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

Hello, and welcome to the Denali Therapeutics webinar highlighting our therapeutic programs addressing amyotrophic lateral sclerosis or ALS and frontotemporal dementia or FTD. I am Laura Hansen, Vice President of Investor Relations, and I'd like to thank you for joining us today. Please note that the press release we issued earlier today and the slide deck for this webinar are available in the investors section of our website, denalitherapeutics.com. Before we get started, I'd like to note that the presentations given today and the responses to 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. Should any of these risks or uncertainties materialize or should our assumptions prove to be incorrect, our actual results could differ materially from those forward-looking statements. These risks, uncertainties and assumptions are more fully described in our filings with the SEC, including our latest quarterly report on Form 10-Q and our latest annual report on Form 10-K. Any forward-looking statements are based on information available to us as of today; and we disclaim any obligation to update any forward-looking statements, except as required by law. On the webcast today, I am joined by members of Denali's management team: Ryan Watts, Chief Executive Officer; Carole Ho, Chief Medical Officer; and Joe Lewcock, Chief Scientific Officer. It is also our pleasure to have Dr. Gene Yeo as our guest speaker. Dr. Yeo is a professor of cellular and molecular medicine at the University of California San Diego, UCSD; a founding mender -- member of the Institute for Genomic Medicine; and a member of the UCSD stem cell program and Moores Cancer Center. I would like to take a moment to review the agenda and Q&A logistics for today. We have scheduled approximately 1 hour and 45 minutes for the webinar, including the presentations and a Q&A session at the end. [Operator Instructions] We will do our best to answer as many questions as possible during the Q&A session. And now I'd like to turn the program over to our Chief Executive Officer, Ryan Watts.

Laura, thank you for the introduction. It's very exciting to be with everyone today to share some of the progress on our key programs. Denali was founded to defeat degeneration. We have a broad portfolio of therapies across multiple degenerative diseases. We've spoken recently quite a bit about our lysosomal storage disease programs as well as our Parkinson's programs. We have a number of programs advancing as well in Alzheimer's disease. However, today, we're going to focus on ALS and FTD. I'd like to set the context by reminding everyone of the size of the unmet need in neurodegenerative diseases. Alzheimer's, Parkinson's, FTD and ALS represent a huge unmet need with basically very limited treatment options and very few disease-modifying therapies. We're applying very specific principles to increase our probability of success. We follow 3 principles: first, what we call genetic pathway potential or using degenogenes, genes that are mutated that cause neurodegeneration. Second is engineering brain delivery. We'll speak about engineering brain delivery, about 2 modalities today, both our small molecule as well as our large molecule efforts. And then finally, biomarker-driven development. I'm pleased to announce that Carole will be presenting new data with our programs, clinical data using this biomarker-driven development strategy. We've seen a lot of recent progress that's -- creates momentum for our portfolio. We have 5 clinical-stage programs, 15 transport vehicle-enabled programs. That -- those are biologics that are engineered to cross the blood-brain barrier. And we're well resourced to go after this broad portfolio, including key partnerships to advance some of our more late-stage programs. Here is an outline of our development portfolio. As mentioned, we've spoken recently about our LRRK2 program as well as our ETV:IDS program. Today, we will focus on 3 programs around ALS and FTD. The first program will be eIF2B, our lead molecule being DNL343. The first indication that we're going after with this molecule will be ALS. We'll then speak about RIP kinase, which is also advancing in ALS in clinical studies. And then we'll end with a summary of the progress we've made on progranulin for FTD. However, I'd like to set the stage by mentioning the steps in neurodegeneration that we're targeting. There are multiple steps. However, we're focused on 2. The first is what we call the triggers. These are direct genetic links to disease. The second target, set of targets, are mechanisms that speed up disease progression known as accelerators. Today again we'll focus on our progranulin program as well as our eIF2B program, which squarely fall into triggers being directly genetically linked to disease; and then our RIP kinase program, which is modulating microglial inflammation. This being said, these 3 programs address fundamental pathways within neurodegeneration. Our focus again will be on ALS and FTD. However, there is broad therapeutic potential for each of these programs. We'll spend time on each of the programs, focusing on the rationale around ALS and FTD. However, there are direct genetic links for these various programs to other diseases, including Alzheimer's disease. In addition to using genes to identify the targets that we're going after, one of the major challenges for us is the blood-brain barrier. The blood-brain barrier evolved to protect the central nervous system from toxins, as well as creating a microenvironment for signaling. This has been a major focus of Denali as basically engineering therapeutics that can cross the blood-brain barrier. It's worth noting that both large molecules and majority of small molecules are kept from entering the brain through this endothelial barrier. I'll start, first, by highlighting our approach for small molecules. Our small molecule approach relies on a rigorous assessment of chemical starting points with BBB delivery potential. From there, we have a disciplined application of CNS-amenable physicochemical properties and then derisk these programs with further in vitro and in vivo studies. Finally, we identify compounds that have robust trough coverage and approximately 1:1 brain-to-plasma ratio. Highlighted are a number of small molecules, on the right-hand side, that have entered clinical studies. We have an excellent chemistry team who have invented these compounds over the last 6 years, and we continue to bring small molecules forward to treat neurodegenerative diseases. And today, we'll focus on 2 of those molecules, DNL343 as well as DNL788. With that in mind, I'd like to highlight the progress of DNL343, which is an eIF2B activator, a novel mechanism for restoring protein homeostasis. It's designed to address hallmark TDP-43 pathology as well as RNA stress granule biology. We completed a Phase I study where we achieved our biomarker goals as well as our safety goals; and Carole will summarize the data today, presenting that for the first time. We've initiated a Phase Ib study in ALS and we will begin -- we will continue to enroll this study over the next several months. Our RIP kinase inhibitor in collaboration with Sanofi, known as SAR443820 or DNL788, is designed to address TNF receptor 1 inflammatory pathway. We've also completed a Phase I study where we've achieved our biomarker goals, and it's generally well tolerated. We will be initiating a Phase II ALS study in Q1 of 2022. This is being led by Sanofi and received fast track approval. Just a reminder. ALS represents a huge unmet need. It affects motor neurons, ultimately leading to voluntary muscle control and movement disorder. About 250,000 people are affected worldwide. DNL343 and DNL788 are potential first-in-class treatments for ALS and have potential for expansion into multiple additional indications, as I've already outlined. The last program we'll focus on utilizes our transport vehicle technology, which is a technology that's invented to get large molecules to cross the blood-brain barrier. Illustrated here is the transferrin receptor, which is expressed at high levels on endothelial cells in the brain. Our technology utilizes this transferrin receptor, which is required for transporting iron [ to get ] large molecules in the brain. And our lead program for FTD is PTV progranulin or protein transport vehicle progranulin. This is a TV enhanced brain uptake molecule, which is the most direct way to increase progranulin in the brain. We are planning to file an IND or CTA by the end of this year and begin a human study thereafter. It's the second TV modality and it's -- and it represents the expansion of our platform. A summary of FTD. It's a deterioration of the frontal and temporal cortical areas of the brain, which results in language and planning ability dysfunction, ultimately dementia and death. It's the most common form of dementia in people under the age of 60, and it's affecting approximately 800,000 people worldwide. There are 3 genetic subtypes caused by mutations in either progranulin, C9orf or MAPT. We will be going after the FTD granulin population with our PTV progranulin molecule. Next, I'm going to hand it over to Gene Yeo, who's going to talk about RNA stress granule biology [ in ] disease, specifically in ALS.

