STH - E30 Dr. Kliment Verba - Cryo-EM & COVID-19 virus receptor traps

[00:00:00] Hello everyone, welcome to Scientist The Human Podcast. It's my pleasure to share this

[00:00:06] episode as part of the series, done in collaboration with the QBI or Quantitative

[00:00:11] Biosciences Institute and UCSF, University of California, San Francisco.

[00:00:16] QBI, phosphorus collaborations across the biomedical and the physical sciences seeking

[00:00:23] quantitative methods to address pressing problems in biology and biomedicine.

[00:00:28] I'm excited by the opportunity to connect with awesome scientists here at the QBI and

[00:00:34] share these conversations with you. Let's get into it.

[00:00:45] Backly, the rare younger ladies can't do with the role of mimicking the scientific human

[00:00:50] podcast in the community.

[00:00:52] Welcome to the episode of Scientist The Human Podcast on your host, Simranjit Singh.

[00:00:57] And today I'm here with Dr. Clement Burba or Clint, who is an assistant professor at the QBI

[00:01:05] or Quantitative Biosciences Institute and UCSF. Welcome to the show.

[00:01:10] Thank you. Let me be here.

[00:01:13] I'm really excited to chat with you, having the opportunity to interview scientists at QBI

[00:01:19] as Braden. I'm really excited to talk to you about your research and let's just jump right into it.

[00:01:25] So you have an interesting background. You've done compared to what you started your research on,

[00:01:34] you kind of went in different directions, kind of focusing a lot on cancer research previously

[00:01:40] and then also now into virology with SARS-CoV-COVID-19, right?

[00:01:46] And so, let's just kind of talk about one of the main tools that you use, which is cryoEM.

[00:01:55] Can you talk a little bit about what cryoEM is?

[00:01:58] Sure. Before we jump into that, I'd like to say that I'm also an assistant professor at the

[00:02:06] Silicon Valley Pharmacology Department in addition to QBI.

[00:02:11] So, let's talk a little bit about the questions.

[00:02:13] Yes, I have a like here at UCSF. Everyone is part of everything.

[00:02:19] Yeah, so cryoEM is relatively...well, it's an incredibly powerful technique.

[00:02:28] I was going to say it's relatively new, that's not true. It's actually really old in the 19th centuries

[00:02:33] but there was a number of technological advancements in the past decade

[00:02:38] which allowed it to really become incredibly useful for us biologists.

[00:02:43] And I guess for those who are not familiar with this overall field of what's called structural biology,

[00:02:53] the idea here is to use different biophysical methods to understand how proteins look in three dimensions,

[00:03:02] in atomic details.

[00:03:04] And, you know, to a very simple sort of parallel to help people with an intuition

[00:03:11] and it's like if you look at a tool set and you have Phillips screwdriver and a flat blade,

[00:03:17] they are shaped sort of in the acts of a function.

[00:03:20] And so the exact same thing happens in molecular atomic details with our proteins.

[00:03:26] And so, again, you know, for my standpoint, everything that is interesting biologies happens

[00:03:33] by work done by proteins. And so everything that we think of as our muscles contracting

[00:03:40] are thinking thoughts, you know, things like or just being structurally sort of stable,

[00:03:46] that's all done by different proteins and we have over 20,000 proteins.

[00:03:52] And each protein has a unique shape or even multiple shapes.

[00:03:56] And depending on that unique shape it does a unique function.

[00:03:59] And on a very often in disease, that, you know, single mutation for example can cause a change

[00:04:07] in that sort of structure in that shape or structures more technically cold.

[00:04:12] And that changes in function and so there's because of that there is a sort of whole field called structural biology.

[00:04:20] But it used to be, you know, it's still not easy but it used to be, to me still sort of,

[00:04:27] you know, for example, crystallography was really just technique, the structural technique which sort of

[00:04:33] was the first one to go through. And then that technique, it's mind-boggling to me,

[00:04:40] it relies on the ideas that this proteins and again each protein has a unique shape.

[00:04:44] They're sort of very rough at some ways and, you know, come together in this crystalline lattice,

[00:04:51] which is very periodic. And then you should x-rays at that lattice.

[00:04:56] And so on surprisingly a lot of proteins would not be able to do that.

[00:05:00] And so our sort of, although it's an incredibly powerful technique and still is,

[00:05:04] and sort of there's a lot of biology which it couldn't quite capture.

[00:05:08] And so, cryolectomychroscopy uses electrons to image sort of vitrious samples.

[00:05:16] Vitrious meanings are frozen in their native buffer environment on nowadays actually they can be part of itself.

[00:05:24] And the frozen is so fast that it's faster than the water molecules have time to organize into crystalline lattice.

[00:05:33] And then we can shoot electrons at them, basically preventing the formation of ice.

[00:05:38] Exactly, exactly. And then we shoot electrons at them and then use a lot of, you know,

[00:05:45] and at the end we get this very grayscale images of individual molecules or individual proteins.

[00:05:53] And to step back a little bit so understanding how single proteins work by how their shape looks like

[00:06:00] is very important. But proteins are social creatures and so they actually work as part of protein complexes.

[00:06:06] And so multiple different proteins come together and hang out together.

[00:06:10] And then a lot of, you know, most biology, M-O-G-G-G is driven by rearrangement of those complexes.

[00:06:16] And so seeing how multiple proteins come together is sort of an incredibly exciting avenue.

[00:06:20] And so again, so we basically have proteins or protein complexes in, you know,

[00:06:24] put them in a microscope in this glass-like ice.

[00:06:30] And then we shoot electrons at them and then we get this grayscale images of individual particles,

[00:06:36] individual sort of small images of individual protein complexes.

[00:06:40] And then there's a lot of computation. And that's sort of so one of the major advancement has been in the detectors for electrons.

[00:06:49] Actually, it's driven a lot by here at UCSF when collaboration was a lot of other places.

