For episode 8, we chat with Patrick Hsu, Co-Founder & Core Investigator at Arc Institute. Listen in to find out why Patrick things the future of modality is both underrated and overrated. First In Human is a biotech podcast that interviews industry leaders and investors to learn about their journey to in-human clinical trials. Presented by Vial, a tech-enabled CRO, hosted by Simon Burns, CEO & Co-Founder. Episodes launch weekly on Tuesdays.
Simon Burns: Thank you for joining us today on First In Human, Patrick.
Patrick Hsu: Hi, Simon. Thanks for having me on. Looking forward to the conversation today.
Simon Burns: Awesome. Quick introductions. I’m Simon, co-founder and CEO of Vial. We’re a next generation CRO making trials faster and more efficient. Patrick, we know you well, but give us a quick background.
Patrick Hsu: Yeah, no, sure. I’m a molecular and genome biologist by training. And, my work ranges broadly in molecular tool development, synthetic biology, bioengineering, and more recently, broader ways that we can try to improve how we do science itself. Including finding aspects, but also how do we organize researchers and scientific institutions.
I’m co-founder of the Arc Institute, a research institute I helped start last year. That’s a partnership between UC Berkeley, Stanford, and UCSF. And also a bioengineering professor at Cal. So our research group works in developing new ways of doing genome editing. And we’re broadly excited about the ways that we can take basic science discoveries and translate them beyond simply academic papers to bring them into the real world to make products that can impact patients. So, excited of course about, the exciting work being done a- at Vial. And more broadly better ways of getting medicines to patients faster.
Simon Burns: Love it, love to hear it. Should we talk CRISPR? It’s been an exciting space. Some ups, some clinical holds, some litigation. Lots of next generation CRISPR companies now. been most surprising? You’ve seen it all, from your super early days at Editas.
Patrick Hsu: I think there have been lots of things that we’ve learned over the last, decade that I’ve been working in CRISPR. starting from the time that I was a graduate student doing a lot of the kind of early basic science experiments taking these bacterial Cas9 nucleuses and testing them in human cells for, you doing genome editing for some of the first times. And course after my PhD work, we started thinking about how we can kind of translate these into human therapeutics at Editas and then more broadly, in this very exciting and, fast-paced field of genetic medicines across many different types of modalities. I think there are a few clear lessons that I’ve learned over this time and, and the translation piece of things.
The first is I think is you make a lot of these same mistakes, right? One of the interesting things that I did at Editas was to set up our company seminar series. And our first speaker was John Maraganore uh, who is the founding CEO of Alnylam for RNA inter.. Which was commercializing RNA interference and in many ways had a lot of parallels to the CRISPR story, right? So, we think this is a genetically encoded, genetically programmable system, The undruggable space of targets are now accessible in a programmable fashion, There is, lot of litigation over, the early discoveries multiple different companies trying to commercialize RNAi technology in parallel and to, lots of companies growing really quickly, raising a lot of money, going public at record pace and building multiple different programs across many therapeutic areas in parallel. Because ultimately, we weren’t necessarily sure what would work and so, you would start multiple programs in parallel and do so in a way that may not have been optimally operationally efficient, right? But then, billions of dollars and, you know, many years later, consolidating on a set of applications where you could get easy delivery, And so I think, one of the key learnings is one, it’s ultimately really about focus. it’s really about the people and I think the early choices that you make at the, really the early stages of the company remarkably surprisingly really, really matter. Which, I’m, I’m sure, you know, you’re, familiar with. [Laughs]
So, that comes down to like, you know, how you choose indications. It comes down to, your hiring decisions. It comes down to how you think about delivery. About whether or you’re gonna do multiple different legalities or indications or disease models at the same time or really pick one. Even though you feel certainly with a platform technology that you could be doing multiple things, right? So like, choosing, but more importantly, choosing what you’re not going to do is an extremely important lesson from this geno-ting era, right? Maybe I’m in a new era.
Simon Burns: [ Laughs] Yeah, maybe I’m curious o- on that platform companies, there’s loads of them now who face these challenges. is the advice [00:05:00] just focus, develop a plan, stick to the plan, or partnerships are really, always a push, right?
