Jo Viney, Ph.D., Founder, President & CEO of Seismic Therapeutic, spoke with Precision AQ EVP & Managing Director, Hannah Deresiewicz, at the Jefferies Global Healthcare Conference last week to discuss Seismic's promising I&I pipeline and powerful IMPACT platform enabled by machine learning. Learn more about how Seismic is making a major shift in the discovery and development of immunology therapies in Precision AQ's "Access to the Best and Brightest" series!
Transcript
Hello, I am Hannah Deresiewicz, EVP, Managing Director of Precision AQ's Investor Relations and External Communications team. Today I'm bringing you access to the best and brightest in biotech live from the Jeffreys Healthcare Conference in New York. Joe, it is a pleasure to have you join us today to discuss your latest venture, Seismic Therapeutic. Dr. Viney is the Founder, President and CEO of Seismic Therapeutic, a biotechnology company making a major shift in how immunology therapies are discovered and developed as enabled by machine learning. Joe, AI and machine learning are all the vogue in drug development now. How is your impact? Platform unique in driving the discovery and development of your novel programs. That's a really good question. And actually it's a question that we pondered a lot when we first launched the company about 2 1/2 years ago. So before then, machine learning had really been well applied to the area of small molecule drug discovery, but there was relatively little application in protein therapeutics. There are a couple of companies and a number of academic labs really focused on ML and protein engineering. So when we started the company, we spent a lot of time. Thinking about where's the sweet spot, where will we have most impact? And so we really thought about building a group that intersected machine learning with structural biology and protein engineering. And so that's really how we have built the algorithms that we have. So to give you some example, we and many others are using machine learning to help us augment the drug like properties or the developer ability of a molecule. That's something that's very important for all protein therapeutics, but because we're working on enzymes. An agonist of inhibitory receptors, we also have the ability to use machine learning to dial up or dial down function. But a real cornerstone of our platform are our invisibility and invitee algorithms. And these are designed to invisible Elise, the protein, hide it, mask it, cloak it from the immune system. And this is incredibly important for the approaches that we're using. So we've been really thrilled with how the algorithms have been able to be stitched together. So what it really means in, in sort of layman's terms is that we've really reduced that design. Test cycle time between each iteration of protein engineering. So you still need the protein engineers, but where before they may have had to have worked on develop ability with this line of proteins and function with this line of proteins and chemical liability with this line of proteins and then try to stitch all of those mutations together at the end. Now with machine learning, we can do all of this in parallel process called parallelization. And that's what's made it really exciting. And so for us after launching the company. They were really pleased to see that both our our drug product areas, one being enzymes and one being antibodies. We went from idea to development candidate in just 18 months. And the usual time frames, you know, on average 2 1/2 to three years, we've really cut that time down, meaning we can get drugs to patients more rapidly. No, it's incredibly exciting and that tears me up quickly quite nicely. For my next question, which we'll turn to your to your pipeline. So we saw Biohaven had disappointing data last week with their IG. Greater BHV 1300, I'm curious to know what are your thoughts on the data and how does your PAN IG protease S117 differ and how could it be more effective at lowering IG? Yeah, that's a good question. Well, actually sort of taking us back 2 1/2 three years really the mechanism of reducing IG was still in its infancy and I think we've really seen an explosion in the IG reducing space. And so our Gen. X is out in front with they've got and hydraulic. So they've really. Turn the the benefit of reducing IG levels that that it can be really transformative for patients with autoantibody mediated diseases. So they're well ahead, they're approved. Now we're seeing the next generation and different approaches. So there are a number of next generation FCRN inhibitors including antibodies to the FCRN. So Innovan have this approach, a number of other smaller companies that are a little further behind of that approach. And what's exciting there is the potential for reducing IG levels a little bit. More potentially than they've got, although we'll have to wait and see. By Haven had a very different approach. They were using a liver targeted degradation mechanism to reduce IG levels. These are very different modalities than we're using. We're S 11/17. So what we have engineered is a protease that can very selectively and exquisitely cleave IG in half. So it not only reduces IG, but it can do it much more rapidly and in a sustained fashion. We've put an FT backbone on that protease. But in addition, we've got this added biology, this polypharmacology that actually none of the other IgG reduced the mechanisms have. So we can cleave the bottom end, the effector end off of an IG molecule. So we prevent any complement dependent or antibody dependent cellular cytotoxicity that sort of real mechanisms that cause the tissue damage. But we also cleave the B cell receptor. So now those memory B cells that usually recognise auto antigen, they're disabled, they can't actually see. They're not going to get activated, they're not going to secrete more pathogenic antibodies. And most importantly, they're not going to take up that auto antigen and further exacerbate the immune response. So I think it's just important to remember the the real benefit of IgG reduction for patients. And there's a lot of different mechanisms. There's not necessarily read through from 1 modality to the other other than IG reduction. And now with the polypharmacology with our molecule, we feel that we're going to have the ability to touch. Other pathways that actually just reducing IG alone can't do. Very, very interesting. Now that's incredible. And I know the other thing you're working on, you have a very unique PD1 agonist as well. So hoping you can tell us a little bit about this program. And I think that one's entering the clinic in 2025. Yep, that's right. So both of our programs should be in the clinic in 2025. So the PD1 agonist S 4321, we've used machine learning to really help us understand and model how nature normally agonizes PD one. So this is an inhibitory receptor that's on effective. These cells and on regulatory T cells and so usually you PDL 1, the ligand agonizes PD1 and turns those T cells on. So there's now precedent, there's others ahead