[00:06:55] Which allowed us to really see sort of new features at much better resolutions than previously available in this grayscale images.

[00:07:03] But also another big thing has been in algorithms and so another big development in the past, you know, five to six years.

[00:07:10] Well, it's like 10 years. Been really sort of with the incorporation of GPUs for processing neural networks

[00:07:20] and all kind of other methodologies really allows us to basically go from those grayscale images to three-dimensional atomic resolution shapes.

[00:07:30] And then figure out why, you know, and then using this technology, you know,

[00:07:35] multiple millions of dollars worth of equipment and, you know, and hours of processing highly, you know, very powerful workstations.

[00:07:45] You get this through dimensional shapes and then you can compare, say, cancerous mutated protein versus what we call a wild tide,

[00:07:55] which is like normal functioning protein. Or you can look at viral proteins and complex proteins.

[00:08:01] And then try to understand how they come together and that then leads us to, you know, well, first better understanding of how things work.

[00:08:10] But also potentially gives us sort of a helpful to try to potentially come up with compounds or some other methods of drugging and rescuing stuff.

[00:08:20] Yeah. Yeah, amazing summary. Thank you.

[00:08:23] Very long winded. Sorry. No, it feels good.

[00:08:26] So you mentioned a lot of things there. So you shoot electrons, you get grayscale images or two-dimensional images.

[00:08:34] Yes. And then you have to use computation and algorithms, I guess, as a part of that, to build the 2D images into 3D models of those.

[00:08:45] Absolutely.

[00:08:46] So I imagine between the 2D and the 3D, there's a ton of work involved.

[00:08:52] Could you give kind of like a layman's explanation of how that's done? Like how the 2D is then converted into 3D?

[00:08:59] Sure. So I mean, I guess there's multiple things. So actually going to sort of mathematically going from 2D to 3D has been fairly well figured out and parameterized.

[00:09:15] And then you can do that easily. Sort of projection, background production, a projection computer, you know, tomography, for example, you know, in medicine has been doing this for a long time.

[00:09:27] And then there is sort of different technical ways of doing it. But, you know, basically you sort of, it's not quite a shadow but it's like an x-ray.

[00:09:34] So when you shoot, you know, electrons are simple.

[00:09:39] You know, the images, I mean, the simplest way of thinking is shadows technically is that's not exactly correct.

[00:09:45] And then the idea is that even if we think of them as shadows, if you sort of rotate them and you can imagine collecting a large amount of, you know, sort of rotated images of shadows, then you can back project and get sort of a three dimensional image of your sample, I guess, of study.

[00:10:03] Where it actually is, so but again mathematically that has been well worked out. And we can go backwards and forwards. Most of this is done, you know, in through using the other space called the Fourier space.

[00:10:17] But that's been figured out for a long time. Now what's been much more difficult and where a lot of algorithms play the big role in improving what we get is sort of three dimensional classification.

[00:10:32] Our two dimensional, three dimensional classification. And so because the thing, so the remarkable thing with our processing of overall, of this image and with electron microscopy is that as we image, say, sample, we destroy it.

[00:10:48] And so that fundamentally now limits the amount of electrons we can shoot our sample with. Right? So think of it as, you know, think of it as you being in a dark space and you're having different flashlights.

[00:11:03] And, you know, obviously if you have a really strong flashlight, you'll see very well. But basically we have to shoot with a very dim flashlight because if you shoot very strong flashlight, you destroy yourself.

[00:11:14] And so once you have this, it puts again a very fundamental boundary on how many electrons you can shoot.

[00:11:21] Then every electron really counts. But then also it gives us very noisy images. So as a signal to noise in our images is very low.

[00:11:31] That, you know, actually by untrained eye of often, you would not even probably recognize those as images of some side.

[00:11:40] And so the problem is not going backwards from 2D to 3D. The problem is you have this, you know, you'll have hundreds of views of your sample in eyes and two dimensions.

[00:11:55] And now what you need to figure out is that what were the axes of its rotation basically when it was frozen?

[00:12:04] If you have those three axes of rotation, then the angles where it was rotated and each of the axes.

[00:12:11] Then mathematically you can go into 3D very straightforwardly. So figuring out those axes is very hard.

[00:12:18] Now on top of that, each particle in the protein complex they're not all exactly the same. And so they actually proteins are quite dynamic.

[00:12:29] And the more interesting the protein is, the more dynamic I would argue it is. And so often when you compare them, you know, you have to basically realize that there are multiple different proteins.

[00:12:41] And you have to also be able to group them together. And we have to group them together because of this limited signal to noise.

[00:12:48] Because when you basically what you do is you, what's called average but you can simply just think of adding them together.

[00:12:54] And when you add this particle images together, the random noise will cancel out because sometimes it'll be positive, sometimes it'll be negative.

[00:13:04] So over different images it will cancel out. But what's not random will add.

[00:13:08] And so that's how you get better signal to noise through what's called averaging.

[00:13:12] But you have to align things so that you average them together in sort of, you know, there have to be in register with each other.

[00:13:19] And then you add them without register, outside registers and you just sort of smear everything out.

[00:13:23] And so that computation also it's called alignment and classification. So those computationally are very quite difficult to do appropriately in this very high noise context.

[00:13:36] And so a lot of algorithms again been focusing on using things like maximum likelihood frameworks and sort of dealing working with probability distributions for as in trunk 18 exam arbitrarily.

[00:13:50] Which allows you to sort of estimate the error of your sort of, you know, your precision of this of alignment that's been a lot of sort of a novel development in that.

[00:14:01] That's really cool. So if you have the, so the electron, the electrons that are shot at the protein are in two dimensions.

[00:14:12] Like there are the way that they're being received is in two dimensions not from all directions.

[00:14:18] No, so all directions would be scanning electron microscope. Okay, we're working with transmission electron microscope.

