Autonomous Microscopy with Kevin Dean

Welcome to the Axial Podcast. Axial is an early-stage investment firm based in San Francisco. We partner with great founders and inventors investing in early-stage life science companies, often when they are no more than an idea. Axial is fanatical about helping the rare inventor who's compelled to build their own enduring business. Hey Kevin, well, thanks for doing this podcast. We've had a really fun conversation, um, and always excited to talk to like smart football players, uh, because it's very rare. Um, but welcome, thanks for your time. Maybe we can do a brief introduction and we just start off the conversation.

Awesome. Well, thank you for having me. So yeah, my name is Kevin Dean. I'm originally from a small town in Northern California called Weed, California. It's mainly famous for being in the back of a Sublime album, and, um, Yeah, so I grew up in a small business and, you know, essentially started working in a restaurant at about age 13. And, and I think that was an incredibly motivating experience to go to school and become, you know, ultimately a scientist today.

Cool. Okay. So you grew up in California. Is Weed north or south California?

It's north. It's about 45 miles south of Oregon on I-5.

Interesting. But so long from Long Beach. The only reason I know about Sublime is I was born in Long Beach, so I know about Sublime. It's a so weird connection. Oh, we'll talk about some offline, but was there a moment where you realize either you want to become a scientist or you realize you were good at it? Was it high school? Was there a teacher?

You know, ultimately, I, I think I had to stumble into it. I was always I always enjoyed science at the high school level, but it was a small school and, you know, there's only so far you can go. I remember taking AP Physics without a teacher and I realized then that, you know, I wasn't a physicist. But really, once I got to Willamette in undergrad and had excellent chemistry teachers, did I start to really realize that this is something that would be going after.

Cool. Okay, and so then you go to Willamette, you're All-American football player and a world-class scientist. I think you were studying— what was your major, chemistry, or—

yeah, it was, uh, physical chemistry. And I feel like I have to put an asterisk on that, uh, All-American academic, regional academic All-American, but you know, that's all Division I, II, and III, so I'll take it. And But, you know, so, yeah, so I assume—

Same thing, I'm in a hall of— I'm in a football hall of fame for Orange County, but like for academics.

Like, count it, you know, count it. I like it. I love seeing the nerdy people on the squad, you know, and the amount of work you have to put in to play sports and do something like chemistry was just immense. You know, I had 14— labs, and essentially, you know, you— we would be running from lab to football practice. And really, Willamette was an incredible place because of that. We did have a lot of people, you know, that were there to get a great education, but also we wanted to win on the weekends. So, so yeah, like, got to Willamette, basically had a great freshman chemistry teacher. I knew that I was probably going to go pre-med or science somehow. I didn't really have a clear, distinct vision like so many kids have these days. And, you know, I just kind of found one thing after another. You know, I had a great professor at Willamette named Chuck Williamson, and he offered me a research job one summer. And that was actually after my sophomore year. After my freshman year, I was on the paint squad crew for Willamette. And so random, you know, like I went from one thing to another, from painting to studying phase transitions in a physical chemistry lab. And then that just, like I said, one thing led to the next, and before I knew it, I was applying to graduate schools, going to University of Colorado. I found an amazing PI there, Amy Palmer, who really empowered us to do great things and had exciting projects, and, you know, like I said, I just went from that to running a facility for a year, then going into a postdoc.

And one thing to get in your head during college— so you're doing chemistry, you go to practice and run into organic chemistry labs, and that must have been stressful but really fun. What was the key point where you realized, hey, I can do this professionally? Especially in the backdrop of— I'm assuming most of your friend group and your teammates weren't thinking about PhD programs. They were thinking about going pro— I'm joking— but like they were thinking about like, you know, just a job and maybe not PhD. And so like, how did you— like, was there like a person or like a moment where you realized, hey, I can do this professionally?

Yeah, you know, that what's interesting is that we were surrounded by people that were all very driven. And so I would be studying with my friend Lee Wachowski, who was on the swim team, and Tim Alton, who was another excellent football player at Willamette. And, you know, we, we went through chemistry together as kind of a small cluster of student athletes. Another famous person who was there is Nick Simmons. Actually, the— I mean, the only famous person. I would not consider any of us there yet. Simmons was also in our chemistry group, and he was a, I think, a 3-time Olympian, so the 800-meter dash. And so, like, if looking back, it's pretty ridiculous. I think we had a class of 11 people, and I think Nick is the only one without an advanced degree, but I mean, he's pre-Olympic, you know, like experiences. So it's an incredibly driven, like smart group of people, and I think maybe we didn't realize we could make a living out of it, but we didn't doubt it either.

