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

Hi. I'm Yuko Oku, one of the associates on the biotech team at JPMorgan. It's a pleasure to welcome Denali to the JPMorgan Conference. A quick housekeeping note before we begin, we'll be holding a Q&A session right after the presentation in the Georgian room just across the hall. With that, I'd like to turn it over to CEO, Ryan Watts.

All right. It's exciting to be here again at JPMorgan, especially exciting today because we're presenting new data. A little bit unusual, we're going to go into some depth. I love data, we love data. We're going to share some clinical data, plus progress in the portfolio. But I'd like to start with a reminder, which is the huge unmet need we're trying to tackle. We've seen incredible progress in oncology, rare disease, obviously, new modalities. Many of us, including myself, have worked in the oncology space, and now we're turning our attention to neurodegeneration. And the way we view this is using the molecular genetics to identify pathways and targets that we think are tractable. Going after subpopulations, and then expanding it to the broader populations, as you see in Parkinson's and Alzheimer's. So today, I'll focus on 2 therapeutic areas, primarily Parkinson's disease, and give you an update on our LRRK2 program, which we just announced this morning; the completion of our Phase Ib with positive biomarker data; as well as progress on both molecules, this is DNL201 and 151. I'll also share with you data on our Hunter syndrome program and our blood-brain barrier technology to get enzymes and antibodies and proteins across a blood-brain barrier and introduce several new programs using this technology. So it was actually here 5 years ago that the idea of Denali came about. We launched the company in May of 2015. It's been an incredible 4.5 years. As you can see here, these are some of the filings that we've had in the last 4.5 years and including clinical data. So today, again, the focus will be on our ETV:IDS program for Hunter syndrome as well as the LRRK2 program, DNL201 and 151. All of this work has been enabled through partnerships. We have a number of partnerships, enabling partnerships plus strategic partnerships. And it's great to see many of you here today, as part of our collaborators, trying to solve these big problems in neurodegenerative disease. So the company is built around 2 platforms, the biology platform, which we call the degenogene platform, using human genetics to define pathways and targets; and then the blood-brain barrier technology platform. Here, we've been -- we've seen success in engineering small molecules to cross the blood-brain barrier. We, in fact, just have our third small molecule program enter clinical testing, an eIF2B activator. But in addition to this, we have engineered proteins to cross the blood-brain barrier. And then ultimately, AAV, in which we announced a deal last year with SIRION. And there, we're focusing on engineering AAV capsids to either express proteins across the blood-brain barrier or AAV itself crossing the blood-brain barrier. So these are the 2 platforms that the company is built around, the degenogene biology platform with over 100 scientists focusing on the biology as well as engineering molecules to cross the blood-brain barrier and access the central nervous system. Here's our portfolio. Just a couple of highlights here. We're introducing a few new programs. We have a program known as PTV progranulin, which is protein transport vehicle. It's a progranul infusion. I'll share data at the end of this presentation in which we can rescue CNS phenotypes, rescue lysosomal defects with PTV progranulin. We also now have 5 clinical stage programs, 4 of which are focused on the CNS and 1 last year, which entered clinical testing with our partner, Sanofi, focused on peripheral inflammatory disease. So you can see that the portfolio is progressing. It was a busy year-end with 2 filings and also reading out clinical data. So this is what we've achieved. And I think, by and large, we've achieved almost everything we set out plus more. Last year, when we presented, again, the completion of the LRRK2 Phase Ib study as well as reading out results from the Phase I study for DNL151. In terms of ETV:IDS, we've filed, and I'll show a clinical -- or I'll show you preclinical data, rescue of neurodegeneration with this molecule. eIF2B, we haven't talked a lot about eIF2B. It's integrated stress response activator. A way of basically blocking the integrated stress response through activating eIF2B. And that -- in fact, the CTA has been approved, and we'll begin clinical testing early this year. And then finally, we've actually completed the enrollment of the ALS and Alzheimer's study but have an open-label extension for RIP kinase. And there, with our partner Sanofi, we expect data the first half of this year and make a decision by mid-2020 on where we go with the RIP kinase program. Okay. So let's dive into the data. Let's start first with the LRRK2 program and where we are, and this is important. Again, we announced this morning positive results from this clinical study. I'll remind you of the biology. So LRRK2 is a leucine-rich repeat kinase 2. Mutations in LRRK2 increase risk of both Parkinson's disease and Crohn's disease. So there's both a systemic disease as well as a CNS disease. These