 Hi, this is Kimball Fraser, next to me Clay Foster, and also on the split screen is Gilsung Park. These are two of the co-authors on our recent paper on Unka Target. Unka Target has asked us to sort of summarize the work in this manuscript, and also put it in a larger context of other work that we've done related to this manuscript. I'm gonna try and briefly tell you the story here, and along the way, I'll ask Clay and Gilsung to chime in with specific experiments that they did. Before we get into the Unka Target paper, I should mention that in the prior year, we actually had a couple of other papers that we published in the journal Leukemia. All of these have dealt with essentially the same topic, which is the discovery of B-cell acute lymphoblastic Leukemia in a zebrafish model that was created many years ago, actually about 10 years ago. And this zebrafish model is one that our lab and many other labs have used over the past decade. It uses a zebrafish rag to promoter, which is active in immature B and T cells, to drive a transgene human mix specifically to induce leukemia. Landmark work by Dave Langenau, really almost two decades ago, 2003, I believe, was the first example of driving cancer and zebrafish with a transgene that was done at the time with murine mic in a near identical setup, again, using the zebrafish rag to promoter. And people had been using that model for a long time, a little over a decade or so later. Alejandro Gutierrez came along and made the line that we used that used human mic rather than murine mic, and people continued studying T-cell acute lymphoblastic leukemia in those fish. So I acquired those animals from Alejandro Gali probably almost 10 years ago now, before I came to my current institution. And we had been studying them for some time. And we've been studying particular context with a marker transgene that used the zebrafish LCK promoter, a promoter that's highly active in T cells to drive GFP. And so in our hands, we saw lots of animals that got T-cell acute lymphoblastic leukemia, and it was easy to recognize those leukemias because they were all glowing bright green. A former postdoc in the lab, Chiara Borga, who's also a co-author on this current paper, really made the important observation that these animals were getting dim cancers also. In other words, cancers that when you looked at the zebrafish through the microscope, they were fluorescent, but not impressively so. And I really can't speak to other laboratories that had been using this model, but I can tell you, speaking from my own perspective, having looked at these animals an awful lot over 10 years, I had really discarded these animals figuring that these were either animals that had early TAL that just hadn't blossomed or were maybe just animals that didn't have malignancy at all that just happened to have an accumulation of T-cells. So in these two prior manuscripts, we did lots and lots of studies to prove that they in fact were B-cell acute lymphoblastic leukemia and to show that the genes that zebrafish B-cell acute lymphoblastic leukemia express are similar to the same genes that human BAL also expresses. The other thing that we did was about the same time that we were discovering zebrafish BAL in our model system, the Langana group that I already referred to that continues to use the murine mic model, also discovered BAL in their system. And so we had also written another paper where we compared BALs between these two zebrafish systems. So now the Anka Target paper was really sort of ongoing work that we continue to expand on. And I'm sort of gonna just walk through the paper. You'll have to excuse me, I have the paper over and another monitor and I'm gonna scroll through it and more or less just tell you what I think are some of the important features of the figures. So in the first figure, Gilson did a really awesome experiment that highlights the beauty of zebrafish as a cancer model. He monitored over 600 animals that were carrying this constellation of a rag tube, human mech and an LCKGFP transgene to see which animals got TAL, which would be bright under the microscope and that we could prove by fluorescent microscopy as well as flow cytometric analysis of the individual ALL cells. And rather than tell you the result, I think maybe I'll just invite Gilson to tell you what was the result that he saw when he monitored these 600 animals for the better part of the year. So about day 180, about 40% of fish develop the TAL, which is a bright tumor and about 12% develop the BLL, which is also dim tumor and other 7% of fish develop the both simultaneous T and B ALL at the same time. So, but over the year, about the percentage of the TAL kind of increased to about 64% and BLL is up to 23% and mixed ALL is up to 13%. Thanks Gilson. So I'll make a few comments. The first is just to make sure people who haven't been following this story closely, what Gilson means when he says mixed ALL. So, you know, these fish have B cells and T cells and they can get two