[00:14:24] And so, yes. So what we do is we shoot electrons at the top from the top of the microscope which could be 10, 15 feet tall in big ones.

[00:14:34] And then you have your tiny, tiny sample which is frozen. So it's still just liquid nitrogen temperature.

[00:14:39] So electrons go through it. The electron actually goes through bunch of lenses and get focused. That's why we work with electrons.

[00:14:44] And so, like the overall setup is actually very similar to a light microscope where you have a light source. You have a bunch of lenses.

[00:14:55] You have a sample bunch of lenses and so it's an entity detector or an IP. So this is the exact same idea here. We have a source. So now it's electron because electrons have a much, remember wave particle duality.

[00:15:08] So electrons have a very short wavelength compared to photons. Well, visible light photons. And then yeah.

[00:15:19] And so it goes through your sample and then there's sort of there's a phase shift actually between scattered non-scattered electrons.

[00:15:26] And then you detect that as a two-dimensional image on your detector, each in the camera.

[00:15:33] And so it makes sense knowing the axes that the protein is in the...

[00:15:39] Yeah. The vitrious ice I guess is exactly the important to figuring out the...

[00:15:45] So yeah, exactly. Think about it again. And then two or two sort of thing would be, I know hold up any object like a cup.

[00:15:53] And then imagine that you shine light at it and imagine you have a board in the back.

[00:15:59] And then as you get some image and then as you rotate the cups, that image changes.

[00:16:04] And so fundamentally if you know at what angle the cup was when the image was taken effectively, then you can actually...

[00:16:11] And then if you have bunch of those, you can actually fairly straightforwardly go into 3D shape of the cup.

[00:16:18] By figuring that angles is a hard one.

[00:16:22] Okay, nice. Yeah, yeah. Thanks for the deep dive up.

[00:16:26] So like zooming out a little bit, using this technology to figure out protein structures or the way that two proteins might be interacting with each other.

[00:16:36] You mentioned that it's helpful in a number of disciplines. In cancer research for example, it could be helpful in drug discovery.

[00:16:46] Could you talk a little bit about that and kind of your experience in that field?

[00:16:49] Sure, of course. Yeah, so... Right. So let's see.

[00:16:55] So I mean fundamentally when you...

[00:17:00] So usually when something goes wrong, right?

[00:17:04] It means that easily you have a protein which is supposed to be doing something not doing something or a protein which is not supposed to be doing something doing something.

[00:17:13] Or it's doing something in their own spot or their own time. That's kind of the way things go around.

[00:17:19] And then going back to the shape idea and structural biology is that fundamental sort of dogma of the field is that there's four things like catalysis or bindings that has to be shape complementarity.

[00:17:37] So again, if you have a phillips screw, you want to have a phillips screwdriver.

[00:17:43] You will be hard to put together... you know, put a phillips screw in with a flat plate.

[00:17:49] And so the idea here is then is you have this protein. It has usually a specific say active site as they're cold. That's sort of the site where as a magic happens.

[00:18:02] And for example, if that's often like proteins I work on a lot, protein kinases. They're responsible for transferring phosphate from ATP molecule to their substrates.

[00:18:15] And that changes, that's sort of involved in a lot of cellular decision making.

[00:18:20] And so when cancer happens basically it means that the cells kind of start listening to their rest of the cell society and kind of just do their own thing.

[00:18:34] And that usually means that they start to divide on control or beliefs, they migrate and things like that.

[00:18:39] So those cell decisions a lot of them are driven by this protein kinases, which basically just put some phosphate onto some specific group of proteins.

[00:18:48] That's then get recognized by other proteins and changes completely the cell sort of program.

[00:18:53] So for that to happen there has to be a specific binding event to this protein kinase.

[00:19:00] And kinase again is just a class of proteins which do this first-four relation reactions.

[00:19:05] And then so for example now if you have a protein which is doing that when it's not supposed to often the cancers are driven by this kinase has been hyper activated, what you need is you need to make a compound that will go into that site on the protein which does the magic and will be perfect fit like a glove.

[00:19:29] So that it plugs it up and can't do its chemistry.

[00:19:34] And so getting that compound a lot of it still relies on trial and error with medicinal chemistry, but a lot of it once especially you have a lead relies on being able to visualize this sort of lead compound bound to this protein of interest.

[00:19:52] And seeing that oh like we can make it better in this way, if we grow this group here or add the group here we can make this new interaction which will make it bind even better sort of rational drug design.

[00:20:05] And so that's where structural biology helps a lot.

[00:20:08] And so for example one of our recent work which I've done together with another KBI professor here at UCSF Natalia Jura was working on her two first three receptor tyrosine kinases and which is a specifically very potent oncogenic signaling complex.

[00:20:30] Whereas there is you know for example her two positive breast cancers and usually over expression of her to his you know associate with very poor prognosis.

[00:20:38] And so and then it's specifically just two proteins her two and her tears are come together and then that initiates a very strong sort of cell proliferative program so cells just start dividing really strong.

[00:20:52] But structurally they're never been visualized together and so we were the first ones using cryo to actually visualize them together and it was really cool is that we got the structure of the normal complex.

[00:21:06] But we also got a structure of a mutant cancer mutant complex and it was really cool that this proteins come together when they're together they are activated so you want to sort of have a type control on the other together or not.

[00:21:21] And so part of this protein protein interface in the wild types of normal complex was actually sort of weakened and we didn't actually even see it in our structure.

[00:21:33] But then the cancer mutant single mutation actually really stabilizing the face stabilist sort of stapling the two proteins together really sort of pushing that complex formation and now you had this much more higher so higher activity sort of signaling through that complex eating to proliferation.

[00:21:54] And so again just visualizing that that was totally unexpected the fact that evolutionary you are you know through and actually what school is this element.