Thank you, Ryan, for the beautiful introduction here to the unmet need here in ALS. So let me see if my slides work. Okay, so I am here to tell you about the -- some of the work [ in my group ] that touches on the biology of stress granules and how stress granules are important in neurodegeneration, with the idea that some of the insights we uncover may be helpful in industry and thinking about therapeutic modalities. So just to remind everybody: The central dogma here is represented as genes are transcribed as RNA from the human genome and the RNA is then processed and then translated in the cytoplasm. And all of these involve many RNA-binding proteins that modulate the RNA life cycle -- oops. [ Ryan ], could you go back a couple slides? Thanks. Next slide, please. And the importance of RNA-binding proteins have been showcased. I can showcase here by showing that many of these familiar mutations actually encode [ variants ] in RNA-binding proteins. And RNA processing defects are prevalent both in familial and sporadic disease. And just to remind everybody: that a large fraction of ALS patients display TDP-43 cytoplasmic aggregates. That is now a clear hallmark of the disease. This slide is very clearly indicating that there are many RNA-binding proteins, as I've pointed out, that have mutations linked to ALS and FTD, but it's very -- it's been very exciting to see that many of these mutations are actually found in intrinsically disordered regions in the proteins themselves in [indiscernible] regions. And these regions -- next slide, please. These regions are involved in homotypic and heterotypic interactions that cause the formation of liquid phase transitions and condensates. And so we call these condensates in the cytoplasm stress granules, and these stress granules contain RBP-RNA complexes. And here you can see on the figure below that one of the hallmarks of stress granules is G3BP1 positivity. It's -- G3BP1 is an RNA-binding protein that is typically diffused in the cytoplasm, and when cells are subjected to stress, G3BP1 and its partners condense into these liquid droplets that over time will become insoluble [ fibers ] in disease. Next slide, please. And so the way we think about the interaction between stress granules and neurodegeneration is that all cells require stress granules to form, and this in the context of stress. And stress granules are typically transient. And in wild-type [ healthy ] conditions, stress granules dissipate over time, but in the context of ALS mutations, these stress granules do not dissolve and maintain a pathological state that becomes permanent and insoluble over time. And next we see that there have been many, many reports that implicate that ALS [ see mutations in disordered regions ] that affect stress granule dynamics. Next slide, please. And so in the big picture perspective here, we think that stress granule dysfunction is a clearly shared feature in not just ALS and FTD but also in Alzheimer's disease and many other neurodegenerative diseases. Next slide. And so I'm going to quickly go over how we have been thinking about stress granules and how we model stress granules and to clarify what is the composition of stress granules as well. Next slide. So one of the key features of modeling stress granule biology is using human-derived cell lines [ and that are the correct ] cell types. So we typically use iPS-derived motor neurons that have genetic risk factors, mutations in RNA-binding proteins or sporadic ALS lines; and subject these to environmental stressors that can then perturb the system and [ then interacts with that model ] in these iPS lines to create disease-relevant phenotypes, relevant to, for example, ALS and FTD. Next slide, please. And just to remind everybody again: G3BP1-positive foci and puncta are what we use as a working definition of what stress granules are. And G3BP1 forms this core complex of proteins. And RNAs are in one -- in this particular model here. And RNAs can then recruit other RNA-binding proteins that contain IDR regions. And this growth of these granules is what creates these very clearly visible puncta known as stress granules you can see from the arsenite-treated samples across a variety of different cell types. Next slide, please. And so just to set the stage for what we think of as a relevant model for how we can use stress to model ALS. We use iPS-derived motor neurons here. And we can show that stress drives TDP-43 into stress granules in the cytoplasm, on the far left, top left. And stress conditions actually perturbs TDP-43 binding across the transcriptome. And here we can also show that TDP-43 binding is now lost from key biomarkers such as stathmin-2 exon. Inclusion can happen in the stress and obviously disease state. We also know that stress can reveal a disturbance in RNA localization. And we show that, in mutant cells, stress leads to the mRNA species not recovering and sort of being static and presenting in a perturbed state. Then finally, stress can [ exemplify, amplify ] disease delayed-onset cell death in ALS mutant lines, whereas we don't normally see [ aberrant ] cell death, but in a stress condition, ALS lines exhibit increased cell death. Next slide. And so the -- starting to answer the question: What are in stress granules? We have [ leveraged ] proximity labeling using the APEX2 enzyme here that are knocking into iPS-derived motor neurons and different cell types. And so this proximity labeling allows [indiscernible] of other proteins that are in proximity, right, to G3BP1 in the stressed state. And this allow us to then use [ mass spec approaches ] to identify the proteins that are found in stress granules. Next slide. This represents the first comprehensive effort to look at the human stress granule composition [ in different ] cell lines and cell types. And so we've identified 300, over, proteins that are clearly identified in human SGs. And as you can see in the middle here, a majority of these are actually RNA-binding proteins, as we expect, because the stress granules contain protein RNA aggregates. We find that in neuronal cells. They contain many similar components as in non-neuronal cells, but there are cell type differences. You can see below that there are a variety of different RNA-binding proteins that are identified only in the -- in neurons. And from this compendium of RNA-binding proteins and non-RNA-binding proteins in stress granules, we're able to then deplete these proteins and show that many of them lead to -- if depleted, leads to a reduction in stress granule formation. Next slide, please. And just to amplify the importance of this stress granules protein: We can modify these stress granule -- novel stress granule proteins in fly models with collaborations with Mark Kankel in -- at Biogen at the time and then, just next slide, with Fen-Biao Gao and UMass to show that, if you alter the levels of these novel stress granule proteins, you can prevent and in fact reverse some of the degenerative phenotypes that we find in ALS models, in this case fly models. Next slide, please, one of the other studies we have been conducting to think about how to manipulate stress granules to utilize G3BP1 assays [ that's ] high-throughput, high-content screens. And so we've also identified compounds meant to reveal to us how these proteins aggregates can form and how can we perturb their interactions. So here we're able to identify compounds that you can incubate with purified stress granule-like aggregates. And these stress granules, in vitro, will fall apart with the intercalation of these compounds, whereas upstream compounds like cycloheximide and ISRIB may not affect the direct -- may not directly affect the purified stress granule aggregates. And that tells us a lot about the chemical structure and the physical components of stress granules. Some of these compounds can also be incubated in primary neurons derived from mice models of ALS. And interestingly, they can prevent increased cell death and push it back to control conditions. So we believe that some of these compounds have clarified the mechanism of how some of these stress granules can form and how we can cause them to fall apart or dissipate and recover in primary cells. This was highlighted in a F1000 report that suggests that this study is exciting, as it lends credibility that you can identify drugs that could modify these membraneless organelles. Next slide. This slide highlights another effort in the lab that thinks about what are the genetic determinants of stress granule abundance. And previously, we have articulated that RNA-binding proteins are a major component of stress granules, from our previous cell paper. And here we focus on RNA-binding proteins and ask the question: If we deplete these RNA-binding proteins genetically, which RNA-binding proteins, if depleted, can cause a change in the stress granule abundance. And here we've utilized 2 -- combined 2 different approaches here. One is pooled CRISPR screens with a library targeting RNA-binding proteins. We've also combined a microraft array or assay, where there are 40,000 microrafts, [ one of the slides ]. And the pooled CRISPR approach were able to get single-cell -- or a clonal growth of cells [ when ] RBP depleted in each raft. And then these rafts are then subjected to imaging to ask if the depletion of that RNA-binding protein led to a change in the stress granule abundance. Next slide, please. And these microrafts are unique because, once we identify using imaging if stress granule abundance is affected, we can physically pinpoint the position of the raft, release the raft from this [ slide ] and then sequence the barcode that represent what gene is being depleted or knocked down or knocked out in the cell and identify what are these [ novel ] genes that control stress granule abundance. And so we were able to, for example, show that surprisingly there are multiple steps in the RNA life cycle where proteins, and in this case RNA-binding proteins, can affect the abundance, including both initiation and resolution, of stress granules. Here we show one specific example. SNRNP200 depletion using sRNAs in [ human ] cell lines can actually reduce the stress granule abundance in cells. And then excitingly, just this year, Christine Vande Velde's lab actually showed that there are SNRNP200 inclusions found in ALS patient samples, right? And this indicates that our ability to identify genetic determinants of stress granule abundance may actually be also relevant in patient samples. Next slide. And so I hope -- I've tried to summarize here how we think about strategies to understand and manipulate stress granules. We've been excited to identify proteins found in the formation of stress granules. We believe that manipulating those can also be helpful in thinking about therapeutic approaches. We've identified genes that control stress granule information [ in ] the combination of imaging technologies and pooled CRISPR screens. And finally, we've been excited to think about ways to manipulate the protein-RNA interface to prevent recruitment of disease-associated RNA-binding proteins [ and ] TDP-43 or blocking the transition of these stress granules into persistent insoluble aggregates. Next slide. And with that, I'm happy to end my talk here. And I want to thank [ the lab for most ] exciting collaborations with folks at UCSF; UCSD; Montreal; and of course, our colleagues in industry. Thank you, Ryan.

