My name is Adrian Jubeck. I'm a geneticist at Brigham and Women's Hospital in Boston. I'm actually a molecular geneticist and a cytogeneticist. So I go back and forth between both worlds. And I wish I was as creative as Julie in terms of coming up with questions in advance. I didn't. So we did a lot of these on the fly. I think what I'd like to do as much as possible, let's make this interactive. So if I say something you don't understand, throw up your hand, throw a pen. Let's figure out a way to make sure that if you have a question, we answer it right in real time as much as possible. And if we don't have participation, I'll do my best to pick on people. Just try to encourage some conversation as much as possible. Out of curiosity, just before I start, so I know we just had a question earlier about fish. But how many people have done cancer or currently do cancer-based studies as part of the clinical work? Handful. Yeah, regularly? Or is it? No. I have done. Have done, sure. And so this lecture specifically is designed to really highlight the nomenclature, the nature, the structure of fish with a context of hematologic malignancies specifically. And so we'll really try to beat that as much as possible. Some of this is going to be repetition from some of the concepts that you heard from Julie earlier. So in general, I think I just mentioned this basically. But this is, again, designed to just quick background repeat more pictures on what fish is, how it works, the nature of it, the probe types, and really what their utility is. And that really is the emphasis we're going to spend quite a bit of time on, understanding what different probe strategies exist, why we have them, and what the utility of that specifically with cancer in mind. We'll give some examples. And then how that relates to the nomenclature. And so hopefully, again, by the end of this, you have a better sense. Just you look at nomenclature string, you have a good sense of, I know what's going on. I understand what's happening more or less at this point in time. And then we'll end just talking about, if you're bringing in fish to your own lab, what are the steps that basically are involved in terms of bringing that probe on? What are the things you have to consider? And certainly, I think that depending how we're doing with time, we'll talk about some of the experience we've had, some of the mistakes that we've done, and some of the lessons we've learned through that process. OK. So fish, I think everyone's probably seen this picture about a million times. But I think that as a general overview, I think that everyone, no doubt, especially at the last lecture, is well aware that, again, fish is basically the process through which we, to steal Julie's words, we unzip DNA, take a probe. Typically, a DNA could also be an RNA probe molecule that's labeled. And ultimately, through the process of denaturing and hybridization, sort of DNA complementarity, we evaluate basically where that signal is. Where does the specific sequence of DNA bind to in the genome? And then as pretty as that picture is, again, there's some more painful steps that go into this. And again, we're not going to get into the specifics of all these. But I think really what I wanted to take home is that when we're talking about fish-based studies, again, there's many possibilities in terms of what types of tissue we're working with. From a cancer perspective, certainly from a heme perspective, oftentimes we're talking about working with fresh material. So again, this is a bone marrow sample, a peripheral blood sample in those cases. But certainly with pathology in general, I think the currency is oftentimes fixed material, so paraffin embedded. And we'll definitely get into the differences between the two. General steps are listed here. Definitely won't talk through all these points. But ultimately, what I wanted to highlight again is that it's sort of common steps that basically get initiated. But depending on what they're working with, whether fresh or fixed material, there's a pretty significant difference in terms of the turnaround time. So when we're talking about fresh material, again, we have the ability to potentially get a same-day result. And certainly that's increasingly being the case with some rapid hide buffers. So again, cocktails that are specifically designed for the purpose of being able to produce a result with sort of high-quality signal in a relatively quick amount of time. And we'll show some examples of where that's particularly useful. Whereas for fixed material, because of the process of actually taking a formalin-fixed sample, it has to go through some additional steps that aren't required for fresh material. And that adds on some time in general. And so when we talk about fixed material, again, oftentimes it's for solid tumors. We're really not talking about a next day or even a two-day result. Sometimes it's four to seven days, something like that. But ultimately, why are we doing FISH? And I think certainly a few different reasons are listed here. And I'm sure this group