[00:22:03] There was this element which basically went from being disordered to being ordered upon this cancer mutant and so evolutionary basically the you know this interface has been sort of

[00:22:18] placed at this space where it's partially used or not as whether it could be specifically so it's not too strong so the proteins don't interact very strongly together and cancer completely overrides that.

[00:22:32] And then the any other driving is a signal and what was also cool in that work we were also got a structure with actually a drug called herception which is an antibody which is approved and you know their patients taking it.

[00:22:46] And then it's never been visualized in its context of the sort of signaling complex and so we're able to see how it binds exactly in this structure and also with that hopefully you know I think now.

[00:22:59] We are probably not going to do it but you know medicinal chemists and people working on this can use that structure to make a drug even better having this atomic understanding of exactly where everything is.

[00:23:11] Yeah but that's an excellent example and it's a curious case of cancer mutation leading not just to a hyperactive protein being hyperactive by itself but leading to this enhanced interaction with another protein to drive signaling which is really cool and yeah cool cool work.

[00:23:31] So I have a question about that so her two and her three are their cell surface cell membrane proteins.

[00:23:41] So if you have a protein that normally lives in the cell membrane when it comes to imaging that protein or getting the structure of that protein is that a barrier at all does that come into play.

[00:23:52] Yeah so it's actually right so it certainly makes it much more difficult.

[00:23:58] And so you can break sort of proteins into there's sort of what's called cytoplasmic protein and then their membrane proteins which have been experiencing sort of this explosion of structural biology because of cryoum.

[00:24:16] But then these proteins are actually really hard because what they have is they have a pretty large piece that is outside the cell.

[00:24:23] They have a small piece that's inside the membrane and then they have a reasonably large piece that is inside the cell, cytoplasmic.

[00:24:31] And so it's that presents sort of the worst of all cases because the first thing so for those who don't know the inside of the membrane is hydrophobic.

[00:24:43] It means you don't want there's no wearers there.

[00:24:46] And so when you work with soluble proteins they're all hydrophilic so you kind of have your normal buffers and it's fine.

[00:24:54] When you have your membrane proteins there's a whole set up where you put them in this sort of you know either like a detergent, you know my cell they're called or different like nanogisk environments basically you kind of hide them into this hydrophobic patches.

[00:25:08] But here we have chunks which are you know soluble chunks which are hydrophobic so that's difficulty part number one.

[00:25:15] The second part is the fact that outside is oxidizing and inside is reducing and this has to do with cysteine residues and they're sulfide.

[00:25:27] Cysteine is an amino acid, one of the 20 amino acids.

[00:25:32] And so what this means is that cysteine can either be in a reduced or oxidized state and when it's oxidized it forms disulfide which are often key elements of extracellular proteins.

[00:25:44] And so there basically is this little staples, equivalent staples across the protein structure which hold it together.

[00:25:50] And so here usually if you're working inside the plastic protein you actually add specific chemicals to make sure that there's no disulfides form because it should be a reducing environment.

[00:26:01] When you're working with the extracellular proteins you actually don't add any of those chemicals because you're supposed to form those disulfide for structure.

[00:26:08] But here we have half of the protein having to be reducing and the other half having to be oxidizing which makes it considerably more difficult.

[00:26:16] And so those are sort of the two challenges, that's why you know like look people have been wanting to give this structure for a very long time.

[00:26:24] So it hasn't been done not for the lack of trying, it's just been very hard. And so this is sort of you know it's a large heterogeneous floppy protein which is sort of prevented the crystallography part of it.

[00:26:37] But also having this sort of being able to set up your condition so that the outside is happy, the inside is happy and the membrane is happy certainly poses a big challenge.

[00:26:48] Yeah sounds like it's an extra layer of considerations that you have to kind of solve for.

[00:26:55] Yeah really cool example for applying cryoEM and structural biology cancer research so you've also done quite a bit in virology particularly with COVID virus, so how did you get involved in that?

[00:27:12] Yeah so I mean you know I guess I think as a lot of people especially here you see oh god involved in it is that you know pandemic happened.

[00:27:22] You know you couldn't really do your usual research and also there definitely was a feeling that you know it's like an infringement on our scientific thing where we should be able to help.

[00:27:39] You know we should you know something kind of like protein virus and biology kind of know something but we should be able to do something useful.

[00:27:46] And so actually you know we made this large consortium of structural biologists who are you know who worked following up on the work driven by you know Nevin Krogan on identifying.

[00:27:59] So you know the virus for those who don't know I mean it's actually you know I know working biology before but the more you know I'm involved in it the more I find the fascinating.

[00:28:08] And so one fascinating thing you know we have over 20,000 protein viruses have very small number of proteins.

[00:28:13] So for example COVID depending on the virus depending on how you can't say 26 such proteins.

[00:28:18] So this virus comes in with 26 proteins and the weeks complete havoc on the world.

[00:28:25] Right? And so it's almost insulting that we can't figure it out because it's 26 proteins right it's a very small number.

[00:28:33] And so and so what the one of the way it does it is by basically having its own proteins common to protein complexes sort of you know interact with the human cell proteins.

[00:28:46] And that allows it to completely hijack human cells for its own sort of nefarious needs.

[00:28:53] And so Nevin sort of you know as part of this large you know effort identified sort of the protein complexes that this 26 proteins came together you know what human proteins did they bind to.

[00:29:06] And so the next step was like okay what does it look like how does they actually interact because you know again if you want to start developing compounds you need to know the interfaces where they interact.

[00:29:17] So you can make that compound which will fit that interface like a glove and break it apart.

[00:29:22] And so and so to do that we formed this you know structural biology consortium QC or D structural biology consortium.

[00:29:30] And you know which at some point had over 60 sort of volunteer based effort across the CSF on whoever wanted to just contribute to sort of an open call which I think was really really cool.