Wow, that's so interesting that like you had this cluster of talent in a small school. That's like kind of lucky, to be honest, in like retrospect. You had all these talented people all in one place. That kind of usually talent clusters in a lot of ways. Or sometimes people kind of self-organize almost, and then they find each other. And then, do you feel like that having that friend group motivated you and made you a better—

I mean, I was constantly pissed that Tim was getting better grades than me. So, um, you know, we were competitive, and but we also helped each other. And it was, you know, I think that there's also a bit of a tradition of people just doing well. Same with University of Colorado, you know, you're around talented people and that's one of the most exciting parts about working in this sort of sector is that everyone is smart, you know, like everyone is good. And, and so the creativity and the energy that you can get from that's exciting. And so even looking back at my colleagues from University of Colorado, they're all doing well, you know, because they are all talented. And I don't think it was necessarily like a wave of talent that came in. It was almost like an institutional expectation for talent and ambition. And so, I mean, but who knows? You're right, these things can come in waves.

That's cool. I mean, I just, I remember while I was in college, I would take chemistry courses. I was a biochemistry major, but I'd take chemistry courses, and everybody's like, you know, I call them chemistry bros. They're, they're my friends, but like, there were 8 runners. They're all track and field. Yeah, and they all take chemistry courses together. And now looking back, it's like 10 years ago, they run laboratories, they're artists, they're like big-time executives. And it's like, I don't know, it's like having that friend group and how it self-organizes, it kind of is a self-fulfilling prophecy. We find a really good group, yeah, support each other, we motivate each other, and it kind of— that starting point seems to then like lead to a lot of great things down the line. That kind of— okay, so go to— why'd you pick Colorado? And then what was the experience going from maybe a small school to like a bigger school? Um, oh yeah, last question is, are you excited about the next football season? Deion Sanders is the coach.

Yeah, I can tell you that a lot of my friends are very excited about primetime.

I might go have an investment in Boulder. I'm gonna go, uh, yeah, hang out there and just watch some games.

Well, if you need someone to catch a game with, let me know.

I'll let you know.

Um, yeah, no, I'm curious. Um, it will be exciting to see it. You know, they definitely needed a change. You cannot do much worse, so he can only, only up from where he's at right now. So, uh, Yeah, so it was a very interesting transition going from a small liberal arts school in the Northwest to a big state school in Colorado. When we visited there for the grad student recruitment, you know, they took us to the top level of the football stadium where you're just overlooking the football stadium and the Rocky Mountains and the foothills right behind you. And it was just stunning. I knew at that point that I was like, I was done for University of Washington. Was not going to happen. So there was— I was choosing between UW and CU, and CU is still, you know, in retrospect, that was the perfect place for me. And so it was a big transition. You know, you show up in grad school and it's essentially— it was similar to how I felt when I got to college for undergrad. It's like you just kind of walk into a wall and you just get smashed, right? Like all of a sudden the classes are harder, the people are smarter., and you just have to buckle down and step it up, right? Like, that's the only choice you have. And especially in grad school, because your expectations are that you're, you're taking classes, you're rotating in labs, and you're also teaching a little bit. And so all of a sudden, your, your burden goes through the roof. And, you know, unlike football, where you say, all right, I've got 4 to 5 hours a day where I'm setting decide to do this. Like, if you're in the lab and something's not working, you're just like in the lab until it works. And so that first year, you know, we just, we really went hard. We worked and then when we weren't working, we took advantage of all the great things that Boulder has, the trails, the bars, you know, the football games and just had a blast.

How did you— I have friends of mine who are elite athletes and then they go to grad school or even get a job. How did you feel? I had this experience in a small scale too. How did you fill in that hole? 'Cause like doing a sport takes up like half your day. It's like, how did you like have an outlet? For me, I ended up playing basketball. It took like 2 years to figure that out. But like, was there like a period in time where you had to like figure out a new hobby almost to like take it?

Yeah, I had to figure out a way to be motivated, you know? I am terribly motivated by vanity, which when you're approaching 40 is pretty much the only reason to really exercise anymore. I just don't want to look worse. And, um, and so I had to really get over that, you know, like it's so motivating to be like, I have to get in shape for next season, but now next season is no longer. And so I think I really got into mountain biking. I did join some gyms in And, but there was nothing quite that could ever fill the, the outlet of being around 100 of your friends. And maybe if you have a bad day and kind of blow a test, um, the outlet of running full speed into another human— that's, that's typically frowned upon outside of it.

It took me forever to get— it took me like 2 years to like decompress. I almost felt weird, like, like, what's going on? My freshman year in college, I'm like I was like, what's going on here? I have so much time, but I'm like, feel so like tired. And so I'm getting so much sleep, but I'm exhausted. And it was just like, I had to like learn how to play basketball and, you know, do stuff like that. Just do physical activities. But also the second question is like, I have friends who have a friend, Greg. He was with this small school in Nebraska, went to grad school at Berkeley. How did you, did you get lab experience before grad school? Or was there some sort of learning curve to like be in lab all day and like have to learn how to plan experiments and be a little more independent?