would probably come up with a few other ones. But in general, when we talk about a FISH-based study, ultimately there's a few reasons we want to do this. One is that we are aware of events that are below the level of resolution of the current assays we're using. So we talk about GTG banding analysis. Again, best case scenario, we say maybe events that are 5 megs, 3 megs, somewhere around that region. Depending on the length. Certainly with cancer, it's oftentimes quite a bit larger than that, because the chromosomes are smaller, uglier, frankly, at that point in time. And so we know there are specific aberrations that are cryptic that we necessarily cannot detect by GTG banding. And this is an ideal situation in which we can apply FISH-based studies. Also, it's the rapidity of this, the ability to quickly get a result. So if I have a differential diagnosis and I need to identify a specific change, identify presence or absence in the context of treatment, this is, again, exactly a situation in which you want to use FISH. Again, with cancer in mind. And obviously, the last, probably one of the most critical examples of why FISH is being used, too, again, is it precludes the need for cell culturing. And so suddenly bypassing one of the major, both strengths and limitations, that's associated with karyotyping. Karyotyping, again, we allow it to divide. You get a low resolution, but a whole genome assay, really. But the cells have to divide. If they don't divide, you're out of luck. Suddenly, with FISH, we can overcome that barrier and evaluate a specific signal. Before we go on to the next slide, so what are the limitations, though? FISH. What can it do? What are the problems with a FISH-based approach? You have to know what you're looking for. That was too productive of a question. You have to know what you're looking for. Absolutely. That's the big thing. What's the question you're going to ask? If you don't know what you're going to ask, you're out of luck at that point in time. It's the same point. OK, we've got two for one. That's awesome. And I'd say that's one of the big ones, definitely. I think we'll touch back on that a little bit later on. So obviously, when you talk of FISH, there's actually a few different layers to this that we really have to consider. And so one of the first layers is what, ooh, that's a little grainy. What, basically, is the target? What are we after in this particular case? And so this breaks down to three different categories. We typically think of this as centromeric probes. Again, this is the alpha satellite DNA that was mentioned earlier. We have locus-specific probes. And so oftentimes, again, it's a specific region of the DNA that we're really after, a specific gene oftentimes that we're particularly keen on evaluating, the numerical or the structural component of that gene. And then whole chromosome paint, which, again, is a technique that I think, again, we really should drop off the slide at this point in time. We still do quite a bit on a research side of things. I think from a clinical diagnostic side, it's almost, at this point, completely extinct. I don't know if I know any labs that are currently using this. So the locus-specific probe? Yes. This is a superconcave saturation that's found. Did you do interface with these genes? No, no, no, no. Yeah, this is the same point, I think, that's come up before. So again, just depending on the state of the chromosomes at that point in time, you see sometimes separation between the individual sister chromatids. And that ends up being just a regular process. It's something where it ends up coming up quite a bit, or something that you have to be cognizant of when you start doing validations of anything. Just, what do I see? How does the probe perform? And we'll get to some other points that can act as potential confounders. But no, it's one, basically, locus that just split into these two sister chromatids. And in interphase, you would not see that. Because of the state of condensation at that point in time, you don't have the chromosomes. It's a lot like DNA. Good question, though. Any other ones? And then other aspects, again, sort of layered on top of this, again, what probe we're using. The next aspect is, what are we actually targeting? And what do we need to target? So we have interphase and metaphase fish in this particular case. And so in both cases, we can get information about, basically, the numerical changes that occur. How many copies do I have of each of these individual locus? This is a probe that spans the centromere for chromosome 8 and then a locus-specific region on 8p21 point something. And so we can see all these look like probably normal nuclei in this case, two signals for each. You can also see the variation that happens, that despite the fact these are on the same chromosome, the centromere and 8p region, that depending on the actual interphase nuclei, these can look like they're in quite close proximity or look like they're rather far apart. And so again, there's actually quite a bit of variability that happens on individual