maintain near complete rescue of BMP levels in the brain of these animals. At 3 weeks, there is still good rescue, and even at 6 weeks, you haven't fully returned back to knockout levels. So even though this molecule is cleared relatively quickly, we feel it has a chance to give us prolonged rescue of these lysosomal phenotypes. And then lastly, I'll talk about using BMP as a clinically translatable biomarker. And so much likely, Carole's slides I show on the left-hand side, some mouse data on the right, some human data. And what I showed just very quickly with the mouse data on the left is measuring the CSF of these program in knockout animals, you're able to see the same reduction in BMP and the same ability to rescue it from PTV Progranulin. And in fact, the correlation between the levels of BMP and brain and the levels of BMP and CSF, both before and after dosing are very well correlated. And -- but importantly, then if you look at, again, back to patients with granulin-muted FTD as well as nongranulin driven frontal temporal dementia. You could see that there is a similar decrease in the same BMPs in these patients, really arguing that this biology is highly relevant to the human disease, and it's not just an aberration of the mouse model. Okay. So to close then and running a bit behind, so I'll go quickly, talk about our OTV molecule, again conjugating all of the nucleotides to our TV platform. And here is a quick overview of what the OTV looks like. It's a highly flexible platform. Our current version of OTV, which we call OTV Version 1.0, has a single ASO conjugated to the TV platform. But there are really a number of areas where we can continue to make changes and optimize, depending on the particular therapeutic target we're interested. This includes fab binding, TFR affinity, conjugation site, conjugation chemistry, the number of ASOs conjugated as well as the oligonucleotide chemistry itself. And so I'll just show you one quick experiment, sort of a proof-of-concept for our OTV platform here, where we're looking at dosing and OTV that's conjugated to -- conjugated with an ASO targeted MALAT-1, which is just a gene that's commonly used in the ASO field because it's broadly expressed and relatively easy to measure reductions. And what we can see is that when we do systemic dosing of an ASO only or a normal monoclonal antibody without the TV conjugated to ASO, we see no knockdown of target gene expression, either in Cortex, which is shown on the left or spinal cord that's shown on the right. However, when we look at either a single dose, which is the open circles of an OTV treatment or a multi dose and that's 4 doses of OTV. We -- what we can see is a very nice knockdown of MALAT-1 gene expression in both cortex as well as spinal cord. So we're very excited to build upon these initial results and demonstrating that we can get nice knockdown of our target gene sequences within the CNS through peripheral delivery. And just to close further OTV, again, to highlight really the potential and the potential broadness of which we can use this platform. There is really already been a number of exciting targets for all of nucleotides have been identified within the CNS space that range across indications with potential for more targets possible. And these include highly genetically validated targets that have often already been validated within preclinical models of disease, and sometimes, I've already shown initial exciting clinical data. And with this in mind to really take advantage of our OTV platform, we've recently signed a deal with Secarna to develop 2 OTV targets where Secarna will work to develop ASO sequences to conjugate with our OTV platform. And what I'm most excited about here is really that the benefit of OTV is not just limited to the ability to peripherally dose this molecule, which is really, again, a big plus. But going back to what Ryan mentioned earlier today, there is still real biodistribution issues with intrathecal or ICV delivery, the way that most all of the nucleotides are delivered. And we feel like our better biodistribution has the potential to lead to better efficacy as well as being able to target peripheral disease in addition to central disease as these OTS will have biodistribution outside the CNS as well. So I'll just wrap up there and say thank you for your time, and hopefully, I tried to show you some real benefits across a number of different modalities for the TV program; on the ATV, showing how we can enhance the activity of these molecules; and then how we can use the TV platform to unlock a number of different therapeutic modalities. So thank you for time, and I'll turn it over to Alex now.