Yeah, there's definitely growth, you know, like you go from an undergraduate position and you're kind of— there's different levels of expectations. And there's some students that, you know, don't perform at the expected level but perform much better. And so I think at an undergraduate level, you're basically expected to be technically proficient, right? Then when you're in grad school, all of a sudden it's like you're not supposed to be just technically proficient. You're supposed to be more thoughtful about how you're doing things. You have to manage your own time. And so there is a learning curve there. You know, it's really like at a postdoc level, you're supposed to be fully independent. You're supposed to not only be able to come up with your own projects, but you have to, you know, really drive them to fruition and explain them. And I you can immediately write it down and get someone to give you some money, or at least someone to publish it in the meantime, right? And so each one of these comes with learning curves. And I think just having the right environment is key to that, right? Like, you need to— you don't want to learn these things the hard way necessarily, although I, I do learn quite effectively the hard way. Good advisors that and colleagues that are willing to invest in you and teach you the right ways is important.

Let's talk about that. Okay, you're in grad school. How did you pick your lab?

I think you had a dual lab, uh, like maybe in Ralph Himness.

Yeah. And then how did you pick them as your lab? And then talk about picking the right mentor too. It's always kind of the most important decision, especially as a scientist, is find the right lab and mentorship to kind of help guide you along some of these things you're talking about.

Yeah. And so, you know, we did rotations our first year and I did, I worked with Amy and another guy named David Nesbitt and a third named Artie. And so it just became so immediately clear that my interests were more aligned with Amy Citrus. Yeah. And so it was just a no-brainer. She was a great person. She was, uh, totally supportive, creative, excited, um, the type of person that makes— builds you up and doesn't make you, um, you know, there are some people that will chop your knees off a little bit, and she had none of those traits. So And she had big expectations for you, so she really did foster that growth. And so when it came down to joining the lab, she had just gotten a grant with Ralph, and the whole idea there was we just need to develop better ways to identify mutant fluorescent proteins. And it was involving microfluidics and doing spectroscopy and using lasers to grab cells and move them around. In retrospect, it was perhaps the most complicated way to do that project, but it had all the cool stuff I liked. It was just a no-brainer. I, I— she asked if I would be interested in joining the lab, and I didn't even leave the office. I told her absolutely, I'm in.

That's cool. So in my opinion, you're one of the best people in microscopy. And, and so it seems like grad school is a point in time where you kind of got down that rabbit hole. Yeah. And just to maybe backtrack and find the starting point, did you get interested in imaging in college or was it grad school or like what, at what point did you realize, hey, this microscopy imaging thing, I like it, I'm gonna, I'm gonna continue to practice this and focus on this? Was it just at that point of joining the lab or? Yeah.

So Amy's lab was very much about using microscopy as a method to evaluate how things move around the cells. How does zinc go from, say, the mitochondria to the cytosol, and how does that influence cellular behavior? And so, she creates these very nice sensors that report out at a subcellular level fluctuations in these factors, whether it be zinc or calcium or, you know, a GTPase activity. And so, I think it's just a natural thing for you to look at this and be like, this is insanely powerful. I do think if I go one step back, I always loved photography. And so now I just got the fanciest photograph— like, camera in the world, right? Like, I've— I would never be able to afford this camera. It can see single molecules. And, and so there's a really beautiful aspect of it. Like, you see nice microscopy, and it is art, but it's more importantly, it's also like a spatially— quantitative readout of, say, protein abundance or enzyme activity or gene expression. And there's no way you can infer that information other than ways like methods like microscopy that give you both space and time insight.

What's the term? Nothing is A picture's worth a thousand words. Absolutely.

Sublime.

So that's kind of the value prop of microscopy. Well, they have a lot of friends who work in microscopes, and so, and when I talk to them, they're almost like car mechanics. You know, they have to like, you know, use a monkey wrench, or there's always a story. Any funny stories during grad school? Like you broke a microscope, had to fix it at night. I usually, my microscope friends usually like, I like in grad school, I I had these European guys, like 4 of them, they would maintain the latest sheet microscope. And they'd always be smoking at midnight 'cause they're just stressed out. Like, they're smoking like, oh my gosh, we gotta fix this microscope. And they're trying to talk it through. Any funny stories like that when you were in grad school?