interphase nuclei, just based on positional effects and how condensed or decondensed the DNA is. Whereas with metaphase-based fish analyses, in these cases, again, we have the ability to both evaluate numerical and structural changes in this particular case. So how many copies do I have? And again, you can see maybe it looks like two faint dots there. Again, it's just the sister chromatids that are separated. But we have a better ability to evaluate specifically what's going on in terms of where the specific signals are in those instances. And obviously, the limitation with the latter. So great that we get this structural information. But then again, this actively requires dividing cells. So now we're back to the same paradigm that we were with GTG banding analysis, where the cells aren't dividing, we're out of luck. Out of curiosity, why? Well, we'll get to that. Yes, we'll get back to that. So starting with interphase nuclei, and then we'll get into some probes in a moment. When you're looking at interphase nuclei, I think there's a few things that you want to consider when we're doing the evaluation. One is, again, when we look at what are the probes that we're hybridizing. So in this particular case, again, we've gone through a little bit of nomenclature. So nuke-ish, nuke, again, it's interphase nuclei. We have three different probes here in this particular cases. So what chromosomes are we evaluating? You can tell based on that nomenclature. Awesome. Yeah, 12 and 13. It's exactly it. And this is just a slightly different nomenclature, but LSI is a locus-specific one for chromosome 13 also. And this is actually a cocktail that's used for CLL. In this particular case, there's parentheses around this whole region, indicating that, again, it's all these probes are hybridized together in a single reaction. So within each individual nuclei, we're evaluating all those probes together. And then here, from a cancer perspective, there's 182 out of 200 nuclei that had this specific hybridization pattern. So what happened to the others? Normal, exactly, perfect. And again, cancer, fair enough. We specifically speak to the abnormal cases or the abnormal patterns when they exist in this particular instance. So from a leader's perspective, if you're actually looking down the scope, a few of the things that you really want to be mindful of, again, how many signals? That's obvious. Counting dots, fine. What's the signal intensity? And so in that amplification and not quite amplification, we'll get to that point. Amplification, to some extent, will come down to how many signals you saw. But intensity will matter. And again, we'll get to that point. So maybe keep that in the back of your mind if you don't skip ahead a couple of slides. But the intensity of the probe, and ultimately, sometimes, again, the probe location. And so while I did mention earlier that, again, if you look at individual interfaced nuclei, you can see quite a bit of variability in terms of how probes are spaced, even if they're on the same chromosomes. Having said that, if you see consistent patterns that keep on emerging, that can oftentimes be incredibly informative about some of the underlying genomic changes that's happening. So what will be the public and the medical control for these probes? Mm. So if it's three, do you think it's going to be three? So again, that's a good question. A lot of that comes down to actually how you validate a probe. So typically, as long as you have two different probes that are hybridized, co-hybridized together, one serves as the control for the other. You expect to see basically just hybridization, yes or no, at that point in time. And so in the context of a validation, we'll get to that. You know what the typical characteristics, you know what you expect to see in terms of how many hybridization signals. And now at this point in time, you're looking at a relative difference in the sample. Hope that answers your question. But if not, we definitely have some examples that hopefully should get to that point specifically. OK. Ooh. Jumped ahead. Probe types. So what type of probes? What do you use probes for? I'm a cancer, well, cancer anything. What do you use probes for? What type of probes exist? Who can name a probe? Breakaway probes. Break apart? Yeah, that's one for sure. Oh, wow. For sure. Let's throw ahead. I heard like four voices. But what was that? Yeah. So yeah. So and I would say those oftentimes, you could consider those enumeration probes. If we're looking at for a specific signal, we're also looking for how many copies or how many numbers of that sequence. Is there any other ones? There's one other. Pardon me? Yeah, yeah. And again, I'd say absolutely correct. Also, I think what I'm bending that into a general enumeration category. It's how many copies basically do we have. There's one other one that has to do with rearrangements, and we'll get to that in a little bit. Yes, dual color, dual fusion. Exactly. Awesome. So those are two different types. Both are rearrangements. Break apart, dual color, dual fusion. We'll get to