As a reminder, this call is being recorded, and a replay will be made available on the Equillium website following the conclusion of the event. I'd now like to turn the call over to your host, Stephen Connelly, Chief Scientific Officer at Equillium. Please go ahead.
Thanks, [Tara]. Good morning, everyone, and thanks for joining today's webinar, titled "Insights to the Aryl Hydrocarbon Receptor and miR-124 Axis." Before we start, it probably makes sense to give a little bit of context on this work. From the outset, both the aryl hydrocarbon receptor and miR-124 are two distinct biological molecules, and indeed, therapeutic strategies modulating both of them have been shown to be effective in treating ulcerative colitis. However, data exists from cells, animals, and humans that point to a more direct and functional relationship between these two molecules. Despite the biological and therapeutic relevance, the paradoxical relationship between these two has remained elusive and controversial. Understanding this relationship has critically important implications for both mechanistic interpretation and rational development of therapies targeting the AhR-miR-124 axis in inflammatory disease.
Presented here today, we'll do a cross-section of work aimed at elucidating the functional relationship between AhR and miR-124, specifically providing a unifying mechanistic framework linking ligand-dependent AhR signaling to miR-124 expression. To ensure we have enough time for Q&A, we won't go over all of the data in the deck, please be sure to download the full deck after today's call. Why don't we just dive straight into it? Let's do a quick recap on the aryl hydrocarbon receptor as a signaling hub. The aryl hydrocarbon receptor is a xenosensor. It's a ligand-activated transcription factor that resides predominantly in our barrier tissues. When ligands come into the cytoplasm and bind to AhR, that causes translocation and dimerization within the nucleus.
The recruitment of cofactors or binding to DNA leads to the induction of multiple pathways responsible for detoxification, immune modulation, anti-inflammatory cytokines, antioxidant proteins, and barrier function and repair proteins. Let's walk you through a few of the top ones in each of the different functional areas. Historically, the aryl hydrocarbon receptor has been largely known for its ability to initiate xenobiotic responses, so metabolism genes, things that will destroy molecules that have come into the tissues, given their signals, and then need to be cleared. Things like the CYP1A, CYP1A2, CYP1B1, a number of other genes, including enzymes that glucuronidate molecules, all with the sole purpose of eliciting a response to remove those molecules from the cell and turn off the signal.
More recently, in the last decade, areas relating to immune regulation and barrier function have been more widely studied, and this has changed the paradigm for how we think about the aryl hydrocarbon receptor. Here there are a number of different genes modulated, but also cofactors recruited in a cellular dependent manner. Cofactors recruited in the immune cell typically modulate immune responses like NF-κB, RORγt, and then cofactors recruited into the epithelial barriers like Wnt, HIF, AP1, et cetera, modulate things like barrier function repair and regeneration. Depending on the molecule and depending on the signal and the cell, you can get what we call biased agonism. Not all AhR modulators are created equally. These different gene sets will show a different signature, largely dependent on the ligand itself.
Importantly, xenobiotics or external unnatural molecules like pollutants or drugs typically elicit very strong xenobiotic responses, whereas dietary ligands, those elicit responses that are more related to the immune regulation, barrier function, and repair, things that are more important from a therapeutic perspective. An area of emerging research has been the microRNAs that are induced or modulated by AhR. I'm just showing the ones that we know from the literature are induced from AhR activation, but there are a number of miRs that have been implicated that actually modulate the levels of AhR, and one of those in the literature is miR-124. Let's talk about the AhR-miR-124 axis. What's known between this relationship? There's a nice set of papers by Zhao, and what they report is that there's an inverse correlation between miR-124 and AhR protein levels in tissues from active celiac disease patients.
Overexpression in animal and cell models have shown overexpression of miR-124 decreased AhR expression and had a net pro-inflammatory effect, while inhibition of miR-124 increased AhR expression and had an anti-inflammatory effect. This suggests there's some sort of negative or regulatory relationship between those, where overexpression of miR-124 can then downregulate AhR as potentially a signal termination mechanism. Vice versa, when levels of miR-124 are reduced, AhR levels return. This is meant to be perhaps a way of balancing the anti and pro-inflammatory responses within cells. There is another set of data out there. I like to term this the sort of obefazimod paradox, and that's data that would suggest that induction of miR-124 does not elicit increased cytokines. It's actually anti-inflammatory.
The key challenge with this data set would be that we don't really understand from this data the levels of activation or expression of AhR as a way to reconcile this with the data presented by Zhao. However, this model can be reconciled if, in fact, obefazimod is itself an AhR modulator. Now, to tackle that, there are three key questions we have to ask. Is ABX464 an AhR modulator? Does AhR modulation itself induce miR-124? What are the functional consequences of AhR blockade on the induction of these downstream pathways? To tackle this, and largely spurred on by questions from the investor and academic community, we looked at in silico modeling as a way to actually just check a priori, is there a potential binding mechanism here of ABX464 to the PAS-B binding domain in AhR?
We actually took a number of AhR modulators and antagonists and tested them in a validated AhR luciferase reporter assay, also looked to confirm that these ligands do indeed bind to the PAS-B binding domain. Largely driven by the pioneering work by Abivax in terms of identifying upregulation of miR-124 and providing a plethora of very well-established and validated protocols to monitor miR-124. We essentially copied a lot of this work to show does AhR modulation also similarly induce miR-124, and what is the relationship between these in terms of ligand-dependent AhR signaling and miR-124 expression.
Lastly, in a number of what we think are very well-validated systems of regulatory cells, namely regulatory macrophages, M2 macrophages, or regulatory T cells, we use those systems to essentially activate AhR using EQ504 or obefazimod, and then looked at downstream induction of CYP1A1, miR-124, IL-10, IL-22, or claudin in cells and model systems where we could look at regulatory effects, epithelial barrier repair and function, as well as regeneration. Let's talk about the first of those, in silico modeling. Let's take a second to think about what the binding domain of AhR looks like. It's shown here on the right-hand side. We call this the AhR PAS-B binding domain. It's where the canonical ligands bind to AhR. Think about this as kind of a baseball glove. It's got this large hydrophobic cleft that's able to accommodate the binding of a number of different chemical scaffolds.
Quite a large variety indeed. On the left-hand side, you can see endogenous ligands, natural ligands, and xenobiotics and pollutants. This covers a wide variety of different chemical scaffolds. In fact, there are, as of 2017, 13 approved drugs that we know bind and activate AhR. Last year, a computational study looking at similarity-based screenings of the PubChem library found there were over 77,000 potential AhR binding molecules. This highlights the fact that the AhR binding domain as a xenosensor, taking chemical signals and turning them into biological functions, can accommodate a wide variety of chemical scaffolds. With that in mind, we took the core scaffold of ABX464 and took a look at it for its properties related to AhR modulation.
It does tick a number of those boxes in that it's flat, planar, conformationally constrained, has extended aromatic or heterocyclic systems, and has moderate to high lipophilicity. We tested that not just against any compound, but against known AhR modulators. There are three that will come up here, l aquinimod, which is a Teva and Active Biotech-studied molecule. This molecule has been in Crohn's patients, MS patients. This molecule is a potent AhR agonist, but unfortunately has poor tolerability in patients. What you can see here is the overlay of obefazimod, ABX464 behind laquinimod, and we can see that the quinoline core overlaps very elegantly, there's also an extended aromatic system to the right here. You can see the same with AQILION's AQ312, which is essentially a laquinimod analog.