Ooh, I think just 1 or 2 horror stories. You know, like we would have environment chambers that maintain temperature, CO2, humidity. So if you're trying to do an experiment, you want to keep these cells happy for 24, 48 hours. But, you know, this is basically a chamber sitting on top of a half a million dollar microscope, and if condensation forms, it could be lights out. And so we came in one morning and there was just water running down the side of this brand new microscope. And so there's always things like that. You know, and you are right that we are essentially a bunch of like tool heads to some extent. Like we like, you know, there's certain types of science that draw a certain type of person, I think. And microscopists really love to tinker with things. We call it, we build things out of fancy Legos all the time, right? Like just Legos that do the coolest things. Like if you want to move this thing 1,000 times per second or whatever it may be that you want to do. And then you get all these Legos functioning together in a synchronized fashion and they can just do stunning things. And so, yeah, we're constantly building and playing with things and you gotta love that aspect of the job. And I think that's why we do it in part, right?

I think just like I have for my micro— I always have, like I have friends who are car mechanics. Growing up, I have friends who like have a, friend whose dad is a car mechanic, and he always had these tools in the garage and even at home. And I go visit my microscopy friend, they always have similar bags of tools, and just kind of reminds me of like car mechanics.

Yeah, I mean, Reto Fiolka, my colleague next door, you know, he was in aerospace engineering before this, you know, like, um, you, you're like Eric Betzig who won the Nobel Prize building like a crazy like forklift-like machine for his father's business back in the day. And so it's like, yeah, these are people that like, it's like applied physics with like lasers and cool gadgets. And so, and then I think on the most rewarding thing is like, yeah, you know, I'm sure it's really nice when a car runs well, but like, man, is it beautiful when you see like something that someone's never seen before in a sample, right?

That's absolutely true. And so you're in grad school. Maybe you talk about the projects you're working on and then how that set you up to do a postdoc, and then you got to set up the BioFrontiers imaging facility and all that stuff.

So like I said, we were developing a microfluidic cell sorter that could rapidly measure things like quantum efficiency, which is how, how efficient a fluorescent molecule is. If it absorbs 100 photons, can it emit 90, or does it emit Trying to develop brighter, more stable, ultimately, photons out of proteins. But the big— as a whole was really interested in developing sensors to read out molecular activities in cells. And so as a part of that, I was around a lot. The microfluidic cell sorter microscope we built. And, and so, you know, as I hope to be a good grad student, you just read a lot. And as a consequence, I was exposed to someone named Galen Denooser, who has been a leader for probably the last few decades developing ways to actually extract from the images. So I had been focused on wiring the— and I knew that we had to like extract this biological information, but there was kind of a level, and he provided really the most elegant, beautiful, rich ways to like extract information from these cells. And, and so I, I just kept migrating closer in that direction. So That's ultimately where I went for a postdoc. But like you said, there, there was a little bit of a layover there on the way. And so essentially, while I was wrapping up my work, I'd basically, during my PhD, I'd helped establish this microscope facility in the newly founded BioFrontiers Institute. And, you know, just one thing led to another, and I had, you know, one or two papers I still had to wrap up from grad school, and I'd basically already started this facility. I could get a pay raise. And so I just took over operation of that facility that I kind of founded as a grad student. And I really did enjoy, like, pushing people towards more rigorous, better biological imaging. But deep down, I knew that, like, I wanted to be the one doing the pushing. Like, I wanted to push push other people. And so that's why in the long run, and try to get up with Goudens.

Cool. So you're at Colorado, you get your PhD. Um, let's see, let me— did you, um, did you like— how did you— when you got your PhD or maybe got your first paper published, did you ever like— did you like— do you have like an email list of your college buddies or something? Or do you like— are you in a group chat and you like— do you brag? Or Did you kind of like, or did they come to your graduation? I'm just trying to connect the work you did in grad school to maybe this crew of people you had in college who, you know, kind of seems formal then.

Yeah. So I was really, so Lee, my like study partner from Willamette, you know, she came to University of Colorado for school.

She was my roommate.

So especially on the first part of the grad school, school, you know, you know, journey, she was there with me. Cool. She ultimately went back to Berkeley to get her PhD, but you know, when it came to like, you know, my buddies are somewhat interested in the stuff that I do, but don't really care. Like, we, we're just talking trash about this 15 years That's funny.

And so, okay, so you're, you're running this institute at BioFrontiers. How did you think about your next decision? Like, do you, do you want to become a professor? It seems like that was the case. And did you have other options? Maybe get a job in industry, maybe a biotech company hire you?