Great. Thank you, Gene. We appreciate you going through the RNA stress granule biology and its link to ALS. I'm now going to focus in on eIF2B. We'll bring back in the stress granule biology as it relates to eIF2B and specifically to DNL343. So I'll start by summarizing where we are with eIF2B. This is a novel first-in-class approach to the treatment of ALS and other degenerative diseases. DNL343 is a brain penetrant small molecule activator of eIF2B. And I'm going to describe a little bit the molecular cascade and pathway around eIF2B, also happy to present preclinical data showing robust rescue of pathology as well as key insights into pathway modulation that we plan to translate into the clinical setting. Carole will present the Phase I data study, which has been successfully completed where we've achieved both biomarker and our goals around safety. We've initiated the Phase Ib study in ALS, as I've already mentioned. And as we mentioned before, our first therapeutic area is ALS. However, there is a direct genetic link to other diseases such as vanishing white matter disease. We see a broad role for the integrated stress response in Alzheimer's and other degenerative diseases. And it's important to note that this small molecule was discovered and is fully owned by Denali. And just highlight on the left-hand side the broad therapeutic potential around eIF2B. So I'd like to start first with just a general description of the integrated stress response. You'll see throughout the slides that I'll link it to RNA stress granules. And I appreciate Gene showing a close relationship with various ALS genes and RNA stress granules, but I'm going to start first by describing integrated stress response. And importantly, ISR is a consequence of stressors, cellular stressors. There are actually 4 conserved kinases that sense cellular stress and signal the ISR to essentially shut down protein synthesis transiently. I'll highlight on the left-hand side that in healthy cells the eIF2B cascade is on, and in fact, that's what our small molecule does. It restores cells to a healthy state. In the context of a healthy cell, you have normal protein synthesis, no stress granule formation. And the cell is in a homeostatic state. In the context of both acute stress and chronic stress, you see that eIF2B is turned down. You see stalled protein synthesis and the formation of these reversible stress granules that Gene had just described. This transient acute stress, especially when overtly large, results in transient protection. However, what we see in chronic diseases such as neurodegenerative diseases is that this stress cascade is elevated. And it's constitutive, which results in impaired protein synthesis, persistent stress granule formation and deleterious ISR. Importantly and, I think, as Gene already highlighted, mutations in genes that associate with stress granules that result in the stress granules being locked in place will also result in this chronic stress and ultimately cell death. So there are a number of publications, extensive evidence around the ISR pathway modulation across many diseases. We're focusing today on neurodegenerative diseases, but of course -- what was already described by Gene in ALS, but there are many others, including cognition-related connection to the eIF2B signaling pathway. So again I'd like to focus on our therapeutic hypothesis: So what does eIF2B stand for? It's a eukaryotic translation initiation factor 2b. This pathway is required for protein synthesis in healthy cells. On the left-hand side, we're illustrating that in neurodegeneration eIF2B signaling is reduced. And our goal with DNL343 treatment is basically to restore cells to normal function. On the right-hand side is a molecular architecture of eIF2B and an approach for eIF2B activation. As you can see, in the off state or in a stressed environment, you see 2 heterotetramers. These tetramers come together and form a decomer or a heterododecomer, which can be activated with a eIF2B small molecule. Intracellular protein complex that regulates protein synthesis is this eIF2B complex regulating the amount of protein synthesis. eIF2B activator stabilizes, in fact, this eIF2B complex. And it overrides the off state. In other words, in healthy cells, this complex is already -- this decomer complex is already formed. In cells that are under stress, this is blown apart, and we're able to bring it back together with small molecule. So DNL343 is a novel therapeutic approach designed to restore cells to this healthy state in the context of chronic stress, which is obviously relevant in neurodegenerative diseases. I'm going to take it one step further and focus in on ALS, as Gene has already done. Let's start, begin with genetics. So roughly 50% of ALS degenogenes are associated with these stress granules or RNA homeostasis, as already brilliantly highlighted by Gene. These ALS mutations have been shown to alter stress granule dynamics; and you've seen some of that data, in fact. TDP-43 associates with these stress granules and can form insoluble inclusions. And you can see the graph in the bottom, basically genetic discoveries over time using stress granule of -- linking the stress granule biology as well as RNA homeostasis. So in summary, the ALS-associated genes discovered to date highlight the importance of stress granule biology and TDP-43 in ALS. I'm now going to focus in on DNL343 specifically. So as Gene has already shown, about 95% of ALS patients have TDP-43 inclusions. We see that roughly 50% of FTD patients and approximately 30% of Alzheimer's also have TDP-43 pathology. On the right-hand side is a cellular assay in which we induced stress to form RNA stress granules, again labeled by G3BP1, which Gene has outlined. We see a colocalization with TDP-43 as part of these protein inclusions. Importantly, when we form these stress granules and they're preformed, we can add DNL343, turning on eIF2B and essentially rapidly dissolving these RNA stress granules. This also correlates with protection from cell death. In summary, the eIF2B activator DNL343 reverses these preformed stress granules and TDP-43 inclusions. So similar to the data that Gene had presented, we're just going to highlight here an iPS-derived neuron assay of C9orf72. And here we're looking at propensity to form RNA stress granules in the context of stress, with or without C9orf72, using an isogenic [ line ] in which we used CRISPR to knock out C9orf72. And what you can see in the image in the upper left is basically formation of stress granules quantified in the middle graph. When we add DNL343, we can dissolve these stress granules in these human iPS-derived neurons from an ALS patient. In other words, C9orf patient-derived neurons display increased stress granule formation compared to our isogenic control, and DNL343 rescues the formation of these stress granules. We next went to tissue samples from patients with ALS. And here we were looking at gene expression. And specifically, downstream of the integrated stress response, there were a set of genes that were upregulated as part of the ISR. As you can see on the graph on the left-hand side, there is an increase in expression in the ISR-related genes from tissue samples from spinal cord of ALS patients. We highlight that some of the top genes on the right-hand side are these ISR-expressed genes, so in summary, gene set enrichment analysis of target ALS transcriptome database indicates upregulation of the ISR pathway. So now I've basically outlined the integrated stress response, the role of eIF2B and the connection to ALS and specifically to TDP-43. I'd like to spend more time talking about DNL343 itself and the actual molecular cascade and how DNL343 works. I'll start by using this molecular diagram. And I already highlighted that there are 4 kinases that can be activated in the context of various stresses, including nutrient deprivation or protein dyshomeostasis. When this happens, eIF2 alpha is phosphorylated and eIF2B is turned off. What we're going to, what I'm going to show you in the subsequent slides is that eIF2B can be turned on with DNL343. And as a result, we can restore protein synthesis. We can decrease stress granule formation. And a set of genes that are upregulated in response to stress, such as ATF4 and CHAC1, are reduced with DNL343 treatment. So I'll start here with some of the in vitro work highlighting the mechanism of DNL343. As shown here on the left-hand side, we're looking at protein stability of eIF2B in the context of DNL343. What we see is that, in a dose-dependent fashion, we increase the stability of this complex, which essentially turns the eIF2B on and restoring cells to healthy protein synthesis. DNL343 activates eIF2B by directly interacting with the eIF2B protein complex. We next ask, what happens downstream of activating eIF2B with DNL343? And what we can see is ATF4 both -- at the protein level as well as CHAC1 at the gene transcript level are robustly inhibited in a dose-dependent fashion with DNL343. And then to tie this together with RNA stress granules, as we've shown, and human iPS cells and other cell lines, including colocalization with TDP-43, we looked in the same cell lines the ability of DNL343 to dose-dependently reverse RNA stress granule formation in the context of stress. And this is shown on the left-hand side. So importantly, we've -- now understand the mechanism of DNL343 in a cellular context and the molecular cascade. We next wanted to study DNL343 in vivo. Here we generated our own mouse (sic) [ mouse model ] with a mutation in Eif2b5 known as R191H mutation. This mutation causes chronic ISR activation, and this mutation is linked to vanishing white matter disease. This is an ideal model for DNL343 proof of concept, so importantly, to establish the ability to engage eIF2B, we wanted to show a dose-dependent increase in both plasma and brain of DNL343, as shown here on the right-hand side in the graph. Notably, DNL343 exposure in plasma and brain were tightly correlated across dose levels, suggesting essentially a 1:1 ratio for brain exposure. We next asked 2 questions. Could we engage the target? And could we modify the subsequent pathway? On the left-hand side is again looking at stability of eIF2B now in vivo both in spleen and in brain, and again you see a dose-dependent increase in eIF2B stabilization. This is critical. Essentially, in this chronic ISR model we're able to restore signaling even with mutations directly in the eIF2B complex. This led to a dose-dependent reduction in CHAC1 gene expression in brain. So in summary, DNL343 treatment led to an increase in eIF2B stability, attenuation of CHAC1 gene expression in the brain. And we've now asked the question of what about broader ISR pathway engagement. Similar to the data that we obtained from spinal cord of human ALS, we can look in brains of these mouse models. And we see a broad set of ISR-related genes expressed. And as you can see in this graph, with increasing doses, we returned these genes' expression level back to wild-type levels with DNL343. I'm now going to focus on the functional results using DNL343. So just to step back and talk a little bit about the dosing paradigm: Here we dosed mice for 13 weeks at the following -- at increasing doses up to 10 mg per kg. And what you see at the 2 highest doses, both the 3 and 10 mg per kg, we restored body weight in these mice in this vanishing white matter disease mouse model. Importantly, we also restored motor function. Looking at time to cross the beam, foot slips or falls, we see a robust rescue with DNL343. Chronic DNL34 (sic) [ DNL343 ] treatment dose-dependently restored both body weight and motor function, in summary, in this mouse model of vanishing white matter disease, which is specific to the eIF2B pathway. So now I'd like to summarize the in vitro and in vivo data. We've shown that increase in eIF2B protein stability results in suppression of ISR gene expression and reduction of RNA stress granules in vitro. In vivo, we've seen a dose-dependent target and pathway engagement as well as brain penetration, rescue of both body weight and motor function. In summary, DNL343 proof of concept has been demonstrated in preclinical models and supports studies in ALS and other neurodegenerative diseases. And with this, I'm going to turn it over to Carole to share some of our new clinical data.

