Structure Club
Structure Club invites scientists to present one of their papers to a broad audience. Listeners will hear about cutting edge science from the scientists themselves.
Structure Club
Justin Kollman
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Justin Kollman discusses his recent paper, “Evolutionarily divergent Mycobacterium tuberculosis CTP synthase filaments are under selective pressure”
Welcome to Structure Club. I'm Scott Stag. I'm a professor at Florida State University where I use cryo EM to determine the molecular mechanisms of membrane remodeling.
SPEAKER_02And I'm Ashwund Francis, an assistant professor at Florida State as well. My lab focuses on structural biology of how viruses and host cells interact with each other. Together, Scott and I created the Structure Club as a journal club, podcast and YouTube channel, where the papers are given by the authors themselves.
SPEAKER_01Today's speaker is Justin Coleman.
SPEAKER_02Hi, Justin. Welcome. Hi guys.
SPEAKER_00Great to be here. Thanks a lot.
SPEAKER_01Justin got his PhD in biology from University of California, San Diego in 2005, and he completed his postdoctoral training with David Agard at University of California, San Francisco. And from there, he moved on to start his independent research program at McGill University in 2012. Then he moved his lab to the University of Washington in 2014. And he's been there ever since. In 2025, he had the honor of being appointed to the chair of the Department of Biochemistry. And he's going to be talking with us today about his recent paper, Evolutionarily Divergent Mycobacterium Tuberculosis CTP synthase laments are under selective pressure. The manuscript came out in uh the journal Nature in 2025. Justin, take it away.
SPEAKER_00Okay, thanks. I'll correct you to say that it was the journal Nature Communications. Oh, thank you. Yeah. But thank you so much for the introduction and uh and for having me on the on the podcast. This is hopefully going to be really fun. Um so the the the paper that that uh Scott just introduced, um uh it's it's a a close collaboration with our friend Hesper Rigo, who's at Yale University, uh, who's an outstanding microbial cell biologist. Um and the the work in my lab was was uh shepherded by Eric Lynch, who was uh a staff scientist in the lab. And uh Jinho Park and Hesper's lab did a lot of the um the cellular imaging that we'll I'll show you. Um so I wanted to start with just sort of setting the um uh setting the table a little bit about things we think about in my lab. And a lot of what we are interested in is how how intermediate metabolism is regulated, how you how you uh extract energy from from from molecules, and how do you use that energy to build up all of the small molecules of of living cells? And one of the really interesting things that's become a uh sort of a new paradigm in recent years is the physical organization of metabolic enzymes into larger scale cellular structures uh as a response to changing physiological conditions. So um these kind of take two flavors. We're gonna focus today on metabolic filaments, which are usually a single type of enzyme that forms linear sort of micron scale polymers inside of cells. There are also metabolons, which are multiple enzyme complexes that come together usually to accelerate um activity through a pathway. But the the point is that these structures assemble dynamically and reversibly so that they can give you very rapid responses to um to changes in in metabolic demand. They can also be used to localize outputs, so localize the metabolic intermediates in specific places in the cell. Um and the other important thing here is that these are very widespread. So these kinds of metabolic filaments, like we'll discuss today, are happening all kinds of pathways, lots of different cell types, tissues, and organisms from bacteria to to vertebrates. And so this is a very widespread phenomenon. In our lab, we've mostly focused on enzymes in the coupled pathways of uh glycolysis and gluconeogenesis. This is sort of extracting uh energy from sugars or building sugars, uh, which is linked to nucleotide biosynthesis through the pentose phosphate pathway, uh, sort of de novo nucleotide biosynthesis, where you build all the or create all the building blocks for RNA and DNA, also obviously important for energy and everything. Um, and I'm showing here uh images from a bunch of different papers from other labs. None of these are from our lab, but these are sort of the things that motivate us, is seeing uh these these kinds of images, uh, for example, uh here IMPDH, which is the regulatory enzyme in peering biosynthesis, forming these cellular polymers on uh in this case on activation of T cell lymphocytes. Um and so we were sort of inspired by other people's work to understand like what are these structures doing? Um and in particular in nucleotide biosynthesis, the three key regulatory steps are the formation of uh PRPP, the um the highly phosphorylated sugar on on which you build the uh the ribonucleotides, uh, and then IMPDH, as I mentioned, and CTP synthase, which is the subject of today's paper, which controls perimidine biosynthesis. All three of these form filamentous polymers in cells. And it turns out this is a way of regulating the activity of the enzymes. Um our lab's role is really to look at the structures of these things, reconstitute them in vitro, do the biochemistry, the structural characterization to try to understand how do these filament systems assemble uh or or metabolons assemble, and um what are the functional consequences when when they