You got that. This, uh, the decision tree aspect. So All right, so complex thing. I have never been purely an academic lifestyle, although I do think, you know, if you can land a spot like University of Colorado where you got the, the Flatirons behind you and you've got Prime coming through and, you know, that's an amazing life. So I was always open to it, but, you know, coming from a family that was purely business and being already the black sheep, I was still intrigued by you know, maybe going into a tech sector myself. And so it's very interesting in retrospect. So let's say I'm— I go from grad school, killing it, making $30K about a year, you know, living large, lots of roommates. And then I transition to this faculty or this kind of instructor-level position. Running the facility, salary doubles. I'm now feeling very flush, uh, you know, at $60K in Boulder, not a cheap place, but still feeling better than $30K. And so at that point, you know, I was looking, what was important to me is that I'd not limit my opportunities. I think that's always been one major theme for me is if I start to feel like there's a you know, a barrier above me, I start to get really, you know, nervous. And so I went into this faculty-like position where running the facility, and it was very clear to me that like, there's no way that they're going to move me up from this. Like, this is what I'll be doing for a while. So I think that was one of the first drivers for me to say, I got to look elsewhere. You know, in retrospect, I don't think they had any intention for me staying. I think they wanted me to go elsewhere, right? Like, they were good mentors and that they had bigger expectations for me at the time. That's at least what I think. So, so yeah, I then looked around like, all right, do I want to jump into local tech industry? Boulder's got a great tech industry. And, you know, ironically, I saw some of my friends from grad school go into these jobs and I was like, these jobs are like they should be getting paid more. It's like, there's no way I'm gonna jump into the tech sector and basically, you know, the way people always feel is that like, if you leave the academic track, it's pretty much irreversible. That's less of a thing these days, especially with a few more progressively minded institutions. I know in Boston, they're now going back and forth. UT Southwestern is trying to do that here in Dallas as well. But at that time, it felt like you leave academia, you're done, right? Like, you're going to go in tech. And I just wasn't quite willing to jump for what I saw their salaries going at during that time. And so I decided that I would look into a postdoc because at least that, you know, advance my knowledge, learn something new, maybe marketable after that point. Now, the ironic thing is that every one of my friends who did go into tech at that time, they're all doing very well now. So, um, they may have started low, but they have made up for lost time since then. And, and so, like, I do think that they made the right choice. For themselves because they knew they wanted to go into tech. And then I think the best thing you can do is just go straight into tech if that's what you want to do. Like, don't waste your time. Don't tread water in a postdoc. Just go do what you want to do. And so they're all doing pretty well now. The most annoying of my friends from grad school is retired and he's like 38. So I don't want to talk about that. That's a sensitive subject.

But Don't worry, you're gonna— all the stuff you're doing now is going to be really valid. You got to figure out how to like productize and get rewarded.

So yeah, I, I just need to find a way to be able to like, um, yeah, that's not— and I'm not going to catch up anytime soon, so I got other things.

Um, you will. I don't feel it. I think you're working on something really important and you gotta figure out how to— you have to get tenure first. That's—

see, that's another— tenure, then I'll focus on selling something for over a billion.

Exactly.

Um, no, I'm super— I'm totally just being sarcastic there. But, um, all right, so yeah, like, so, all right, so I decided where I'm at, I'm gonna move somewhere else. And so one, you know, I sent out a couple postdoc letters, like, put like cover letters, and this is why, who I am, this is why I find your work interesting, and I'd like to join your lab. And, you know, I got a few returns back. And, but I think one of the most important parts of the story is I actually went to a conference called Seeing is Believing, which is hosted at Imboul in Heidelberg. And that's where I ran into Gaudens one-on-one. And so chatting with him, talking, you know, basically about the technical things in science. And he basically at that point realized that, like, you know, if I send him a letter, maybe he should respond, I hope. And that's really how it happened. I interviewed at Janelia, which was pretty new at the time, with Howard Hughes Medical Institute. And they kind of wanted me for kind of this hybrid postdoc facility management position. And then I came down to Dallas and Gowns was just moving his lab from Harvard down here. And, you know, I got to kind of realize that, you know, he had big ambitions for the space and we could essentially kind of start something new again. And You know, it's one of those things where I made up my mind. I basically told him the same day, like, yes, I'm in. And a few months later, I was here.

Well, Genelli would have been sweet too. I'm a big fan of Luke Levis. I'm sure you know who Luke is. Yeah, Luke's amazing work. Uh, he's one of these—

yeah.

And then Eric's another— I don't know, I don't know Eric at all, but Eric seems, you know, he seems— he worked at a family business, he said set for 10 years and then won a Nobel Prize and then joined Genelia. It's crazy story. Uh, we'll have him on the podcast later, dude. Yeah, he's, he's awesome. So, okay, you're at UT Southwestern. How did you go about transitioning? And did you have to pick your own project? What were you working on? And, and, and how did you like start thinking about like forming your ideas and you reached the research directions?