cancers at the same time, just like people unfortunate people could get two cancers at the same time. So when he refers to mixed ALL, he's talking about his zebrafish that has the misfortune, if you will, of having both a B ALL and a TAL at the same time. And we think these are, you know, totally independent events, obviously related by the transgenic human MEC driving them. But to me, an animal that gets B ALL is an animal that got B ALL whether it happens to have TAL or not. Okay, so that was sort of clarification. So the two important things that I think are apparent from this study, the first I already alluded to, and that is that, you know, this is a huge study, 600 animals, people who study cancer using mouse models could never do this. It would cost too much, not to mention, you wouldn't want to have to bleed mice every three weeks to look at them or if you were using Luciferase and CCD cameras, you wouldn't want to anesthetize mice every few weeks to look at them. But it was relatively trivial, easy for me to say because Gilson did all the heavy lifting. But it was relatively trivial for Gilson to take a large number of animals, sedate them on a regular basis every several weeks, look at them under the microscope, find which ones had bright tumors, which ones have dim tumors and do all the downstream processing. So I think this really shows something that you can do in zebrafish that you can't do in other cancer models. The other important thing that it shows, and in terms of sort of maybe solving the unspoken mystery of why didn't we know about TAL and zebrafish for over 10 years but we didn't know about BALL? I think this is at least one of the answers to that mystery. Probably not all of them, but it's the answer to the mystery in my lab. And that is not only are the BALL dimmer, making them harder to see, but also the BALL happen about one fourth or one fifth as often as TALL. So on balance, you're getting a lot more animals that have TALL and have BALL and the TALL is easier to detect than the BALL. So I think collectively that sort of accounts for Gilson's results. Now I will tell you that we have some interesting and what I find to be exciting results that all comes from Gilson's work that's gonna expand on this part of the story a bit in the future. I'm sorry to say I'm not ready to disclose that information just yet, but maybe that'll be in a forthcoming on Katarget paper. The other thing that we show in figure one, I'm not gonna take you through every figure in every panel, but at the end of figure one, we also show a couple of new lines that Gilson and Amira Hassan, another co-author in our paper, in my lab also made and we're using these a lot going forward. So, I mean, it's really cool to me that the LCK GFP transgene allows you to see both T and BALL in the same animal or in different animals and that the brightness of the fluorescence allows you to get an early snapshot of whether you think it's TALL or BALL that you can confirm in the future. But I've already pointed out how difficult it is to detect BALL in these animals. Yours truly looked at these animals for probably five years without ever detecting BALL. It took somebody smarter than me like Kiara to come along and recognize that BALL was present. And so for the old and infirm and blind like me, I thought it would be really useful to have lines where BALL would be obvious and easy to recognize. And I thought that would be useful not only in terms of making the disease easier to detect, but also in terms of being able to find drugs or test new drugs to show that animals were responding. And so this wasn't a complicated experiment. In fact, it's not even really an experiment. It's really just sort of a construction. But we took two other transgenic lines that had been created by another laboratory that again use the proficient promoters, promoters for either CD79A or CD79B. These are two different proteins that are both highly expressed by B cells, but to our knowledge are not expressed by T cells or other cell types and using them to drive GFP instead of LCK. And so also in figure one, you can see examples of animals that have BALL with either of these marker transgene and they just look like TALL in the LCK background. The tumors are bright and that makes them easy to detect and it also makes it really easy to see if the tumors are responding to treatments. So that said ways nicely into what I think is figure two if I remember correctly, and yeah, it is. So figure two is a composite of just six different animals. Three of them that were treated with dexamethasone, a glucocorticoid that's routinely used in both T and BALL treatment in human beings. Gilson did the experiments with dexamethasone. It also shows three other animals that were treated with radiation, a therapy that used to be used a lot in ALL that is still used in some cases, but because of toxicity we tried to replace it with drugs. And it shows animals, most of them are in the LCK background, but one example animal is in