do. And so these are examples from our lab of structures that we solved in all of these different um these different parts of the pathways. Uh as I said, we'll focus today on this enzyme CTP synthase. But the upshot is if almost all of these that we've looked at, what we've seen is that that polymerization into these filaments provides an additional mechanism of allosteric control. It's a way of shifting um the conformations of proteins in order to change their response either to allosteric regulators or to directly um change the um the activity state of the enzyme itself. So this is the the star of the show today, CTP synthase. This is a deeply conserved uh enzyme. It does what it what the name implies, it synthesizes the ribonucleotide CTP. To do that, it it uses ATP and glutamine to basically phosphorylate and then do a um uh a transfer of an amine group onto UTP. Um and the important regulatory element here is that CTP directly binds back to the enzyme to inhibit it. And so the the allosteric change between an active and an inactive state is sort of well characterized now. This is the homotetramer of CTP synthase. Each of those chains is identical. And I'm gonna just show it here morphing between uh the active conformation where it's kind of compressed around the active sites. I've got the the thing that the red things that aren't moving are the substrates. So this shows how the the the enzyme changes its shape around the active site or sort of splays out into an inactive conformation. And that that both this heterogene this um uh tetramer structure and the the allosteric mechanism seem to be very deeply conserved from from bacteria to to eukaryotes, um, all sort of have this same functional unit of CTP synthase. Um but they all form filamentous structures that are different. So this was our the first filamentous enzyme that we started working on. Uh back when I started my lab, we were working on the E. coli version of CTP synthase. We've since uh solved structures and done a ton of biochemistry on the enzymes from budding yeast and from humans. And in each case, the the core tetramer is preserved, but they're stacked together into these filaments that uh sterically constrain them in different conformations. So in in E. coli, you can see that the the sort of arms of the X shape are kind of interdigitated in a way that prevents the conformational change I was showing on the previous slide and locks the enzyme in an inactive conformation. Um budding yeast have a very different uh assembly interface. You can see that they're stacked uh sort of more gently on top of each other, they're not really interdigitated. Same with the humans, although the interface is quite different between human and yeast. Uh, but in both cases, the enzymes are locked in one specific conformation. In the human case, it's active, and in budding yeast it's inactive. So we think this means that uh um uh that evolution has has invented this assembly mechanism multiple times as a way to meet specific demands for for CTP synthesis in in different types of cells. So this is where we were when we started this study. This was sort of the the state of knowledge of the enzyme and and what it's what its polymers look like. Uh and so collaborating with Hesper's lab, we looked at uh we tagged CTP synthase in in mycobacteria. Uh in this case, we're using the model system Mycobacterium smigmatus, which is a uh an easier-to-use um uh model system for uh mycobacterium tuberculosis. And here Hesper could see immediately we we saw these uh and Jinho saw these um these puncta of CTPS in um in in the growing cells, uh, and they often seem to localize at septa during division. They would sort of appear in the at the septum during during division events. What was really cool was Hesper and Jinho were able to do structured illumination imaging of of CTP synthase in in growing cells. And here you can see that the puncta actually are kind of elongated. I've I've zoomed in on a few of them to show that they really are short filamentous structures uh within the cells. And in um in cross-section, they could see that the the puncta were always localized near the edge of the cell, um, suggesting that they were they were probably affiliated with the plasma membrane, um, and and and which suggests maybe some some interaction with cell wall building. Um so the other thing that they noticed was that that growth permissive conditions, and here they're testing a number of different condition conditions, they see puncta of of CTPS, but that those become diffuse under conditions where the cells are not proliferating. Um, suggesting there's some uh connection between forming these larger structures and uh and and growth of the cells. This is a time lapse on the bottom here, showing um as cells go from stationary phase into recovery, as you plate them onto growth media, uh you start to get um increased uh uh formation of the CTPS uh puncta here. What was cool is they could they could they could quantify that. Um this plot here is showing that as a function of growth rate of the individual cells, uh they could they could measure the the this is a cluster score, this is essentially a measure of variance of the of the fluorescence signal across um uh across uh the the whole cell. Uh whereas the and and that didn't correlate with increase in the in the amount of CTPS, the the blue dots here. So really as the cells started growing faster, you got uh stronger punctate formation of the of the CTPS. So this is where we started looking at the wanted to look at the the structure in vitro of CT of microbacterial uh uh CTPS.