mutations are kinase hyper-activating. So they increase pathway activity. And what we've shown and others have shown is that when LRRK2 is hyper-activated, you get lysosomal dysfunction. In fact, it's probably on the spectrum of lysosomal storage diseases where extreme lysosomal dysfunction leads to lysosomal storage disease. And moderate or modest lysosomal dysfunction, causes Parkinson's disease. We also learned this from GBA mutations in Parkinson's disease. And I think a very important point, there's a broad set of human genetics now, including protective variance as well as causative variance, that basically link Parkinson's disease broadly in lysosomal dysfunction in the case of sporadic Parkinson's disease. And in fact, what you see is region-specific activation of LRRK2 in sporadic Parkinson's disease. Now there are a number of studies looking at basically this pathway being activated in cells that are degenerating. Now I think it's a key point here. This region specificity makes it a little bit more difficult to read out biomarkers, but that's basically where we're going with our program is understanding the role of LRRK2 in both sporadic as well as mutation carriers. So just a little bit of background on the program itself. We've taken 2 molecules into clinical testing. These are chemically distinct molecules, with different PK profiles, different off-target profiles. And the goal is to basically advance both of these programs through patient studies, biomarker studies and identify the molecule that has the legs to go forward into Phase II/III. Happy to report today that both molecules have the leg. They still -- they have unique profiles, which I'll share. And a reminder that we ran 28-day studies focused on biomarkers of Parkinson's disease and lysosomal dysfunction. Ultimately, target engagement, pathway engagement and I think what we're really looking for are changes in lysosomal biomarkers, okay? So that's the summary, and here's the data. So what I'm showing here is target engagement, pathway engagement as well as lysosomal function changes in both the low dose and the high dose. We set the low dose to the dose that we imagine that we would go forward with into later-stage studies and then we pushed the high dose, as you would expect. And consistent with our goals, we see that the low dose has greater than 50% inhibition throughout the dosing interval. I'm showing you here day 1, so this is after a single dose. And then 28 days after -- this is the last dose after 28 days of dosing. You can see sustained inhibition when looking at LRRK2, phospho LRRK2 as well as pathway engagement, looking at phospho-Rab, which was -- which is one of the downstream effectors, both of which show robust inhibition. And then importantly, we see lysosomal function correction. We see improvement in lysosomal function with inhibiting LRRK2, both at the low and the high dose, as you can see here. Now we see this in both sporadic as well as mutation carriers. And we decided to do 1 additional experiment. We have a number of ongoing biomarker studies. Recall, we've read the study out at the end of the year, and so here, I'm just showing you some of the top line data. And we ran this, I think, very interesting experiment in which we looked at biomarkers that are differentiated between mutation carriers and sporadic. So just selecting for hyper-activated biomarkers in the mutation carrier population. And what you can see here, basically, there are a set of biomarkers. Now interestingly, these are all -- majority of them are lysosomal in nature, basically lysosomal lipids. And here, I picked 3 of these biomarkers. And you can see in a dose-dependent way, albeit we have 9 patients total that are mutation carrier, and maybe a point here, this is the first experiment run in LRRK2 mutation carriers ever with a chemical inhibitor, and we can see a rescue of these lysosomal biomarkers, bringing them back to normal levels. Now when we look at the broader sporadics, we don't see this change. And recall, we're selecting just for the sporadic -- just for the LRRK2 carrier biomarker changes. Now we're going back and looking at changes for both sporadic as well as mutation carriers. So this is exciting data, showing target engagement, pathway engagement and improvement in lysosomal biomarkers with LRRK2. Now that's not the whole story. Here, I'm showing data for DNL151. This molecule has -- is basically about 6 months behind 201. We've considered them co-leads because they both have unique properties. And what you can see here is robust target engagement, pathway engagement, and then similar to what I showed you with 201, improvement in lysosomal function when looking at year-end BMP, which is, again, a lipid biomarker of lysosomal function. So what does this mean for the program? And let me tell you around -- about the safety profile for each of these molecules. So for 201, similar to what we saw in the Phase I, at the low dose, it's generally well tolerated. Most common AEs were headache and nausea. However, in the high dose, we had a higher incident of moderate adverse events. But again, this is headache and nausea. And I think the take-home message here is that they're manageable and reversible. And so we can take the low dose forward, and we're working on the high dose and in particular around a slow-release formulation. We think