those two in terms of those differences specifically in a second. Yes? So I'm just curious, right now in cancer, what percentage of oncology will alter patients that are electrophagistic? Good question. I don't know if it's always an either. Oh, yes. Good point. Excellent. I remember your reminder of that. The question was, I guess in our practice, my practice, what percentage of cases is fish being ordered instead of a molecular test? So that's a good question. I would say disappointingly, a lot of times what happens is that both tests are ordered. And so we've actually taken a pretty active role in our group in terms of saying, we try to eliminate redundant testing wherever possible. And so we'll oftentimes, we're, again, there's a few of us who sort of cross both labs. And so again, we have a pathway we've developed to make sure that when we have cases in which sort of both a molecular and a cytogenetic test are requested with the same potential probability of identifying a result, we'll hold one and run the other in those instances. Yeah. And the situations, and I think we'll get to some of them. If not, I'll definitely give you some examples in the next lecture I give. Then at the highlight points where I think, had you started with a molecular test, you may have missed a result. And again, sometimes with the exception, sometimes the rule, and vice versa with fish in those instances. OK. So let's start with enumeration probes in this particular case. Again, this is typically just a probe to say how many copies will have a specific region in these instances. Typically, again, as sort of per ACMG, there's a number of guidelines. They have a specific guideline on fish performance and sort of criteria and things that you should consider. And one of the comments in that guideline, again, is making sure that when you have a probe, any probe, you always co-hybridize it with a second probe, never one by itself. And the reason for that, again, is that if that probe fails for whatever reason, if there's a homozygous deletion, by example, in a cancer case, again, you'd have no ability to differentiate sort of the absence of signal from a poor hybridization. So two different probes, mix them together, oftentimes the centromere with a locus that you're specifically interested in this case. So this would be an example of, again, chromosome 8. We have a set probes, alpha centromere DNA for the centromere of chromosome 8, and then MYC here at AQ24.21. And again, for something like this, again, we'll expect the normal hybridization signal pattern to be exactly that, fairly straightforward. And then again, I think we've gone over at this point in time what the nomenclature is, so number of nuclei, number of signals in this particular case. And again, I think maybe just to sort of go back to the point about the Julie mentioned about sort of differences between plus and time. So again, for metaphase fish-based studies, you can think of plus and minus as presence and absence. In this particular case, again, it's how many copies, and it's typically done through, again, multiplications of the X, the multiplication symbol in those instances. And then an abnormal result. So again, this particular case, we see three copies of this D8Z2. I'm Canadian, and it's Canada today, so D8Z2, I should say. MYC, and then 52 out of 200, they see this specific pattern. So I've sensed at the bottom of this first abnormal result, polysomy versus polyploidy. How can I tell the difference? Which one is it? Someone makes eye contact, I gotta pick on them. What do you got? In this case, polysomy, you got first three centimeter copies, and then the three, four and 10. So certainly, I think I would say that that's, certainly could be suggestive of. The reality is, again, ask a targeted question, you only get a targeted answer. We actually have no idea what's going on with the rest of the genome. So either is a possibility, and frankly, certainly you can make inferences based on what we know about the tumor type we're looking at, but ultimately, we really don't know. And so that's left up in the air in a situation like this, specifically. Yeah. Oh, definitely not amplified. We'll get to that point in a second. So this would be a gain. Oh, well, not even a gain. This would be just, again, exactly this. A polysomy or a polyploidy result. If the specific nature of the test was presence or absence of MYC amplification, you would say there was no amplification. We'll get to that distinction in a second. So in this case, you rate 200 cells, right? So 200, so in this example, it says that 200 individual interphase nuclei were scored. And of those, 52 showed that pattern. Okay, so another 150, another 148 shows normal? Correct, correct. Which could either be, you know, it's a subpopulation of the disease that doesn't have the specific binding. You get intratumoral heterogeneity, or more likely in the context of cancer, it's that you have normal admixture. But necessarily, we're not working with a pure subpopulation of cells, but the cancer represents a subpopulation of the total