[00:29:43] And then yeah and I had a privilege to sort of co-leading that effort and trying to use what we do best sort of the structural biology tools and all we have to really bring molecular understanding into those protein complexes.

[00:29:57] And so and we got a number of school new structures and I mean and what school is the ones who start games or structures.

[00:30:05] It you know I it became sort of from a biophysical standpoint quite exciting to me because viruses you know you can single viruses is sort of a compression algorithm right where viruses have to you know human cells have all the shrubbery to you know all these genes and circuits and things like that to be able to do.

[00:30:26] All kinds of things right. And then the viruses have to come in and sort of find very key points of those you know in the cell and sort of pressure points and press on them to rewire the cell to its own needs so that's one level of compression by another cool thing is that because it has limited genome you know it has a limited number of proteins.

[00:30:54] But it has to do all this functions and so what ends up doing happening is that the number of the spread teens end up actually adopting multiple different shapes depending on what they interact with and from a biophysical standpoint the sort of plasticity I think is really exciting.

[00:31:10] And sort of again going back to the you know Philip screwdriver here have the screwdriver which depending on what it it's actually not committed to being flat blade or Phillips it's kind of like intermediate state.

[00:31:21] And when it sees a Phillips screw it's naps into the Phillips shape and has sees a flat blade screw it's naps into that shape.

[00:31:28] And that to me the fact that biology can do that is actually fundamentally really really cool.

[00:31:33] I think a lot of that happens in mammalian cells and better understanding of these sort of pleiotropic states that I find very fascinating.

[00:31:43] So is that less common in mammalian cells this kind of plasticity of protein to come?

[00:31:49] I think so. I mean mine the standing is that it's true and again I mean one could imagine I mean I think it's so there's certainly a ton of proteins which are cold and

[00:31:57] I think that's why I think the protein is a lot of protein that's not really good for me.

[00:32:04] And so there's quite a bit of those and I certainly are enriched in sort of transcription factors and regulatory proteins.

[00:32:14] There's definitely a lot of proteins like I think I would argue that a lot of protein kinases are in this sort of metastable states where they're not totally intrinsically

[00:32:23] but a lot of them are not totally folded either.

[00:32:28] But I think viruses again are under unique pressure to really get the most function out of each protein that they have.

[00:32:38] And so I think I would imagine mammalian cells, human cells are not under such an incredible pressure.

[00:32:47] We have so much junk DNA, well annotated DNA already that realistically if you need a new function just get a new protein.

[00:32:58] Duplicate a protein you have, tinker with it like no big deal right?

[00:33:03] Where if a virus you kind of keep the genome size in check and if it becomes longer you need to get better proofreading and all this kind of stuff.

[00:33:12] And so there really seems to pressure is there much stronger and so you know there's much so I think there's definitely more of the sort of shapeshift of proteins in viruses.

[00:33:23] So it's part of this larger consortium you got involved in COVID research.

[00:33:29] And so I saw one thing I wanted to ask you about is you worked on recently published on something called receptor traps.

[00:33:42] It related to the ACE2 receptor which is a COVID binding molecule that's the molecule on mammalian cells that COVID viruses is binding to.

[00:33:55] What are receptor traps?

[00:33:57] Oh yeah so cool so this is actually a work which was you know really driven by Anum Glasgow and Tanya Cortenius lab and Sumia within the remission Jim Wells lab.

[00:34:11] And as part of the consortium we did a lot of structural biology on that.

[00:34:14] And so the idea of the receptor traps is that you so the virus you know spike protein on the bars interacts with ACE2s and that's how it interacts with human cells and gets internalized.

[00:34:28] And so a lot of antibodies for example would try to break that apart you know basically again you antibody would bind that the interface at the site where the spike protein would try to interact with ACE2 and you know prevent that binding therefore the virus would not be able to enter the human cells.

[00:34:46] And so the idea here is that can you make this human ACE2 protein? Can you actually mutate it so that it would bind to the coronavirus spike protein tighter.

[00:34:59] So if it binds better to the coronavirus spike protein then it would theoretically outcompete the ACE2 that is on human cells and therefore all the virus would come in bind all the you know all the

[00:35:14] ACE2 receptor traps and then would not basically be able to interact with the human cell and dogenous ACE2 that is on the cell surface and normal cells.

[00:35:26] And so you know cortemium and weld lab they basically do the combination of computational re-engineering of the interface using Rosetta, a software called Rosetta.

[00:35:43] And then they also use sort of actual library panning to get better affinity sort of experimentally to mutate the interfaces and get better binders.

[00:35:54] And then they came up with a couple of these binders of this receptor traps where they're literally like four mutations which led to way better binding to spike protein and actually in culture equities that it would neutralize the virus and things like that.

[00:36:09] And the question was you know how does it work in molecular details and then also sort of on the technical level was was there a computational modeling similar to the actual experimental structure.

[00:36:21] And so we got the structures of the receptor traps with the spike protein and then yeah we saw how you know it was really cool.

[00:36:32] How like a tiny tiny shift of you know one to two Armstrong's of the interface allowed sort of much better repacking of the interface leading to a much better binder.

[00:36:43] And so yeah I mean something that is you know such a tiny tiny tiny difference leading to such a big effect. So that was ever.

[00:36:51] Yeah that's cool.

[00:36:53] And one sort of a wonder, just to add one cool thing about the receptor traps is with antibodies you know the sort of an orthogonal system.

[00:37:01] So if you have an antibody there like this you know therapeutic antibodies when they come in and the bind they you know theoretically that interface.

[00:37:09] Some of them do bind in a similar interface as the H2 but otherwise they are sort of set interface might not be evolutionally constrained to stay the same.

[00:37:19] But what school was a receptor traps is because it's basically the same exact thing as recognizing the human normal H2 the virus can't just will annul in the new phase interface away because then it won't be able to come into the cells.