SPEAKER_02Dustin, may I ask you a question there? Yeah, please. Uh so just to clarify and keep this uh going, so uh the as a bacteria grows, the cluster forms and becomes larger. Is that what the interpretation is? Okay, good.
SPEAKER_00That's right. Despite the fact that it and it's not driven by more CTPS being synthesized, right?
SPEAKER_02But it's being locally organized in some way.
SPEAKER_00Being locally organized, yeah, exactly right. Um so to try to understand what these structures were, uh, we tried to reconstitute this in vitro. And um here showing this is Eric's work where uh in the presence of substrates, the active enzyme forms these linear filamentous polymers, uh, whereas in the presence of its inhibitor and product, CTP, uh it's it's it's monodispers tetramers. This was uh but this was very surprising. These did not look like any of the CTPS filaments we had we had encountered before. Um and in fact, when we we wrote that in the in the in the text of the manuscript, and I got really annoyed because the the the the reviewers didn't want to just take our word for it that these didn't look like typical uh CTPS filaments. They wanted a comparison. So that that's where these panels came from. Where here this is the CTPS filament uh and comparing it with these much more distinct X-shaped cross sections of projections of the of the structures we'd solved previously. Um and so so Eric took these negative stain samples, took the sample into cryo EM and solved the structure. It was fairly straightforward in terms of structure determination, and this is showing just a segment of filament that is reconstructed. This is three stacked tetramers, and they had the the canonical um uh uh CTPS tetramer organization. Here he's highlighting that the the the structures on the periphery, the the dark blue domains, are the glutaminase domains, which catalyze one of the uh the the reactions, uh, but those sort of lie in the periphery of the filament. Um and what we what Eric saw then was looking at the at the assembly interface where this blue circle is. Um it's a here's showing the density in the in the top, rotated to the side. It's a relatively small interface here with loose packing between these phenylalanines and a and a pair of um uh ionic interactions between an arginine and a spartate on either side of sort of have a two-fold symmetry axis um uh in the horizontal uh line here. I should actually that that reminds me, so it might not be immediately obvious, but the the structures are um uh the the tetramer is a D2 symmetric tetramer, um, and the the that's sort of arranged in the filament with one of the twofold symmetry axes coincident with the helical axis. So these really have all of their one of the twofold axes are all aligned along the filament. The other part of the interaction was the N-terminus, which was fully ordered all the way out to the first methionine, uh sort of nestled in up against the uh a partner across this assembly interface. So we had a really good um uh view of how these these came together. And this interface is completely different than anything we've seen with other structures. This is just showing that we had good density for the um for the substrates in this active filament. And in fact, because it was active, this is actually a phosphorylated intermediate of UTP. Um, and and so we it the density suggested to us that we had this this phosphointermediate. Um we're doing a lot more with cryo EM is looking at enzymes during active turnover, uh, which has been really informative, especially with things that have a lot of movement during during turnover. Uh so this is comparing, these are the three structures I showed earlier. This is showing the MTB structure on the left here, uh, where at the bottom we're showing that we've sort of painted the the assembly interfaces in orange. You can see that the structure in the filament is rotated 90 degrees relative to its its sort of configuration in the other existing structures. So this is really completely novel. And I think this really um sort of puts the nail, or what is it? Sort of, I think this is great support for the idea that this is arisen independently multiple times um in multiple different different species.