Patrick Hsu: Well, you know how the thing with advice is. always tends to be a little generic when it’s giving out too high of a level. Certainly platform companies today at the, turn of 2022 into 2023 is going to be quite different from the razzle dazzle, game of the last three years, right? Where we had these, massive sprawling, capital intensive platform builds in many ways have structurally strained basic business things like their cap table construction to straining the real estate availability for, actually, building these labs to the, relatively narrow spec for the scientists that can actually staff all of these companies. Plus the clin ops and tech ops piece of things, right? So I think a lot of what we’re seeing now is a lot more operational focus.
Simon Burns: The founder led bio movement is having, kind of a moment, right now. There’s some good right now, right? You could talk about virtual CROs virtual model using CROs, capital availability, technology and tooling and infrastructure. First, why do you think founder led bio’s having a moment in, Dubai, the general, I guess, theme of the movement.
Patrick Hsu: I think… So, I finished my PhD in 2014 and was interested then very much in how we could, try to translate basic science discoveries, right? on one hand, this wasn’t necessarily that long ago. On the other hand, it very much feels like in a different cultural time, right? I w- go around, the Kendall Square area and meet with the, biotech investors that you’re supposed to talk to, And was sort of kicking the tires on. Oh, maybe I should, go into VC and learn how the pros do biotech company creation because seemed to be really different culturally and a very different set of players and fundamentally a different type of jargon, even from the way that you think about entrepreneurship i- the Bay area. where, you know, I to high school and college.
And the number one question folks kept asking me is, have you done an MBA, kind of business experience do you have? And, I think, today that’s fundamentally different. We have, multiple examples of inventors, like primary inventors from advancing from their PhD or their post docs, Building companies, the shots to be leaders, being able to build a senior executive team, being able to raise money, being able to build a pipeline or build exciting platforms. And I think, in many ways this is, a reflection of how palpably technologies are advancing the pace of biology. It does feel fundamentally different from, you know, than 10 years ago. And I think nativeness by which folks today finishing their training are able to navigate these things as, leading to certain types of companies that are able to, use platforms and use computation and use genetic tools uh, in high throughput approaches i- in ways that have been really rate limiting for previous technologies.
I think that being said, right? I’m a big believer in balance and calibration and I think being able to really rapidly get, putative compositions, Or really rapidly targetive foreign diseases only matters actually so much. End of the day, you are going to need specific programs, and then you’re gonna run into, the real or, very real, I don’t know if they’re the real that, real challenges of drug development, right? How do you actually, do CMC? How do you do in vivo pharmacology? How do you develop an IND-enabling package? How do you enroll in a clinical trial, right? And execute on them.
Simon Burns: Right.
Patrick Hsu: And through those things, are really rate limiting.
Simon Burns: It seems like every time I’m talking to you, you’re thinking about rate limiting steps in drug development infrastructure. You’re spending a lot of time with Arc thinking about that. what are some of the key infrastructure components do you think are, one, will unlock an acceleration to what’s been the most exciting in the last few years of, seeing companies like, twist and just, rapid acceleration of tools.
Patrick Hsu: One of the things that, I asked myself back then was, “Should I, work in industry?” Or should I, go back to academia and run a basic science lab?” And this is back when I was at Editas thinking about my future. And, you know, I think, trainees today grapple with those very same questions, but in a very different landscape of possibilities, And, there’s a lot of, talks today in the culture about the post doc crisis or exodus that fundamentally reflects the really attractive opportunities for folks to, work on exciting science and industry, right? And I think there are, you know, a number of fundamental reasons why that’s the case. I think partly, the things that academics care about and I think that industry cares about have never been closer, j… There’s just a lot of cross appreciation for, what is important. And I think are also able to, you know, raise, large amounts of capital in a commercial setting to, just take ambitious shots, right? Versus, spending, a year writing an R01, maybe a year working on your second submission, another year to get the money and then you can start hiring people. It happens at a very different pace. That’s [00:10:00] genuinely true, On the other hand, right?
I think people today underrate the things that we get in basic science that we get in the university research setting, I feel really fortunate every day to work in, right? Freedom the ability to really do things right and be curiosity driven. And fundamentally, Working in basic science today means that you’re betting on frontier technologies, being able to leapfrog current generations.
One of the things that I learned very early on in therapeutic product development is, the way that things get locked in really early. As soon as you have some sort of, product, rather that’s like an AV or this is my modality, this is how I deliver it, this is my disease. All of those parameters get locked really early on and then you do the slow stuff, All the in vivo experiments, PKPD tox, et cetera, Another technology, if it’s sufficiently better, can fully leapfrog that if it is uniquely enabling, so, I think, thinking through those different timelines is just a really interesting exercise to do these days because I think those [ inaudible 00:14:07] cycles are getting faster and faster.