Yeah. So I came and I partnered up with Reto Fiocco right away, and our goal was to establish better ways to evaluate cells in tissue context. And so this is kind of the big umbrella that Gauden said, this is where I want to put my focus, and you guys can do whatever you want. Just do it in this, under this big umbrella, right? Like, and so we just took off like kids in a candy shop. We just started building absolutely everything we could, trying to find a way to, to image better. And, and so, yeah, like, within that, we came up with, um, you know, what's still one of our most highly used microscopes is Axially Swept Light Sheet Microscopy. It's a mouthful, but this allows us to essentially decouple this common trade-off in light sheet microscopy, which is that if you want a large field of view, you're going to have bad Z resolution. And so we, we came up with a mechanism that allows us to achieve pretty much the highest Z resolution in light sheet microscopy and do it over a large field of view. We give up a couple things to achieve that, but, you know, you know, everything has its own space. And we were able to find some great applications, just, you know, and it goes to show how long it takes to kind of support a project. But that microscope really just kind of came to fruition recently when we got a Nature paper from a postdoc in Gown's lab, just came out like maybe 2, 3 weeks ago. And so, yeah, like that was essentially how we got here and, you know, We, we built a bunch of microscopes. We dabbled with a lot of things. We've had light sheets. We've come up with oblique plane microscopes, 2 and 3 photon laser scanning microscopes. We've done some adaptive optics, which allows you to correct for distortions in the light that you experience as you go into a tissue. And today, yeah, that, that just continues. So, You know, it didn't really answer your question very well. It's like, how did I choose my project? It was just, it was, we just got this major umbrella that we could operate in almost free of responsibility and just meshed well and we just went crazy.

Cool. Yeah, it seems like you seem to be the go-to person at UP Southwestern and everybody wants to collaborate with you and then kind of say, oh, you have this really fancy tool, what, what if, what, what can I use with it? And now you you're actually doing stuff for your tools. Maybe you could talk more about like microscopy, maybe light sheet in particular, like what have been the key drivers for the gains you've been able to achieve over the others? Has it been better optics, probes, a bunch of other, maybe other stuff, maybe a combination of all that stuff? The last maybe 5, 10 years, these gains in microscopy we've seen, you know, what can you kind of credit them to?

Yeah, so my, my opinion is that we're, you know, all of us have the same optics, and so you can only capture light with an objective so well. You know, we have all the same cameras. All the cameras have gotten better and better. They've gotten lower noise, bigger sensors, um, and, you know, higher quantum efficiencies. So where do I really see, like, the gains? I think It comes down to almost like kind of synergy of all these different technologies. Like, we got really good FPGAs that can control the microscope deterministically, incredibly fast. We've got great lasers, you know, like essentially the Legos that we're building out of have just gone to be really, really good. And And so you, you can add and remove modules and do so very, you know, not easily, but you can, you can add functionality to a microscope and then all of a sudden you add all these things together and it becomes something greater than the sum of its parts. And so like for us, one of the biggest realizations was that if we could somehow control the the position of the light sheet along its propagation direction and do so deterministically, then we could, you know, come up with a solution. And by piecing together different papers that have existed in literature for the last 10, 15 years, you know, there was a way. And you take all that creativity and these tools, and that's just kind of why science is so fun. We wouldn't have been able to come up with that independently, but because we had such great people around and, you know, you're chatting and tossing around ideas and someone says, well, what about this? Or how about this? And then all of a sudden you have something that's just like, yeah, that's providing insight into, you know, an important process in cancer. And that's just amazing.

Awesome. And then maybe transition towards this kind of concept of you have all these various tools and they kind of add them together like Legos and you create this whole new way to, in a better way to image. I don't know if you came up with this concept, but you definitely are a pioneer in this idea of like a self-driving microscope. And so could you talk more about like why is that possible now? Yeah. And kind of the importance and then What's state of the art right now and what are the big problems you need to solve? And there's more nuance, right? You're not going to automate really super resolution microscopes. Maybe you can correct me there. I'm not the expert. You are. But is there nuance in terms of what type of microscopes can be automated to what extent and stuff like that? Yeah.