[00:37:30] And so what was really cool about this work was that the receptor traps which were sort of designed for the original Washington variant of the spike protein worked actually better on delta non-machron than work on an original one because the delta non-machron actually spiked proteins bound to the human is too better and so it bound to the receptor trap.

[00:37:52] And so theoretically it shouldn't be able to mutate away the interface like it would like it has done with most of the antibody.

[00:38:01] Yeah is the H2 receptor the only protein that the spike protein binds to in order to get in the cell?

[00:38:09] Yeah, so that's a big question. So that's one way to do.

[00:38:13] So there definitely been reports of other proteins and entry points but it definitely seems to be the major one.

[00:38:22] The other ones certainly much less well in this too.

[00:38:27] Even in cell culture you sort of in cells which don't have H2 you get much better viral infections in sort of laboratory when you have to put H2 into the cells for that one.

[00:38:41] It would say that it definitely is a major but not a lot because it definitely been reports about the proteins also.

[00:38:46] Yeah I mean that stuff's really cool so that there's a potential therapeutic strategy could be developed.

[00:38:52] I mean I think a lot of it, I mean again it's conceptually which certainly would work.

[00:38:57] Now the question is H2 is an enzyme so with having a ton of it floating around, it'd be bad as terms of side effects.

[00:39:05] I mean I was asking you know how much so this would be a biologic and so how much of you know once you have biologic it's just producing a lot of them.

[00:39:18] It's hard because proteins are much harder to make than small ones.

[00:39:24] It's storing them exactly.

[00:39:26] I think conceptually 100% in terms of it would take a lot of effort to bring it to market.

[00:39:35] Yeah really cool research though.

[00:39:38] So you've talked about a couple things so as I said at the beginning some really different, you've been involved in some really different areas of research and how did you get into this?

[00:39:48] How did you get it interested in either structural biology or science or research in general?

[00:39:54] Yeah I mean yeah I mean it's a good or so you know some of it but you know fundamentally I've always been interested in how the world works.

[00:40:08] So I think that sort of a fundamental trait that you know especially like kids should be on a lookout for, do you want to understand how the world works?

[00:40:19] So do you want to make new stuff or do you not care about either one?

[00:40:24] I was ever saying.

[00:40:25] And so I always was curious about how things work and you know how I said on biology specifically, you know I...

[00:40:35] that is you know I actually probably been driven by a couple of really good teachers I had which really got me fired up about this.

[00:40:44] And then at some point you know but I could imagine...

[00:40:50] yeah I know at this point it's hard for me to imagine being as excited about something else.

[00:40:56] So I probably would say that sort of you know converted the general curiosity into biology specific curiosity.

[00:41:04] And then you know I went to a small liberal arts college, we didn't even have like biochemistry as a major.

[00:41:10] There was biology so I ended up doing biology major and chemistry minor. I also did a philosophy major but...

[00:41:17] for whatever.

[00:41:19] But does that come into play in your everyday work?

[00:41:22] I mean I think it does.

[00:41:24] Yeah hugely so.

[00:41:27] Right I mean I think philosophy is actually... I think everybody should do some philosophy.

[00:41:31] I mean I think it's super useful because it really flashes out sort of what are premises, what are nouns and like in terms of rigorous thinking about systems I think it's really really good.

[00:41:45] But anyway so I wasn't like this liberal arts college and then I wanted to do research.

[00:41:53] And then at that point I guess the question was like what level of understanding...

[00:42:00] where did I call it quits right?

[00:42:03] Where you know quite memorizing or remembering self biology and sort of feeling like I had to memorize a bunch of circles and rectangles, weird names and how...

[00:42:13] where's the arrows point and I found that not satisfactory enough.

[00:42:18] So figuring out where how these circles and rectangles actually looked like and how that worked,

[00:42:26] I felt that was sort of that once I saw all the structures I sort of felt like I understood and I was happy.

[00:42:32] And going beyond that into you know I know like quantum modeling of active sites was probably a little bit too much from detail.

[00:42:42] And so that's where you know and so I actually remember I think walking into the office of small college and so you can do these things,

[00:42:52] I walk into like the department chair office and said hey I think I saw more.

[00:42:57] I would like to do some research and I'm kind of interested in self biology biochemistry.

[00:43:03] And I think that's where I was at.

[00:43:05] I was in a new faculty there, Miriam Shebord, told me she was like I'll go talk to her.

[00:43:12] She's looking for a student and so I talked to her and you know she seemed to you know she liked me enough to let me work in her lab.

[00:43:22] And then as I was working in her lab there was more on the sort of self biology aspects.

[00:43:28] And through that year of work became clear that again I kind of always went on the sense of mechanism and sort of how did it look like,

[00:43:36] how did it actually work.

[00:43:38] And so she recognized it and so she connected me with a structural biology lab, Carl and Mottos lab at Noscural and Stating University.

[00:43:49] And so I went there and did some crystallography as an undergrad.

[00:43:54] And that was, that sort of was my first experience of structural biology and I really liked that.

[00:44:00] And so I continued sort of going between the two labs and working on these two you know on those projects.

[00:44:06] And I guess when I was applying for grad schools you know I really became fascinated with this idea of protein folding.

[00:44:12] And so I guess for those for not familiar with this you know we've been talking about proteins having specific shapes.

[00:44:19] So it's a cool part about it, maybe the coolest part about it is the fact that this proteins are made as this shapeless polypeptides are called sort of strings.

[00:44:28] And they sort of self organized in this specific shapes.

[00:44:34] And so nobody chisels them out.

[00:44:36] They fold that process called folding and they adopt a specific shape.

[00:44:41] And so that was to me super super cool understanding how that happened.

[00:44:47] And so with that idea I came and I joined David Ager's lab here at DCSF working on actually trying to understand how so the proteins adopters specific shapes

[00:44:59] and then the process is again pro process called folding.

[00:45:03] And then sometimes they don't quite reach their specific shapes.