Very interestingly, there's some overlap here with Eli Lilly's recent patent portfolio of AhR modulators, and this is termed compound number 68. Again, overlap of that quinoline core and extended aromatic system, and good shape, lipophilicity, and hydrophobicity complementarity. Moving forward from this, we actually did some molecular docking, and we used MOE from Chemical Computing Group to do a number of induced fit docking simulations. We took a number of experimentally solved structures, more specifically the one with indirubin, which is the PDB7ZUB, identified some of the key elements of the binding pocket, and then ran docking simulations with ABX464. What we can see from this is that on the left-hand side, indirubin in its experimentally resolved structure binds very well into this hydrophobic pocket termed by the green colors here.
You can see on the right-hand side, phenylalanine 295, which is a conserved residue that creates stacking interactions with many known AhR modulators. On the right-hand side, the model of ABX464 also shows this similar binding pattern, and if we overlay the residues here, you can see very strong hydrophobic stacking interactions with this Phe 295, very similar to what we observe with the experimentally resolved structure of indirubin. If we take a look here and ask, what are the effects of glucuronidation on ABX464 and its potential binding to the PAS-B domain? If we take a look on the right-hand side here, you can see the two-dimensional structure of ABX464. On the left-hand side, you can see that when you rotate this molecule through 90 and then another 90 degrees, it's very flat, planar, and fits very well into the binding pocket.
Unfortunately, when you do glucuronidation to this molecule, this planarity is lost, steric hindrance is introduced, and a lot of polar steric bulk is also a potential hindrance to the binding in the AhR PAS-B domain. We believe that ABX464 would likely be active as an AhR modulator, but the N-glucuronidated, the metabolite form in humans, would likely be inactive. To actually test this, we moved on to using the AhR luciferase reporter assay to compare potency of known AhR modulators. This is an assay that provides a highly sensitive, quantitative, and physiologically relevant readout directly linking ligand binding to transcriptional output. As you can see on the right-hand side here, we tested ABX464, indigo naturalis, tapinarof, indirubin, ITE, and ABX464 N-glucuronidated. These molecules are available for purchase.
These molecules were either purchased online and/or, in the terms of ABX464, we actually synthesized the N-glucuronidated version using human liver microsomes to get the metabolite. What you can see is the EQ504, as has been disclosed before, is a very potent and selective AhR modulator. You see indirubin, indigo naturalis, tapinarof, and ABX464 is a moderately potent AhR agonist with an EC50 of about 400 nM. ITE, the precursor molecule, the very well-studied AhR modulator derived naturally in human cells, ITE, has an EC50 of about 460. The N-glucuronidated version, as predicted from the molecular modeling, does not bind. Of course, this is a cellular model, and the purpose of N-glucuronidation to molecules is to typically retard entrance into cells as part of a metabolism and clearance mechanism.
This could be either lack of entry of the glucuronidated form into the cells, or if it does get into the cells, a lack of binding to the aryl hydrocarbon receptor for the reasons previously disclosed. Another important element here is to really prove that the molecules bind to the PAS-B canonical binding domain of AhR. For this, we use a high-affinity AhR antagonist called GNF351, has an IC50 of about 62 nM. This has been widely used as a research tool to specifically block AhR transcriptional responses in vitro to confirm AhR-specific attribution in multiple cell types.
On the right-hand side, you can see that when we take the highest concentrations of either EQ504 tested or ABX464 tested in the potency assay, and then we add to that 500 nM of GNF, and that is a 1 : 5 ratio of drug to antagonist, or a 2: 1 ratio of ABX464 to antagonist. You can still see very effective inhibition of the signal and very effective competition of the AhR modulator and resulting in its lack of transcriptional responses. It's a very elegant way to demonstrate that this is not only an AhR modulator, that it binds to the PAS-B domain, and that it can be competed in terms of signal proliferation. Let's switch gears. We know that AhR modulation can be induced by ABX464, but not by its glucuronidated metabolite.
Let's take a look at does miR-124 get induced by activation of the AhR pathways. Now, for this, as I pointed out previously, we largely leveraged the Abivax data from the Apolit 2022 paper, which does an excellent job of really defining protocols for which you can study the upregulation of miR-124 in a number of different cell types. We've selected some data here to show you where it was originally shown to be very strong for ABX464, and then we've also shown you side by side EQ504. Now, on the left-hand side, what you can see is the level of induction of miR-124 compared to DMSO control of EQ504 at 5,550, as well as ABX464 and the N-glucuronidated version of ABX464, again at 5,550. It's very clear that ABX464 can induce miR-124, that the glucuronidated version here does not induce miR-124.
Similarly with the other AhR agonist that we're testing here, EQ504, it is also and is strong inducer of miR-124. New data here that would hint towards miR-124 being a pathway downstream of AhR activation. These results correlated actually very well with the reported data in that Apolit paper here shown by figure one. We also took a look at the cytokines that are inhibited, again, really leveraging the data from that Apolit paper where IL-17 and IL-6 were strongly inhibited by ABX464 at 5,000 nM. We tested the 50 nM EQ504, which is typically around where we see the EC50 of our molecule across multiple cell types, ABX464 at the originally tested concentration, and then for a side-by-side comparison, ABX464 also at 50 nM. What you can see here is a nice inhibition of IL-17 and IL-6 in a statistically significant manner.
Again, very correlated to exactly what was seen in that Apolit 2022 paper. What's interesting is the dietary ligands of AhR, such as the indoles, tryptophan derivatives, are known to be very strong inducers of IL-22. We've seen this across multiple cell types previously, where we see very strong induction of IL-22. In this system, where you have activated PBMCs, so this is not just normal PBMCs, these are PBMCs that have been pre-activated. What we can see is that EQ504 is a very strong inducer of IL-22 compared to ABX464 at either 5,000 or 50. Again, this comes back to the idea of biased agonism, and that not all ligands or activators of AhR elicit the same transcriptional responses or signatures. One of the things that we were drawn to EQ504 was it's a very strong inducer of IL-22.
It was nice to see this here once again replicated in the same system. Narrowing down within PBMCs, the Apolit authors focused in on CD4-positive T cells. They were able to sort and isolate CD4 T cells and show that miR-124 could be induced with ABX464. We replicated that experiment, but in this experiment, we also introduced the AhR antagonist. We used 50, 500, and 5,000 as the concentrations for ABX464 compared to EQ504 at 50 nM, and we kept a steady, fixed level of AhR antagonist. The reason we do this is because this way we can be sure that the AhR antagonist is not introducing any idiosyncratic effects. More importantly, what this does allow us to do is create some quasi dose response in terms of the ratio between AhR modulator and antagonist.
What we can see is very similar to what was previously reported as EQ504, ABX464, at 5,000 and 50 can induce miR-124. Very importantly, that AhR antagonism in a dose-dependent manner can actually reduce the levels of miR-124. This would be the first set of data that really highlights that miR-124 is downstream and dependent on AhR activation. We further went on to actually look at the effects on IL-17 expressing cells in isolated CD4 cells. This was, again, another statistically significant finding from the Apolit paper. We see a very similar commensurate decrease in the IL-17 expressing CD4 cells when either EQ504 or ABX464 at 5,000 or 50. Again, really just demonstrating that there's a lot of complementarity between our experiments and the previously published data.
Moving on into macrophages, a very similar experiment here in terms of inducing miR-124 through AhR activation, and then the addition of an AhR antagonist to show the effects on the AhR and miR-124 axis. What we can see on the left-hand side is compared to a DMSO control, we tested EQ504 at 50, ABX464 at 5,050, and we introduced GNF at 500 nM. That represents a 1:10 ratio with EQ504, a 10:1 ratio of drug to an antagonist with a concentration of 5,000 of ABX464, and then a 1:10 ratio when we look at ABX464 at 50 nM. Very interestingly, in all of these concentrations, we see strong miR-124 induction and then strong inhibition of miR-124 induction when we block the AhR activation.