this is related likely to the off-target effects of DNL201, which we've reported in the Phase I study, which is PDE off-target effects. In terms of 151, it was basically generally well tolerated. Majority of AEs were mild, no SAEs, no discontinuations. And in fact, what we've done is we've modified the protocol. We've done all the dosing we would like to do in the Phase I, but we've added higher doses just to explore that window. And in terms of the Phase Ib, the same, we've enrolled the original study, and we added another arm at higher doses, basically exploring the window at DNL151. So in terms of next steps, we have 2 molecules that could be ready to go into Phase II/III. Certainly, DNL201, with the totality of the data, we could go forward, especially at the low dose. However, we're continuing with 151. And where we're spending most of our time is preparing for the II/III by engaging this global network of collaborators. Centogene in terms of genotyping, bringing in LRRK2 carriers. We launched a website, engaged Parkinson's to bring patients basically into these studies. And we expect by midyear, we'll have finalized our decision between 201 and 151. In terms of 151, basically now looking at the patient data, and in terms of 201, doing further biomarker analysis, similar to what I've shown you around these lysosomal biomarkers. We also have developed a slow-release formulation for 201, which we think will manage the headaches and nausea and, as a result, sort of blunt the Cmax and give us a great profile and target inhibition. Okay. So that's it on LRRK2. We look forward to answering questions after this presentation. I'm now going to turn my attention to our blood-brain barrier platform and the filing, recent IND filing for the ETV:IDS program for the transport vehicle. So just a reminder that the blood-brain barrier represents a major obstacle in treating neurological diseases and specifically neurodegenerative diseases. There has been a belief that the blood-brain barrier is disrupted in neurodegenerative diseases, but there's a lot of data that suggests that's, in fact, not the case. You don't have over disruption of the blood-brain barrier. And so what we set out to do with Denali 5 years ago is basically to build platforms and approaches to engineer crossing the blood-brain barrier. And one area in particular where this is needed is for antibodies, and we're all familiar, of course, with aducanumab and other Abeta-related antibodies and the exposure that these molecules get in the brain. And what you see is about a steady state ratio of 1,000 antibodies in blood to 1 in brain. And so using the transferrin receptor technology, we've built our own what's called transport vehicle, which is engineering the Fc region of an antibody to bind to this blood-barrier receptor to get increased uptake into the brain. And we've now applied this. I'm actually going to show examples for each of these. We've applied this technology to enzymes, antibodies, proteins and ultimately to ASOs. So I'd love to share with you data with ASOs. Those studies look very promising, and those are ongoing. But let's start first with the enzyme technology. So I'll use this molecular animation to describe the technology. So in dark orange is the Fc region of an antibody fused here to iduronate 2-sulfatase, or IDS, for Hunter syndrome. This Fc region, a portion of it, 1 foot of the Fc, we've engineered binding to the transferrin receptor, which is highly expressed on the vasculature in the brain. When molecules like antibodies or proteins and Elaprase, for example, for IDS is -- are injected systemically. Very little -- very little of drug gets across the blood-brain barrier. You get a, sort of, nonspecific uptake. The idea here then is that you can latch on to these endogenous receptors that are transcytosed across the vessel wall and brought up into the CNS. I'm going to show you a series of experiments in which we show that this is the case. This transport is constitutive, it's not disease related. We've looked at Alzheimer's, Parkinson's, various models as well as in human tissue, and we see that transferrin receptor is highly expressed and doesn't diminish with age or with disease. So the first test of this hypothesis, which we're entering clinical trials this year, is for Hunter syndrome. And here we have the enzyme transport vehicle fused to IDS. And just a reminder that these lysosomal storage diseases are monogenic diseases. There are approved therapies that treat the systemic disease but not the central disease. And similar to what we're seeing with LRRK2 in many of our programs, the heart of the biology is basically lysosomal dysfunction. In this case, it's robust lysosomal dysfunction. So here's one of my favorite experiments in which we've engineered mice to have the human transferrin receptor, crossed them to the Hunter syndrome mice and compared head-to-head with Elaprase. And what you can see on the left graft is basically liver GAGs are rescued equally with Elaprase and ETV:IDS. So the goal here is to basically replace Elaprase. However, in the middle graft, you can see these GAGs, which are heparin sulfate and dermatan sulfate. They're basically the substrate for IDS, accumulate in brain, Elaprase does not reduce base GAGs, however, ETV:IDS robustly inhibits or reduces GAGs. We also see a subsequent rescue of lysosomal