All right. Thank you, Joe. It's a great pleasure to round out the presentation section of this R&D Day with an outlook into the future, how we can grow and expand our TV portfolio and how we work with our partners to unlock the full potential of this platform. I want to start with a look at the universe of the opportunity. At the center of all these diseases and indications is the need to get the drug to the site of action in the brain. And this is what we believe we can do with our modular technology where we can match each individual target with the optimal modality. Shown here on the left-hand side are the indications that we're currently focusing on. So this is the broad field of neurodegenerative diseases, Alzheimer's, Parkinson's and ALS. And also the large number of lysosomal storage diseases that have a neuronal component where enzyme replacement therapy for the brain can provide a real benefit for patients. On the right-hand side, shown here in blue is the universe of future opportunities where we believe that a number of targets and pathologies and cell types can be contractable if we can get large molecules into the brain. And this ranges from other neurology indications such as pain or neuropsychiatry or neuromuscular diseases to oncology, of course, with primary brain tumors such as glioblastoma, but also brain metastases, for example, HER2-positive breast cancer, and Joe showed some very encouraging early data. And all the way to certain infectious diseases where addressing a virus in the brain with an antibody may provide huge benefit for patients. A quick look from the future back to today. Joe already showed our broad portfolio. We currently are working on 12 programs. The most advanced, of course, our ETV:IDS program with data by the end of the year. And shortly behind that, our 2 programs in IND-enabling study for TREM2 and Progranulin. I just want to mention on this broad cohort, the battery of early-stage programs that we have with 9 programs in discovery stage, the majority of this portfolio is, by the way, for Alzheimer's and Parkinson's disease. And that for the majority of these early-stage programs, we already have in vivo data in a disease-relevant animal model. So shifting to our partnerships. Collaborations have been central to our strategy at Denali from day 1, and we've been very active, and I think we can say successful on the deal-making side. Shown on this slide here are those collaborations which are directly relevant to building and executing on our TV technology. Grouped or color-coded here is, in blue, those which we call enabling technologies where we work with partners on the engineering and the development of the technology or certain targets. And in these collaborations, we hold all development and commercial prices. And then in orange, are our 2 big strategic collaborations with Takeda and Biogen, which I will go into more on the next slide. On this slide here, I just want to highlight 2 of the enabling collaborations, our oldest partner, Star; and our newest partner, Secarna. Star based in the U.K., in Cambridge. They are pioneers in FC engineering. As you've heard throughout the day, the FC, the part of the antibody in which we have introduced the binding site to transferrin receptor, which then shuttles the protein through the blood-brain barrier. And through this collaboration and the subsequent acquisition of a subsidiary of F-star, called F-star Gamma, we now hold a broad intellectual property portfolio, which provides platform, IP protection technology. Joe mentioned our news partner, Secarna. This is, in fact, as of this week. We're very excited to kick off the work with Secarna based in Munich. Secarna, they're experts in ASO discovery. They have a proven and proprietary technology for ASO selection, and we are working with Secarna to identify ASOs for our 2 -- for the first 2 targets, one for Alzheimer's and one for Parkinson's disease, and to combine this with the TV to use IV-delivered ASO therapies for brain treatment. Looking at our 2 strategic collaborations. We're very excited about our 2 partners, Takeda and Biogen. I do also want to say here that we have a third strategic collaboration with Sanofi, which is around our small molecule RIP-kinase inhibitor program, which is outside of the scope of this presentation today, which is focused on the footprint barrier. So our first partner was Takeda, back in January 2018. We're very excited about the relationship with Takeda. It's a very close dialogue on the science side on the 3 targets that -- for which Takeda has an option: Progranulin TREM2 and tau. And Takeda is the option opt into these programs formally at IND, at which point this would turn into a true co-development, co-commercialization agreement with 50-50 sharing of costs and profits globally. On the right-hand side, we're very excited. We just closed the deal in recent weeks with Biogen. In fact, we're excited about every aspect of this deal. First and foremost, about Biogen as a partner. Biogen, just as we are, deeply committed to defeating degeneration. And Biogen also, of course, has deep experience and expertise in using antibodies for CNS indications. So Biogen really understands the challenge of the blood-brain barrier, but also the opportunity of increasing exposure, increasing exposure in the brain for more target engagement. So this deal has 2 components to it for which Biogen paid us a total of $1.025 billion upfront between cash and equity. One component is our small molecule LRRK2 inhibitor program for Parkinson's disease, which is, again, a co-development and co-commercialization agreement, which is outside of our scope here today. But then on the TV side, Biogen has an option to 2 of our TV programs, one of which is ATV:Abeta. And then in addition, Biogen also has an option to -- for a first right of negotiation if and when we seek a partner for future ATV or ETV programs in a certain set of indications. So moving ahead at this point. The last slide, I do want to bring everything together with this slide, bringing together the past and the future. Key message here, we are at a very exciting point in time at Denali. A point in time that has been many years in the making through work here and with our collaborators and is based on decades of blood-brain barrier research. We are eagerly awaiting the data on the ETV:IDS program, which we expect by the end of this year. And if positive, these data could mean a real inflection point. If these data, in fact, show that we can safely deliver large concentrations of a protein to the brain and have a sustained pharmacological effect, then this technology holds great opportunity for significant impact and value for patients. So on this positive note, I will close my section, and I will hand it back to Ryan for the rest of the Q&A.