Thank you, Ryan. And thanks for the overview of the preclinical data, which provide an excellent foundation for the biomarker-driven development strategy that we have applied to the design of our Phase I healthy volunteer study. The study design and results of this Phase I study were presented today at the Northeast ALS Consortium conference as well. The Phase I healthy volunteer study is a single-center, double-blind, placebo-controlled, first-in-human safety, tolerability, PK and PD study of DNL343 in 95 healthy volunteers aged 18 to 50 years of age. The study included a single ascending dose portion, a multiple ascending dose portion and also a food effect study to enable this data to support moving [ into ] patients for the next clinical study. The key end points were safety, PK and PD readouts of the integrated stress response pathway using an assay of ex vivo stimulated stress conditions in PBMCs or peripheral blood mononuclear cells that were collected from healthy volunteers dosed with DNL343. The Phase I study is completed and met all safety and biomarker-driven development goals and supports moving forward to Phase Ib in ALS patients. In the next few slides, we'll share the results of this study. So I'll start with the safety summary. DNL343 was generally safe and well tolerated across all dose level studies -- studied. There were no serious adverse events or discontinuations due to study drug. Of those participants receiving DNL343 that experienced treatment-emergent adverse events, all were mild in the single ascending dose and mild or moderate in the multiple ascending dose. Moderate AEs in the multi ascending dose portion of the study were also more frequent in placebo than DNL343-treated participants. No clinically important findings were observed in vital sign, safety labs or other safety assessments. The most common treatment-emergent adverse event experienced in DNL343 participants are reviewed in detail on this table. So this table provides an overview of treatment-emergent adverse events reported in 2 or more participants in the active or placebo groups. There were no dose-related increases in overall treatment-emergent adverse events observed across the studies. As you can see, all treatment-emergent events in subjects receiving DNL343 resolved without intervention, except analgesics for some headache and procedural pain. Notably, postural dizziness treatment-emergent adverse events were mild and not associated with clinically meaningful changes in systolic blood pressure, with a similar frequency observed across 343 and placebo participants. In summary, 343 was generally safe and very well tolerated in both the single ascending dose and multiple ascending dose cohorts. Now we'll review the pharmacokinetic profile of DNL343 in this first-in-human study. DNL343 exhibited well-behaved clinical pharmacokinetics. Following once-daily administration for 14 days at steady state, DNL343 exhibited prolonged oral absorption, low fluctuations in plasma concentrations and a long terminal half-life of 40 to 50 hours. Extensive distribution into the CNS was observed based on a mean CSF-unbound plasma ratio of approximately 0.65 to 0.89. This pharmacokinetic profile supports once-daily dosing of DNL343 and extensive penetration into the CSF. The well-behaved pharmacokinetic profile of DNL343 was also associated with robust peripheral inhibition of integrated stress response biomarkers as measured by ATF4 protein and CHAC1 gene expression. Data from the single ascending dose portion of the study were used to select dose levels for the multiple ascending dose study that were anticipated in stress-induced peripheral blood mononuclear cells to reduce the integrated stress response pathway, and this is exactly what we saw. On the left graph, you can see that peripheral blood mononuclear cells collected from healthy volunteers treated for 14 days with DNL343 demonstrated a reduction in ATF4 protein after ex vivo stimulation of PBMCs at all multiple doses tested. After completion of the dosing after 14 days, a recovery of stress response is observed after day 15, reflecting the long half-life of DNL343. Similar effects are seen on the right graph, where PBMCs collected from healthy volunteers treated with DNL343 demonstrated a reduction in CHAC1 protein expression after ex vivo stimulation of PBMCs from treated participants. These data support that robust inhibition of the integrated stress response pathway at all multiple doses tested were achieved. And the data support dose selection for our Phase Ib in participants with ALS. With Denali's biomarker-driven development strategy, we always aim to utilize preclinical model data and human data to understand and predict effects in the CNS and the brain. Ryan earlier shared data from the vanishing white matter eIF2B mutant mouse, which exhibits overactivity of the integrated stress response pathway. In the mouse model experiment, a dose-dependent effect on brain pathway engagement as measured by CHAC1 was observed with [ near normalization ] of ISR biomarkers at the highest dose. The relationship between exposure and pathway engagement from the DNL343-treated mutant mouse is shown in the orange and green lines here demonstrating that both plasma and brain PK similarly predict mouse brain integrated stress response pathway admission as measured here by CHAC1. We did a similar experiment in humans, where we can only obviously collect peripheral tissues, but as you can see here, the data from the Phase I healthy volunteer study demonstrates a similar relationship between exposure, and integrated stress response pathway inhibition, in the blue line. Together, these data support the translatability of human peripheral PK and pathway engagement for rationally predicting inhibition of the integrated stress response pathway in the central nervous system. These data enable Denali to select the correct DNL343 dose levels to test in humans in our Phase Ib study. We hypothesize that DNL343 dose levels in humans that maximize the peripheral ISR biomarker response will also inhibit the integrated stress response in the -- activity in the brain and spinal cord. So here is the design of our Phase Ib study in ALS. This Phase Ib study is a randomized, double-blind, placebo-controlled 28-day study to evaluate the safety, tolerability, PK and PD of DNL343 in adult participants with ALS. This is followed by an 18-month open-label extension for collection of additional biomarker data. The key end points include safety, PK and similar pathway engagement biomarkers that we studied in the Phase I healthy volunteer study, along with exploratory biomarkers of integrated stress response and neurodegeneration. Key inclusion and exclusion criteria included diagnosis of ALS less than 3 years from symptoms onset; and a slow vital capacity of greater than, equal to 50% predicted. We'll study 2 dose levels selected based on our Phase I data that I shared. And we'll select one of these doses after the 28 days based on biomarker data for the open-label extension. Enrollment was initiated in Q3 of 2021 at global sites, with an anticipated readout of part 1 in Q2 of 2022. In summary, the data shared here validates the DNL343 CNS distribution and activity on the integrated stress response pathway in healthy volunteers. Based on this data, we have initiated a Phase Ib study in ALS, which is currently enrolling. While our first clinical indication is ALS, DNL343 has broad therapeutic potential to address the ISR pathway that is active in multiple neurodegenerative diseases. I'll now segue to share an update on the progress of our RIPK1 program, which also has broad therapeutic potential across multiple neurologic diseases. Denali identified the RIPK pathway as an important therapeutic target back in 2015, recognizing the importance of microglial inflammatory pathways in Alzheimer's disease and other neurodegenerative diseases. Denali advanced the first CNS penetrant RIPK1 inhibitor into the clinic and also generated peripherally restricted RIPK1 inhibitors that have also advanced to the clinic in autoimmune indications with our partner Sanofi. In late 2018, Denali partnered with Sanofi on a strategic co-development and co-commercialization across a portfolio of RIPK1 inhibitors for both autoimmune and neurodegenerative indications, including ALS, multiple sclerosis and Alzheimer's disease. Sanofi is leading the SAR 820 or DNL788 development in ALS. The joint development team successfully completed Phase I healthy volunteer testing for SAR 820 this summer. The first-in-human single ascending dose and multiple ascending dose study demonstrated a good pharmacokinetic safety and tolerability profile, excellent central nervous system penetration and strong target engagement as measured by phosphoserine 166 RIPK1 kinase activity. We now have plans to advance to Phase II in participants with ALS. As a reminder. The genetic rationale for RIPK1 inhibition in neurodegeneration is linked to the microglial risk in Alzheimer's disease and familial mutations in ALS that interact with RIPK1, like TBK1 and optineurin, and trigger disease. We hypothesize that RIPK1 is an accelerator of disease, whereby RIPK1 activation via TNF receptor 1 results in microglial dysfunction and metabolic dysregulation through inflammatory-related cell death and microglia, astrocytes [ and neurons ]. RIPK has broad applicability in neurodegenerative diseases, but consistent with the theme of our webinar today, we will focus on the development in ALS. Along with Sanofi, we asked the question of whether RIPK1 activity was upregulated in the spinal cord of sporadic ALS. Data shown here demonstrates that RIPK1 expression as well as kinase activity is elevated in ALS tissue. Based on this and preclinical data, we have initiated and completed a Phase I study with SAR 820, a CNS penetrant inhibitor of RIPK1. Leveraging our experience with our portfolio of RIPK1 inhibitors, including a prior CNS penetrant inhibitor that was evaluated in a previous Phase Ib study, we plan to advance to a Phase II clinical end point study in adult ALS participants. Sanofi is actively planning this Phase II study, and the study will evaluate safety and clinical efficacy. Here is the design of the Phase II ALS study, which has been named the HIMALAYA trial, with SAR 820 or DNL788. This is a multicenter, randomized, double-blind, placebo-controlled 24-week study to evaluate efficacy, safety and tolerability in adult participants with ALS. It is followed by a 2-year long-term safety extension. The key primary and secondary end points are clinical in nature, including a primary end point of change in ALSFRS-R score; and secondary end points including a combined assessment of function and survival, respiratory function, muscle strength, quality of life and also biomarkers of neurodegeneration. Key inclusion and exclusion criteria include a diagnosis of ALS less than 2 years from symptom onset, with standard-of-care medicines permitted; slow vital capacity of greater than, equal to 60% predicted. And ability of -- the ability to swallow tablets is also required. Enrollment is anticipated in Q2 of 2022, with 50 to 60 global sites in North America, Europe and China planned. In summary, the joint development program for the only CNS penetrant RIPK inhibitor in the clinic, to our knowledge, has demonstrated target engagement and a well-tolerated safety profile. Sanofi is leading the development in ALS, with plans to begin the Phase II HIMALAYA ALS study in Q1 of 2022. DNL788 has broad potential in neurodegenerative diseases. And evaluation and planning for other indications, including multiple sclerosis and Alzheimer's disease, are also in progress. And with that, I would like to turn over to Joe Lewcock, who will take us through the development program in FTD for the PTV progranulin program. Thank you.