Similarly, looking at the cytokines that are induced, two of the cytokines that were highlighted in the Apolit paper were TNF-α and IL-6. I think we see here a very strong reduction in those cytokines, and we see that map across to our experiments here with EQ504 and ABX464. Very statistically significant finding in terms of levels of reduction. We went beyond just TNF-α and IL-6 and actually looked at reductions in IL-23, IL-12, and IL-1β. We saw a very strong, significant decrease with EQ504. In the context of IL-23, we saw similar decreases of ABX464. It did appear that EQ504 was a more efficient blocker of IL-12 and IL-1β. Again, these could be differences from ligand-dependent biased agonism or signature responses. This just does highlight the ability of AhR modulation in dampening down multiple pro-inflammatory cytokines.
With that in mind, I think that data does very effectively demonstrate that miR-124 induction is downstream of AhR activation. The next question we wanted to understand was, in cells that are inducers of CYP1A1, miR-124, IL-10, and IL-22, what is the functional consequence of blockade of AhR activation on those systems? Today, we'll just talk about regulatory T cells, but in the larger deck, you'll see the data from the macrophages. These are two systems that Equillium has published quite extensively on, including more recently at the AAI conference. If we take Tregs, these are cells that have been differentiated into T regulatory cells, then treated with either EQ504 or ABX464, and then with the addition of an AhR antagonist. Here we're looking at the relative responses to CYP1A1 and miR-124.
As is the signature of AhR activation, we can see induction of CYP1A1 in these T regulatory cells, and we can also see induction of miR-124. Now, when you add to this the antagonist GNF at 500, which represents a 1: 10, a 10 :1 , and a 1 : 10 ratio across these concentrations, you can see a nice, clear inhibition of induction of CYP1A1 induction at an activation marker, as well as a reduction in miR-124. Again, highlighting that miR-124 is a downstream effect of AhR activation. Now, when we do the same experiment, and this time we actually look at the effects of induction of IL-10 and IL-22, we see again a very similar response in the induction of IL-10 and IL-22 is AhR dependent. It happens with both EQ504 as a known AhR modulator, as well as ABX464.
The addition of GNF at concentration of 500, again, representing a 1: 10, 10:1, 1:10 ratio, you can see a nice inhibition of induction of these anti-inflammatory cytokines. Again, we won't go into necessarily the macrophage data, but the short story is there it shows exactly the same relationship. I urge you to maybe download the deck later and take a look at that. What about the induction of CYP1A1 on miR-124 by AhR activation in intestinal epithelial cell lines? Previously reported by Abivax in the Apolit 2022 paper, is that miR-124 induction is not seen in epithelial cells, and that's what's replicated here on the right-hand side. Regardless of the concentration, and we show here a nice wide dose response, we do not see induction of miR-124 by either EQ504 or ABX464.
You do still have AhR modulation as demonstrated by induction of CYP1A1 on the left-hand side. A nice dose response with EQ504 and you see at the higher levels of ABX464 induction of CYP1A1. Moving forward, we take the 100 nM concentration here from EQ504 because this is largely where the molecule tends to plateau out, and then we took the 1,000 nM concentration, which is the maximal concentration here tested for ABX464. If you look at those cells, again using 100 of EQ504 or 1,000 nM of ABX464 and add to this 2,500 nM of GNF, we use a lot more inhibitor here, needed to be optimized, likely because the inhibitor has poor transit into these epithelial cells.
What you can see from a perspective of cause and effect is that activation of AhR and induction of CYP1A1 happens with EQ504 and ABX464, and then the addition of GNF and AhR antagonist commensurately decreases that. With EQ504, that's a 1: 25 ratio, and with ABX464, that's a 1: 2.5 ratio and that's why we see a sort of quasi dose response here. What are the functional consequences of AhR antagonism in epithelial cells? We're not seeing miR-124 induction, but we are seeing activation of CYP1A1. What does that do functionally within these cells? Again, there's two sets of data, but I'll be just presenting the wound healing data here, but I do urge you to take a look at the barrier integrity data by TEER assay at a later date.
This is an experimental model that we've published quite widely on, including more recently at the American Academy of Immunology, where we actually tested this against indirubin, another clinically validated AhR modulator. As a wound healing assay, you take these layers of cells, you propagate them, and then you score them to create a scratch, and then you monitor the healing rates over time. What you can see here on the right-hand side are the percentage wound area compared to the DMSO control after four days. We can see after four days, the DMSO vehicle has about 40% wound area. EQ504 or ABX464 at 100 and 1,000 nM respectively, significantly improve wound healing by about double of that.
When you add an AhR antagonist into this experiment, you worsen wound healing, suggesting that the wound healing via AhR is also AhR dependent and is worsened when you inhibit the AhR dependent pathways. This is shown in this data specifically when we look at the biomarker for activation of AhR, CYP1A1. We take these T84 cells that have been scored, and at the end of those four days, we take a look at the induction of CYP1A1 compared to the DMSO control. As we've shown in immune cells, EQ504 or ABX464 at 100 or 1,000 nM respectively, induce CYP1A1 as a biomarker of activation.
Similarly, when we add in the antagonist at either a 1: 25 for EQ504 or a 1 : 2.5 ratio for ABX464, we see a very strong inhibition of induction of CYP1A1, again highlighting the AhR dependency of these molecules. Lastly, one of the important biomarkers for epithelial barrier function and repair is claudin-2. Elevated levels of claudin-2 have been associated with leaky gut or impaired gut barrier function. It's increased in ulcerative colitis and DSS colitis models. AhR activation has been associated with reduced claudin levels in intestinal models, and decreased claudin-2 may reflect restoration of epithelial barrier function following AhR agonist treatment. A lot of this data is already published, and those publications are listed there in the footnote.
What you see here, similar to what happens with wound healing and CYP1A1, the marker of downstream induction claudin-2 is reduced when you add an AhR modulator like EQ504 or ABX464. You can see those nice reductions, statistically significant reductions in the levels of claudin-2. When you add an AhR antagonist, those levels shoot up. Interestingly, those levels get even worse, suggesting that AhR has sort of tonic signaling properties in modulating barrier function and further inhibition of AhR activity worsens barrier function. That's an observation that's been replicated elsewhere. We know that if you knock out AhR, barrier tissues actually become leaky, claudin levels rise, this is a very important finding here in terms of looking at a functional readout for AhR modulation via either EQ504 or ABX464.
It also goes a long way to perhaps shedding light on why Abivax's ABX464 is effective in the treatment of ulcerative colitis, because not only its modulation of immune cells, but its modulation of barrier function repair and regeneration in epithelial cells. I think this is a new, really important finding and one that I think further expands the value proposition of AhR modulation and the miR-124 axis. Let's just quickly go through the summary and conclusions. Is ABX464 an AhR modulator? Well, our results would suggest yes. It's a moderately potent AhR modulator with a potency of about 400 nM that binds to the canonical PAS-B domain of AhR. The ABX464 and glucuronidated version does not bind nor modulate AhR.