And so I think that there's a lot labs working in this smarter automated microscopy sector. And so what is it that made this possible is I would say like computer programming, I've gotten to be— there's so many rich tools that you can just grab out there. So Python is a great example. If you just want like this deep learning module, just download PyTorch or TensorFlow, right? And, or maybe you want this object detection, you can just grab this YOLO detector, as they call it, you only look once. And so now you have all this rich computer vision that's out there. And then you can just start to connect the loop, right? Like you can acquire data, you can instantly start to gain insight from that data, and then feed feed that back into the acquisition of the, the, you know, the images by this, you know, computer. So to do this well does require some sophistication. I've learned more than I've ever learned about software in the last year and a half simply because I now have to help maintain proper, like, engineering practices and have a lot of people working on a project that you want the software to be sustainable. And to make it sustainable, you have to have good practices. So, um, before I deviate too much, I think that's really one of the key drivers. Now everyone has a little bit of a different take on what they want from the smart microscope. Jan Husken is perhaps the first person to really term this smart microscopy, and he's done things like automatically rotate a zebrafish to identify the best ways to image it. There's a, you know, recent paper from Wes Legon's lab at UNC where he identifies cells that are undergoing certain stages of mitosis, and then you can switch imaging modes and look at only the cells that are actively dividing. There's a lot of different examples out there. Philip Keller's lab came up with ways to dynamically track embryos as they grow and adjust the light sheet positions. I was with Loïc Royer, who's at Chan Zuckerberg now. And, and so there's a lot of different avenues now. Where we're going is slightly different. And so all these applications I told you were kind of like identifying in a live imaging system, adapting the microscope to the specimen itself. And we're doing a slightly different approach where we are taking a biological problem that's challenging and then kind of adapting the microscope for that. So in particular, we're trying to look at how cells, when they are metastasizing, how do they actually find the target tissue that they want to go and invade? And how do they grow and proliferate in that context? And so if we want to look at that, we have to find an ASAP. We want to be able to find a single cell in an entire mouse lung, kidney, or liver. And if we try to do that with a standard confocal microscope in the core facility down the hallway, you know, back-of-the-envelope calculations suggest it will take us 4 years to image one. So can we come up with a smart microscope that can then image at, say, lower resolution, adapt to the specimen, automatically identify the boundaries of the tissue so that we don't image empty space? Can we have it automatically track the alignment and focus of the microscope so that every single image that's acquired is in focus and perfectly aligned? And perhaps on top of that, if we're trying to study this needle in a haystack and there's only going to be 1 or 2 cells per tissue, then there's going to be quite a few tissues that have 0 cells. So we actually have to also come up with ways to make sure that this is essentially high throughput. So we're looking into robots now that can automatically load specimens into these things. And so it's just, it's really fun when you get to tinker with a microscope and, you know, say like, well, let's get into robotics now. You know, like I have zero experience with robotics. It's actually terrifying to think that we might have a robot near such an expensive piece of equipment. But, you know, the biology justifies the risk. And so, or just jump right in and try it.

Cool. Yeah, I think microscopy has kind of moved in terms of the biology, like fixed cells, fixed samples, dead cells, to live cells and to live animals. Next step. Yeah. So you need more automation because these samples are sensitive and you need more throughput. And so on that point of the app, you building applications and focusing on cancer metastasis, Could you talk more about like the throughput you guys are at? You know, like what's your— what are your core metrics? Is it like experiments per day? Is it samples per run? And then could you talk more about like why microscopy is uniquely suited for cancer metastasis and maybe much better than say like a single cell genomic method? Okay.

So yeah. So what's throughput is the question, you know? So let's say we have, you know, we built a microscope, it provides 300 nanometer resolution in XYZ. And so that microscope is what we— it's not smart, you know, we call it a dumb microscope. Now that thing just images like, you know, one thing after another, after another, after another. And as a result, it doesn't adapt to the specimen at all. So we might put a single tissue on there. And walk away, and 3 days later we have the data. And that's assuming everything goes well, right? So, and I'm talking that these chemically cleared samples that we've made transparent so that we can image them. Now in our new microscope, we can do that same experiment where we, we basically screen the entire specimen, but we can do it in now like 3 or 4 minutes and, and find the events that we're interested in and switch and and then do that 300-nanometer resolution imaging locally. And so instead of looking at 1 or 2 samples a week, we're now— we can consider doing 10 to 20. And that's just because we're just now getting started. And there's one other side effect of that is, like, if we have to image everything at our highest resolution, the amount of data that we generate is insane. You might have a single experiment with 10 terabytes of data. This is actually not bad by some standards, but, you know, going through that data is a major problem. By adapting the microscope to change the resolution that we're imaging at and only focus at the highest event, or like the most— like the candidate events at our highest resolution, which are most data hungry, we can then drastically reduce our data overhead as well. And so, that was the first part of the question is kind of like, what is our success for throughput? Our goal really would be to get to the point where we can image, you know, 100 lungs in a week and a half, 100 livers in a week and a half, and 100 kidneys. And so, that's just the level of throughput that has yet to have ever been obtained, but it seems possible. The second part of your question, what was that again? Can you remind me?

So, what makes, like, cancer metastasis uniquely suited for microscopy?

Okay. Yeah. Yeah.

Yeah.