Excellent. Thank you, Alex. I'm going to end by where we probably should have started, which is an enormous thank you, first to the patients who have participated in our studies or who are currently participating in our studies from Hunter syndrome to ALS to Alzheimer's and to Parkinson's. We know that it takes a collaborative effort across patient communities as well as our employees. And so I think the biggest thank you is to those who worked on these programs within Denali, and an enormous amount of work over the last 5 years and an incredible team. We look forward to the time we can get back together and have a photo just like this. But with that being said, everyone has continued to have a huge impact, especially during this highly unusual year. So with that, we're going to turn to the Q&A, and I'm going to ask my team, everyone to join me, and we're going to try to rapid fire here. We have over 50 questions. I don't think we're going to be able to cover all 50, but we're going to look for -- try to cover as much as we can in the next 25 minutes. So great. So now that the team is on, let's just make sure we have Joe, Alex, Carole and myself. I'm going to begin the Q&A. And Joe, I'm going to focus on a set of questions for you. I'm actually trying to often pick 2 questions that are related to each other so that we can answer more of them. So let me start with this one. So Joe, please speculate on a potential mechanism of action for enhancing activity and potency beyond just enhancing brain penetration. This was shown in the TREM2 example of ATV. Does the ATV enhance ability for TREM2 binding to target tissue and therefore, is the activity higher? So that's one question. And then a related question, what is the hypothesis for why the combination of ATV:TRAS and PER more effective than the original antibodies in a regular xenograft model that it's not necessarily directly leveraging the TV technology. And I think what they're probably asking is brain exposure. So the question is really centered on the mechanism for enhancing activity of antibodies, independent of brain exposure? What's your -- what do you think, Joe?

Yes. And great question. I love one question when you say, please speculate. So then I can go -- I can really go after it. But I mean, I think based on the data we have, it's basically what you said, Ryan, which is our ATV sequence has relatively low affinity binding to TFR. But it does seem like when you're binding target on the surface of a cell, that binding TFR likely insist does increase the effective -- or affinity, pardon me, through vividly tapered effect and increases the relative potency of the molecule. So we think that's what's going -- likely going on with both our TREM2 and our ATV:HER2 molecules. And that's been shown, and we've seen that in cell-based assays that correlate with our in vivo data, showing that they actually do have a broader effect when added to cells.