of us actually who have PD1 agonists in this space. How we're differentiated is we're really focused on trying to use an affinity of PD1 with our antibody that mimics the natural ligand and that allows us to really get that fast off rate that on, off, on, off on that PD1 receptor. So that we can agonize and turn down those effector T cells and we can activate the regulatory T cells. So the top end of the molecule is a little bit different than some of the other PD1 agonists, but actually it's the bottom end of our molecule that truly differentiates. So we weren't happy with just agonizing PD one on T cells. We wanted to target another inhibitory receptor called FC gamma R2B that's on antigen presenting cells and B cells. And so we use machine learning to really understand how the FCC. Region of an antibody binds to FC gamma R2B, the inhibitory receptor and FC gamma R2A, the activation receptor. These are very, very similar. So engineering NFC to buy one versus the other has been a real challenge. So we've found some very nice mutations that mimic again the affinity that's similar to wildtype IG. What we've done is get that same binding affinity to FC gamma R2B and we've eliminated all the activating FC gamma R2A binding and what we've. Actually also found very recently is that that binding can actually agonize eczema ARTB. So now we can send an inhibitory receptor signal into into the other end of the molecule, into the other cells. So we've got Jewel cell bidirectional activity. We can inhibit T cells on the top end of the molecule and B cells and antigen presenting cells on the bottom end of the molecule. SO2 pathways, two different cell types that we think will drive greater efficacy. No, it's really, really incredible. I mean, both pipeline programs sound amazing and really. Taking advantage of your platform. So congratulations on that. I think wanna switch gears a little bit and talk about you and your experience. So I know you previously founded Pandion Therapeutics and manage it through a successful acquisition by Merck. So curious how that experience compares or contrasts with building and biotech company in today's environment. Yeah. So I think the lessons learned from Pandit, I should say actually panting was my first ever exposure to the startup world. I'd spent most of my career, actually all of my career. Up until that point in large biotech, so panting was very exciting. It was wonderful to be the first employee and it was really a little bit of a dream to go from launching the company and then acquisition by by Merck in four years and one day. It was the exact time frame that the company was in existence. But it was lovely to I really got the thrill and excitement of building a company from scratch, going from idea to development candidate. And what it really taught me was the power. Of thinking hard about what biomarkers you want to build into your early clinical trials to show that differentiation and their efficacy, particularly in the autoimmune disease space where there's a lot of a lot of activity. So back when we launched Pandian in 2017, autoimmunity wasn't necessarily in vogue. I think the difference now is that autoimmunity or now rebranded as INI is very much involved. So that's refreshing to see. I think I learned about the importance of. Bringing together a really experienced and diverse team in the in building Pandian and then the importance of executing and really that, that calculated risk taking to get you to the next inflection point and generate the data that is meaningful and could be appreciated by others. So hopefully I've taken all the best practices from Pandyan and I, we have actually a lot of the Pandyan team, the R&D team with us at seismic. And so it was great. We really hit the ground running at seismic and we're hoping to bring more. Right. This time we have two development candidates headed to the clinic and we're excited to see what difference that can make the patients. Yeah, absolutely. It, it is amazing how quickly it shifts what people are excited about and how much that momentum can be valuable in building a new a new company. So last question for you. You sort of alluded to this, right? Over the course of your career, you've worked on innovative drug development at biopharma companies and all of all sizes from larger ones like Amgen and Biogen to much smaller startups. Curious if you can comment on the people. Side of what it takes to build teams and companies to do pioneering drug development. Yeah, so it's really important to work with smart people, but I think they're smart people around the world. What makes a team really excel is the different perspectives and backgrounds. And so the team that we had at Pandy and I just mentioned a lot of them have come to come to Seismic. We also have some additional team members at Seismic. And the real key to it is having people with different backgrounds, different experiences. Different education and that really enriches how you sort of mix things up and come up with innovative ideas. So I'm a big fan of making sure the team is diverse. You know, my team, I wish that we were all the same and we could get through meetings and everyone could agree with me, but that's not how it is. So we have a lot of spirited and heated discussions, but that's what makes us make the right choices. I'm also really passionate about increasing the diversity of the next generation of scientists. I was very fortunate that I had a lot of opportunities provided to me, but I think there were many other people who were females or underrepresented groups in in biotech who, you know, just didn't have anyone to champion them and help find opportunities for them. So I spend a lot of my time mentoring the next generation. We do a lot of mentoring within the company, but actually most of my leadership team mentors outside the company as well and particularly looking for people who come from different socioeconomic. Backgrounds, different racial backgrounds, different educational backgrounds. We spend a lot of time with bringing in interns from some of these underrepresented or underserved groups, just exposing them to biotech. And actually we end up hiring some of them too. And that really enriches our team. And So what I'm looking for, you know, in the next 1020 years is, is, you know, an industry that is looks like the general population and not like some, you know, ivory tower of elitism that you know. Historically, it was probably a little bit more biotech and far more about, yeah, no, absolutely right. And it's it's true. At the end of the day, the whole industry is working towards the same goal of helping patients. And so making sure everyone has access to the same resources is thinking about the whole population that we want to cure. There's a lot of benefit there. So that was all my questions for you today. It was great having you with us and we'll look forward to all that's to come. Great. Well, thanks so much and thanks for your interest and looking forward to continuing the conversations.To view or add a comment, sign in