function back to baseline using BMP, similar to what we used as a biomarker in the LRRK2 study, okay? Now the next experiment is a chronic dosing experiment in which we can look at a marker of neurodegeneration known as neurofilament light. Neurofilament is up-regulated when cells die. And similar to what you see in broad neurodegenerative diseases, in lysosomal storage diseases, about 70% of patients develop neurodegeneration. And you can use a biomarker for neurodegeneration, such as neurofilament. And here, we see in this long-term dosing study, 80% to 90% reduction in GAGs and complete rescue of neurodegeneration. So obviously, this is the ultimate goal of our ETV:IDS program. Now there are other approaches where we're seeing competitively about 30% reduction in GAGs. We've set the bar at, at least 50% reduction after short-term dosing. And we expect to see, after long-term dosing, GAGs brought down about 80%, 90%, and complete rescue of lysosomal dysfunction as well as block of neurodegeneration. So where are we? We've filed the IND. We're now preparing for clinical studies, and our goal is basically to have an interim readout this year. Now we hope that we don't need many dose escalations. We expect that some -- our initial doses may be efficacious in terms of reducing GAGs. That being said, it's a rare population, a handful of patients. We should have a clear insight if this molecule is working later this year. This also reads through broadly to the transport vehicle technology, which we're applying to other modalities, and that's what I'd like to share with you next. So first is the antibody transport vehicle. Now we recognize in these rare diseases, there's a huge unmet need. However, if we could apply this technology more broadly to diseases like Alzheimer's disease, we could see broad benefit. And I'll give one example here, which is our antibody transport vehicle. And a recent paper that we published in Neuron, highlighting the role of a receptor known as TREM2. TREM2 loss of function increases risk for Alzheimer's disease. And in this particular set of experiments, we showed that loss of function basically in microglial cells gives these sells an inability to turn over lipids, again, a lysosomal dysfunction, as many of these diseases, and we show it here. So when you treat mice with cuprizone, you get degeneration of myelin. These microglial cells engulf the myelin. And then you have the TREM2 loss of function, you have an inability to turn over these proteins. And what we see, sort of, remarkably compared to a standard anti-TREM2 antibody, after a single dose, we can robustly activate TREM2. In fact, in this case, the ATV TREM2 is given out 1/5 the dose of a standard antibody. And even at this dose, we see no effect with the standard antibody. However, ATV TREM2 activates this cholesterol metabolizing enzyme. So our expectation here is that we basically can rescue this microglial dysfunction broadly in Alzheimer's disease. So this program is advancing towards clinical studies as well. Now in addition to fusing enzymes and creating antibodies using the transport vehicle, we can also fuse this to various therapeutic proteins. And here, I'll show for the first time one example of that, which is PTV progranulin. So progranulin loss of function causes FTD. And the goal here is basically to restore granulin functions in the brain and rescue lysosomal defects. Now interestingly, here, when you have progranulin loss of function, you see a drop in BMP. And our goal here is basically to rescue this back to normal levels or normalization of lysosomal function. And what you can see is if we take progranulin fused to an Fc versus progranulin fused to a PTV or the transport vehicle to get across the blood-brain barrier, we rescue lysosomal function in brain whereas both of them rescue in liver. So you can see a brain-specific rescue using PTV progranulin. So in summary, we've now made progress with our TV platform, moving towards the clinic. We're looking forward to clinical data later this year validating the platform. We are applying this already to antibodies, to other proteins. And as I mentioned at the beginning, we have initial data looking at ASOs fused to the transport vehicle that looks very promising, and we plan to expand our efforts here using basically a systemic injection of what we call OTVs or oligonucleotide-fused transport vehicles to basically knock down gene expression in the brain. And with that, let me highlight what we plan to do in the coming years. So first is basically to continue preparation for the Phase II/III trial for LRRK2. We'll finalize our molecule mid-2020, as both molecules have the legs now. In terms of ETV:IDS, we'll begin dosing and look immediately for biomarker data. We plan to have this data at the latest at the end of 2020. And in terms of eIF2B, we will initiate a healthy volunteer study now. That CTA has been approved, and I look forward in future conferences to share the biology around eIF2B activators. And then finally, results from the ALS and Alzheimer's study. As both of those studies are now enrolled, we expect those results as well as the open-label for the RIP kinase program. And I'll just mention that we've gone very broad with the BBB platform, applying it to various modalities, and we look forward to seeing that advance in the future. And with that, I thank you.