Simon Burns: Interesting. I saw recently the Broad is now 15 years old which feels almost surprising that it’s done so much in 15 years. I know w- Arc, you were very deliberate in thinking about making it last for generations and being an institution that fostered science for an incredibly long time. what from the Broad, did you take inspiration from? What did you try and, take a new approach to? And just, I would love to hear more about your thought process on institution building for, scientific discovery.
Patrick Hsu: As you know, did my PhD at the Broad and really lucky to, work in a place where there were incredible scientists amazing resources and you know, just a, a lot of ambition to take big swings. And the Broad really, you know, I think popularized and was successful with a certain style of science. Taking advantage of really important bio-technologies since Single Cell and CRISPR and computational methods for human genetics. At a time when all of those things were really taking off, Think, at Arc, in terms of thinking about, research institutes, there have been a number of, fundamental first principles that we’ve been going off of.
The first is all of these really exciting technological trends, cell therapies, CRISPR, mRNA. They promise the potential of historic medical breakthroughs in the years ahead. Yet, it feels that their systems and processes by which we do science makes it feel like we’re really just kind of default likely to undershoot this potential, the question is, what can we do to, make this cycle of progress faster? I think the Broad has pioneered, One particular model for doing first rate science. It actually turns out that there’s a really illustrious history of biomedical research organizations in, for example, the United States over the last century. But also, internationally with the Max Planck Institutes in Germany, the Crick Institute in London Howard Hughes Medical Institute and Janelia Farm in Virginia and so on. The Broad, the Whitehead, Salk and Scripps in San Diego, et cetera.
I think most of these places though, if you look at them, are fundamentally they have world-class research professors and groups that are working on their own problems, And modern research has become increasingly dependent on complex biomedical tooling. Yet, I think certain super labs grow to effectively own internal platforms. So I think there’s an aspect of centralization of all of these complex capabilities that we’re doing at Arc, Being able to provide long-term, full-throated research funding so that our investigators can really work on their most important ideas. And to support them in a way that feels really operationally effective, Algae’s very slow, right? And one of the things that we learn in a field like mine, right, is operational efficiency really matters.
Simon Burns: [ Laughs]
Patrick Hsu: So, if you’re, if you [inaudible 00:17:39] something 20% faster week on week over the course of a biology project that actually gives you massive compounded returns, and so these are the kinds of things that we’re excited about and in terms of building our technology centers in addition to our core investigator pillar at Arc, where we’re bringing together trained research scientists organized in industry-style teams to do focus-batched technology development. We’d like Arc as an organization to be able to, work on research programs that we think are really important.
And one of our initial institute-wide strategic initiatives revolves around Alzheimer’s Disease but we’re broadly excited about understanding and treating complex disorders like cancer, auto-immunity and others over time. We have like significant growth plans ahead. Today we’re 85 people. it’s a really unique privilege to, be able to really operationally plan for the growth that we’d like. We have [00:15:00] lots of hiring that’s happening. But it’s been a fun year.
Simon Burns: Wow. Congrats on 85 people, that’s awesome. Let’s go into a segment of overrated and underrated. I want to get your take on some potentially controversial topics. First off, new ways of doing gene editing base editing, [ a lot of talk. DNA integration is getting a lot of talk. Overrated, underrated, the future of modality?
Patrick Hsu: I would say both overrated and underrated. on one hand, it’s self-serving. On the other hand, I think you have to you have to be able to, smell your own something, right? [ so like, on one hand, they’re overrated because all of these things are fundamentally bottle-necked by how you do the delivery, And there’s a very small number of ways that we can actually get these large macro-molecule machines into the right cell type or the right tissues in the body in order to make these changes that we want. that’s ordinarily in the therapeutic setting. all of these, exciting gene editing platform companies require the ability of actually get them in, yet on the other hand I think extremely underrated because I don’t think we have fully grasped what the ability to, really program biology in this way is going to enable, that sort of, the one gene editing of knocking out genes is really fundamentally only making changes on things that are already there. ability, for example, of DNA integrations, which we work on actively in the lab to put in large pieces of code is going to let us, put in synthetic programs into biology in a way that’s never been possible. I think that’s going to drive like, new generations of syn-bio these new capabilities will drive new ideas.