SPEAKER_01Justin, quick question. So um the uh so the the the higher order structure is different, but how how different are the individual monomers across and then also the tetramer? Like does it diverge as it gets further from monomer or does it stay pretty consistent?
SPEAKER_00So the monomers and the tetramers are really well conserved. Um uh it's it's a little hard to see that here, but I guess if you compare so the the TB structure on the side view here on the top and the human structure on the bottom, I think you can see that these are very similar conformations in addition to being uh conserved at the sort of domain level. The um the actual conformation of the protein is very well conserved. Um, and in all of these cases, we've got structures of like in the filament and outside of the filament. Uh, we have both conformations, and that allosteric transition between active and inactive conformations is really, really conserved across all of these things. So it really is like this deeply conserved unit. Uh, and then you're layering this additional kind of regulatory mechanism on top of it uh by by inducing polymerization. Does that make sense? Yeah, absolutely. I have a lot of questions at the end, but I'm gonna hold them. Okay, I'll try to get through the rest of it quickly. Uh oh, and this is showing just that. So here Eric showed solved the structure of C T of inhibited uh mycobacterial CTPS, which is in this uh sort of more compressed inhibited conformation, showed that it it looks identical to to, as I said, to other species, but also explained why it's a free tetramer in this conformation. In that conformation, you can't simultaneously be making assembly contact sort of on both sides of the of the polymer. Here he's showing if he aligns on the left, everything aligns nicely. But then on the right, you can see that the the in gray is the CTP bound structure where the the changing of the conformation of the tetramer now would disrupt the interface on the other side, weakening the whole structure and and causing it to fall apart into inhibited tetramers. Uh so this is the point in most studies where we would start engineering mutations to start breaking filaments and see what happens biochemically or or physiologically. But luckily, nature has done that uh work for us. There's a there were a number of um mutations that have been positively selected for in clinical isolates of of TB. Uh, and two were really interesting to us. One was in the active site near the ADP, or where the ATP binds. Uh the this is a structure that has ADP in it. Uh, and one is right at the assembly interface. There's a histidine 264 in TB that forms this uh this interaction with it uh, I think it's an aspartate across the, yeah, aspartate across the interface. And that mutation of the histidine totally disrupts filament assembly. The mutation in the active site doesn't have an effect on on filament assembly. And that was recapitulated in sigmatis. Here, these are images just showing that the the mutation that still forms filaments still forms filaments in cells, and the one that disrupts it in vitro also disrupts it in cells. So this in addition to to um, I guess the the thing we like about this, right, is that it it suggests that uh the the structures that exist in cells are are likely the same things that we're seeing in vitro because the mutation that disrupts it in vitro does the same thing in vivo. Um and importantly, there were biochemical differences uh in in these uh these in these two mutants. They both had reduced activity, especially obviously the one on the active site really reduces activity. Um the one in the at the assembly interface reduces activity significantly, but also shifts the sensitivity to the inhibitor CTP. So in the present, we get about a tenfold uh uh decrease in the amount of uh CTP it takes to reduce activity by half, the IC50. And so um this but importantly, these were both uh mutations that were positively selected for in cells that were being exposed clinically to antibiotics. And so that suggests that there's some positive advantage to the cells of reducing the activity of CTP synthase. And here they found two different mechanisms through two different mutations to achieve that. Um so what happens to these mutations in terms of cell growth? Well, here we were particularly interested in the the non-assembly mutant uh H267R. And what Hesper saw is there's a huge variation now in the lag phase and how long it took the cells to re-enter, re-enter growth. Um, this is just uh colony size as a function of time. And uh they could see they could watch the uh oh sorry, so so one hypothesis we had was that one of the things that could be disrupted in um in cell growth is