Great. So along these lines, as we develop the TV platform or the question is written as you develop the TV platform, what is the fastest or best way to provide proof-of-concept for enhancing target. And again, back to HER2, Abeta, and maybe I can answer that one. And Carole, you can add to it. So this has been an ongoing debate within Denali since really the origin, especially when we had our first nonhuman primate data with the TV platform. And we have constantly debated how to utilize the platform across different modalities. And there was no question in our mind that going after enzymes first allowed for the fastest and most robust proof of concept. It also allows us to derisk TFR and prove that it's a path to the brain. The question then became, what would be the first antibody. At this point, we have several antibodies or proteins neck and neck. Our TREM2 program and our Progranulin program are most advanced programs. And there, as you've seen with Carole's presentation, we're investing heavily in biomarkers that will allow us for a quick biomarker proof of concept. But I think probably the way that this question is asked is really around clinical efficacy, and I think that's a debate that we've had around HER2. It's not trivial. It's not exactly quick. And I think as you pointed out here in the question, Abeta may be another path there we could look in Phase 1 at amyloid reduction. So I think that will probably be the answer to that question. I know Carole or Alex or others, if you want to add to that. Okay. So I guess a follow-on to that, it says, "How can you do this rapidly?" Are there areas you'd rather partner out such as oncology or HER2 seems logical for a cancer company. So Alex, I'm going to hand this to you.

Yes. Thank you. Looking back at the universe of opportunities side for the foreseeable future, our focus will be on the left-hand side, will be on orange, will be on the neurodegenerative diseases and lysosomal storage disease with the neuronal component. That's why the company exists. That's our mission to defeat degeneration. With respect to the opportunities on the right-hand side, these are some that we may pursue ourselves in the future or probably more likely, together with a partner who already has deep expertise and infrastructure in developing these indications, and I think oncology is a great example for that.

Great. Well, Carole, I'm going to hand the next question to you. And this is a very specific question, but I think it's brought up several times in the Q&A around bar that we have set for 50% reduction. So the question is, if SHP609, which is the intrathecal delivered by [indiscernible] with 70% reduction. Why would 50% production work? This is a great direct question. So Carole, I'm going to hand the team for this.

Great. So yes, it's an excellent question. I think it touches on the concepts that both Ryan brought up in his talk in terms of where the CSF is collected when measuring the gags as well as I put up Slide 56 from my talk and looking at the differentiation between programs. And I'll just start by noting what Ryan highlighted in his presentation where so taking out CSF to measure GAGs at the same site of the administration of intrathecal therapy is going to overestimate the reduction of GAG that's embed in the brain. As you can see on this slide, there has not been published data demonstrating the correlation between the GAG reduction in the CSF and GAG reduction and downstream effects on lysosomal function and neuroaxonal function, for example, looking at neurofilament in the brain or in the CSF correlating with brain. So I think that that's maybe the first point that I'll make. I think the second that I'll make is just in terms of the patient population. And this goes to why we are very focused on collecting natural history data in this disease area, analyzing patient biomarker samples because we want to who are the best patients to enroll and who will be most likely to respond. We saw that clinically, these patients, they're cognitive and developmental function drops off as early as the age of 3 years old or so. And so we are really looking to treat as early as possible and look at enrolling both younger patients as well as patients that are more advanced to see if we can demonstrate in a reasonable length study of benefit in cognition.

A related question around the initial data is the following. So what aspects of the initial ETV:IDS data do you think provides the most support for alternative TV modalities. In other words, what gives you confidence that ETV:IDS data would translate to success with ATV or other TVs? So I'll take that, and then Joe, I think you should be prepared to supplement my answer to this question. And so I think without any question, the ETV:IDS data is most validating for other enzyme approaches. In fact, because of the dose frequency and the modality, it's clear that if we have success with one enzyme replacement therapy, it's likely that others can work using the same TV technology. That being said, progranulin -- and Joe just showed a large amount of data, acts very much like an enzyme, and it has a long duration of response, and therefore, we think it's also highly validating for progranulin. Along those lines, it's pretty clear that the OTV platform, or the Oligonucleotide transport platform, which brings ASOs -- I mean, it's actually pretty amazing to imagine how this would work. You have an antibody fused to an ASO injected systemically is able to cross the blood-brain barrier, enter cells and knock down gene expression. And so I think the ETV:IDS data is essentially validating for rain crossing and cellular uptake as we've shown a large amount of cells. So also, in some way, validating the OTV. I think the antibodies are somewhat unique. There is no doubt that the ETV:IDS data proves that transferrin receptor is a viable path to brain, but antibodies have a different duration of response, likely dosed once a month. And as Joe already showed, there is a dynamic around additivity or even synergies between transferrin receptor and the target for the antibody. And so I think in those cases, we have a lot to learn. Obviously, it's still validating for antibodies, but they're each going to be unique in the way that we approach them. I don't know, Joe, do you want to add anything to that?