Again, this could be a lack of entrance by the cell, given that these polar glucuronidations are meant to inhibit cell transfer, or it could be that if it's in the cell, that polar steric bulk and loss of planarity precludes its binding and activation of AhR. The next question was, does AhR modulation induce miR-124? Again, using a lot of the pre-published work by Abivax in the Apolit 2022 paper, AhR modulation by EQ504 or ABX464, but not by ABX464 and glucuronidation, does induce miR-124 and decreases pro-inflammatory cytokines in multiple cell types. That's very interesting that here this data would also suggest that it's AhR-dependent signaling leading to the induction of miR-124. Whether you use an antagonist or a modulator like ABX464 and glucuronide that does not induce AhR activation, you do not see the induction of miR-124.
There is no miR-124 induction in epithelial cell lines. That's a previously reported finding. It seems that miR-124 is contextually important just for immune cells. From a functional perspective, we know that AhR modulation is effective at inducing downstream pathways, including miR-124 in immune cells and epithelial cells. What was really nice here was that using functional models of inflammation, tissue barrier function, and regeneration, we show that AhR antagonism inhibits the induction of CYP1A1, miR-124, IL-10, and IL-22 in T regulatory cells and macrophages, and it also inhibits CYP1A1 and claudin-2 in epithelial cells, worsening the healing. Again, not shown here, but I urge you to take a look at the barrier function assays included in the larger deck, is that we also see that AhR antagonism reduces and inhibits CYP1A1 in epithelial cell lines in worsening the barrier integrity.
That sort of moves us on to really this last slide, which I think moves us towards a more unified mechanistic framework for how we think about ligand-dependent AhR signaling and miR-124. This is an agglomeration of work that's published by ourselves, by Abivax, but also the academic community. That would suggest that when AhR modulators bind to AhR and activate it induces downstream pathways. It induces modulation of immune and barrier function proteins. It induces expression of xenobiotic enzymes that terminate the signal and remove the ligand from the system via CYP1A1 or glucuronidation. It also induces a number of microRNAs, shown here specifically for today's context, miR-124. We know that other miRs, such as miR, I think it's 355, are also related to T effector cells and IL-10 signaling.
The fact that miRs are downstream markers of ligand-activated transcription in the AhR pathway should not be a surprise. What's interesting is that the CYP1A1 or glucuronidation negative feedback mechanisms leading to ligand metabolism, or the induction of miR-124 that leads to lowering of AhR signaling, could be parts of the regulatory framework that help modulate AhR activation in cells. Overall, what this shows is that modulation of the AhR pathway via AhR modulators such as EQ504 or ABX464 reduce inflammation and fibrosis and increase barrier function and repair mechanisms. With that, I'd like to thank the key acknowledgments here of the Equillium research team and pass this back to the moderator for Q&A.
Great. Thank you, Stephen. At this time, we will be conducting a question-and-answer session, so please hold for a brief moment while we pull for questions. Our first question comes from Steve Seedhouse at Cantor Fitzgerald. Please go ahead, Steve.
Great. Thanks so much. Very interesting studies. Thanks for hosting this. I wanted to work through a couple interesting questions I had here. Just first regarding the mechanism behind miR-124 induction, do you think it's transcriptional? The working hypothesis for obefazimod, I think, is around splicing regulation, but it seems just more likely that it would be transcriptionally induced. I'm just curious what you think.
We're not experts at miR-124, so all I can do is give you my sort of opinions and sort of hypothesis here. When AhR is activated, it can bind to DNA elements, which is where we'd likely see genomic elements like miRs be generated, but it can also induce cofactors. I think what we can tell from this is that it's very clearly AhR modulated. Is that the only mechanism? It seems like from this data, it would be very much a dominant mechanism here. Is it transcription or is it something related to AhR activating something and then that leading to modulation in miR-124? I don't think these experiments go to that level of detail. I think from our perspective, the key question was, when you modulate AhR, does that modulate miR-124?
The exact mechanism here, I believe, is transcriptional, but I think further work would be required there. I think the alternative mechanism here, in that it binds to the CBC binding domain, which has been demonstrated, and that changes splicing. It would be very interesting to identify ways to be able to blockade that pathway and actually look at the impact on cells. That would be a way to sort of tease out relative contribution there. I think we can all appreciate that using an AhR antagonist that then blocks downstream miR-124 induction does very, very strongly suggest this is a downstream-induced marker of AhR activation.
Yep. Great points. Just on the negative regulatory activity that miR-124 has on AhR. In your experiments that you've shown here, did you look at AhR expression over time? What is happening there? Is it getting down-regulated? What do you think will happen in vivo? In other words, what are the implications here for the durability of the efficacy, like in a maintenance setting in IBD, for instance?
Yeah. We've looked at this in immune cells, and that data is in the broader deck. We did this in the T regulatory cell, and we see a reduction in AhR expression by about maybe 20%-30%. It doesn't cause the whole cell loss of the receptor. We should maybe think about how the Zhao research was completed. They very heavily overexpressed miR-124. We can't really reconcile the concentrations of miR-124. We've already known in the literature that when you activate AhR, the ligand and the receptor, the ligand gets metabolized or shuffled out of the cell and moved away. There are repressors that turn off the signal, and the protein of AhR can be moved on to the proteasome for degradation. Loss of AhR after activation has been previously reported.
The miR-124 axis is sort of interesting because it would be another element that is sort of a regulatory element in response to activation, in that it's generated and then it sort of dampens the signal. We do see a reduction in AhR expression, total protein expression, but this doesn't, we believe, result in a meaningful drop in protein available for signaling. You also see this that in Crohn's patients, there does appear to be a reduction in AhR, whereas ulcerative colitis patients, there does not seem to be much of a change in the levels of AhR expression.
Lastly, I'd say that we've done animal models that have dosed for prolonged periods of time, and we can continue to see sustained levels of activation and efficacy, suggesting that there's plenty enough AhR still present as it sort of is made, used, recycled, made, used, recycled, et cetera, that would allow us to continue to stimulate that pathway in a durable fashion for, as you point out, maintenance dosing in ulcerative colitis.
Yep. Thank you. Just as a follow-up on that. This is the last question from me today. Do you think that, given that miR-124 may have this negative regulatory effect, but it doesn't seem to be sort of fully abolishing what AhR agonism is doing that's constructive. Do you think miR-124 separately has its own independent anti-inflammatory effects separate from that regulatory action, such that maybe on net, its induction is helpful? Just because I'm just curious if the degree of miR-124 induction is relevant here for the efficacy translating into the clinic, or if you'd just be better off having an AhR agonist that doesn't induce miR-124 at all.
Yeah. Working backwards, we've tested a number of different AhR modulators, indirubin, et cetera, tapinarof, and these do also induce miR-124. We focused in on EQ504 for this presentation. It does seem that AhR modulation broadly does induce miR-124. Your question is, does miR-124 on itself induce an anti-inflammatory effect? Kind of like, if you induce IL-10, you induce IL-22, these are all functional proteins, claudin. There are a laundry list of things that are regulated by AhR that contribute to therapeutic efficacy here. This study, and we should be fair to what this study helps us answer and what it doesn't help us answer, is it shows that miR-124 is induced by AhR, but it doesn't actually levy the contribution in terms of anti-inflammatory effects from miR-124 or other components of the AhR pathway.
I think that experiment, one would want to just add in miR-124 in the absence of an inducer, like an AhR activator, et cetera. That would be a key experiment where one could really ascribe some anti-inflammatory, anti-fibrotic, tissue barrier function effects to the molecule itself, miR-124. The challenge we have in our experiments, of course, be it the original Apolit paper or our work here today, is that you're always giving an AhR modulator. It's very hard, if not impossible, to tease out the relative contribution. How I generally see it is that if it does have effect, which is great, it's induced by AhR activators, right?