the machinery for segregating chromosomes. And there's a there's a protein called PAR B. And PAR B is interesting because it actually uses CTP the way that a lot of like eukaryotic cytoskeletal elements use ATP or GTP. It uses CTP to promote its polymerization and cycling. And so and PAR B um segregates chromosomes to the poles of the mycobacterium micobacterial cells, uh, and they can see that in this wild type structure or these wild these wild type cells. They they always see the part B at the at the at the uh poles. And in this chymograph, that's an average of of 54 cells, you can see there's a there's a moment early on where there's no uh par B at the new pole, but then some of it, but you know, one of the structures migrates to the new pole, and then you get these nice sort of processive tracking of the poles with PAR B. That seems to be disrupted in both of these cases where you're you're reducing CTP synthase activity in these two clinical isolate um uh uh uh mutations. Uh and so there's some disruption of of normal chromosome segregation. We thought that this might be uh due to some local co-localization of the par B with CTP synthase. That turns out not to be the case, that PAR B is always at the at the poles, and CTPS is is sort of floating around, usually near the near the center of the cell. So that we didn't see a correlation between those things. But this there's a lot more work to do here, I think, to try to figure out um how defects in CTPS activity are um are leading to these um this strange localization of PAR B in the cells or the slower relocalization. So that's sort of the end of the cell biology and the and the and the um structural biology of uh MTB uh CTPS. But that none of this was what we set out to do. Um we actually have a project looking at the Human enzyme where we've got um have been characterizing inhibitors. And there's a class of inhibitors that we have structures of. Um this is a our highest resolution structure of one of them uh bound to the human enzyme. And um these are now in in clinical trials for um for lymphoma treatment. Uh but but we were interested to know there were some some small molecules that have been described for the mycobacterial protein. So we were wondering if we could use the determined structures of those as a starting point for doing some medicinal chemistry to try to improve compound binding uh to see if we could find small molecules that that targeted the microbacterial protein with with high affinity. And so these these two compounds were reported in a prior study uh at from the uh Takeda Corporation, uh, where they found that they inhibited the um uh they they actually inhibited mycobacterial growth uh and specifically discovered that they they inhibit CTP synthase. Um and we showed that they do, in fact, inhibit the mycobacterial CTP synthase, and they don't inhibit um the human isoform. So these uh plot on the right is just showing that, and that's kind of ideally what you would like, right? Something that targets this conserved protein in the in the uh bacterium, but doesn't have an effect on the human enzyme. We're still curious about why one of them activates the the human CTPS one. That actually could be very interesting, and so we're gonna go back and characterize that at some point. So we got structures, I won't belabor all of this. The point is just that the the um the uh the compounds bind roughly in the same pocket as the as the human CTPS inhibitor uh that I'm showing in the lower right here. The purple is the human CTPS inhibitor and orange is one of the structures that we solved here. They we think they function by blocking binding of ATP. You can see in the upper right here that the pink, oops, the pink ATP uh would be displaced by the binding of the small orange compound. So we think these are relatively low affinity compounds, um, but since we know that they we know that they directly uh target CTPS, that they affect uh mycobacterial growth, and we know CTPS is essential for for in um in mycobacteria, uh, we think that we have some opportunity here to prove these compounds to to better target uh uh TB.
SPEAKER_01Um Justin, can I can I pause? Um so uh you these compounds were uh discovered, right? You didn't this wasn't engineered.
SPEAKER_00That's right. These these were um these were discovered in in screens, and I as I understand the initial screen was just dumping a compound library on growing uh mycobacteria and looking for things that slowed or eliminated growth.
SPEAKER_01Yeah. So philosoph philosophical question, or I guess uh thought question. So that's targeting the enzyme. Wouldn't it be cool to target the assembly? Like, is that is that too wild of a thought? Are you thinking about that?