No. I think that, that was a great answer. And I think it's really about using a pharmacodynamic market, so not just measuring drug concentration, but looking at what's happening in the brain of patients to get a real definitive readout that we're increasing concentrations in brain and delivering the target cells, right? And I think that, that has broad readthrough for the platform overall just to have clinical data saying, BFR mediated uptake using a TV will work in patients, right? I think that, that part alone will really derisk a lot of our early-stage program.

So Carole, how should we think about CSF GAG reduction in non neuronopathic MPS II patients? Will the data presentation be stratified based on CNS involvement? I'm just going to, I think, maybe add a question to that, how strong is the evidence around CSF GAGs in non neuronopathic patients in terms of predicting if they have a cognitive deficit.

Yes. So just in terms of looking at CSF GAGs, even in non neuronopathic disease, there are still going to be increased gags in the CSF compared to normal individuals. And so we will be looking at both of those. In the first cohort that we are studying for proof-of-concept, those are all neuronopathic patients that we'll be looking at for GAG reduction. In terms of the question regarding correlation of GAG reduction and CNS effects, I don't think that there is enough data to really provide a clear correlation between the CSF GAGs and clinical end points. I think what we do know is that in neuronopathic disease at autopsy, there is clearly evidence of this attenuation of gangliosides and other lipids in neurons. And so the work that we've done looking at pathway engagement and that the secondary storage that is associated certainly with development delay in these patients, I think it's going to be very important to look at in this program.

Great. Now we'll turn our attention to neurofilament. There are 2 -- at least 2 questions on neurofilament, and I'll ask both of them, and Carole ask you to answer these questions. So how do you envision neurofilament adoption in your clinical plan? Should we presume this is related to ETV:IDS? Should we presume this will have most utilization at a patient phenotyping stage how about in terms of monitoring clinical outcomes as a surrogate? Do we have sufficient understanding of the natural history of Hunter with neurofilament? And I'm going to ask a related question. I think, Carole, you can wrap this up, but basically, how will you control the variability in neurofilament levels in smallest trials? Do you think you need to measure CSF levels of neurofilament to get that usable signal? And I guess they're referencing maybe other small trials where it was measured in blood rather than CSF.

Yes. So these are all great questions. And in terms of the natural history of neurofilament in Hunter syndrome, to our knowledge, we're the first to actually publish data on neurofilament levels in Hunter's patients. So this is still very early days, but regarding the potential value of neurofilament on multiple fronts, I think this could be a very powerful biomarker. And our approach right now is to use it in our Phase I/II study to look at evidence of neuroaxonal injury that we think may correlate that clinical outcomes. But again, there is no data yet correlating these levels yet with clinical outcomes. And that is one of the reasons that we have embarked in our natural history study that I mentioned earlier that we started enrolling in October last year to generate long-term longitudinal data, not just professional data, so that we can understand that further. I think that neurofilament could be very important also in a monitoring tool in the clinic for looking at when neuronopathic disease may start. So you heard from Dr. Jones earlier that the course of the neuronopathic disease is somewhat unpredictable and not always laid very clearly to genotype it. And so based on that, there may be value in having a tool to measure changes in neurofilament over time. But these are all things that we need additional natural history data in order to understand.