Thank you, Carole. I'm excited today to provide an update on our PTV progranulin development program, also known as DNL593. DNL593 is a CNS penetrant progranulin replacement therapy. As such, we're initially targeting progranulin-deficient FTD as the initial indication, but we do think that this program has potential for expansion into other indications based on our exciting preclinical data, suggesting we're able to achieve broad CNS biodistribution of PTV progranulin as well as rescue both lysosomal phenotypes and disease-related pathologies present in these preclinical models. We are planning an IND or CTA filing for this program at the end of 2021. And a reminder: that this program is a part of our strategic partnership with Takeda for joint development and commercialization. As Ryan mentioned in his introduction, the PTV represents an expansion of our TV platform designed to deliver biologics to the brain. PTV stands for protein transport vehicle and its ability to increase the uptake of these proteins into the CNS. And in this case, we're using a full-length progranulin molecule conjugated to our TV. Much of the data that I'm going to show in my update here today was recently published in a manuscript in Cell, shown on the right side of the slide. And so although I'm only going to be able to show highlights today, if folks were interested in more detailed information, I would refer you to this manuscript. So first, let's dive a little bit into the biology of progranulin deficiency. So what has become well accepted now in the field is that multiple different mutations in the granulin gene result in decreased levels of progranulin protein both in the plasma as well as the CSF. And some DENALI data to show that is shown on the bottom-left side of the slide here. This results in a number of disease-related phenotypes in these patients, and that includes significant microgliosis or glial activation as well as TDP-43 pathology and significant neuronal atrophy that correlates with increases in neurofilament light in the CSF. What data from our group at Denali as well as other labs in the field has recently identified is that it seems like the mechanistic underpinnings for these disease-related phenotypes are a result of lysosomal dysfunction that is caused by progranulin deficiency. And this lysosomal function then relates into disease-related phenotypes that can be observed. And I'm going to go into a little bit more detail on some of these lysosomal phenotypes today and how we feel like they can be rescued by our PTV progranulin molecule. So before I dive into the data, I wanted to start with the schematic to talk people through exactly how we feel like progranulin functions in the cell. So progranulin is present in the extracellular space and is taken up into cells through cell surface receptors such as sortilin and targeted to the lysosome. When it reaches the lysosome, the full-length progranulin is broken down into individual granulin peptides. Once these granulin peptides are in the lysosome, they seem to bind specifically to a relatively lysosome-specific lipid, known as BMP. And progranulin binding to BMP stabilizes BMP. And this becomes important because BMP is a major constituent of the intraluminal vesicles in lysosomes. And these intraluminal vesicles are important areas to enable proper function of lysosomal enzymes. Therefore, when you have a progranulin deficiency, your BMP becomes destabilized. You get less functional intraluminal vesicles in the lysosomes. And you get lysosomal enzymes that then become less functional, and therefore, you get accumulation of lysosomal substrates and broader lysosomal dysfunction within the cell. And this core mechanism is really important to understanding what we think about how progranulin contributes to the disease and therefore what we want to do about it therapeutically. So let's dive into a little bit of the data. And we'll start here by looking at super-resolution microscopy of cells and now labeling the lysosomes, in green here, shown with the LAMP-2 stain; as well as BMP, shown in red; and progranulin, shown in aqua. And what you can see, as mentioned in the introduction, is that both progranulin and BMP are localized to these lysosomal compartments. In cells lacking progranulin, what you see is not only, of course, a reduction in progranulin itself but a significant reduction in the levels of BMP, this lysosomal lipid. We can then add our PTV progranulin molecule to these cells. And what we're able to see is that we can first direct the PTV progranulin appropriately to the lysosome, and this lysosomal localization of progranulin is able to rescue levels of BMP within the cell. And quantification of both of those metrics is shown on the bottom side of the slide. So what does this mean for lysosomal function? When we look more closely at lysosomal function here using a DQ-BSA assay in order to measure the amount of lysosomal proteolysis, what we can see is, as compared to granulin or wild-type cells on the left side of this slide, that the granulin knockout cells have reduced lysosomal proteolysis and, therefore, abnormal lysosomal function. And that's shown in the pictures on the left and quantified in the bar graph on the right. Encouragingly, when we add our PTV progranulin molecule but not a control IgG, we see that this is able to rescue normal lysosomal function within these cells. So therefore, we could show that not only do can we deliver progranulin to cells, rescue BMP but then also rescue normal lysosomal dysfunction. So now let's move in vivo. And what I'm showing here is -- Ryan, can you go back one slide, please? Thank you. So now we're looking in vivo. And what I'm showing here is that PTV progranulin following systemic administration is able to achieve broad biodistribution within the CNS so that our TV sequence not only enables enhanced brain uptake of our PTV progranulin molecule but also effective delivery to multiple of the key CNS cell types. And that's shown on the image on the right-hand side of the slide, where if you focus in on the pink, that is our PTV molecule that's been effectively delivered to multiple cells within the brain, including neurons, which are co-stained in green; as well as microglia that are co-stained in aqua. And both of those are indicated with arrows on the right-hand side of the slide. This is in contrast to dosing of a control molecule, which is an Fc progranulin. The only difference is it lacks our TV sequence. So you can see in this case, systemic administration of a progranulin molecule alone is insufficient to achieve brain uptake or biodistribution and targeting of these multiple cell types in brain. So then we went on to look at the ability of PTV progranulin to rescue some of the phenotypes that are associated with granulin deficiency in vivo. We did a number of experiments to address this, and I'll show one of them here, which is it involves, first, a method of separating each of the CNS cell types from brain into discrete -- into ability to look at them independently. So that includes neurons, microglia and astrocytes. And then we went on to look at the presence of some of these lysosomal lipids in each of these cell types. So first, I'll show that we were able to -- in microglia, the microglia showed the most prominent changes in lysosomal lipids. And so what you're looking at here is a heat map of multiple lysosomal enzymes. So shown on the left-hand side is a wild-type animal. And comparing that to a granulin knockout animal, you see changes in multiple different lysosomal lipids including BMP as well as some other lysosomal substrates. And so this is consistent with significant lysosomal dysfunction in these cells. However, through weekly or biweekly treatment with PTV progranulin, we are able to rescue these lysosomal accumulations that occur in microglia. We see a similar story in other cell types. So shown below is in -- first in astrocytes as well as in neurons that although these cell types have less significant lipid changes, which is consistent with the function of these cells in the CNS, but that we were able to rescue these lipid changes in all cell types consistent with a rescue of lysosomal function throughout all cell types within the brain. So now let us take one step further and look at the impact of granulin deficiency on the function of certain lysosomal enzymes. And I'll focus on one lysosomal enzyme here, GCase, which is encoded by the GBA gene and well known for its links to neurodegeneration. And what you can see is in progranulin-deficient animals shown here that there's a significant reduction in the amount of Gcase activity in brain as compared to wild-type animals. Although this cannot really be effectively rescued by an Fc granulin alone, when we treat with PTV progranulin, we're able to rescue the amount of Gcase activity back up to healthy wild-type levels. Correspondingly, we see that an accumulation in knockout animals of glucosylsphingosine, which is one of the major substrates of Gcase enzyme in granulin-deficient animals, and we're able to rescue the accumulation of these defects through treatment with PTV progranulin. Measuring these types of lysosomal enzymes is not only an encouraging way to measure a functional effect in preclinical studies but may also have the potential to correlate to clinically relevant biomarkers for the program as well. And I show that here on the right side of the slide, now looking at brain samples from patients with progranulin-deficient FTD. And what's very encouraging is these patients show a similar increase in glucosylsphingosine levels in brain, arguing we may -- not only is this biology that we're looking at quite relevant to the clinical situation, but we may be able to use this biomarker as a metric of drug function in our clinical studies. So then moving downstream of this lysosomal dysfunction, what is the impact on disease-related phenotypes, and I'll show an example of a couple of those here today. The first is looking at lipofuscin accumulation that occurs in neurons both in FTD patients as well as in the granulin knockout mouse model. And that's shown in the sort of gray inclusions here that can be seen in the middle panel of the pictures. And these accumulate in age-dependent fashion and granulin knockout animals. But through either weekly or every other week treatment of these animals with our PTV progranulin construct, we were able to rescue these lipofuscin levels close back to -- back to close to wild-type levels, showing that within neurons, we're able to not only rescue the lysosomal defects but the downstream impact of these defects. Then let's move on to look at the microglial activation that's present in the brains of these animals. And as I mentioned in the introduction, this is also one of the hallmark pathologies present in FTD progranulin patients. And what you could see in these 3 images is now looking at IBA1, which is marking microglia; GFAP, marking active astrocytes; and CD68, which is the marker specific for activated microglia. In the granulin-deficient animals, you see significant gliosis, both microgliosis and astrogliosis, compared to the wild-type control animals on the left. However, like with the other phenotypes, dosing with our PTV progranulin molecule is able to result in significant rescue of both of these phenotypes, as shown in the panel on the right. And then I'll just include some quantification of that data here. Looking at IBA1 on the left, again, a complete rescue, a significant rescue of both CD68 and GFAP levels as well, it's suggesting that PTV is able to rescue these gliosis phenotypes that are present both in patients as well as in the granulin knockout animals. So just to wrap things up from here then, I wanted to bring it all together in one schematic that tries to capture all elements of how progranulin impacts disease biology and how different therapeutic approaches may affect this biology. So first, when we look at the healthy situation, we see normal levels of progranulin both in the blood compartment, on the left-hand side of the panel; as well as in the brain, on the right side. This results to normal lysosomal function in CNS cell types and a lack of any neuroinflammation. However, in the case of progranulin-deficient FTD, the reduced progranulin levels in these patients both in the periphery and in the CNS result in altered lysosomal function and eventually accumulation of lysosomal substrates that results in neuroinflammation, neurodegeneration and other phenotypes. Now there's multiple ways to approach this biology, and one that's become popular in the field is through blocking the internalization of progranulin and therefore increasing circulating levels of progranulin. And this can be done through blocking cell surface receptors such as sortilin. And it has been proven both preclinically and clinically that blocking sortilin does increase extracellular progranulin levels both in the blood and, to some extent, in the CNS as well. However, the potential concern with this approach is just by the fact that you're blocking the receptor and therefore increasing extracellular levels, you have the potential to impact the internalization of progranulin and the direction of progranulin to the lysosome where it's thought to be most functional. So we feel like our PTV progranulin approach is the most direct way to increase progranulin levels in the periphery, then using our TV technology facilitate uptake of this molecule from the periphery into the brain. And then conveniently through both sortilin receptors and some of the native progranulin receptors as well as through TfR, it facilitates uptake of this molecule and targeting to the lysosome where it can improve the function of lysosomal enzymes and rescue the lysosomal dysfunction present in these patients. So we're quite excited about this program based on its ability to enhance the uptake of peripherally administered progranulin and the preclinical data suggesting that we're able to rescue a range of lysosomal defects with this molecule. That -- in addition to that, we're able to rescue both microglial dysfunction and neurodegeneration in progranulin-deficient mice. And some of the mechanistic work we've done preclinically, we feel like it sets us up well to have a biomarker-driven development strategy based on being able to measure lysosomal function, gliosis and neuronal health using biomarkers that are translated from our preclinical studies. So overall, we feel like our TV-enhanced brain uptake is the most direct and effective way to increase progranulin in brain, and we're excited to advance this program into human studies in 2022. And so I will stop there and transition back over to Ryan to go over some of the near-term milestones.