It comes back down to the original point here, which was when we set out to do these studies, it was to ask a lot of questions from the academic and investment community is, what is the relationship between AhR and miR-124? That miR-124 seems to be induced by AhR in an AhR-dependent fashion. It doesn't necessarily tell you what level of contribution that has to therapeutic efficacy. What I would say in the last point is that people have taken AhR modulators, put it into animal models, and used anti-IL-10 antibodies, anti-IL-22 antibodies, and they see a very strong reduction in the efficacy in those models by neutralizing IL-10, IL-22, and other pathways. I'd say that if it is contributing, I would posit that it's probably a minor contribution compared to the others.
Again, caveated with this is not a set of experiments meant to tease that out.
Thanks so much.
Thank you.
Thanks for the question, Steve. Our next question comes from Thomas Smith at Leerink. Please go ahead.
Hey, Steve. Good morning. Thanks so much for putting together this really comprehensive event. Really interesting data sets. Just a couple from us. Just on the wound healing assay, it looks like you aren't seeing miR-124 induction in epithelial cells, yet you have an AhR-dependent relationship on tight junction proteins. I am wondering, what do you think is driving that? Have you looked at other tight junction proteins beyond claudin-2?
We have, in other systems, we've done this in barrier function tissue TEER assays. I think we've looked at the claudin-1, claudin-2, a variety. We can certainly follow up with that information at a later date. There are a number of those known in the literature. Now, miR-124 was previously disclosed to not being upregulated in epithelial tissues. This was from the Apolit 2022 paper. We were not necessarily surprised to find that it wasn't induced. We do see the canonical activation marker of AhR. When we look at the other miRs that are related to AhR activation, they tend to be very contextual, meaning that they show up only in certain cell types. I think the miRs here, as we think about the part of the downstream transcriptome of AhR, would be cellularly contextual, right?
This comes back to Steve's question, do we think it's driving efficacy? Maybe in a certain cell type or its contribution is limited to a certain cell type, that's reflected in this data here. I think learning from what we've seen with other miR molecules is that it tends to be maybe important to certain cell types. Here it could be that it's a very important regulator of the AhR activation pathway in immune cells, because it seems to be very predominant in immune cells, but it doesn't seem to have an effect in epithelial cells. Conversely, the levels of induction of CYP1A are much, much higher in epithelial cells than you see typically in immune cells, unless you really hone in on a specific immune subset.
It could be that depending on the cell, they have their own specific intrinsic mechanisms for regulating downstream AhR activation.
Got it. That makes sense. Maybe I could just ask, maybe taking a step back, given what you're seeing mechanistically here, you've done all of this mechanistic work now. Maybe talk a little bit about how you expect that to translate clinically relative to obefazimod. Maybe another corollary to that, obefazimod isn't gut restricted. Are there any implications, I guess, to your approach here with gut-restricted AhR modulation that we should be thinking about?
Yeah. Super important question and one that we're early in our sort of clinical journey, right? A lot of this will be better answered in humans, but I'll give you my opinions and hypothesis here on sort of the relative differences here, maybe from a molecular perspective, and then potential advantages through EQ504 and Equillium's approach. Now, we should first talk about the pharmacology. When you agonize something, you can decouple PK and PD. When you antagonize, you largely want to be an IC90, sit on the receptor, because every time you come off that receptor, it can signal. When you agonize, you should think about this as you pushing a ball down a hill. Once it's going, it's going.
You often see disconnect between the PK/PD relationship, but you also see cumulative responses in which that the downstream markers have durable pharmacodynamic effects, even when the AhR receptor is no longer occupied. Now, that's important because it sort of gets at one of the key questions I think we get asked a lot, which is, if the N-glucuronidated version is not an active AhR modulator, but it has been hypothesized to be the active component in patients, the parent molecule, which is an AhR modulator, is in a minor component. How do we reconcile that? How would you reconcile that? There are three main ways we can think about this. I think the first one is that, again, with pulse agonism, you don't need to be on the receptor all the time. It's not a total exposure-sensitive system.
It's really time on the receptor to turn it on, once it's on, it's doing its thing. You don't necessarily need high levels of continuous exposure to elicit those pharmacodynamic responses. Perhaps even just the shorter pulses of obefazimod before it becomes N-glucuronidated and accumulates as what we believe is a non-AhR modulating moiety. Maybe that's enough. The next one is that the N-glucuronidation can be deconjugated in the intestinal tissues. There's macrophages, neutrophils, they do produce beta-glucuronidases. I think that's likely to be a minor contribution here, is that the molecule would find its way out into those tissues, be deglucuronidated, become active in those tissues. It's certainly plausible. I think perhaps the most interesting point to think about here is enterohepatic recycling.
Now, the concentrations of the N-glucuronidated version of obefazimod appear to be a bit of a biological and pharmacological outlier, and that they are five to 20 micromolar of concentration. This is going on, unfortunately, the sort of scant data that I can glean from the literature, I think this might be a better question posed to Abivax, is that at those very high levels of molecule, if even a fraction of that is excreted into the bile, which is a very natural mechanism for glucuronidation. Remember, adding this polar group is meant to excrete the molecule. The fact that we're seeing very large and extended PK times does suggest that there's some enterohepatic recycling.
Even if a fraction of that gets enterohepatically recycled, gets pushed into the bile, into the colon, where the long residence time in the colon allows deconjugation from the beta-glucuronidases there to the parent AhR activating molecule, and then reabsorption. That does give you a potential mechanism for gut enrichment, whereby you're actually shuttling the molecule out of the peripheral blood, into bile, into the colon, deconjugated, reuptaken, and when that activates AhR, elicits its pharmacologic response. As that moves back into the plasma, it gets glucuronidated, and that cycle continues. We don't, in the PK data that I've seen, see a classical enterohepatic recycling second peak as they call it. The likelihood is that it's such high concentrations in the peripheral blood that I point out, even if a fraction of this is enterohepatically recycled, it's pharmacodynamically meaningful, but it may not show as a sharp peak.
It may just show more as a blunted sort of tail to the PK. I think what would be really interesting is understanding the levels of ABX464 as its parent molecule in intestinal tissues and answer the question, how much of AhR are we actively modulating in those tissues? You could understand, is this fully saturated? Is this as much efficacy that we can get out of the molecule, or is there more to be had? Which is why our approach has been largely to get the molecule directly to colon tissues because it's the epithelial cells, the resident lamina propria cells that we're modulating, that those are the pharmacologically relevant cells, and to avoid systemic exposures by creating these high local, low systemic ratios.
Got it. That makes a lot of sense. Maybe last question on my end. Can we get an update on some of the clinical preparation and formulation work that's ongoing? Just remind us, what are the outstanding gating factors to getting 504 into patients here? Thanks so much, Steve.
Yeah, no problem. Thanks, Tom. Well, look, the short of that is there's been no change. I defer you to the corporate deck there. The guidance still holds that we expect to initiate this study here in mid-year, which we believe brings up to Q3. There'll be future updates pending on that as and when they're available, but I'll defer to the corporate deck for all of the deferred guidance.
Great. Thanks for the questions, Tom. Our next question comes from Alex Thompson at Stifel. Please go ahead, Alex.
Great. Thanks so much for taking our questions. Appreciate the presentation. To follow up a little on Tom's question. As you're thinking about 504 and gut restriction in particular, is there a rate or do you expect 504 to be glucuronidate or metabolized more rapidly or less rapidly than obefazimod? Is it really the time to modulate AhR that matters here, or is it the way that you are modulating AhR as it relates to the downstream signal that you're seeing in cells at this point that might matter clinically?