the CD79A background and it shows that all six of these animals as well as many, many others that aren't shown in the figure respond to these same two treatment modalities just like human ALL does. And so I think that already shows how the CD79A animals are actually probably a lot more powerful than the LCK animals if you want to look at BALL. Probably nothing else to tell you about figure two. Figure three. So figure three is an expansion of some data that we already published, I guess in our first leukemia paper about this story. And I'm gonna call Clay in in just a minute, but by and large this is an experiment looking at gene expression. It's using a really cool technique called nanostring that I won't go into details in, but this is experiments that Gilson and Chiara did together probably about three years ago. And again, like I mentioned, we'd already published some of this in one of our prior leukemia papers, but in the leukemia paper, mostly what we did was look at the genes that distinguish BALL from TALL and zebrafish. And then we looked at the genes that zebrafish BALL express and compared them to genes that human BALL express. Excuse me. So this figure, which I guess is figure three in the Anka Target manuscript sort of breaks things down a little bit more. I don't know, I'm sort of putting Clay on the spot. I didn't tell him I was gonna ask him to do this, but the three different panels of figure three look at every sample, but a large number of different samples that we did. And then breaks it down in a way that we didn't in our prior manuscript and also include some samples that we hadn't included in our prior manuscript. Two or figure three B looks at marrow samples of lymphocytes and figure three C looks at thymus samples of lymphocytes. And I'm gonna shut up. I don't know if Clay has anything to chime in, but even if he doesn't, I have something else where I'll get to talk to him. I mean, that's pretty much it, yeah. These are, yeah, just the nanostring data that Chiara had looked at a little bit. A lot of these are sort of hand-picked genes that we think diagnostic between the two different groups just to see the differences. Yeah, I mean, there's not much more to that figure. So there's not much more to that figure, but I hope there's gonna be a lot more to this story. So figure three C looks only at thymic lymphocytes. And some of these thymic lymphocytes, the ones that don't express very much GFP are B lineage, we think. And some of them, those that express GFP highly are T lineage. And so this is a nice snapshot. Like Clay mentioned, these are sort of hand-picked genes that we thought would be good at distinguishing B and T cells. And in our hands, they proved to do that. But just like everybody else, we're evolving into single cell RNA seed studies. And two people in my lab, both co-authors on this paper, primarily Amira, but also Gilson are now doing lots of single cell sequencing of both B and TAL cells and B and T cells from thymus and marrow and spleen and other lymphoid organs. And so many of these same genes shown here are proving to be sort of the diagnostic genes that we're using to identify cells in a single cell context and work that we hope to publish in the relatively near future. I think that's it for figure three. Figure four is really all the work Jessica Burroughs Garcia, another former postdoc in the laboratory couldn't be with us for the Skype call. So Jessica, I guess sort of the intermediate between doing experiments like nanostring or bulk RNA seed where we looked at a bunch of different leukemia cells or a bunch of different lymphocytes all at the same time. And then the single cell RNA seed studies that I just alluded to that we've done and are doing but haven't published on yet. So in this particular figure, what Jessica did was she took individual cells and rather than doing single cell RNA seek on them, she did single cell QRTPCR to see which genes were expressed by individual TLL cells or BAL cells. And really this result didn't show anything surprising but it showed something that was nonetheless pretty important to us. She did a lot of different single cell experiments and this is a small subset of those. But what this figure is showing is just that 10 BAL cells, all of those cells individually are expressing lots of B cell genes but not many if any T cell genes. And conversely, 10 TLL cells are expressing lots of T cell genes but not many are any B cell genes. So exactly what one would expect. But I mean, as we all know in science sometimes what you're expected is what you get. So that was a pretty important experiment for Jessica to do. And like I said, we're sort of sitting on some other unpublished work that she did that's related to this that we hope to publish in the not too distant future. Figure five, we can largely skip by. It's not really an experiment. It's just an alignment between human and myrimic and the reason we chose to include it here was just to show readers that human and myrimic are