Rich McCormick: [00:00:00] Hi, I’m Rich McCormick, executive vice president of clinical strategy here at Vial. Joining me today on our First In Human podcast is Alex Federation, co-founder of Talus Bio. Thanks for joining us today, Alex.
Alex Federation: Thanks so much for having me.
Rich McCormick: So we’ll jump right in. Can you share just your personal journey that led you to co-founding Talus?
Alex Federation: For me, a lot of this goes back to my earliest training in grad school. Talus is focused on this problem of the undruggable genome. I still remember one of the first days I trained at Harvard, walking into these big fancy buildings where they invented the field of organic chemistry. You go into class and one of the first things they say is, “Oh, we have all these druggable targets.” 90 percent of the genome is undruggable, which was stunning and hard to believe that they just had resigned so much of biology to this area that they didn’t want to touch.
I was lucky enough to find a lab there where I could train that was not afraid to go after this undruggable space. We did some cool stuff there. Since then, my whole career has been trying to find the right technologies to plug in at these different parts of drug discovery to really try and tackle this problem of undruggables. Eventually that all led to what we’re doing at Talus.
Rich McCormick: I love your company’s philosophy of, “nothing is undruggable”. Heading in that direction, can you guide us through that approach to developing those therapeutics for transcription factors?
Alex Federation: Transcription factors are one family of proteins that are in this undruggable space. They’re some of the oldest proteins we know about. Even before we knew what DNA was, we started to learn what transcription factors were, because when they go wrong, the problems they cause are just so massive. A fly doesn’t grow its wings or a mouse doesn’t grow a tail.
And, as you might imagine, when these go wrong, their job is to find DNA, interpret the information in that DNA, interpret the information inside and outside of the cell, and make decisions about what genes should be on or off.
You might imagine that in cancer or in other diseases, when that decision making process goes wrong, it’s really easy for these proteins to lead to a state where it will just tell the cells to grow and grow, keeping the growth genes on forever. That’s what often happens. So the problem for these proteins is that since their jobs are so complicated, when you take them out of the cell, out of the nucleus where the DNA is and RNA is, and all of their collaborative partners, they don’t fulfill their function properly.
Most other drug targets, we can take them out of the cell, put them in a test tube, and do all sorts of things to find molecules, and they still maintain their function. Transcription factors don’t. What we’ve had to do at Talus is bring new technologies forward that let us understand and observe how these proteins work in the natural unmodified human cell, and then use those technologies to find molecules that can bind to these proteins and block their activity. Often in cancer, these proteins are hyperactive. We’re finding molecules to stop their activity, turn off the genes for growth, and cause these cells to either die, or stop growing.
Rich McCormick: That’s interesting. So then how does Talus bridge that gap between the high number of transcription factors, and then the limited number of approved drugs?
Alex Federation: You’re right. There is this huge gap. There’s about 200, or so, transcription factors that we have found in the field to be associated with at least one type of cancer. It’s about 10 approved drugs to target those. A Huge gap.
The way we approach this at Talus is in a data dependent, driven way. Also an unbiased way. Our technology uses something called proteomics. You’re probably familiar with genomics. Sequencing DNA in order to understand biology. My co founder, Lindsey, is really a leader in, “How do we do this next step? How do we sequence proteins at scale?” That’s what we use at Talus to try and bridge this gap.
When we look at a molecule, since we use proteomics, we sequence all the proteins. We see what that molecule does to every transcription factor in the cell. Unlike the traditional approach where you have one target, you test many molecules. We actually can test many molecules against all the transcription factors at once to really let us rapidly iterate, discover, and optimize the molecules that we find.
The two dimensional approach to drug discovery can be a lot more efficient from a discovery point of view, but also a lot more challenging on the back end because then when we have all this data, we have to use computational tools to wade through it and prioritize what we’re going to do next.
Rich McCormick: Interesting. So then maybe that’s where you were headed. So I was going to ask about the validation process for the newly discovered TF inhibitors.
Alex Federation: This is a problem that’s really top of mind, right now. We just finished one of our biggest discovery efforts and have literally hundreds of molecules that all look interesting and triaging and prioritizing how we approach that is a new problem for us. The process right now, it looks a lot like what I was mentioning.
How do we [00:05:00] use computational tools and machine learning to help rank all of these molecules based on all the data that we have. How potent does it hit one of these transcription factors? How selective is it? We see what these molecules do to every TF in the cell. We want to start with something that’s pretty specific and only modulating. The TF that’s contributing to the cancer growth.
There’s other things, too. Like, the actual chemistry we take into consideration. The commercial viability of these molecules going forward. How many other people are trying to work on these targets? It’s a multifaceted approach, leading on the computational side of it to help us make those decisions.
Rich McCormick: How does your Marmot platform factor into that drug discovery?
Alex Federation: The Marmot is essentially letting us do that first step. As well as, the follow up that we were just talking about. Marmot lets us take a molecule, see what it does to all the transcription factors. Then, allows us to really turn the crank. When we have that initial molecule, it’s still pretty far away from being a drug.
But, like you mentioned before, these targets don’t have structure. They’re hard to purify outside of the cell. We need to use these advanced technologies while we’re optimizing these molecules, as well. That’s what Marmot lets us do. We can take a molecule, synthesize all sorts of analogs, see how those changes to the molecule affect the potency of the drug, how they affect the selectivity in the cell, and other parameters that we’re trying to optimize for, like pharmacokinetics and stability and other drug like properties.