Yeah. Each AhR modulator has its own unique properties, right? I think by and large, what you want to induce is less of the xenobiotic, more of the immune. Inducing CYP1A1, there are 13 approved drugs. Inducing CYP1A1 on itself is not a toxic attribute of the molecules or anything else that. Things that we've seen with the dioxins or even the signals of indigo naturalis are largely idiosyncratic to the molecule and unlikely to be a class effect of AhR modulation. When we think about giving a molecule directly to the colon, we're not thinking about this as gut restricted. I tend to avoid using that term because I don't want to give the impression that this has some active restriction mechanism that keeps a molecule in the GI tract and forbids it from going out anywhere else.
I think that would probably be an overreach. We know that when molecules get into GI tissues, that some will get out into the plasma. In our toxicology studies and in the literature, there's no specific predefined level for which you want to stay under. This is really just about tipping the scales in our favor, which is the cells that we want to activate are the epithelial cells, the immune cells that are in the lamina propria. We don't believe we're getting any bang for our buck from modulating cells in the peripheral blood. Sometimes you can see changes in the peripheral blood because of what you've changed in the colon. That's always been an area of controversy. We want to deliver the molecule directly to the colon tissues.
The same properties that allow you get into the tissues are likely the same properties that are going to get you into the plasma. We've seen it with steroids, multiple other different molecules, 5-ASAs, is that if you can deliver the molecule to GI tissues directly, you see high tissue concentrations and low systemic. That allows us to drop the total dose. In our modeling, which we can give an update on later in the year, our modeling, as would be expected, shows that we can get high tissue concentrations, lower systemic concentrations. Again, we believe we're going to be well below the level of important activation of AhR in the peripheral blood. Certainly, very much adequate in those colon tissues to activate AhR and get all of our signal of efficacy.
I think we'll learn more out of the phase I in terms of how much those ratios pan out. I think it's hard to say anything here about comparative efficacy to obefazimod. That certainly wasn't the intention of these studies. It was really about what is the relationship between AhR and miR-124. I just think the advantage we have is a molecule de novo designed to modulate AhR. It's based off the naturally occurring strong immune and barrier function modulating endogenous AhR ligand, and we're going to deliver this and monitor its levels in colons to make sure that we're maximizing all of the efficacy on target for AhR. That, if anything, would be one of the blind spots with the obefazimod and ABX464 dataset is we just don't know how much AhR it would be modulating in tissues.
Great. Maybe as a follow-up, as we think about the phase I study and you've kind of framed it as a proof of mechanism study, what kind of biomarkers, PD or otherwise, should we be thinking about in that sort of a dataset?
I'd say a win for us would be getting tissue concentrations into the levels that we expect for EC50, EC90. The next one is CYP1A induction. That's a canonical biomarker. The reason that's so widely used is because it happens in all cell types. Differential levels, but it happens in all cell types. It's very predictable, and it's a very clear signal above the background noise. It's a very transient biomarker. That allows us to really look at dose response. The other downstream markers, which we may or may not see in normal healthy volunteers, because you typically require activated cells, right? It's typically more important for cells in the gut or cells that are activated, is things like IL-10, IL-22, changes in other biomarkers. We will do some pretty broad transcriptomic analysis on an exploratory data basis.
What we know from all of the translational work completed by ourselves and others is that activation of CYP1A1 is indicative of downstream transcription of important genes to modulating the immune and barrier function. This is meant to really bridge what we know about the indigo naturalis experience, where they've dosed it to many hundreds, potentially thousands of patients. In those datasets, they've shown on-target activation of AhR by induction of CYP1A1. I think that's how we think about this from a proof of mechanism perspective.
Great. Thank you.
Great. Thank you for the question. Our next question comes from [Jacob Ferberg] at Raymond James. Please go ahead, [Jacob].
Thanks so much for taking the question. I really appreciate how thorough you were on the slides today. I guess just a quick follow-up to one of the last questions. In the event that you do measure microRNA-124 in the phase I healthies, would you be able to help frame for us if there's a fold level of induction that you would see as de-risking?
For miR-124, we're not the experts there. The caveat with all of these experiments are really meant to tease out the relationship. Mechanistic pathway profiling, less perhaps comparative pharmacology here. What I would say is we would most likely now, knowing that EQ504 modulates miR-124 and that it's a good biomarker in blood, but perhaps less so in epithelial tissues, which we'll be collecting, is that I think it would behoove us to collect that and take a look at it. The increased levels of miR-124, they did not appear to correlate dose responsively with CYP or even the downstream markers. That's where I would caution is that we just don't know enough, perhaps, in terms of how they correlate to other markers.
That's why CYP1A1 is so widely studied and so widely relied on, is that it's just a very well-characterized cause-and-effect system for pharmacodynamics of AhR modulation. I'd say we'd collect it, try to make sense of it, and perhaps between now and getting that clinical data read, we can perhaps look to get a better understanding of those levels. I would say that it so far, because it looks like it's a marker that might accumulate, so multiple doses lead to accumulation, and then you see a plateau. Those types of biomarkers typically don't give you the best read on dose response, because depending temporally on when you look for them, we see that a little in our data today. We're looking so late down the line is that you've saturated and accumulated all of your response.
I think markers like CYP1A1, which are more transient and happened earlier and in a more dose-responsive fashion, might just be more enriching from a dose selection perspective than miR-124, which may accumulate and persist, saturating the signal, so you might lose your ability in dynamic range.
That's very helpful. Thanks. I guess just one follow-up on the glucuronidation. I apologize if I missed this detail. Is EQ504 glucuronidated? If not, would you expect that to be an advantage versus obefazimod in terms of potentially having a longer duration of action?
I'm not sure we can ever call it advantage or disadvantage at this stage, because we don't know really the effects of humans. Typically, glucuronidation should cause the rapid excretion of a molecule. To see it accumulated suggests either high plasma binding, high plasma protein binding, so it just sort of sits around in the plasma, so you create this larger sink. There's perhaps little to no benefit of that happening, if it's not the active species, which is what the data here would suggest. It doesn't mean that it can't then release it as a sink to become pharmacologically active elsewhere. It's hard to really tease out whether that's an advantage or disadvantage for us in our molecule. What I can tell you about our molecule is that we don't have extended persistent accumulation in the plasma.
We have good tissue retention in epithelial cells. Then we have fairly rapid plasma clearance. That's likely a contribution of CYP enzymes, glucuronidation, et cetera. From our early animal models, we have not seen, in our toxicology PK models, a mechanism like obefazimod, where it is almost entirely or predominantly glucuronidated. Then that changes the pharmacodynamic profile. We have a very well-behaved, typical sort of decay curve, again, in animals, of EQ504. That's essentially what we're looking for, is activation within the tissues. It gets out to the plasma. It gets metabolized. No active metabolites, no quirkiness, no additional accumulation mechanisms. What would be interesting is if what you see with obefazimod, which I do believe is an advantage for the molecule, is that you don't really want AhR activation systemically. Ideally, it just happens locally. That's largely just to avoid over-metabolizing your molecule.
By actually modifying the molecule to not be an AhR modulator, might just be a protective masking mechanism, which again is inherent to AhR as a pathway. There you could argue perhaps could be an advantage, but it's not a prerequisite. That's not what is absolutely required. You don't need to necessarily design that in. I think what you want to see is rapid clearance from the plasma, rapid uptake in the GI tissues, and this asymmetry in tissue versus plasma concentrations that leads to activation in the tissue, but not activation in the plasma.