SPEAKER_00It it is. I mean, so so um we've thought about that in a number of contexts, and one of the challenges that is that like the assembly interfaces aren't like typical druggable pockets, right? They're not like some deep hydrophobic pocket or something, but but there are lots of you know, like macrocycles and things that you can imagine developing that would uh, and there's a lot of work going on here at UW to to engineer um using protein design tools, macrocycles, to be able to target sort of undruggable pockets like that. So we we are considering that for some, you know, here in this particular instance, it seems like under selective pressure, the mycobacteria can in some cases um themselves opt out of forming filaments. And we we think in in this case that's that's a way of slowing growth so that they're not susceptible to other um antibiotics that say hit um hit the cell wall. So there might actually be a uh an advantage to accelerating the activity of the of of this enzyme to to increase growth because the the cells that are actively growing are gonna be more uh susceptible to like cell wall inhibitors and things, right? Um and so there's I think there's lots of interesting ways you can imagine um attacking this. Uh so that's really that's really it. And as I said, like this is where we started. And the interesting, the cool thing about the way this paper came together was was um we did not expect we would see filaments uh of of mycobacterial C TVS because none of the none of the interfaces that we had seen in other species, like the E. coli enzyme, none of those seemed to be conserved. So we thought we would just see tetramers and we would do this sort of um structural analysis of the compounds and then move on from there. Uh and when Eric first put these on grids, he saw filaments and he was. I was just talking to him about it yesterday as I was thinking about uh doing this podcast, and um and uh he reminded me that they looked so strange. We weren't expecting to see filaments, and these didn't look like any filaments of CTPS we'd ever seen before. He thought initially that somehow he had switched grids with another person in the lab who was working on a filament of a different enzyme that looked kind of similar to this. And so he had to repeat it several times to convince himself that he was actually looking at CTPS. Um and and so so uh it was this sort of serendipitous discovery, uh, which led us actually then to contact Hesper, who I knew from from my time at UC San Francisco, and and sort of initiate that part of the the in vivo part of the of the collaboration. Um so that's yeah, so that's uh so we sort of end with this thing, and it's a funny, like as I was rereading the paper, it's this weird little appendix to the paper, it seems like, where uh like where did this come from? But this was actually the thing that motivated the whole the whole paper to to begin with. Right. And then uh yeah, we have some some insight into what we think is why these compounds target the mycobacterial protein, but wouldn't have an effect on the uh wouldn't bind to the human enzyme, but um and the mechanism by which they're they're inhibiting. Um, but but I think that that's kind of a detail that we can we can skip. So that's that's kind of it for the for the for the paper, um and happy to to discuss. Um great.
SPEAKER_01Yeah, because I I definitely have well that that was that was great, by the way. I really enjoyed that. So but I do have some uh you know I'm bristling with questions. So one is so you know, these these metabolons are uh you know sort of been exciting as potential targets for cryum for for a long time, right? It's like okay, it's great, big machine, you know, and the one sort of my favorite is pyruvate dehydrogenase, right? And it makes total sense, right? Because you have three different enzymes, it shuttles from one to the next to the next, and it's like you know, makes it like a make a little machine. This this is not the case for you, and it's so counterintuitive because I would have thought assembling it into a filament would shut it down because, like, there's you you know, you have uh substrate inhibition, right? Or product inhibition, right? Because it's just you got to get all those substrates there. So what I I don't understand, right?
SPEAKER_00Yeah. So um that's funny. You sound a lot like many of the reviewers of our early grants who are me, it wasn't me. You found it. I got this all the time, like, oh, this is just a storage form of the enzyme, it's uninteresting, it's boring. Yeah, um, but and you know, maybe in some cases that's true, but but here actually, I I didn't put the figure in. Um we can actually see turnover happening in this structure, um, where uh the the enzyme actually has to have movement between the two its two domains, like within a monomer, there the two domains have to rotate relative to each other. And the way that these are stacked on top of each other allows that free movement to happen. Um, and that's actually in the in the paper we show um uh at the time Eric was using like 3D variability analysis in Krausspark to show that like one side of the protein would be in a more closed active conformation with substrates bound, and the other s the other side would be the two active sites on the other side would be open up and empty. It's as if there's like this this switching back and forth, but there's enough uh flexibility within the polymer to allow that. There are other geometries that that don't allow that, right? But in this case it did.
SPEAKER_01So this is wild. So it's like the um it's like classic cooperativity, but it's classic cooperativity in the filament, right? So the filament is turning all those things on in your in in sort of your uh that's cool.
SPEAKER_00We had another really cool paper about one of the human isoforms of of this specific enzyme, where we saw this this massive increase in cooperativity in the fill in the filament. It had, you know, um uh I think the I think the Hill coefficient was I'm probably gonna get the number wrong, but I think it was a close to eight, and the you know, the subject is a tetramer, right? So it was a way of getting us because of the coupled conformational changes within the polymer, it gave you that. Yeah, yeah.