[00:45:06] And so cells have made this whole slew of other proteins and circuits to help as a proteins fold and this proteins are called chaperones.

[00:45:16] And so what I studied in David's lab was how as a molecular chaperones actually help as a proteins fold

[00:45:23] and the proteins they well not only help them fold but what turns out is that there is actually a number of this protein chaperones and this is called substrate protein of client proteins for these chaperones which seem to interact long post initial folding sort of really starting to questions idea.

[00:45:41] Usually canonically chaperones would be thought of as you need the chaperones during this initial process of folding or if you end up misfolded somehow.

[00:45:51] But in terms of that there's a ton of very important proteins including protein kinases which is that you know, which actually sort of always are dependent on chaperones.

[00:46:02] And so they can access this strange metastable partially unfolded states all throughout their lifetime.

[00:46:10] And so how is that sort of fits evolutionary? You know why would you want to have this sort of proteins which are tasked with life for death decisions in this metastable states?

[00:46:21] To me it's sort of the still you know it's a more exciting question.

[00:46:26] And so I guess that's kind of yeah. So I went through you know, so by teachers pushed me towards biology I sang and then the experience and I guess just you know as a mindset I end up sort of going towards more mechanism structural biology.

[00:46:41] And then you know probably again the most like this connection of protein folding to function to resulting function to me is incredibly exciting.

[00:46:53] That's why again I sing this viral proteins which can adopt multiple shapes is really cool. And that's why again I sing protein kinases are really cool because a lot of them actually quite pliable so they don't quite have as committed of this sort of shape as they should.

[00:47:08] And so whatever you know how evolutionary tunes there are to be at this middle state and metastable states I think it's very interesting.

[00:47:15] Yeah I think it's amazing that sometimes all it takes is a teacher or two that's excited about something to get you excited about something.

[00:47:25] I have kind of a similar story. I mean in college I wasn't particularly sure which direction I wanted to go in and then I was pre-med and then I was taking biochemistry as part of that.

[00:47:36] And this wasn't until senior year of college and I had a professor who was just super excited about biochemistry.

[00:47:42] He also was doing research and so I was you know we will talk to this guy so excited about it and he got me excited and then you know you're out.

[00:47:50] Yeah, so it's awesome that's it sometimes it's just a just takes a teacher or two that's excited to kind of nudge you in a direction and then you find what you're passionate about.

[00:48:01] So outside of research what are your passions?

[00:48:05] So I word on the street is that there might be a bit of an artist.

[00:48:10] Yeah, artist there's a generous again. I would say you know Tinker or maybe.

[00:48:17] So yeah Tinker is probably the most appropriate thing so I like making stuff.

[00:48:31] Let's start with that.

[00:48:33] So you're doing kind of both from now on?

[00:48:35] Yeah, exactly.

[00:48:36] So you're doing stuff out and you're making stuff out.

[00:48:38] Yeah, exactly.

[00:48:39] I would go crazy if I wouldn't make stuff because I say you know often in science, there's just so much failure shit.

[00:48:48] Like 99% of time stop just doesn't work.

[00:48:51] And so saying steak forever to materialize and so it's good to go and do something with your hands.

[00:49:01] So you know intense sort of intense work for three or four hours and then you have some physical thing manifest itself in front of you which you may.

[00:49:08] And so that feeling is really good.

[00:49:11] And so yeah, so I actually wanted my big passions as cars actually.

[00:49:18] So I actually have a buddy of mine from Grasco and I built a rally car together.

[00:49:24] We were in Grasco and we raised a couple of times and we still have the car we haven't raised in a little while.

[00:49:28] Been a little busy and we need to fix a bunch of stuff up but I definitely started a lot of Tinker rain with the cars.

[00:49:35] And then also you know in addition to that, because of actually a lot of work on a car we've had to build a rally car and you're on how to weld.

[00:49:45] And that's also like a really cool you know I feel like once you can make stuff out of metal it's unclear why you would make it out of anything else to me.

[00:49:55] Because you know it's especially any metal you know aluminum steel stainless titanium.

[00:50:02] And so and so with that you know being contributing to a number of our projects.

[00:50:08] And also sort of tinkered with electronics and so actually I helped contribute to a number of.

[00:50:15] So yeah, so I helped contribute to this thing we built with a friend of mine Jarosco where basically it's like three dimensional gimbal I guess where you know person can stand inside and rotate in three axes.

[00:50:33] And so we built it together with him a while back now.

[00:50:37] And and then again another thing is I helped make lights for what's called an art car.

[00:50:45] So it's the same where like a car is converted to be made look something else and not like cars.

[00:50:51] And so I helped build one and we put like 10,000 LED lights on it which yeah it's like a lot of wiring and soldering.

[00:50:59] And because you know you have to it's like micro welding.

[00:51:03] Exactly yeah yeah exactly but it's remarkable how much power you need to supply for you know once you know each LED doesn't draw that much but once you have 10,000 of them you need a lot of juice.

[00:51:18] And so we you know we another friend of mine and I actually contribute to that.

[00:51:23] So you know kind of tinker with all I think tinker is probably the most accurate you know kind of make stuff whatever it is.

[00:51:30] Then you know trying to get into more like fiber optiki type stuff now which I think is like really cool sort of it allows you know problem with the ideas is that they're very sort of you know you need to diffuse them well.

[00:51:45] So you know you can do it in mix and wells to look good.

[00:51:48] And so I think it's fiber optic you know is a really cool medium where you can make it sort of very sort of continuous you know well diffuse light.

[00:51:59] And so it's something I've been playing around right now.

[00:52:02] Yeah tinker yeah it's a cool way to put it because you got deeper toes and a lot of different tools and you know whatever you enjoy.

[00:52:09] So I think interesting that I'm noticing is that so in your tinkering you're enjoying making physical things.