Great. Thanks, Steve.
Thank you for the questions, [Jacob]. Our next question comes from Leland Gershell at Oppenheimer. Please go ahead, Leland.
Great. Thank you. Thank you, Stephen. This has been a terrific presentation, and we appreciate the updated data. First, wanted to ask, I guess teeing off the point you made earlier about with a sort of an agonist conditioning, you need to have the molecule on the receptor for very long, it can have durable downstream effects. Have you looked at the time course between obefazimod and EQ504 in terms of the time course of action and expression of miR-124? Also wanted to ask your opinion, obviously, you've gone a long way here in terms of the science, delineating the pathways between AhR and miR-124. Do you think that this negative feedback mechanism is sufficient to sort of help us explain and resolve the conflicting evidence between some of those preclinical data and the clinical data we've seen from obefazimod?
Yeah. Working backwards, I'm, again, not the expert of all of that data in hand. Like I say, we really set out this to sort of do mechanistic attribution and pathway profiling. There are some disadvantages or limitations, deficiencies in using cellular models. That is when we add the drug, it's on the cells, right? Actually in the Apolit paper, the drug is added over multiple times. You assume in all of those cellular models, drug is always on the receptor because it's always in the system. That doesn't reflect what happens in humans. The drug is dosed, it gets on the receptor, then it's cleared and cleared from the system. It doesn't necessarily reflect the cellular models well in that way. We should just acknowledge the deficiencies and limitations of cellular models.
We have done experiments where we compared this to indirubin, which is the relative induction of CYP1A1 versus IL-10 and IL-20 and IL-22. This isn't about how much time is required on the receptor, because in all of these cellular models, the molecule's always on the receptor, essentially, in those scenarios. What we did see is that indirubin induces a lot more CYP1A1 per call it pound of, or whatever metric you want to use, of IL-10 and 22. That's sort of known in the literature as ITE, the precursor molecule for EQ504, is a very strong inducer of IL-10 and IL-22. It's less of an inducer of CYP1A1. In both of those experiments, the IL-10 and IL-22, or both of those molecules for indirubin and EQ504, the IL-10 and IL-22 goes up, and it stays up for quite a long period of time.
Whereas CYP1A1 comes up and comes back down pretty quickly. I'd say that demonstrates the sort of temporal effects that when you agonize the pathway, the downstream effects can be durable, but the limitation would be that we'd have to wash molecule off or clear that system of molecule to then look at really the time decay there. I really can't answer in a sophisticated fashion about what it is the specific time one wants. I think from a clinical perspective, we'd be looking to see that we could durably induce CYP1A1 in the tissues in a dose-responsive way, and we'd be looking for choosing doses that were either on the curve or into that plateau, and then we'd be moving that on into efficacy studies.
I think there's limited we can do on the bench with these agonists to tease out trying to thread the time on receptor loop.
Thanks. Lastly, just a question. As you have all this wonderful work you've been doing on the science and looking at different ligands of AhR, have you looked at or do you plan to perhaps look at the Formation Bio compound KMR301, which has been reported in the literature under, I think, a different name, but that is also a small molecule miR-124 inducer. It may be more potent than obefazimod has based on the same kind of square structure. Just wondering if you have looked at any mechanistic activity of that compound. Thanks.
Yeah. Again, I think this comes down to sort of what our objectives internally is. I think uncovering the relationship between miR-124 and AhR is really important. I think that moves the needle. It helps us rationalize therapeutic effects. It can help you think about the design of molecules. It can help you think about biomarkers, as we've discussed on the call, looking at miR-124 now. Given that's a clinical biomarker, right, there's a lot of value in that. I don't think what we're interested in doing is making analogs of other compounds. I think we're very confident in that we spent a lot of time looking for AhR modulators. We settled on the analogs of ITE, again, because they're naturally derived, right? The parent molecule is the molecule that circulates in the lungs and guts of humans. It's a very strong and non-toxic agonist of AhR.
We've developed analogs of that. I think that we believe that we have a potent selective AhR modulator with good drug-like properties. It has all the attributes that we're looking for. We now know that it can induce miR-124, and similarly, we now know that ABX464 is an AhR modulator. I think that relationship has been strengthened. I think from here on in, it really doesn't change what we do with EQ504. We have our plan. We understand what our molecule does. We understand how we want to deliver it. We're currently formulating that for the phase I. Generating other molecules or whatever, I think has limited value to us and our development programs.
I think from here on out, we'll just be focusing in on what downstream pathways, such as miR-124, for instance, and how that relates to clinical efficacy, how that might inform our own drug development program.
Thanks very much.
From my understanding, Formation Bio has a very similar molecule to ABX464, and I'm not aware of any other molecules developed and sort of marketed necessarily as pure miR-124 inducers. I think from today's presentation, one can assume that if you're an AhR modulator, you're also a miR-124 inducer.
Absolutely. Great. Thank you very much, Stephen.
Thank you.
Thanks for the question, Leland. Our next question comes from William Wood at B. Riley. Please go ahead.
Thanks for taking our questions, really appreciate the thorough data here today, Steve. Maybe just thinking about, we've discussed a number of cytokines today, which you presented here, priorly, a lot of focus been on IL-10, IL-22, but also Abivax has been more focused a bit more on IL-17, IL-6. With just sort of the number of families here and the number of cytokines that you presented and discussed even just today, how should we think about what combinations or even what individual specific cytokines we should be thinking about, possibly even families of cytokines, which may provide the best support for downstream clinical benefit? I have a follow-up.
Yeah. There's probably buckets, and that's why I put the top four, because in reality, there's a lot more than four. The other part of this is that if I promised 50 things to show you if 50 things moved, you'd hold me to it, right? We've really tried to distill this, I think, to IL-10 and IL-22, and I'll walk you through the reason why. Naturally derived AhR ligands from the tryptophan metabolites indoles, right? They're very strong inducers of IL-10 and IL-22. Those are typically referred to as hallmark cytokines of AhR modulation. They're also very widely studied. ITE and indirubin, two naturally occurring, one endogenous, one exogenous AhR modulator. Two most widely studied AhR modulators in the literature. Very strong induction of IL-10 and IL-22. When you block IL-10 and IL-22, you lose efficacy, suggesting they're pretty dominant, right? They're also pleiotropic cytokines.
There's things that happen downstream. They're also very proximal to AhR activation. We think about IL-10 and IL-22 as being directly transcriptionally derived from AhR activation, whereas inhibition of pro-inflammatory cytokines is probably a little more distal, right? You're probably modulating NF-κB, and that's modulating other things, and T cell differentiation. That's further downstream. I like to pick biomarkers that are very proximal to the activation event and that are really important to think about clinical efficacy. We know that if you knock out IL-10, IL-22 in animal models, they undergo a lot of gastroenteritis. There is some conflicting data with both of those, but I think in the context of looking at the barrier function and immune regulation effects, it's very clear they're very important.
We also know that in patients who have IBD, they have lower levels of IL-10 and IL-22, and when treated with molecules that derive efficacy, even Skyrizi, I think there's some data here, too, is that they tend to improve levels of IL-10 and IL-22 because the system is starting to repair and restore. I tend to think about things is that we have AhR, IL-10 and IL-22 because they seem to be centrally important to mucosal biology and health and immune homeostasis and barrier function. You've got things like claudin, which are clearly very much regulated by the AhR axis. People have studied those very widely. I think that's the second set of tissue markers we'd be looking at. You've got your phenotypic changes. Do you reduce levels of Th17 cells? Well, that wasn't a direct consequence of AhR modulation.