Rich McCormick: You mentioned cancer in an earlier answer. What about challenges in developing therapeutics for rare diseases like chordoma?
Alex Federation: There’s pros and cons to every decision like this. Markets can change. The obvious con, from a business point of view, for a small disease is that there’s a small number of patients. So it’s harder to, sometimes, justify the investment from a venture side or from a strategic side in a rare disease like chordoma.
But, that being said, we’ve seen some massive successes in the rare disease space, even just in the last few months. SpringWorks comes to mind, just had a drug approved in a rare desmoid tumor. There’s other folks now pursuing that indication because they paved the way. That’s one of the pros for going after rare diseases like this, is that the development path can be pretty streamlined.
The resources can be pretty rich, so we are lucky to collaborate with the Chordoma Foundation, who’s not just built out a lot of support for people like us on the preclinical side to help us optimize our drugs faster. But, also on the clinical side to help set us up for efficiency success in first clinical trials for the molecules we’re making.
That’s the way we think about it. There’s also the obvious thing I didn’t even mention, but there’s a huge unmet need. That is what drives the decision at the end of the day. There’s no approved drugs in chordoma. Even chemotherapy is ineffective here. It’s a disease that’s managed by surgery, radiation, and then waiting until the cancer inevitably recurs. We’re eager to bring these molecules forward quickly as we can for these patients.
Rich McCormick: How might a breakthrough with brachyury inhibitors impact treatments beyond chordoma, especially in other cancer types?
Alex Federation: Yeah, that’s a good point. This is an interesting transcription factor. Brachyury is only normally expressed in very early embryonic development. Essentially all of the tissues in adults, and children too, anyone that’s been born, brachyury is shut off. In chordoma in particular, when brachyury is turned on in these particular cells that are around our spinal cord. It drives these chordoma tumors. It’s the sole driver of those cancers.
If we look across many other cancers, colorectal, triple negative breast cancer, lung cancer, even some other rare cancers like ewing sarcoma, brachyury is also reactivated there. We think it’s less well studied there and it’s not the sole driver, but seems to really play a collaborative role in metastasis, and resistance to chemotherapy. It’s something that we’re starting to explore, now that we have more advanced molecules is where is the next best place to start looking at brachyury inhibition to help other indications as well.
Rich McCormick: You mentioned pediatric cancer in that response. It seems like you have some grants that have become available to Talus. Do you plan to use those grants to tackle the unique challenges with childhood cancers?
Alex Federation: Child cancers are really interesting, right? If you think about non-small cell lung cancer. This is often a disease associated with smoking or exposure to some carcinogens. People have decades of time being exposed to these mutagenic substances to accumulate all these mutations that eventually, once enough hits happen, can lead to cancer.
That’s not the case with kids, right? They haven’t been alive long enough to accumulate that [00:10:00] many mutations. What tends to happen in pediatric cancer is that there’s really bad luck. They get one mutation in something like a transcription factor that can get them to go and regulate the activity of thousands of genes downstream to make up for the fact they haven’t had time to accumulate this genetic damage.
They get one mutation in a TF that can have widespread effects. The way we’re approaching it, the thing about this is, is that since these mutations are so profound, and essentially they’re also recurrent. They tend to happen, over and over, in different patients. We know what they are. So we can, now, go into models where we have these new mutations that happen in pediatric cancers, use a platform like marmot to try and find molecules very specific at modulating these mutated forms of transcription factors.
That’s a big focus of our grants for especially these childhood tumors, these solid tumors where chemotherapy is not really a solution, yet.
Rich McCormick: We’ve mentioned the Marmot platform a couple of times. Could you elaborate on what sets it apart from the competition? What do you see as the vision for future impact of Talus in the realm of transcription factors?
Alex Federation: The traditional approach to the undruggable space and one that’s had some success is, bringing forward different methods that allow us to find molecules that stick to undruggable proteins. In the past, when people have done this, they take these proteins out of the cell into an in vitro system or into a test tube. Or, they engineer the cell in some way to allow them to study those proteins. For transcription factors though, both of these approaches just inherently stop the transcription factor from being able to be in the right confirmation in the right shape to do its job.
As a result, the molecules that people have found using these traditional approaches have just tended to be ineffective. They’re binding to an irrelevant confirmation of the protein. What we do instead, (the fundamental hypothesis of what we do) is we need to be discovering these molecules in the most native system we can: unmodified live human cells.
That’s the differentiating factor for Marmot is bringing new technology to the earliest stages of drug discovery to let us do that. We’ve been mostly focused during this conversation on discovering molecules, but if we take a step back, one thing that we’re starting to look at more now is we can use these technologies to measure any sort of changes in the transcription factor at proteome, or the transcription factor activity in human cells. There’s a lot of interest and activity now in cellular reprogramming, or tools to try slow aging.
These are all phenotypes. Things that are driven by expressing new transcription factors. Same with cell therapy. Can we figure out new transcription factors to help us manipulate cells to have new behaviors that we want as clinicians, as biologists? Tools like this can help us understand what’s going on under the hood in those sorts of applications, as well.
Rich McCormick: Alex, it’s been a pleasure meeting with you today. Thank you for being a guest on Vial’s First In Human podcast. The team here at Vial wishes you and your team at Talus Bio, nothing but future success.
Alex Federation: Thanks so much for having us and being interested in our science.