Great. This next question is highly technical, and I think it relates to the difference between intrathecal and other enzyme approaches. So are there non brain tissues bordering the CSF that accumulate GAGs that could be well corrected by intrathecal enzyme or an IV enzyme fusion that does not cross the BBB. If so, is it true that correction of these non brain tissues, for example, [indiscernible] or [indiscernible], would not be expected to correlate at all with neurocognitive improvement. And so I think I'll answer this question. It's a really -- it's a fascinating question. And indeed, one could imagine if a molecule was vascular targeting, let's say, similar to the bivalent JCR pharma-like molecule, it could reduce GAGs simply through targeting, for example, the [indiscernible] or vascular tissues that are basically producing CSF. And as a result, you could have a complete disconnect between GAG measured in brain and GAG measures, to some extent, in CSF. Although we know that some GAGs are actually produced in cells and brain as it clears. So again, a reminder for CSF. That being said, it's a great question. And I think in some ways, we think of the intrathecal trial from Shire less around the lack of correlation likely that you're not receiving GAG levels in most parts of the brain, but you're actually reducing it just near the intrathecal delivery side. Okay. So we're going to rapid fire here. So Alex, this one is for you. What TV programs do you owe payments back to Star?

Yes. Well, thanks, Ryan, happy to jump in here as well. So we do owe payments. We own the own milestone payment to Star on a number of programs, and I cannot disclose the exact number. But on the small number of programs that go through clinical development. There is a total of $77 million in milestones between preclinical and clinical milestones. Just to make sure that we -- just to reemphasize, we are not owing any royalties to star. The total amount, including regulatory and commercial milestones is $427 million.

Great. Okay. So here is a question. Any thoughts -- this one is for you, Joe. Any thoughts on prospects for OTV linked to a largest gene such as Glucocerebrosidase or AADC? And then, I guess, the question here is on timing of identified next clinical candidate for the TV platform beyond the 3 pointed out. So IDS:Abeta, TREM2, and could this be an OTV or more likely, PTV? So let's start first, Joe, with the OTV. What do you imagine? What's the maximum size of oligonucleotide we can actually link a gene as opposed to maybe an ASO that's knocking down.

Yes. And that's a great question, and the short answer is, it's yet to be determined. And I don't think that we've reached the size limit where we say that's too big, it no longer works. Our OTV work is still at its early days. And as I point out in this slide, there is endless parameters that can be optimized. But one of the things we'd like to understand is how far can we push it, and what can we deliver using the TV both all of oligonucleotide and elsewhere.

Great. And then the second question is timing of identifying next clinical candidates. So where are we bringing the next set of TV-enabled molecules forward?

So I mentioned in my slides, I might have lost over it quickly, but we plan to file an IND in late 2021 or early 2022 for our PTV:Progranulin molecule. So that would be the next IND that would be expected from the TV platform. And then hopefully, the ATV:TREM2 molecule is not too far behind.

So Progranulin and TREM2, and then, of course, we have Abeta and OTVs. And as I think we just want to highlight that we signed a recent deal with on the OTV platform and moving forward of course with that..

I was just going to say, with OTV, I mean the part of the benefit is you can move relatively quickly, right, because it's coupling ASO sequences to the TV platform. We're not talking about timelines quite yet, but there is a potential to move rapidly with that type of platform.

So the next questions. There are 2 questions again related, so trying to go after 2 at a time here. So is your program in Progranulin, different from the elector program in mechanism of action. Don't both raise extracellular progranulin under the assumption that enough will get into the lysosome. Doesn't sortilin need to be the key uptake receptor for progranulin for the programs to differ significantly? So that's actually one question. Let's go ahead and answer that one, Joe, around how we see...

It's an excellent question. And I think it comes back to the idea of why does the sortilin antibody from alector increased extracellular progranulin in the first place, right? And the mechanism that they've described, which makes sense to us that sortilin is a receptor for progranulin that brings it into the lysosome. And so by blocking that receptor, you increase the amount of extra cellular progranulin, but you're doing that at the expense of delivering progranulin to the lysosome. And so idea could be that you're actually doing the opposite of what you want to do by not allowing progranulin to be delivered effectively to the lysosome where we believe it has function.