Thank you, Joe. Appreciate the update on the programs. Let's now highlight some of the milestones that we should expect coming forward. I'll just focus first on our first -- our 2 programs, DNL310 and DNL151. So these 2 programs are the most advanced programs. Our focus is essentially initiating late-stage trials for both of these programs including to continue to collect data from our ongoing Phase I/II from the DNL310 or ETV:IDS program. Today, we focus on 3 programs: eIF2B or DNL343; the RIP kinase program; and PTV progranulin. In terms of next set of data for these programs, we were expecting Phase Ib top line results in mid-2022 for eIF2B, and we'll get into some of the questions and discussions around this program. RIPK program, we're initiating a Phase II study with Sanofi. Sanofi, of course, taking the lead on that study, and that's slated to begin at the beginning of 2022. And then we'll be filing an IND and/or CTA for DNL593 by the end of the year. Just highlight that also ATV:TREM2 is on its way to also filing an IND or CTA. Both of these programs are opt-in programs from our partner. Beyond that, we continue to expand our portfolio, especially our ETV portfolio, with a focus on building clinical manufacturing and commercial capabilities. We recently had Katie Peng join the team, and Katie will be leading the build-out of our commercial team around rare diseases. We continue to partner with our collaborators on Alzheimer's and Parkinson's and look forward to advancing those programs as well. And as you can see today, we're really focused, in the center here, on ALS and FTD, where we have both partnered programs but also wholly owned programs. So with that, I'm going to ask our presenters to come back, and we're going to go into Q&A. We have a number of questions here and look forward to addressing these questions.

Excellent. Welcome back. All right. So I'm going to start with a set of questions. And I think Gene and Joe, you're going to be -- I'm going to be asking you a few questions at the beginning here. And then Carole, definitely some questions on the clinical development as well. So the first question is what is an intrinsically disordered region. And I think this is actually a great question because it's really important when we think about the mechanism of these ALS mutations. So Gene, why don't you tell us what an intrinsically disordered region is? And Joe, maybe you can follow up with what you see the relevance of these regions are.

In the proteins that we study and also RNA-binding proteins, these are typically very glycine-rich regions in that -- that often are sort of floppy and unstructured. They enable a lot of protein-protein interactions with themselves or with other proteins with similar domains. And so they are intrinsically disordered only because they tend -- if you express them recombinantly, they will aggregate and they will then -- they separate from the rest of the constituents in a test tube. If there are probably better definitions for the -- or working definition for these, but in our minds, they are unstructured regions, yes.

And Joe, maybe the relevance of these intrinsically disordered regions in ALS.

Yes. And I think Gene hit on it really well, which is most RNA-binding proteins contain both an RNA-binding domain and one of these intrinsically disordered domains. And as Gene mentioned, they're important for protein-protein interactions. They are also required for stress granule formation, right? And so that's why most of these RNA-binding proteins have them. But as Gene alluded to briefly in his talk, most of the mutations that are associated with ALS also occur in these intrinsically disordered regions and make them more aggregation from, right? So it's some of the best data linking some of these RNA-binding proteins functionally to disease is really showing that when you have disease-associated mutations, your stress granule dynamics are changed. And that is really -- impacts that equilibrium that Gene mentioned in his talk.

Great. Thanks, Joe. So I think I love this next question because it's like the fundamental questions, I think, in most neurodegenerative diseases. So in the presence of TDP-43 inclusions in ALS and FTD patients, is it a driver of disease or a marker of the disease? Is it a driver of pathophysiology? Is it sufficient on its own? How far downstream or upstream in the disease pathology is it? And so I would love to answer this, but I think, Joe, maybe I'll hand it to you and just comment that TDP-43 itself can be mutated and, as a degenerative gene, can essentially cause neurodegeneration. But Joe, I'm going to hand it to you.

Yes. No, and I think that's the strongest evidence, Ryan, is that mutations in TDP-43 can cause disease. Now that's not -- it's only a small subset of patients. But when you look at the RNA-binding proteins in aggregate, mutations, again, in these disordered regions of multiple RNA-binding proteins are sufficient to cause disease. These are rare mutations with high penetrants, right? And so that really argues for a functional contribution to this. And I think there's growing work looking downstream of TDP-43 and showing a functional impact of disrupting TDP-43 biology through these aggregates on downstream biologies. And I don't know, maybe Gene can comment on that. He's done a lot of work in that area.

Yes. I mean I think there is a consensus in the field now that TDP-43 loss of function, because it's being sequestered and trapped in these granules or in the cytoplasm's inclusions, may lead to severe downstream consequence, right? So for example, I think Don Cleveland's and Kevin Eggan's manuscripts on stathmin-2 being a downstream target where the exon inclusion and this cryptic exon that leads to a loss of -- presence of the stathmin-2 protein, it's a -- it contributed to disease, right? And in our data and also consistent with your data, if you lose TDP-43 from the cells, either by knocking it down or showing that in knockout experiments; or in our case, by stressing the situation in the cells, TDP-43 is no longer binding the exon to prevent it from being suppressed, right? So it's no longer there, it's included and then cause the down-regulation of the stathmin-2. So I think it's -- we are leaning towards an eventual loss of function of the protein. Now whether it is the only driver of the disease, I don't think that's really clear yet. And is it the -- is this something upstream, and I think that evidence in the field will probably show up over time to indicate there are other upstream misregulated events. Yes.

Yes. Maybe I'll just add a simple point, which is it's a question of necessity and sufficiency. And definitely, TDP-43 is sufficient, and I think the genetics implies that. In the context of patients that don't have TDP-43 mutations, the question is, is it necessary? And I think the fact that it's 95% of ALS patients have inclusions suggests that it's playing a role and it's certainly sufficient on its own. Okay. So what is the evidence that breaking up stress granules can actually reverse disease pathology in ALS and FTD? And maybe, Gene, I'll start with you. And Joe, you can comment.

Right. So there have been experiments in the field that have shown that if you alter some of the key proteins that are implicated in stress granule information and stability, you can see reversal of phenotypes in cells and in mice models, right? And one of those experiments have come from the ataxin 2 depletions from Aaron Gitler's work. Ataxin 2 is an important stress granule protein or at least part of this stress granule interaction. And there's been also work showing that you can create stress granules using OptoGranules or technologies which with light induction creates these aggregates. And that can cause [ phenotype ], and then you can turn off the light and let these things dissolve, and then the phenotypes go away. So these are good in vitro models that suggest that these have consequence and removing them can have reversion to a normal state.

Yes. Great. Joe, any thoughts on stress granules around reversing pathology and its association with cell death in particular?

Yes. No, and I think that the -- Gene described it very well. There's data from a number of labs suggesting that if you can't dissociate stress granules, you'll have toxicity, but then being able to dissociate them rescues that toxicity. I think he gave a couple of good examples. Paul Taylor has done some excellent work in this area as well. And then I'll also comment that the ISR itself, so the chronic activation of the ISR will lead -- is sufficient to lead to neuronal apoptosis. And this has been seen in experimental situations where even if you look at stress granules or not, and we've published a paper not so long ago showing even in acute neuronal injury, that inhibition of the ISR is able to protect neurons from degeneration. So it does seem to be a broadly relevant pathway in controlling neuronal survival.

Which is actually in one of our early papers a couple of years ago now, right, and we showed that some of these cleaner compounds, which are 2 compounds, do seem to disrupt stress granule stability and recovered TDP-43's localization back to its normal state. And in cells' mutations, we treat these cells, these cells survive longer and actually have much decreased cell death, right? So I mean these are, again, maybe not completely direct experiments, but they do point to SGs being an interesting therapeutic modality that can be corrected, yes.