SPEAKER_01That's really cool.
SPEAKER_02Justin, so uh how about the lengths of these polymers? Do they change in different cells? Uh I thought in your first image they were they appeared very distinctly in the light microscope, right?
SPEAKER_00Yeah, yeah. And we think so, you know, it it varies between species and everything, but but these guys tend to be quite long in in vitro on the order of you know half a micron to a micron. Sometimes we see filaments that individual filaments that were are that long. We haven't actually measured them, but but uh typically they're they're shorter than that and they fit like within one um cryoen hole. Uh but in cells they get quite long. And what we do know is that for in especially in the eukaryotic systems, it's not as well characterized in bacteria. In eukaryotes, CTB synthase actually co-localizes with other enzymes. Like I showed at the beginning, these these three um uh regulatory enzymes, IMPDH and CTPS in particular, form filaments on their own, but those filaments can co-localize with each other in cells. So we think in cells what we're getting is not a single long filament, but a bundle of a bunch of them coming together. And there's some some good evidence that suggests that. And we'd love to we're going to the cell.
SPEAKER_01Are you gonna do tomography?
SPEAKER_00Oh my gosh. We uh we actually right now are are milling cells down in the basement, um, uh with the hope of of being able to do this. Uh well, we'll see. Tomography is still still scary to me, but uh we're trying.
SPEAKER_01Absolutely. Yeah, but it's the coolest stuff, right?
SPEAKER_00Yeah, that's that it really because we get this question all the time like, how do we know that the the thing that you see in vitro is is really the thing that you see in the cell? And the easiest thing to do would be able to show the structure in the cell is the same as we see in vitro rather than indirect mutagenesis and stuff, yeah.
SPEAKER_01Right, right. All right, well, I'm looking at the clock, so I I wanna um uh I don't want us to take uh you know drag this out too long. So um, but let's uh move now to the um uh interview portion of the podcast. So we have a few questions for you about uh the paper, how the paper came together and sort of your brought your journey as a scientist. So we'll we'll go back and forth. So Francis has the first question, uh, and then we'll then we'll take turns.
SPEAKER_02So uh Justin, you know, uh we asked this uh all the time. So what was the breakthrough moment when you knew, oh my god, this is how we I I remember you you're telling us about oh, we started here, but that's actually the last figure. So at what point did you figure, okay, this is going to be worthwhile pursuing and stuff?
SPEAKER_00So I mean, there were a couple of those. Like the the the first, like aha was like, oh my god, these things are making polymers in vitro. Um, and then immediately we're skeptical and we're like, okay, maybe this is some in vitro artifact. Like, is this a real phenomenon? And so then that's when we contacted Hesper. When Sper was able to tag it and see, so the next aha moment was when we could see uh in vivo that these structures actually existed. That was and then and then when we knew when we learned about these clinical isolates that that had these mutations that affect the structure, uh, that to us was really exciting because it meant that not only would were the filaments likely doing something in, but that that it was it was tied to the um ability of the cells to to persist in infections. And so I think it I don't know which one of those was the, but that that the the sum of those things got us to the point where we're like, oh, this is actually a really interesting and exciting thing.
SPEAKER_02So cool. Yeah.
SPEAKER_01Yeah. So I love these aha moments, right? That's like the things are so exciting. So um you uh as you know, as scientists. So in your career, do you have like a piece of data or an image or a plot that's sort of burned into your memory? It's like, you know, something so exciting you didn't anticipate. You have any stories like that?