You're at the [indiscernible] breakout. Maybe to start off we can have everyone introduce themselves.

Okay. I'll start. Steve Krognes, Chief Financial Officer.

Yes. Ryan Watts, CEO.

Carole Ho, Chief Medical Officer and Head of Development.

Alex Schuth, the Chief Operating Officer.

Okay. So I want to kick it off with my question first. You had a lot of updates today, particularly on the LRRK2 program. Could you walk us through the rationale why you are evaluating higher doses for DNL151 despite meeting all your safety [indiscernible].

Yes. So I will hand that one to Carole. And the question -- just so people may not have heard it or the recording may not have heard it. The question is, why do we continue to evaluate higher doses for DNL151?

Yes. So it's a great question. And I think that this is a new therapeutic area. As you know, we are the only company that has small molecules LRRK2 inhibitors in the clinic, and we are the first to move these into patients. As we start to understand the profile of both of these molecules, we do feel it's very important that in early development that we fully explore the entire dose range, that we understand the therapeutic window fully and also understand the effects on biomarkers. In Ryan's presentation, it went very quickly, but I think it's worth noting that in the 151 study, we have not reached the maximum tolerated dose for the Phase I healthy volunteer study. And I think in terms of good practice in early development, we feel it's important to explore that. While we are doing that, as you know, we have, at the same time, already initiated our Phase Ib. So we have already begun dosing in patients in the Phase Ib based on this positive data so far. In terms of the target engagement and safety profile, we are expanding the Phase Ib study to also add an additional higher dose. And again, the goal of this is to fully understand the therapeutic window as well as understand biomarkers that are affected by LRRK2 inhibition.

I think I'll just add a point as well. So when we started the LRRK2 program, and entered into partnership with Genentech in which we brought the program in, specifically, DNL201, we engineered a number of molecules that we thought had diversity from DNL201 and, in particular, around the off-target effects. And so what we've reported in the Phase I for DNL201 is basically a hypotension, increase in heart rates consistent with PDE. And pretty much what we see in the Phase Ib is similar to that consistency. 151, the goal in engineering that molecule is to eliminate that risk. And in fact, that appears to be the case. And so we want to continue to explore that window by dosing it, obviously, higher doses. And I think as Carole pointed out, we hit the biomarker goals for 151, started the patient study and then continued to explore that higher dose and then initiated also a higher dose in the Phase Ib. And part of this is this is a brand-new target, strong genetic rationale, really strong rationale building in sporadic PD, but we want the flexibility to move either molecule forward. And maybe the last point, and Alex can probably add to this, but -- and we've mentioned this previously, the financials are different on 201 and 151, 201 being invented at Genentech and 151 being invented at Denali, roughly half in terms of milestones and royalties. And so that being said, that's not a driving force for our decision, but we want to take the best molecule and have the flexibility to explore higher levels of inhibition, even though the genetics and the cellular data and the mechanistic data support that about 50% to 60% -- 50% inhibition is sufficient to basically bring LRRK2 levels back to normal activity. Alex, if you want to add anything from -- on that.

I think you summarized it very clearly, the financials that we owe Genentech are very modest. It's mid- to high single-digit royalties on 201 and half of that for 151.

Yes.

Do you have a strategy with regards to your BBB technology in terms of licensing?

Yes. So the question was, do we have a strategy with regards to licensing the BBB technology? I'd like to hand that one to Alex, and then I may or may not add anything to it.

Thanks, Ryan. So partnering is a core part of our overall corporate strategy. Partnering is -- we look at partnering as a way to continue to execute on a broad portfolio. We have 14 programs in total that we're working on right now. And as you've seen from Ryan, the blood-brain barrier technology has potential opportunities in many different applications, both from a modality perspective, but looking down the road, also from an indication perspective. So absolutely, yes. So partnering does play a key role in our overall strategy. We do have one partnership already here in this field with Takeda. We're partnered on 3 ATVs, or antibody transport vehicle molecules with Takeda. They are preclinical. Just, again, to point out that our most advanced programs, the ETV:IDS program and also the -- on the small molecule field are unpartnered at this point.

I think part of the strategy early on is not to license out the technology, rather do molecule-specific deals is what we did with Takeda. And that hasn't changed. I mean, we've built the technology to enable molecules, and we're not at a point where we're licensing the technology.

Maybe just to summarize the Phase I results for DNL201 and 151. Where are you seeing the big points of differences between the results of those 2 studies? How does it look in terms of magnitude on different biomarker changes that you made here?

Carole, do you want to take that?

Yes. So I'll just start with, I think, probably the biggest differences I see it is that the dosing frequency. DNL151 could support once a day dosing or possibly twice a day dosing, but there is a possibility for once a day dosing. And the data that was presented this morning was in the Phase I with once a day dosing. DNL201 is a twice a day dosing drug. I think as Ryan already mentioned also in terms of the off-target effects with respect to PDE3, PDE5 off-target activity, which is leading to some of the adverse events that we see in the higher dose level, which are fully manageable, reversible, I'll also note, in many cases, are actually self-resolving despite continued dosing. That is something that will limit any further increase in dose level beyond what we have presented today. With 151, so far, as noted, the safety profile is very clean. And even at the highest dose levels that we dosed, the AE profile is very manageable. Majority of those adverse events are mild and there have not been any serious adverse events.