specimen we're looking at. And oftentimes, from hematologic malignancy, you really don't have the ability to, by fish alone, differentiate what's cancer and what's not. It's simply looking at a big, bright blue ball, that DAPI image at that point in time, okay? I did have a couple questions, I'm glad. So the question I'd like to hear again, so how few nuclei do you have to have before you say this is a reportable finding? So if I see three cells, we have 52, we said. If I saw three cells, do I report it? Do I not? Yeah, awesome, perfect, I don't know who said that, but that's exactly it, what's the cutoff? What have I established? And that ends up being the central question, is that to some extent, that ties your hands, to some extent, there's always borderline results that can sort of worry you into being more descriptive, but certainly, it's what have you established in the context of setting up that probe, what's your confidence that this result is something real or not, yes? Does it matter if your result is positive or negative? Because if you count 100 negative versus 100 positive, you may, 100 positive, you're fine. If you count 100 and you found a positive, in 80%, you may not need to count to 200. But if you counted 100 and it's negative, you may need to count to 200 to identify that's wrong. And so, the challenge, your logic is correct. Typically, what happens is that, again, in the context of the validation, you say, I'm going to count 200 in this situation, or 500, you pick a number, and you go through it. How do you stick with it? As much as you can, because the challenging part is that if you just say, the first 100 are negative, I'm gonna keep on trying, at some point in time, you're actually increasing your likelihood that you may find one that's an artifact of the actual hybridization, and again, that's knowing how the probe performs, that's knowing the kinetics, that's knowing what the experience is like, but certainly, it's not uncommon for probes in the context of validation to see one probe that has three signals, or sorry, one nuclei that has three signals, in a normal disease-free individual, or one with one signal. So, out of 200, how many need to be positive to be reported? Ah, so that's back to the, that's my question. That's a- Out of 200, so now, you always count 200 and then see- So, you establish a cutoff. I wasn't sure how much total you do count, regardless if it's positive or negative, versus- No, no, no, no. So, typically, you go in saying, I'm going to count this number, based on the specific condition that I'm working. So, if it's initial diagnostic workup, again, 200's typically a number for hematologic malignancies. If it's follow-up for remission studies, you go 500, and I think that, again, Julie brought up a really good point, which is, in this day of molecular technologies, again, one of the reasons that 500 was initially established as a number of probes that sort of made sense for minimal residual disease is because there wasn't other modalities to test to that level. Now, increasingly, there's, frankly, more sensitive molecular-based techniques. So, I think it's figuring out, what are the situations where I really have to go that far, versus other situations where, frankly, this isn't the best way to do it at that point in time, where there's other ways to get to MRD that's, frankly, more sensitive than a fish-based assay. Okay. And we asked, someone had a question about amplification. So, now, I think we're trying to get there at this point in time. So, now, two different abnormal probe signals, in this particular case. On the top example, the actual letters AMP, AMP is being used. And so again, that's finished nomenclature that particular case for it. What's being defined as an amplification event. And then we list again the number of nuclei with that signal. And the bottom, we see a relative increase in MIC. So it looks like there's three copies. You can see the difference in these particular instances. Again, if there's different number of signals, again, this number goes from outside the parenthesis, if this was all the same number. If they both had two copies, it would be times two on the outside. But now again, that it's specific probes of specific signals. Again, it goes and follows that individual probe. But so another question again. So one I'm calling an amplification, the other I'm calling relative gain. So what's the number there? What's the magic number? Is there a magic number? How do I define amplification? Now we're getting to the point of the lecture, people are looking away at this point in time. So amplification generally, the truth is there really, there isn't one single definition that people use. The different papers have described different things. I think that there are instances in which specific genes will have a defined guideline. We'll talk about HER2 later on in one of the lectures. HER2 is a very defined, has a definition of, this is what we call an amplification event in that specific disease with a specific probe. Otherwise, I think you see different