nearly identical at the protein level. And we didn't publish this but Clay, this actually does give me another opportunity. I know Clay isn't gonna remember any of this off the top of his head, but Clay did a really methodical amino acid by amino acid and domain by domain comparison of myrine and human mick. He flooded my inbox with this. He wasn't in our group physically at the time. So I don't know if he's gonna, if he can really tell you anything but I could tell you what my gestalt from his extent of analysis was. And that is we'd expect that human and myrimic should do the exact same thing. You know anything to add to that? There was one, if I remember right, this was quite a while ago, but there was one key change. I think in the dimerization domain, there was one key change in an electrostatic interaction that I think my conclusion was that that was probably pretty important into what H mick and M mick, the difference, whatever difference there is, that was probably pretty key to it. I don't remember exactly which residue it was. It was too long ago, I'd have to look back. Well clearly Clay remembers it much better than I because I didn't even remember that Hallmark residue. Maybe we should have highlighted here or my oversight in the figure. But I think the important point is, I mean, we don't have any experimental mental validation. Now that I have Clay here, maybe we'll do some experiments that can experimentally validate it. But I think one of the other mysteries that sort of came out of this project wasn't only the fact that Mick could drive T versus BAL and the fact that one was really easily forthcoming and the other came a decade later. The other thing that we saw, and this was a collaborative work with the Langenau group, was we saw that the BAL that were driven by Miri Mick and the BAL that were driven by human Mick actually had very different gene expressions and that's actually gonna be the last figure that we'll get to in just a second. But there are several potential explanations for why that might be. I think I could speak for Dave Langenau in saying that he felt that probably it was the strain background that might be the key difference that if we put Miri Mick in our background, maybe they would look more like the gene expression profiles they saw and that conversely, if we put human Mick in their genetic background, maybe the opposite would be the case. That's certainly a possibility. For me personally, I sort of felt like, well, maybe it's just like an expression level difference. The RG2 promoters, we think are the same but they undoubtedly landed in different places in the genome because they were just put in by random total transgenesis so they could sort of have landed anywhere. So maybe Miri Mick and human Mick are expressed in different levels. Not only in the two different genotypes but also maybe in B versus T cells. So that's a potential explanation. Clay just alluded to a potential molecular interaction. Mick has to either hetero or homodimerize to do its thing transcriptionally. There are also several other proteins that either can bind Mick and sequester it from doing its thing or combine with Mick's partners and sequester them from doing their things. So you can imagine that if dimerization were different and the it is even finished between human and Miri Mick that might account for differences. So we don't have any answers here but that's the mystery that could be solved going forward if we or others step forward to do it. So the last part of our figure, sorry of our manuscript is figure six. And figure six is sort of a redo of some experiments that we published in a prior leukemia paper. It's redone in a couple of ways. One is the samples that were used and aren't the exact same. And the other is the bioinformatics tools were not the exact same. This is really, I kind of twisted Clay's arm to make him participate in this interview. He was a little camera shy, but if he'd be willing I'd like him to tell us a little about the bioinformatic analysis that he did. Because like I said, it was very similar to the collaborative work we did with the Langana group in the prior leukemia paper but he more or less just used different tools to look at gene expression. To me, what he showed was that, well, when you use different tools a lot of your results are the same but they're not exactly the same. Sure, I mean, we were certainly trying to cast maybe a little bit wider net here than we had before in the leukemia paper. So some of our thresholds were a little bit lower and stringency, but basically this is the compare, we compared the HMIC, the ALLs which we were calling IGHZ positive. We found this in the leukemia paper with the Langana group and we compared it with the MiriMIC, we called IGHM positive. So Z is the HMIC is the M and we did just a basic re-analysis of those two groups using, there were a few technical differences. We had a little bit more of an updated genome. As I said, our thresholds and parameters were a little bit