I mean I think the ultimate proof will come in our treatment study is where we're looking at disease patient survival and various biomarkers.

That's right. Yes.

So which is more important, reversing preformed stress granules or the inclusion of TDP-43 or both? Maybe, Joe, I'll have you answer this one. I think it's going to be a tough question to answer, but I think the question is really around the relevance of stress granule or TDP-43 or both or are they just intimately linked.

Yes. I mean -- and Gene has done some work on this in terms of -- one answer to this question is when does TDP-43 appear in stress granules, right? And I'll let Gene comment in more detail, but the data suggests it's not a core stress granule component, but you have chronic stress granules and then you get TDP-43 accumulation, right?

That's right, yes.

And before I let Gene kind of go into more detail there, I will say, I mean, I think we do also have to think about ALS on sort of a like cell-by-cell level, right, where it's not like you get TDP-43 pathology in your entire nervous system all at one time. I mean it's well known that ALS move spatially through the nervous system during the progression of a patient. And so even if it's already existing in one area of the CNS, it's likely still forming in other areas. And so I think that there -- regardless of what stage you intervene, I think that there's potential for benefit in terms of progression in these patients based on that mechanism.

Yes. When we treat iPS-derived motor neurons with a long-term chronic stressor, what we see is that we have stress granules that form. And then TDP-43 begins associating these stress granules. So it's not a core stress granule partner but associates. And then in wild-type cells, when you remove the stressor, TDP-43 recovers and leads. But in cells which are from ALS patients, they don't recover. And this stickiness, I guess, is retained, right? And so we do think that if you can find ways to put the interaction between stress granules and TDP-43 accumulation, that will enable stress granules to -- or TDP-43, sorry, to then depart and return to its normal state. So I think there's different ways to think about that problem. I think resolution is, I think, a key part of how we're thinking about from a therapeutic opportunity, yes.

Great. Okay. Now a series of questions on DNL343. I think, Joe, probably you and I on the mechanism can cover these. So do you have any evidence that DNL343 is effective at targeting inclusions containing phosphorylated or ubiquitinated TDP-43?

Yes. It's a great question. And I think the reason for asking, I'm assuming, is because phosphorylated or ubiquitinated TDP-43 is associated with more pathological granules, right? And the answer is we've done some work there. It's been hard to generate phosphorylated TDP-43 or ubiquitinated TDP-43 that we really believe in some of our cell models. And so we're attempting that, but it's something that we haven't looked at as much as we could because of technical reasons at this point.

Right. We'll do some rapid fire. We have a large number of questions here. So can DNL343 prevent the formation of stress granules? If so, is there a path of development for 343 in early stages of the disease to prevent progression? And I think I'd just comment that, yes, it can prevent stress granule information. What we show is a higher bar, which is dissolving pre-existing stress granules. So the answer is both. Can DNL343 activate all types of eIF2B alleles? Or is there a specific mutation for which DNL343 doesn't work? And obviously, we showed in the mouse model one mutation in eIF2B in which it obviously works robustly. Most of the time, we're targeting wild-type eIF2B, so I don't know, Joe, if you've done any other work on some of these very rare mutations of eIF2B beyond the R191H.

What I can say is it should -- it would be predictive to work on all disease-related alleles. We've tested some of them in the lab. But others, based on the prediction of the mechanism of how it works, we'd expect it to be equally effective in all cases. There are specific mutants that we engineered with eIF2B that disrupt the binding of our molecule, and so we can identify what residues are key for its activity, but none of those are the same as the disease-associated mutations. And for all of those, we see equal activity in the ones we've addressed so far and can predict for others as well.

Okay. Does DNL343 mimic a naturally protective mutation in eIF2A? So I'm reading it exactly, so I want to make sure that I get this correctly. So the question is around are there naturally occurring protective mutations. Does DNL343 mimic what would be a naturally occurring protective mutation?

So I guess what I could say exactly is we've never done that exact experiment, but theoretically, it would be very similar. So if you were to take the phosphorylation site in eIF2-alpha, that's phosphorylated upon stress. And you mutated that, so it couldn't be phosphorylated. That would probably act pretty similar to our compounds. Again, that's -- that mutant has been published in some groups. We haven't done a side-by-side comparison. But the prediction would be the case. And what's important mechanistically there, if we dive a little bit deeper into that question, is -- and you spoke about it a little bit, Ryan, as we call the molecule an eIF2B activator because it's a convenient way to talk about it, right? But most accurately, it's blocking the repression that it gets upon stress, right? So when -- in a healthy cell, the molecule is essentially inert, right? It doesn't further activate eIF2B upon more-than-normal levels. But when you have this stress-induced repression of the system, the molecule is able to block that stress-induced repression and bring it back up to normal levels. And so I think that's important and can get to some of the safety-related elements of the molecule itself as well.

And that's a perfect -- that actually answers the next question. Is there any theoretical risk of over-activated eIF2B? And I think I'll just highlight what Joe said is that it's not your traditional activator, right? It's essentially restoring normal function. When function is normal, you're obviously not overactivated. And so I think you nailed that one. So Carole, I'm going to turn it to you on some development questions. So are there any on-target expected AEs?

So based on our GLP studies in both mouse and nonhuman primate, our definitive studies, we did not have expectations of adverse events that we were looking for in a Phase I study. But I think notably, given that this is only 1 of 2 yet to be activators that have gone into the clinic and there's been no published safety data, we absolutely were looking in our single ascending dose and multiple ascending dose for any adverse events of interest, which, as noted, this was very well tolerated. And we don't have any adverse events that we clearly feel are mechanistically related.

It looked like the PD data showed -- for 343 in healthy volunteers showed that the doses were essentially overlapping in the pharmacodynamics. How do we think about it? Or how do you think about the high and low dose for Phase Ib relative to levels of MD1 through 7 have shown in the data we just presented?

Yes, great question. And we do, when we look at single ascending dose data particularly at the lowest dose, see evidence of dose response. And if you look at the exposure response, we see a relationship between exposure and response. But I think essentially what we're seeing with those multi-dose levels is that we're nearly maximally inhibiting the peripheral ISR response even with the lower dose. So I think that the take-home for me from that data really is that it has a fairly wide therapeutic window. And so looking at the Phase Ib study in ALS participants, we'll study 2-dose levels that allow us a broader window for safety if there is any safety issues that should emerge but then also allow us to look at other endpoints -- exploratory endpoints in that study, that would be with different dose levels.

Thanks, Carole. So why is the Phase Ib ALS study placebo controlled?

Yes. So -- and I think this is always a question particularly because we're very cognizant about participants with ALS enrolling in a study and the lack of a desire to be on randomized placebo. However, given our biomarker-driven approach and collection of CSF looking at biomarker changes, it really is essential for looking at these changes that you have a control given that actually the instrumentation of a lumbar puncture, as you can imagine, does create inflammatory cytokines and other changes in biomarkers that are very difficult to interpret without a control. And then I think with any study having particularly, for the first time, a mechanism is going into a population, having a placebo for control and safety issues or cohort-related clinical issues is really important.

Great. All right. Series of questions, let's cover each one. And I think, Carole, probably you and I can cover these. What short-term data would come out of a Phase Ib that could give you significant insight?

Yes, so the key endpoints that we're looking at are the markers of the integrated stress response pathway and demonstrating that in a patient population where the integrated stress response pathway is actually, naturally, we assume to be elevated, whether we can see that we're still able to reduce that integrated stress response pathway as we have in the healthy volunteer studies. In addition to that, we will be looking at exploratory biomarkers of neurodegeneration and other biomarkers to give additional confidence that we're having an effect on the disease pathology. Now some of that may come after the longer-term extensions. So you noticed that part 1 is only 28 days, which enables us to look at those pathway biomarkers. But for some of the exploratory neurodegeneration biomarkers, for example, we would be looking at those in the open-label extension portion.

Yes. I think you answered the second question essentially, really. So can you measure ISR genes or stress granules in CSF or elsewhere in ALS patients after treatment with 343? What time period would be necessary to measure significant changes? And Carole, you can comment here. I'll comment as well.

Yes. So we can't really measure the ISR response directly in CSF. And that's why we have done a very rigorous comparison of the effects of inhibiting the integrated stress response pathway and effects in brain in animal models. So we're really using the animal model to demonstrate to ourselves and to you that the exposures are very much correlated with inhibition of the integrated stress response pathway that's really independent of tissue. And so for us, when we can demonstrate that we have good exposure in the CSF, we can then use that relationship to predict brain target and pathway engagement.

I'll just add to that, that we have work on proprietary biomarkers and CSF. And I think to be fair, this is a highly competitive program. We're showing very robust data in both the animals but in particular the Phase I data and target engagement. But we'll continue to invest heavily in these biomarkers and I would just say stay tuned. We have to consider that it's a long development path with a lot of competitors. In fact, along those lines, the next question is how do you compare and contrast your eIF2B agonists against your competitor also in Phase I. And there will be -- obviously, there will be a desire to compare in due course. Are there differences in the mechanism or the approach? And I'm just going to assume that our competitor we're talking about here also has eIF2B activator. And we haven't gone into great detail or actual structure of our compounded, likewise, unaware of the structure of their compound and their mechanism, but we're guessing that it's going to be highly competitive with a similar mechanism. And so again, I think this gets back to the sort of biomarker answer, which is there's going to be some data that's going to remain confidential to some period of time as we advance our program. Okay. So next question here, so for the 343 Phase Ib, you'll try to focus on sporadic ALS patients. The question, are you completely agnostic of the genetic driver of disease? How do we view the population for the study? Carole?