definitions. And it typically ranges like something like three to one fold increase of a locus to a probe. So I'd have two copies of my centromeric probe and six copies, by example, of my locus-specific probe. And somewhere around that level or greater ends up being amplification. I can tell you there's an incredible amount of variability in terms of the reports that we see in terms of how that word's being used. I know that there is an effort by the ACMG, I think, to try to move towards standardizing this, both from the clinical side, I think, to try to move towards standardizing this, both from an array and sequencing perspective and a phish-based perspective. But right now, I can say that there's instances that we see reports where we see a gain that people call an amplification. I think that, certainly, I think cytogeneticists, specifically, that being my first love, I think that there's a focus, and I think an important focus, on sort of distinguishing those terms. There's a big difference from a biology perspective on a gain versus an amplification in terms of what we think it does in terms of driving the tumors. Okay, next type of probe. Break apart probes in these cases. So, in this case, again, now, rather than have a single probe that spans a locus, and then, again, a relative control that we're using, now, basically, we have two probes that span a single locus. And, ultimately, again, it's designed to detect the presence or absence of a specific rearrangement of this particular locus. And so, this would be an example of the IGH break apart probe here. Here, showing sort of the Watson, oh, sorry, well, the IGH break apart probe. Again, normal pattern, we expect to see those two signals to be fused together. They should be intact. We don't expect to be rearranged. These, again, they're not overlapping probes. It should be noted that there's always a little bit of a gap between those probes. But, typically, but the way they will, ultimately, appear on interphase and metaphase will be, again, a single, basically, or an overlapping, at least, fusion signal at that point in time. So, we'll call it two fuse signals. Normal, as expected. Then, the ICN has given you just of a, just enough of a rope to hang yourself with in terms of there's two different nomenclatures that you can use. One, the top, being fairly sort of, in which we basically just describe the probe, the break apart probe, as a single unit. So, it's IGH. We don't talk about the two different regions of this probe, the three and the five prime region. And, there's two copies of that and the number of nuclei. Versus, we can be more descriptive in this particular case, but this is exactly the same thing. And, in this instance, we're basically, we're talking about the probe. So, there's a three prime IGH region. There's a five prime IGH region. We have two of these signals. And then, of those, actually, two of the signals are connected. Three prime IGH is connected to five prime IGH twice. So, just being, frankly, again, more, this is long form at this point in time, basically, for the way to fish base. And, hopefully, that makes sense. But, if not, this will be the time to start waving your hands. I can keep on going. As listed here, generally, again, and I think Julie mentioned the same thing, again, chromosomes are, the probes are listed from P to Q, head to toe, at this point in time, the chromosomes. And so, the region that IGH is listed, three prime first, again, it's on the antisense strand. So, the three prime end, basically. It's a negative sense strand, or it's a negative gene, basically, if you'll call it. So, it's three prime is sort of closer to the centromere, five prime is farther. Fine, normal. I think we can all get around that. Abnormal, though. So, now we see a signal pattern like this. So, it's a separation. This is what we expect to see, classic, fine. And, now, you can see from a nomenclature perspective, again, the total number of IGH signals we see, and here we're talking about units. So, the total number of units of IGH breakpoint signals is two, still. That hasn't changed. But, now, it's whether they're intact or not. And so, here we have CEP, again, is the ICN nomenclature to describe, again, there's two probes that are separated, and how many of those two probes. So, we have two IGH signals, and of those, one of them's separated. Classic, rearrangement, straightforward, hopefully. And, I should say, below is sort of a list of, we'll get to some atypicals in a second, but below is a list of some of the other genes that sort of classically, from a cancer perspective, are used from a break-apart perspective. Why break apart? Why not do a color-dual fusion? What's the other one? We haven't got to that, so we're jumping. We're taking a leap of faith here. But, what about ALK? Do you guys know? Maybe not. And, if not, that's fantastic. Pardon? Pardon me? Ooh, we got four voices. Yeah, that's exactly it. It's how... Correct. Perfect, perfect, exactly. And, it's basically how specific is the rearrangement you're trying to detect versus is it simply an