I think in addition to that, in terms -- as it relates to LRRK2 itself, its inhibition and its downstream lysosomal changes, we see a lot of similarities between 201 and 151. So that's actually the beauty of having 2 molecules that are chemically distinct, different off-target profiles, actually different PK/PD profiles, as Carole mentioned, or frequent dosing for 201, it has a higher peak-to-trough; where 151, in addition to the PDE off-target effects, we also engineered to have a more sustained exposure, hoping for q.d. dosing. And a very good point Carole made, that what we presented today was q.d. dosing for DNL151.

I think maybe the one thing that I'll add is that we are very pleased with the DNL201 data. And given that we have all of the data from our Ib patient study, we do feel that this supports a molecule that could move forward to Phase II/III. In terms of the Parkinson's community, the twice a day dosing is not really viewed as being a major barrier, given the dosing frequency of other drugs that are taken, including Sinemet.

So 151, there's no chemical risk compliance, the equipment compliance in terms of dosing?

I think it was originally engineered around the phosphodiesterase off-target effects, which we now know are manageable, reversible. That's its original design. In addition to that, getting a q.d. dosing. But we don't really think about, as Carole said, compliance is not likely an issue in Parkinson's between twice a day versus once a day dosing.

Yes. I think my point there is that 201 is a very viable molecule moving forward into later-stage studies. And my point was just that, that twice a day dosing, while once a day may be more desirable, twice a day in the Parkinson's community is very well accepted because of the other therapies that they're on.

So they're justified.

Yes. Yes, yes.

Right.

For the DNL201 Phase Ib study, you enroll both the sporadic PD patients and also the LRRK2 patient population. Did you see any differences in the P 935 phosphorylation site changes and all markers and lysosome activity? And then lastly, do you think that LRRK2 plays a different role on those 2 [indiscernible]

Okay. I love that question, really around the differences between carriers and sporadic in terms of target engagement, pathway engagement and subsequent lysosomal changes. So what we see, and I think we reported this previously, is that if anything, our inhibitors are slightly more potent with mutation carriers. What we've seen in our previous studies, what we've presented before, is if we take blood from LRRK2 carriers, okay, compared to basically just healthy volunteers, there's maybe a leftward shift in potency by about twofold. The data we have here, we have 9 carriers. There's 28 patients total, right? So about 1 -- basically 1/3 of the study are carriers. And we see that actually both are very potent in inhibiting LRRK2 both in carriers and noncarriers when you look at phospho LRRK2 and phospho-Rab. So not a major difference. We don't have enough data, basically, to say is it a leftward shift of twofold. That's step 1 in like a dose response curve, right? What I did show which we find fascinating and important as you think of really focused approach on mutation carriers, we looked at sporadic versus mutation carriers in terms of biomarkers that are, in fact, differentially expressed in cerebrospinal fluid. That was the intent of that experiment. Let's take -- we have this 28 patients, we have the CSF, where are there actually LRRK2-dependent changes, different from sporadic? And there, we saw that the majority of those biomarkers are, in fact, lysosomal in nature. By the way, it doesn't mean that it isn't changed in sporadic, it just means that we're comparing it directly to sporadic. And when you look at just those in CSF, you see this LRRK2-dependent inhibition. So as I mentioned in the presentation, the next step is how about biomarkers that are changed in both. Now we already presented the BMP data. It's changed in both. That's changed essentially equally in both carriers and noncarriers from urine, and the next step is doing it, obviously, CSF-based biomarkers. And so what I can tell you is that definitely mutation carriers have a hyperactivation of the pathway, that's throughout the entire body. Every cell that expresses LRRK2 will have a two to threefold increase in activity. That includes in kidney, that includes lung, brain. So there, you can see some pathway-dependent changes. When you're looking just at sporadic Parkinson's disease, that's a regional specific degeneration. And even markers of like neurofilament are not up-regulated in Parkinson's disease because it's such a small region. And what's been shown is that LRRK2 is activated in that region, okay, in sporadic Parkinson's. And so at this point, we're confident that we're inhibiting LRRK2. It's essentially equally potent, probably within twofold in sporadic versus mutation carriers and we'll continue to explore biomarkers in sporadic that are shared with LRRK2 carriers.

Maybe moving on to blood-brain BBB platform [indiscernible]. You introduced a number of modalities today within the blood-brain barrier platform. When -- could you talk about each one of those and when do -- what are the advantages of one over the other? And just when you might want to opt for one?