differently. The tools that we used were a little bit different. We did an over-representation analysis using a little bit more modern gene sets just to see what differences we could find and whatever jumped out at us. That's what we were trying to see was to get a clue at some sort of biological process that was different between these two. So one of the challenges that we ran into when we were working on one of the leukemia papers was that, we did the discovery of these DALL population was unexpected and the samples that we were using here, it was questionable how pure, I guess, of a population that was. Yeah, I mean, we thought, I thought that these things were, this was all bulk RNA-seq analysis. And I think that 95 plus percent, maybe 99 plus percent of cells were BALL cells, but maybe not 100 percent of cells. And I think that's happened for other people's bulk studies, too, when you purify BALL, you're also purifying non-malignant lymphocytes, too. We had not only that concern, but that when in purifying our BALL that we might have been purifying some T cells, too. And we thought that that might have sort of come down to, yeah, exactly. We had a preliminary gene list from results for samples that we thought were more homogenous, a more pure, I guess, BALL population. And we tried to use those, that gene list is sort of like a checklist on these samples to sort of exclude things that we thought, well, perhaps this gene actually only seemed up because of maybe the small population of T cells that we saw in one of our samples versus that one of the in-mic, the mirroring mixed samples. So we tried to exclude genes that we thought were a result of that to try and, I guess, remove those, that confounding effect. We think this is a more authentic list. Yeah. Or I guess we think that it's a more authentic list except for all the genes that carry an asterisk in figure six. So I know that Clay didn't want to include these because Clay's a hardcore bioinformatician and biostatistician who wants to do everything the right way. And I'm kind of an old school, old fart. And so I see trends. And so I really pushed hard. I don't know, probably all the other co-authors around here didn't like it too much, but I'm the boss. So we got to do it my way and reviewers didn't object. So I really wanted to show something that didn't pass statistical threshold criteria, but to me was intriguing. And I would invite reviewers and readers to look at the data yourself and you can decide whether you think I'm right or whether you think we should be hardcore statisticians. I think it's really intriguing that in the human mick-driven BALL, we see that an awful lot of other, what I'll just loosely call mick family genes are upregulated in those tumor cells. What I mean by that is so both the human zebrafish genome have other mick family members besides what used to be called CMIC. And now we just say mick. There's also in mick, there's also lmick. In zebrafish, it turns out there's actually CMIC A and CMIC B. And then there are lots of max and mad other heterodimerization partners, not every single one of these genes. But what to me seemed like a non-random assortment of mick and mick related genes are all much more highly expressed in the human mick model than in the nearing mick model. And I don't know, maybe that dovetails nicely with Clay's point about potential changes in the two protein structures that might account for different dimerization, partnering interactions. Having said that, this is RNA data, not protein. So maybe it doesn't mean a darn thing. Anyway, I just wanted to throw it out there because I thought it was an interesting difference even though I recognize and the asterisks denote that not every one of these genes met criteria for statistical significance. Now, in my defense, I would say that, this is a very small sample set. We had four human mick driven BALL and really we only had two mirroring mick driven BALL from the Langenau group that they had sequenced independently twice. So to me, the fact that they didn't meet statistical significance wasn't that surprising when you're only looking at four versus two samples, something's gotta be nearly black and white to jump out of statistically significant. So I would argue, if and when we get around to doing 10 human mick BALL and 10 mirroring mick BALL that all of the genes that I think matter will hold up. But until we do that, I don't know. So I don't know, that kind of brings us to the end. We didn't go through every little dotted I and cross T but I think those are sort of the main points in the paper, both what we can conclude and what we can't yet conclude and what we hope to do experiments to learn the answer to in the future. So I don't think I have anything else to say. I've rambled on. For those of you who've actually viewed this whole video I can't believe it. You must be interested in the topic. And if you are, feel free to reach out to me or any of my co-authors. We'd be happy to talk to you about it more.