SPEAKER_00I do. The one I always talk about. I have one that is just absolutely I still get I still get goosebumps sometimes thinking about it. I was a postdoc um and I was working on uh the machinery that nucleates microtubules. And it uh the the sort of fundamental subunit are these little 300 kilodalton Y-shaped complexes called the gamma tubulin small complex. Uh, and we had known that like multiple copies of those come together in some way to form a larger structure that uh that nucleates microtubules. And I was playing around with conditions to try to promote, I was just hoping I could get two of these little Y-shaped things to stick side by side. And um uh I was playing around with the conditions, and I was down in the in the microscopy suite on the ground floor at uh the uh the um genentech hall at Mission Bay in San Francisco, and um was on the microscope looking at these negative stain images. And you know, back then I was it was always on the on the eyepieces on the microscope, right? Uh so I have the screen down, it's like glowing green, I'm looking in the eyepieces, I'm scrolling around, I'm focusing on what the hell is that? And I ooh, and I saw these structures that were like these elongated spirals of those Y-shaped things coming together in these very fat, sort of looked like uh pasta, these these fat sort of noodles. And I I saw that, and the the first like within three seconds, it was like, what the hell is that? Oh my god, I know what that is. These are the the larger assemblies, this is the thing, and it was this moment, and it was just like uh uh I when you're when you're like the only person in the world who has seen this thing, right? And then there was that, and then it was coupled to like, oh now we're gonna be able to understand how these things work. Like, this is such an obvious next step is I'm gonna get the cry VM structure of this. Uh, and then and then uh and then like also like oh my god, and maybe I'll have a career in science, maybe I can get a position for this. Like, yep. All of that went through my head in like you know, 15 seconds basically. And it was, I was so that it was such a thrilling moment. And uh and all of that came to pass. We did the cry VM structure, we we got the paper, I got a faculty position. We uh I think largely because of that. So it uh it was very exciting.
SPEAKER_02All right. So, you know, so when did you uh realize uh that you wanted to get into science, like uh moving back uh a few more years earlier than your graduate school and stuff? Yeah.
SPEAKER_00Yeah, it's funny. I always think about my um my sophomore year in high school. I took uh I was in like that was when we took biology in my uh in my high school. And uh I was kind of bored by like we did a lot of anatomy and stuff early on, and then uh but central dogma, like like thinking of like machines that do all this stuff, and it's happening all the time inside, right? Like transcription, translation, uh replication, like and and we were like it was gosh, that was a long time ago now. That was like the early 90s, and so it was things like uh genetic engineering, right? Like the that it was it was I think it was I I was learning that biology could be more precise than the kind of this is a liver and this is uh this is a uh it was like there are these machines and they're like doing this precise cutting of DNA to give you sticky ends that you can then use to to recombine DNA to to to make new things. Like that that got me super excited and and really sent me on a a different uh uh sent me down that path of of thinking I wanted to study uh you know uh biology and biochemistry.
SPEAKER_01That's great. Yeah, so that sort of uh relates to the next question about like sort of inspiration. Like we're all sort of the product of various inspirations. And so one thing I like to think about is like who are our scientific heroes. So do you have like a scientific hero and like who would that be and why?
SPEAKER_00It's funny. For me, it's it's pretty personal, I think. It's um uh and I've been really lucky. Like uh my graduate advisor and my postdoc advisor are both people that um who are very different people, but I admire and I learn so much from them about how to be a scientist. Uh in fact, Russ Doolittle is my graduate advisor. That's him on the wall behind me there.
SPEAKER_01Oh wow.
SPEAKER_00Um uh you know, I didn't, I when I came to graduate school, I was not prepared. And uh and I I uh was lucky to have a patient person uh who really took the time to like teach me how to be a good scientist and to think like a scientist. Um and so it's I think that's different than maybe like having a hero who's done something big and exciting. It's more I mean Russ did plenty of big and exciting things, but it was more like the way that he like lived with the science. Uh and it was taught me that like it's personal, right? And it's it's it becomes part of who you are. And and I really uh I really appreciated that about my time with him. And then and then my my postdoc advisor taught me to think bigger and to and and to to be to be bold and in and creative um in ways that uh were really helpful to my career. Yeah, Dave Agar, he was great, or is great.
SPEAKER_01It was a great story. Um thanks for sharing those. So, Justin, thank you so much for joining us today. We really appreciate you taking your time to tell us about your science.
SPEAKER_00It's it's really been a pleasure, and thanks for thanks for listening. It's it was fun to talk about this for sure. Absolutely. It was a joy.
SPEAKER_01Thanks, Justin. Thanks. Thanks a lot, guys. Well, that does it for this episode. We all hope you we hope you will all join us again for the next episode of Structure Club.