event where the rearrangement itself is all you need to know, and oftentimes, again, that's driven by the fact that those are typically promiscuous genes. There's a number of different partners, and ultimately, to our knowledge at this point, those partners are less relevant or irrelevant in most cases. Okay, so now let's get a little bit more crazy. So, two more abnormal patterns. And so, both of them are both shown and then described in this particular case. And so, hopefully, again, you can sort of follow the logic. Again, this should be just fairly descriptive. It's either con or sep at that point in time. So, how many probes do we have, whether they're connected or not. So, in an instance like this, either of these instances, well, let's start with this one. Is this rearranged or not? Do we have a rearranged locus? What are the possibilities? How do we get that pattern? It's mumbling. Mumbling sounds like it's bubbling to the top. Deletion. Yeah, one could be exactly that. It could be a deletion, absolutely. Could just be a straight up deletion. Any other ways you could get to this? Could be a deletion coupled with something else. Yeah, the other option is that you first have a translocation and then a deletion. And so, this is where, again, the only reason I'll leave up this slide is that what ends up being critically important from a break apart fish strategy is know the business end of a gene. Know what gene, or know what part of the gene, and IGH is a tricky one, not a great example for this, but ALK, ROS, RET, all kinases. It's the three prime end of a kinase that ends up being important. Ultimately, if you end up losing the kinase domain, you're not gonna have a fusion that's important. So, I took exactly this example with ALK. And I said, I have lost the three prime region of ALK. There's no rearrangement. Can't be a rearrangement, because the three prime region is the region that basically drives the fusion. So, again, any time you think about a break apart probe, try to think about, again, within the fusion, what would be, and you can think of this from a molecular perspective, too, what's the five prime, what's the amino, versus what's the carboxy terminal of that fusion? And so, what would I expect to see in those cases? What has to be there from that gene of interest for that fusion to be functional? It would be the translocation with the speed over here, because it's not lost. Ah, so, this would have to be a translocation followed by a deletion. And so, there's actually situations in which that can arise. And so, it's not uncommon in some disease types in which you have a balanced rearrangement as a first event, which causes a gene fusion product, and then loss of one of the derivative copies. And the derivative chromosome, again, that you lose one half of that balanced rearrangement or that initially balanced rearrangement, typically, again, is the non-functional element of that. So, again, ALK rearranged. The three prime ALK is what you really need for that fusion to be formed and present. The five prime region doesn't matter. That can go away. And so, again, part of this, I think, is, and we'll get to this in some later lectures today, just understanding mechanism and historical understanding of some of these rearrangements that happen. So, most of these changes will lead to gene fusion in the coding sequence? So, the question is, are they interrelated? So, the tricky part about that, so the question, again, is does this involve the coding sequence? Depends on how the probe's constructed and depends on the nature of the gene. So, you know, ALK, a gene like ALK, we're talking about in-frame gene fusions is what we're looking at. We know where the break point is. The actual probes span exactly, or more or less, where the break point is fairly well. IGH is a slightly different game in this case, in that, again, it's an enhancer element that ends up being switching positions in those instances. But I think that point we'll come back to later on, so hold on to that thought. Okay, yes. Oh, I hope it does. Do-do-do-do-do-do. No, it doesn't. It doesn't, that's why. Yep, so that should be, what should it be? Yep, exactly. Yep, yeah, it should be basically three, this should be three green, two red, and then dot times one. So this was wrong on like three different counts. Good eye, good point. All right, last probe strategy, and then we'll get onto some other examples. So dual color, dual fusion. So in this particular case, again, to compare and contrast this to what we just talked about with break apart probes, again, two different loci that we're evaluating, and now we're looking to see when they come together. And so again, from a normal perspective, this is fairly intuitive. From an abnormal perspective, you end up seeing a pattern that looks like this. So suddenly a dual color, dual fusion will create basically three signals for each individual locus. One from the normal chromosome, and then basically a break, basically the two halves of a translocation that happens. And so again, this is both genes coming together