Right. Yes, great. So just so you know, we dialed back the data that we were going to present. Believe it or not, there is a lot of data in that presentation. We dialed back some of the examples, for example, with the ATV and obviously the progress we're making even around antisense oligos fused to the ATV. So it's going to be on a target-by-target basis. Let me share general principles, okay? What we see is 20 to 30-fold increased uptake across the modalities. So you take something like Elaprase or the 9, maybe it's 10, 11, 12 marketed enzyme replacement therapies. We see now basically opening up that entire lysosomal storage disease market by fusing -- making Fc fusions, obviously, starting with Hunter syndrome. We also disclosed today ETV -- sorry, SGSH. That was one of the examples when we pull the data out. It looks essentially identical to IDS. We can -- with MPS III, we can see a similar, if not identical complete rescue of lysosomal defects with the ETV SGSH. So those are the general principles, we apply it on a modality-by-modality basis. Where I think is the most fascinating opportunity is with antibodies. An example we shared today was TREM2. And there, actually in that exact study, we had a standard anti-TREM2 antibody. We know there is a TREM2 antibody being developed right now, so we wanted to compare to a standard TREM2 with ATV, looking at various doses. And at 1/5 the dose of ATV TREM2, we can basically, robustly, activate that pathway. In the case -- we show an example of actually morphological changes in microglia, plus an associated upregulation of this cholesterol enzyme, which is required for turning over protein, which is interim to loss of function, you see that. We think we can enable, obviously, TREM2. And we have other antibodies in the portfolio. And maybe the last point, progranulin is a protein. So various protein fusions. So at this point, we've disclosed IDS, SGSH, TREM2, progranulin. Where we're still exploring is around synuclein and Tau. They were less enthusiastic because of the recent negative data on Tau in particular. And the need to basically engineer our platform. I think in those cases, there's a very small fraction of Tau in synuclein that drive this spreading. And the antibodies that have been tested are actually very high affinity, so they should get pretty good target engagement. So we've become a little less enthusiastic about those programs and are working on different ways to approach that. And maybe the last point and the last modality that we shared, which is the ASO fusions. Here, what we see in the future is the ability to give a systemic injection of OTV and basically reduce gene expression in brain substantially with this systemic injection. This, I think, opens up fields like Alzheimer's and Parkinson's, where doing direct injection in a million patients may be very difficult, cost, lead time, consuming them with some risk, right?

So you also expect it [indiscernible]. Just back to the [indiscernible]. Are you expecting to come out with your own product? Or are you [indiscernible]?

So that -- maybe if you want to add that?

Yes. So maybe let me make one more point, which probably wouldn't make it as clear early on. So the goal of Denali is to build a fully integrated company that makes and ultimately sells drugs as well. We do partner and plays a key role in our strategy, but it's also very important that we wholly own a large part of our portfolio. Ultimately, the goal is to market ourselves certain drugs. We have not made the decision which one of those. We currently have many candidates with the 2 additional INDs or CTA IND that we filed this year, we'll have 4 in the clinic, only 1 of those 4 is partnered right now. So there is opportunity to continue development on our -- by ourselves. And ultimately, market. Our marketing strategy is to start with the U.S. You will also see this in the 2 deals, strategic deals that we have with Takeda and with Sanofi that we have the ability -- even on the partner programs, that we have the ability to co-commercialize and share profits 50% in the U.S.

And as it relates to TREM2, specifically, we -- that is partnered with Takeda, okay? So -- and you can imagine the development times in Alzheimer's and caution that's -- it's, for us, a brilliant partnership in terms of 50-50 on profit share, but majority of the cost covered by Takeda. And PTV progranulin is also part of the Takeda partnership. Okay?

So in your partnership that's [indiscernible]?

So yes. So the Takeda partnership is around 3 targets. And within those 3 targets, we will not partner with someone else. It would make no sense. We're together, it's a strategic collaboration. We share costs, we share profits. So we're in it with -- on those programs?

So it's target-specific not compounds?

It is target-specific, that's correct.

That's right. Yes. And with disclosing PTV progranulin, that is the third of the 3 targets with Takeda. So it's Tau, TREM2, progranulin, okay?

Regarding the program -- on the program, you showed the biomarker [indiscernible] how much protein you're getting into [ brand ] with [indiscernible] the development time line.