Yes, so in this case, we are looking at sporadic ALS. And I think that largely comes from the fact that the mechanism, more than 95% of ALS patients have TDP-43 inclusions that we think are linked to the stress granules that Gene and Ryan described today. And so we don't feel that patient selection from a genetic perspective makes sense for this target. We are looking at -- in terms of as we think about future development and selecting for patients and in terms of progression rate and other clinical features, that may be important in seeing effects on clinical outcomes.

Okay. Great. So any reason to think there may be more or less activity in certain ALS phenotypes such as bulbar or onset? And how do you -- and how about fast versus slow progressors? Also, how does the patient focus of the HEALEY program impact your trial? And do you apply to be involved in that program? So really around patient phenotyping and then a separate question around the HEALEY program, Carole.

Yes. So I don't know that we have specific scientific data that would tell us about responsiveness in some of these subtypes of ALS. And in this study just in terms of looking at a more homogeneous patient population, we are looking at people that have a diagnosis of ALS and are within a particular time frame, within 3 years of onset of ALS to try to have a slightly more homogeneous population. In terms of the question about the HEALEY platform trial, a great example of, I think, a collaboration with academia and industry across a number of programs, we did not apply for that study. And I think largely, our biomarker-driven development and the approach that we're taking may not easily fit into that structure of that trial at the current time.

Great. Okay. So for 343 Phase Ib, let's see. Looking down the line, where do you think is an appropriate place for 343 in the treatment paradigm of ALS and FTD? So this is in the future here, Carole.

Yes, I think as always, we want to start with treatment of patients that have definitive diagnosis of the disease based on current clinical criteria for the clinical trial. But we would see this being used as early as possible in terms of the diagnostic paradigm and actually also could be used in combination with other therapeutic agents for ALS. So I think just in terms of the ability to combine with other mechanisms, this may be very helpful. And so we don't see this as being something that would be mutually exclusive of other therapies.

Okay. We're going to switch our attention to progranulin for a minute here, and I'm sure we'll get back to eIF2B. So can you comment on the intra, extracellular progranulin and its impact on disease? And I think I'll just add a second question to that, Joe, as you're preparing. Given the other progranulin-targeted agents in development, what do you see as the most critical points of differentiation for DNL593?

Yes. Great question. And maybe to hit the first one in terms of intracellular versus extracellular, some of the best data to describe that, in my view, is in our published work but I didn't get a chance to present today, is showing that the half-life of our PTV progranulin molecule in a mouse is on the order of hours, yet the pharmacodynamic response is on the order of weeks. And what this really suggests is that it's not the circulating progranulin that results in the biological activity, but it's those individual granulins present in lysosome. And so you can do this both in a mouse model or in a cell, where you add progranulin and you can get a lasting effect just by that lysosomal component. And I think that argues to us that the major function or the predominant function of progranulin is actually impacting the lysosome rather than the extracellular progranulin and the interaction with cell surface receptors. And so I think that, that's a core element of the answer there from a biology perspective. In terms of how we differentiate from our competitors, I think it comes down to the elements of the molecule itself, the first of which being effective CNS uptake in biodistribution. I think that, that's critical. It does seem like the progranulin -- the most of the progranulin in the CNS is likely to be local such that getting progranulin to the right place is priority #1. Second is that we're really directly addressing what the genetics tells us about the biology through replacement of progranulin. Targeting other mechanisms that indirectly change progranulin levels runs the risk that you're impacting elements of biology that you are independent of progranulin as well. And I think because we have confidence that we're not doing that with this molecule, we feel like it's the most direct way to bring progranulin to the right place in brain. And in the long run, that's going to differentiate us in the clinical level as well.

Great. Thanks, Joe. And I think related to that, can you comment on the cell biology data suggesting that sortilin is not a key receptor for functional progranulin uptake into the lysosome? Do you find it robust? Have you been able to repeat these experiments?

Yes. It's a great question. And so we have done some of that. I think the data is most consistent with sortilin is important for uptake of progranulin to the lysosome, but it is not the only way progranulin can get into the lysosome. I think there are other mechanisms as well. I do think it differs depending on the cell type. I think certain cell types are more sortilin dependent than other cell types. And so when thinking about modulating that pathway, one has to think about really what's the impact on neurons and microglia and astrocytes, in those key cell types within brain. So I do think there's multiple mechanisms, but I do think sortilin is a major contributor. And actually, the preclinical and clinical data bear that out, where if you weren't blocking uptake of progranulin and through sortilin, I wouldn't expect to see that extracellular progranulin increase.

Okay. So we're down to 5 minutes. This is where we do the rapid fire, and we have at least -- and we have a bunch more questions. So let's see if we can rapid fire through some of these. So have you thought about developing vanishing white matter as a primary indication for eIF2B activator?

Yes, this is something that we're obviously very interested in. And we've engaged with the managing white matter disease community as well around this, and we're currently continuing to evaluate this.

Okay. So for DNL343 and DNL788, they're expected to address an overlapping ALS population? Any differences between patient inclusion criteria for those rolled in the DNL Phase Ib study versus the Sanofi/Denali Phase II study?

Yes. So the Sanofi/Denali study enrolls slightly less severe patients in the sense that they were diagnosed within 2 years as opposed to 3 years. I think just notably, the endpoint is very different where the Sanofi/Denali RIPK study is really looking at a clinical endpoint. And that study, that fairly large Phase II study, was designed also based on the fact that we have had additional clinical experience with that target across both AD and ALS in a Phase Ib study.

Do you have any data indicating the nuclear function of TDP-43 is restored by treatment with DNL343? Joe? We're just going around, yes. We're just going for it.

No, that's good. We haven't looked at that directly. But I think as Gene mentioned, the one thing that we do see is that when we block the granule formation that includes TDP-43, we restore normal TDP-43 localization. So we haven't looked at downstream genes like stathmin, the ones that Gene mentioned, yet. But based on that cell biology, one would predict normal function would be restored. And I do think that those have a lot of potential as potential biomarkers down the line for measuring this pathway.

Okay. We have a goal to answer 8 questions in 2 minutes. Are there any biomarkers of ISR dysfunction that can distinguish between cells and acute stress versus chronic stress? This is a great, great question.

That's a really great question. And I'm supposed to be fast. So what I'll say is we don't know yet. But what I will say is the signature of the ISR differs between cell types as well and between the types of stress. And so it's just going to take some effort to make sure we identify the right pattern there, but I'm hopeful that we can find something.

Yes. This is a good question, and I don't know if we can answer it. But are stress granules present in presymptomatic ALS, FTD patients? How closely does their appearance and their quantity track with onset of symptoms and disease progression? So Gene, is there any evidence that stress granules appear in presymptomatic? I don't know how one assesses that.

I would argue stress granules should appear in most cell types whether or not you have disease, right, because they happen when you have a fever or a virus infection. So these are -- these should appear before. I don't think anyone's actually tracked in presymptomatic tissue or patients whether they are appearing or disappearing. I mean that's very hard to do.

Yes. But I think the point -- I think you're exactly right. The point is that it's the propensity of these stress granules to essentially become persistent, right, that is linked to disease. Okay. Will ISR restoration be sufficient to reverse ALS progression? Or should it be combined with other approaches? Do ALS patients with different subtypes, phenotypes have different degrees of dependency on the ISR? Can you talk about the expression of the eIF2B in cells outside the CNS? Any concerns about overactivation of the eIF2B? Look like these are 4 questions all in one, and maybe I can just comment. We've already discussed overactivation. Activation is expressed in all cells, right? And in healthy cells, it's essentially normal healthy levels, and we're restoring those broadly. For whatever reason, CNS cells are most susceptible to degeneration probably because of their long-lived cells. So I think the first question is really the one. So will ISR restoration be sufficient to reverse ALS progression or should it be combined with other approaches? Who wants to speculate on that?

I think my view is that we want to start to see efficacy in the disease area. And then if we're in the luxury of being able to combine, there's no reason we shouldn't or wouldn't. I think we've seen an oncology combination therapy can be very efficacious.

Okay. So I think we're almost at time. I want to answer one more. So what is the baseline stage of ALS should be enrolling in the Phase Ib trial for eIF2B, Carole?

So those are patients that are within 3 years of diagnosis.

How do you think about timing of intervention with regard to disease severity is all linked together?

I think there, I'll give the more generic example in neurodegeneration, but I think we want to go as early as possible but also need to be able to recruit the trial and show generalizability of the findings that we see in a trial to a broader patient population.

We have a number of other questions but definitely enjoyed the engagement from everyone. Fantastic questions. And we're very excited today to share some new data on our programs as our portfolio continues to advance. And with that, we thank everyone for joining.

Thank you.

Thank you very much. Thanks, Gene, for joining us.

Yes, thank you.

No problem. Thank you.