That's a consequence of the molecules or proteins that were transcribed by AhR and a little bit further distal from the activating event. I think really here, IL-10 and IL-22 are things that we'll focus on very heavily in the clinic. In today's realm of being able to do transcriptomics and get tissues from the colons, we'd probably look to expand that set because it would always be more helpful if we had more biomarkers to think about efficacy or patient population target selection. You can think about IL-10 and IL-22 as the sort of central axis for how we think about AhR modulation and its therapeutic value proposition.
Appreciate that. I know we've been looking forward a little bit, some on our side, thinking about Abivax's maintenance data coming up and then also their data later this year in Crohn's. In terms of what you'll be looking for in those data sets that may give you confidence either in UC or possibly even Crohn's, what specifically might you be looking for, whether at the top line or digging down into the biochemical or biomarker data sets, that will really support your advancements going forward?
Well, look, first, there's a lot of kudos owed to Abivax here and those data sets. The nature of ABX464, I think is game changing for ulcerative colitis, specifically the maintenance data. What's interesting to us about the maintenance data is it's reflected in the indigo naturalis treatment data sets as well. In the indigo naturalis treatment, they had very strong induction of remission rates at week eight. I think it was 40%-50%, depending on which one of those phase II studies you look at. That's very strong. That was still in a patient population that was mixed almost 50/50 between biologic naive and biologic exposed. Pretty reflective of a contemporary population. When they did the longer term studies of indigo naturalis, they saw deepening of those responses. They had more clinical remission in those patients.
Typically for biologics or some of the other small molecules, what you get around week eight and 12 may not necessarily get better. I think the maintenance data's super important for ulcerative colitis patients from something like Abivax, because you see that if you can stay on drug longer, you can see those results improve. I think that's coming from things like IL-22, and that was demonstrated in the indigo naturalis data sets, where IL-22 was correlated with mucosal healing. It's a long-term value proposition which continues to drive barrier function repair, regeneration, immune homeostasis. You'd expect something like IL-22 to have the properties of allowing the tissues to heal. When you use a JAK inhibitor, and you just blunt all of the immune responses, right? I think there's diminishing returns from that from an efficacy perspective.
I think with something like IL-10 and IL-22, you're talking about restoring to a natural balance the mucosal health. That is very intuitive to see why results get better over time, right? We use week eight and 12 often, I wouldn't say arbitrarily, but that's how we test things in ulcerative colitis. The week 52 clinical remission data is game changing for obefazimod ABX464. I would expect that, given the data that we've seen today, knowing that it's modulating AhR, is I would expect that that's likely coming from pathways modulated by the AhR. I hope to see that they are successful in demonstrating very, very deep maintenance data. I would expect that part of the value proposition of EQ504 going forward is that strong induction of early remission rates at week eight or 12, I think that's where we can win.
There's still obvious room for improvement. About 16% for obefazimod. Indigo naturalis in the week 8-12 was getting 40%-50%. There's room for improvement there. I think having that continued improvement to mucosal healing, endoscopic remission, and clinical remission by week 52, I think is the thing that really changes the game for patients and physicians.
Appreciate that, Steve. Thanks.
Thanks, Will.
Our final question comes from Adam Walsh at Roth. Please go ahead, Adam.
Hi. Thanks for taking my question. Stephen, amazing presentation. Appreciate all the detail. Couple here. The first, is there a publication strategy around what you've been presenting today, and what would be the timeline there? Follow on to that, is there an IP strategy that can be derived from this, in terms of composition of matter, method of use, or anything like that?
Yeah. Well, last but not least, appreciate you joining, Adam. It's always good to hear from you. The kudos goes to the team here. I feel like this is a little bit like thanking the Lamborghini salesman for building a great car. It really goes down to the group here who did all this fantastic work. From a publication strategy, we will look to publish this data through additional conferences, potential publications around our own molecule, where we might compare against other AhR modulators like tapinarof, indirubin, and now we can add to that ABX464, and that's important because it's a molecule that's shown great efficacy at phase III. IP strategy. We don't really ever discuss IP strategy here, for obviously good reason. We're always very aware of what we can do from an IP perspective.
I think we're more focused on our molecule than generating IP around anybody else's molecule. I think largely what we're doing is corroborating and further expanding work that's already out there in the public domain. People have tested obefazimod as an AhR modulator. People have shown that relationship. I think what we sought to do was really solidify that with some good mechanistic data, orthogonal approaches, antagonism or whatever, to just sort of really understand the relationship. Now we can learn from clinical data, mechanistic profiling, pathway analysis of other molecules, which I think is to the benefit of patients and general drug development. I think that's all I can say, at least from the publication strategy and IP. Remind me if I missed anything else in your question.
No, that's totally fair. That's great. Appreciate that. Just one final one. In terms of the phase I trial, you mentioned that CYP1A1 is kind of the key induction to look at there with the downstream IL-10 and IL-22 potentially being more seen in UC patients. Does that impact your decision point on whether to include part B UC cohort? When will we hear information on whether that part B UC cohort will be included or not?
Yes. Maybe working backwards on that. We are still working with our advisory board that we more recently put together with Bruce Sands, Geert D'Haens, Florian Rieder, Simon Travis, and Vipul Jairath. We think a sort of really experienced team. We've engaged them early. We've got plenty of time to think about design. I think you'll hear from us probably sometime in and around the phase I what we would be doing afterwards. You raise a point which is, would we learn anything in the phase I that may change phase II? I think that's something that we'll have to carefully consider as part of the long-term development plan. What we want to achieve in the phase I is that we have a molecule that's safe, well-tolerated, activates AhR in colon tissues.
From a CYP1A perspective, this really isn't about hitting a certain level, because every AhR modulator actually induces to a different level its levels of CYP1A1. We shouldn't see this as we have to hit the same as another molecule. Really what you'd be looking for is that plateau. When you hit that plateau, you've fully occupied the receptor, you've fully turned on all the transcriptional machinery. From our translational work in the lab, we know that correlates to maximized induction of all of these other downstream markers. I think that's a win for us, is showing that we can safe, well tolerate, deliver drug to the colon, get to good drug tissue concentrations of the drug, show we can engage CYP1A1. Then I think it'll be about selecting dose, doses, and moving that forward into an ulcerative colitis patient population.
I think more to come there. We're laser-focused on getting that phase I up and running. We have at least received very strong feedback from the ulcerative colitis community that should this get into the realm of treating ulcerative colitis patients, this would be a mechanism of high interest to them given the experience they've seen and/or had with indigo naturalis, which is the natural botanical AhR modulator. Lastly, I think we recently brought on Snehal, who recently came from Ventyx , she's previously at Prometheus , Arena. She really adds to the experience, expertise, and the brain trust here at Equillium as it relates to ulcerative colitis trial. Super excited to have brought her on, and she'll be leading the clinical development. I certainly don't want to take any of her thunder this early in the game.
Stephen, thanks for the clarification. Appreciate it.
Thanks, Adam.
Great. Thanks for the question, Adam. This concludes today's Q&A session. I'll now turn it back to you, Steve, for some closing remarks.
Well, thank you for everybody who joined and perhaps stayed on 20 minutes past. We do very much appreciate everybody's time. Please do visit the website. In the next few hours, the presentation will become live. There will be some additional data. Should you wish to engage in any of the conversations that have been had here, please do reach out via the appropriate channels to Equillium, and we'll be more than happy to take additional questions. Once again, thank you, and good day.