[00:52:16] Yeah it's something that you create you touch right something this tactile and then in your research actually the you know structural biology

[00:52:25] you're also looking at things that are actually there right.

[00:52:29] You might not be able to see it with a naked eye right but the images that are created with the cryo EM.

[00:52:36] It is actually an image of something that is actually there.

[00:52:40] Exactly right it's not something that is being fabricated like it made up.

[00:52:44] Yeah that's kind of cool so you have this passion for like pursuing real things.

[00:52:51] Real things.

[00:52:52] Yeah absolutely yeah that's cool.

[00:52:54] Yeah I remember actually the first time I saw an EM image you might want because like you know again I've done bunch of crystallography and on the grad

[00:53:04] and it's you know we get the diffraction spots and they go just conduit away from diffraction spots, the sweet dimensional structures.

[00:53:11] And then if you do in the more you get the peaks and there's even more convolutive way of going between those two structures.

[00:53:17] But then the first time you take an EM image you're like oh that's my protein.

[00:53:22] That little thing that's how it looks that has like sweet blobs and little things sticking out.

[00:53:26] And it's how it floats out around the cell and it's like and then you would just see bunches I'm hanging out and I thought yeah no absolutely.

[00:53:32] Yeah I think so.

[00:53:35] I mean it's you know yeah I totally agree I guess there's some fundamental draw towards you know things like that.

[00:53:45] I often think that what's really cool though is that biology you know we often compare proteins and things and even like this phillips and fleb blade example right?

[00:53:55] We sort of compare them to our manmade machines but again I think biology is so much cooler in some ways.

[00:54:04] Because again I think some of these proteins A are much more akin to computers and machines I would say protein kinases are like that.

[00:54:14] And what's really also cool is that they are much less discreet than they're much more analog.

[00:54:21] And I think there's actually even computationally you know the people there.

[00:54:26] There have been you know discussions and ideas of non digital computers of course quantum computers are an excellent example of now of the actual non digital computers.

[00:54:37] And so I think it's it's cool that you know we think of computation on computers in a particular way and are like paradigms are limited by some machines we made you know where like here's a wrench this is how it looks like and it's static.

[00:54:49] But in this molecular world of biology you know I think proteins are doing something that we don't quite you know often have a good analogy for because they think often they're much more pliable and the sort of adopt shapes which you need.

[00:55:05] And so you know at the right time in the right spot and what sort of a macro scopic analogy for that I don't you know I don't I don't think we make as humans machines like that.

[00:55:15] And our computers again are a combination of discrete states now we're just trying to get into again I mean our computers actually started being originally made analog computers

[00:55:27] and we went to this you know one non non architecture with discrete states and now you know and then well I guess the non non non architecture discrete states do different things.

[00:55:36] But the point is that I think you know as there's the analogies are very interesting because I think biology is fundamentally doing cooler things than we can draw analogies to.

[00:55:47] Yeah any any hopes for a biological computer? Sure at some point I mean I think so I mean I think we just need to understand better how they do what they do.

[00:56:01] Yeah it's like a general you know process right like at first like I feel like biology is getting to the point where we're starting to understand things enough to start making new things.

[00:56:13] But then I still think we don't understand a lot. And so I think we first need to learn considerably more of how things are done and then at some point we can sure will probably be able to manipulate and do things.

[00:56:27] I mean yeah I mean how at what point will be able to incorporate them into computers? I don't know that's a good question but again there's all kind of exciting computer you know there optical computers computing now happening which is fundamental different from sort of transistor based stuff.

[00:56:43] So I wouldn't you know maybe in our lifetime. Yeah I really cool stuff and you know a lot of the stuff they all the research that you do you know really exciting and if there are some young scientists out there I mean you're young too I'm relatively young.

[00:57:00] That's a younger than us you know that want to do what you do what piece of advice might you give them.

[00:57:08] Boy let's see I mean.

[00:57:19] So you know probably one thing I would say is I certainly seen sometimes cases and personally two where you know don't be shot.

[00:57:34] I guess I would say you know where I feel like often especially really you know people who are really really good really really shy and you know often they can sort of you know themselves you know to you know insecurities you know limits their own growth.

[00:57:57] And so I feel like you know you don't want to be in that position. So you know you better over you know correct there is an under you know to some extent and you know you have to kind of be and also don't be afraid of failure I would say because again I mean I think you should sort of treat this you know your personal growth as sort of a

[00:58:16] high high you could imagine treating it as dialing is a few mixture in the courage right where you kind of go you know up and then once you overshoot you can come down and you

[00:58:27] undershoot you come back up yeah and so you kind of you know don't be afraid of coming too far out and then overshooting and then you know but then you also be able to recognize that you did then come back.

[00:58:39] And you know probably do the good community you know develop you know I think everything hands down everything that you know I've gotten to the place where I am due to amazing people who help you know even you know starting I mean of course starting with my parents you know family but also

[00:58:58] you know to teachers along the way you know friends and you know the mentors all your life and so I feel like you know really treasure you know the

[00:59:11] deep relationships you know be try to develop deep relationships with people and maintain them treasure them and because you know

[00:59:21] ultimately most of our business is you know human based things and and I think you know again I would be nowhere without if I wouldn't meet you know amazing people who are willing to believe in me and you know give me a

[00:59:34] little extra push. I know that's helpful or not but that's I guess but no I don't have no custom mind. I absolutely think so so I mean you know you heard it here Clim said don't be shy.

[00:59:47] Don't be afraid of failure which maybe is another version of your happy shy and develop a strong community which could also be any

[00:59:55] extension of don't be shy go talk to people you know I'm like build these relationships yeah that's great yeah I think we'll leave it there.

[01:00:01] Thank you so much. Shall I put on a great conversation I really appreciate it. Yeah now thank you for taking the path of me.

[01:00:17] I'm the scientist of human air episode stay breathing.