But let's stay on the theme of progranulin, do you have any sense of how biologically equivalent your PTV:Progranulin conjugate is to endogenous progranulin, especially as it relates to function within the lysosome specifically. Do we know how important the cleavage 2 progranulin peptide is for downstream function?

Yes. That's an excellent question. And to get to specifically to cleavage to individual Progranulin, if I were to speculate, what I would say is that a lot of the active protein within the lysosomes is actually cleaved granulin rather than progranulin itself. And that's consistent with some of the in vivo data I showed today and other data we have, suggesting that long after progranulin is flipped from its full-length form, you retain lysosomal rescue. So that's consistent with the idea that a progranulin are having a major function there. I think we still need to get more definitive about that. And so to get back to the idea of how active our PTV:Progranulin is versus regular progranulin, it seems that one of the first things that happens when PTV is delivered to the lysosome is that FC domain is clicked off, right? So for all intents and purposes, our data suggests that the activity of PTV would be very similar to endogenous progranulin. The advantage actually is you progranulin itself, the clearance would be very fast through systemic administration. And so through adding NFC domain, that stays intact through systemic circulation. And you actually prolong exposure through systemically and thus giving cells a better chance to uptake progranulin. But then we think once it's delivered to the lysosome, it's essentially functionally equivalent.

Great. Thanks, Joe. So in the last 3 minutes, I'm going to try to answer 3 questions and others online can correct me if there's anything I say that's incorrect. So first, microglial cells are a clear target for modulation in neurodegeneration. Where are you on targeting these cells? And are there good markers that are internalized after binding? Is this discrimination particularly important for any specific program? So I just want to say that we share data that we rescue, for example, GAG levels in neurons, astrocytes and microglia. This is actually a critical distinction of the TV platform relative to, let's say, a gene therapy where these very -- do not impact microglial cells. So for example, progranulin, TREM2, many of these programs that are targeting microglial cells, we have an advantage using the TFR approach. The second question is, do you think the aducanumab program was hampered by low uptake of the antibody? Are you worried things like ARIA might get worse with more rapid uptake? Is there a good animal model for area to test these? And so we definitely think that dose matters as it relates to Abeta antibodies. I think the data suggests that even for plaque reduction. So the idea here is that at a lower dose or even at a dose equivalent would get far superior plaque reduction. As it relates to ARIA, I think the best experiment to run is actually in clinical studies in Phase I studies in patients with plaque to determine the risk there. We have some hypotheses of why it might be different, but we have -- I think the animal data, it's more challenging to definitively show that. And then there is a question of how do you think about the Denali's approach versus gene therapy? Could they or could they coexist? I think you already made a reference that one of the challenges of gene therapy directly to the brain is the cells that are actually infected. Another major challenge when you inject AAV in the brain, similar to proteins, actually a bigger challenge that proteins is the poor diffusion across brain tissue. So it's okay, if you're injecting directly into the site you want to target, but much more challenging more broadly. So the way that we thought about gene therapy as maybe our third wave of therapies is to engineer proteins across the blood-brain barrier expressed in liver or to engineer AAV capsids that can cross the blood-brain barrier. But for now, really, the major focus is on validating the transport vehicle in patients and then broadly applying it across modalities, give us the ability to determine dose and efficacy with some of the challenges that gene therapy is facing right now, we see this as highly differentiated. So I think with that, just want to thank that the large number of people participated. I think we covered roughly 20% to 30% of the questions. For those of you that put your name with the question, we'll follow-up directly. We really appreciate the attention that everyone's paid at this time, to this webinar, but I just want to end by saying that it is echoed as Alex mentioned, it's a very exciting time at technology. We're enthusiastic about the transport vehicle, very enthusiastic about our small molecule program. I would share a lot date on those programs over the last year and look forward to doing so again early next year. I just want to thank everyone for participating in those that allowed us to do this via webcast, and have a great rest of the day. Thank you.