So, there’s a sort of famous quote from Sydney Brenner, one of the of fathers of modern molecular biology. He says, and I’m going to butcher that, Progress and science depends on, new technologies, new experiments and new ideas basically in that order.”
Simon Burns: Huh.
Patrick Hsu: Right? So these new technologies are fundamentally driving new capabilities that make us imagine what we can differently.
Simon Burns: Wow, that was great. Let’s talk about AI. Lots of ink spilled the last few weeks about the impact of AI in, biotech. Curious of your, take both on the protein, structure, prediction stuff from, metas work in the space. And also just the computational chemistry approach, we’ve seen different and I’m curious y- your take first on what’s enabled by AlphaFold?
Patrick Hsu: There’s… I think sort of like, the previous section, there are aspects that are, again, both over and underrated. And I’m sorry you’re to a professor who’s-
Simon Burns: Great, no. Please.
Patrick Hsu: Who’s split hairs like this, but… [Laughs] I think on one hand overrated, right? And so overrated in the sense that this is the corpus of protein structures that the research community has generated over the last few decades is probably the most well-organized atomic resolution data center that we have in all of biology, all right?
Simon Burns: [Laughs]
Patrick Hsu: So, unsurprisingly, everyone is trying to find what is the next AlphaFold data set? What is it look like? How can we do this again? But I don’t think it exists, so I think it’s going to be hard. Also, in my lab and, in genetic medicines, We sort of broadly ushered in nucleic acid binding proteins, right? Things that bind DNA and RNA. And so AlphaFold fundamentally is useful for proteins, and not their DNA or RNA bound states, right? We just have very few structures of nucleic acid bound states. have enough training data to be able to make accurate predictions, So, I think it’s changing many fields of biology. Has not arguably really been a state-change for us, so that is, another area that is going to require a lot of data that is experimentally bottle-necked to create, right? On the other hand, I think the underrated side of this is we’re able to do things or… You know, in just the last week, I think this work on diffusion models for generative protein design is now very firmly in a territory of like, exploded brain-
Simon Burns: [Laughs]
Patrick Hsu: And like, this is just straight up science fiction stuff, right? The idea that you could write-
Simon Burns: [inaudible 00:23:57]
Patrick Hsu: That’s right, that’s right. Galaxy made it… the, the idea that could write into some like, generative model the way you would for like uh, GPT Chat. Design me a protein that will bind to X, right? that it’ll actually, give it a credible shot that is, experimentally testable is frankly unbelievable.
Simon Burns: Wow science fiction coming for us all. Okay, let’s talk about bio. A lot of technology founders now in biotech, they’re taking a scale approach to things. They’re re-imagining things from the ground up. First principles thinking being applied from technology founders. What do you think the impact of technology founders coming to biotech is going to be with, the tech bio movement?
Patrick Hsu: Overrated.
Simon Burns: Wow, shots fired. Coming for me. [Laughs]
Patrick Hsu: Oh, no. I- I wouldn’t say the rate limiting step for all of these is how do you actually get through clinical trials? How do you actually get drugs into people? And so maybe I was speaking more to,
Simon Burns: Yeah. Yeah, yeah, yeah.
Patrick Hsu: The sort, the, the, [00:20:00] first part of the, drug development stack. But I think the clinical trial aspects are very underrated.
Simon Burns: Thank you. You saved yourself. Let’s talk about automated labs. the future of fully virtual bio tech that’s running on automated labs. What’s your sense of that as enabling technology? Overrated, underrated?
Patrick Hsu: Totally overrated.
Simon Burns: And then lastly, you’ve been on the east coast, you’ve on the west coast. West coast biotech, east coast biotech, who wins?
Patrick Hsu: I think they’re very different. It’s hard to ignore the density, the pace and frankly the regional priority that’s being placed on biotech company creation in the Boston Cambridge area. Every time I go there, I’m, you know, multiple companies there that have spun out of my lab. It’s, it’s just amazing. the area has, a lot of lessons to learn. I- I guess from the [inaudible capital of the west coast need more buildings, right? Fundamentally, this type of work happens in person.
Simon Burns: Well, with that, Patrick, thank you so much. This was great.
Patrick Hsu: This was a pleasure. Thanks, Simon.