Yes. So it's a good question. Each one of the -- what we see are these same-fold differences between Fc and the TV. So TV is a transport vehicle and then a standard Fc fusion. We see this anywhere between 10 or 30-fold. So that's the case. However, each molecule has its own pharmacodynamic profile. So for example, IDS is cleared more rapidly than an antibody, antibody has a much better pharmacodynamic profile. What we show in the data today is basically that rescue of lysosomal dysfunction in brain, where in liver, both the Fc fusion and the PTV progranulin infusion have the same rescue. It's actually exactly what we see with ETV:IDS as well. So what you can calculate basically is that you're seeing 10 to 20 or basically 20 to -- 10 to 20 or 10 to 30 depending fold increase of PTV progranulin brain. What we're not showing is the actual pharmacokinetic profile. That will come as we continue to -- and I mean, it's what you'd expect. It's a fusion protein, has a more rapid kinetics than, say, an antibody does, right? In terms of the development time lines, we're giving no guidance on where we're going. I mean -- but if it's not obvious to you, everything is full speed ahead at Denali. As fast as we can move to get to patients, that's what we will do, right? I think as you look at the totality of the filings, you look at the totality of the clinical data that we've read out in the last 4.5 years, we would love to get into the clinic as rapidly as possible with PTV progranulin. Yes. With the depth of the science, with the clear PK/PD rationale and all the safety profiles as again -- and traditionally we've shared lots of data around that as we go forward. Yes.

So I'd like to ask a question about your [ PD ] platform. So it seems that the trend the receptor has part of our systemic expression, do you anticipate any adverse effects coming from [indiscernible]?

Yes. So we started working on transferrin receptor as a shuttle and we'd be in my own lab in 2006, and we've done really extensive analysis of where else is transferrin receptor expressed. Now I'm going to keep it -- I'd go a little bit deep here and then we'll come back out. But the highest expression of transferrin receptor, really 2 cell types, immature reticulocytes that mature to become red blood cells as they load with iron, and then the other are brain and field cells. So basically -- and some dividing cells. Interestingly, tumors express very high levels of transferrin receptor because they have this high metabolic need for iron. So those maybe are the 3 type of cells. As a result, the majority of risk around transferrin receptor has been in these reticulocytes and all the data that I've shown today, these are effectorless antibodies, so they don't engage the immune system. Stay tuned for how we plan to do that going forward because we have ways around engineering, maintaining the ability to, let's say, reduce plaque in Alzheimer's disease with an ATV molecule by engaging the immune system, but without having any effect on the reticulocytes. So there was a paper we published in Science Translational Medicine in 2013 that essentially looked at exactly that question, where else it's expressed? What is the safety concerns around this? And I suggest you look at -- [ couch it ] all 2013. And that was previous work before founding Denali.

Maybe one last question here. With DNL310 proof-of-concept data expected within the year, could you describe the study that will be started presumably soon? And what constitutes proof of concept's in your mind?

Okay. I'll hand this to Carole and then supplement, yes.

Yes. So we have just filed the IND and the proposed study is a 6-month dose escalation study, where we will start with a dose that we expect actually will reduce CSF GAGs, which is the endpoint that we will be looking for. So the GAGs are the substrate that accumulates in this disease. I just really want to highlight that this is an IV delivered therapy that addresses both the peripheral and the central manifestations of disease. So these patients will not be on Elaprase. So they will be on just DNL310, and so we'll be monitoring also peripheral efficacy as well as looking at central target engagement with respect to reduction of glycosaminoglycans, the GAGs and the CSF. We anticipate greater than a 50% reduction, which would meet our proof-of-concept goal in this short period of dosing.

Yes. And this greater than 50% reduction in fewer doses actually would be differentiated from the competition, number one. The challenge here is that it's a rare disease, few number of patients. It's hard for us to give exact estimates on timing of that data, let's just say, again, we're going all out to give these patients, I think, a critical therapy that will be able to replace Elaprase. And obviously, we'll give updates throughout the year.

Is there a particular threshold on the reduction in GAGs that you would like to see that has been known to show -- translate to a clinical benefit?

Yes. So for us, in our experience, in our models, I think we've set that threshold at greater than 50%. What we're seeing from many of the competitors and just shorter-term dosing is about 30%. We think that, that may not be sufficient to rescue. Once you get greater than 50%, you start to see the lysosomal rescue and the subsequent downstream effect. You do get a cumulative response. So if you have, let's say, 50% reduction over time, it can grow as great as like 80% rescue neurodegeneration, right?

So maybe then I'll just also add that the mode of delivery is important. So in other words, if you administer an enzyme intrathecally, you may only get very regional distribution. So in some cases, you could see a greater reduction in the CSF. That doesn't necessarily translate to actually reduction across the brain. So in our animal studies, we have -- and we presented this data previously, we've looked at broad distribution, not only in reduction of the GAGs in the CSF, but that is actually reflecting total brain GAG reduction as well.

Thank you very much.

Thank you.

Thank you.