So, so, yeah, this is a little bit provocative and, and perhaps step on a couple toes. So, you know, one of the most important things about biology is that there is a, you know, things have developed organelles and spatial distribution on purpose. These things are not bags of DNA, RNA, and proteins, right? There is an architecture to these cells. And so, if you want to understand what a certain enzyme is doing, you need to understand where it is in that cell. That's, you know, that's actually pretty clear to us. You can look at, say, Row A, which changes how a cell is typically thought to be kind of contractile,. And in certain parts of the cell, it actually is doing the exact opposite, polymerizing actin, which is part of the cell's cytoskeleton. And so that's the first thing, is that the cellular levels, these processes are spatially and temporally organized. And you, you know, although for the clear tissue we give up some of that temporal information, it is important that you actually read these things out in a spatial format. And then you think of it as well, One layer up is these cells are colonizing a tissue, and they're not just colonizing any part of that tissue. And so, understanding the context where that cell is, is super important, especially in something like metastasis. You know, we have these patient-derived xenografts from Sean Morrison's lab, and patient A may have succumbed to cancer because it metastasized to their lung. And, but it didn't go to their kidney or liver. And so why is that? You know, these cells are in circulation throughout their entire body, and yet these cells only went to the lung. That's, you know, that's an interesting thing. So what is it about that environment, that context, or that cell that made it only go there? So this is what they refer to as organotropism, where, you know, there's different sites within a tissue as well as different types of tissues that certain types of metastasis will preferentially go to. And so, biggest picture, that's really what we're focused in, is trying to come up with microscopes that can allow you to find what is the molecular feature that allows this cell to only go to the kidney or to only go to the liver, but not the lung, or vice versa. And where in the lung is it? Is it in a bronchiole? Is it in the alveoli? Is it still in the vasculature? You know, these are all important things. If you want to try to develop interventions, you have to understand what are the weakest links in that, in that metastatic process. And so we believe strongly that you have to have, you know, that spatial insight.

Totally agree. And when you study, like, cell trafficking and cancer trafficking in particular, what have you discovered? What kind of new insights have you discovered that maybe kind of are a starting point for understanding cancer a little bit better? And can you also talk about maybe lastly, talk about like, like kind of by studying cancer cell movement, is there like, what do you see the potential? Is it new drug treatments? Is it upending the understanding of like how to actually use existing drugs, stuff like that?

Yeah, so I think, you know, a good example of how the spatial can be valuable is my colleague Andrew Weems, who recently published, could show that, you know, these pro-oncogenic hubs form transiently as part of this septin scaffold. So septin is a curvature sensing molecule. And so that's one of these things that, like, if you just took the entire cell and did proteomics or something, you would not see that. And so that's one aspect is like, even at a single-cell level, these structures are super important. The next is that a lot of these structures can get averaged out in the ensemble. This is a common problem in single-molecule spectroscopy, and the same is true with cells. So if you average all all the molecules in this, like, not only if you average just one protein within a cell, which might be doing two different things in different spots, what you see is a mix, right, when you look at that, unless you have a way that can resolve those populations. And then you take that cell which the has those proteins with another cell which has those proteins in there asynchronous, then those average out too. So these are all just ways that really show the power of looking at things at the single-cell, subcellular level. And there's no disease that, I mean, I don't need to be too bold, but it seems very likely that cancer is perhaps the disease that results from heterogeneity the most. You know, we get, it's not, you know, they'll say it's not one disease, it's like dozens of disease or hundreds of disease because you essentially create this, competitive environment where cells can fluctuate in their phenotype and exhibit different properties. And so we have to figure out which one of those, you know, phenotypes is actually driving the problem of interest. And if we just look at all the cells, you can't see that. You have to see the ones that are actually, you know, caught in the act doing something that they shouldn't be doing.

Really interesting. Maybe to wrap it up, and we'll talk to you now, catch up, and any final lessons, any, any kind of things, things you see on the horizon are exciting, especially for people who are up and coming, you know, maybe the, the up and coming microscopist, the grad students trying to figure out what to do with their life. Any final thoughts?

Yeah, I think they, you know, I've made to demonstrate that I think you can do pretty well if you're just not giving up. And so, you know, just go for it, you know. And the other thing is I never thought I'd actually get to where I am in this position. I never really did so intentionally. I just kept doing things I enjoyed. It's a little bit of a fine thing to tell someone like, just keep doing what you love. You know, just be determined. You know, there were a lot of times when I almost bailed, and, um, you know, I just continue to, you know, stay low and keep my feet moving. As they say in football, you just keep moving forward and good things can happen. Absolutely.

Stay low. That's true. Just stay low and don't stop your feet.

Yep. And I think it's true, you know, you just can't— there's gonna be a lot of people that say it's not possible or that, you know, that you're not cut out for it. You just have to really, you know, just ignore people to some extent, right?

Just keep going. Absolutely. Well, Kevin, thanks for doing this. I really appreciate it. We'll do the catch-up to yours. I really appreciate taking the time.

Awesome, appreciate it.