in those instances, so that's the classic pattern as shown there. But then of course, there's also the possibility of getting more complex. And again, I think that some of these end up being patterns that we actually see fairly commonly. So in the context of BCR-ABL, again, the Philadelphia chromosome, typically CML, but also ALL occasionally again. Having an additional copy of this fusion product is actually most often representative of an extra copy of the Philadelphia chromosome, the DIR22, again, that's the business end of the 922 translocation that happens. So again, this is not an uncommon pattern, it's actually a common secondary abnormality that happens with this disease type. Or we can get some really funky patterns, and this would be just one example of a pattern that could lead, or one example of a rearrangement that could cause something like this, in which you have a three-way translocation. And so a piece of chromosome nine, ABL, will go to chromosome 22, making the BCR-ABL. This piece of 22 goes to 17 by this example, and 17 goes back and sort of closes the loop in those instances. And so that again changes the normal pattern. Again, really the take-home message from this perspective is just, again, think of what the normal pattern is, two fused, one red, one green in those instances for the normal pattern, and then any variation of that, think what that could mean, or just consider that again, that could mean changes in terms of the underlying structure or complexity of the genome. Can you explain one more time how you momentally to normal on the other side? On the last one? Yes. Can you explain that? This one? Yeah, so this actually, I think there's one other thing. So this, basically this should be three signals, three total signals for ABL one. So ABL again on chromosome nine, and so those three signals are basically one here from the normal. Oh, chromosome 29, that's mislabeled. And then two, these are the other two signals, that's the total number of signals. And again, it's the same thing for the green. So every yellow counts for basically one red and one green. Does that make sense? And so that's how we get the total number of signals. And then basically between these two, it's of the three signals we see for BCR and the three signals we see for ABL, two of them, ABL, CON, BCR, are connected. And so that's those two signals there. Does that make sense? Okay. And it's a lot to chew on, and for sure take a look at it, and for sure spot me if there's other ones that I've goofed. But I'll take a closer look at this after. If there's any updates, I'll send them. Oh, when, well, I'll send around one update at least at this point. Okay, so last, we'll try to get this back on track. So the last like 12 minutes or so, 10 minutes or so, I wanna talk about sort of bringing on fish probes, validating probes, and then just understanding sort of what the implications are or the things that you have to consider bringing on a probe. So you wanna bring on a probe. What do you have to do? Fine, step one, fairly easy, obtain the probe. Oftentimes it's order at this point in time. You can make them yourselves, you can label them yourselves. So I think that many of us at this point in time just simply use commercially ordered probes. If you have a specific region, this is actually one that a colleague of mine had designed, they wanted to build their own CDCAN2A probe, so tumor suppressed gene, classically deleted a number of different cancers in this instance. So they ordered CDCAN2A by itself, and they also ordered a centromeric probe for chromosome nine as a control probe in those instances. So now they have to mix them together. Now, if you mix them in a one-to-one ratio, we're going deep here, but if you mix them in a one-to-one ratio, one of these two signals will probably be a lot brighter than the other. Which would you guess? Yeah, why is that? Exactly, so it's more targets. There's just more options, it'll be brighter, so it's just something you have to figure out in the context of initially setting up your dilution for these probes. Fine, established. The next step, before you really get any excited at all in terms of looking at cutoffs or anything else, is that confirm the localization. And this is something I can tell you that we've definitely been burned on in the past when steps weren't followed to the T, is that a probe comes in, you go, oh, I've done this probe before, thank you very much. You start using it, and if you actually haven't mapped back to metaphases and visually confirmed the location, confirmed that this is chromosome nine, and the pattern is happening where it should be, ends up being a critical step that can end up derailing any work you're doing at that point in time. There's been instances in which probes have been mislabeled by the manufacturer, so the colors have been swapped. I can tell you that's a fun one to unravel. Fortunately, it was a research thing, but painful. A few people in my lab learned a few new words that day. So it's step one, again, make sure you know where these probes you're ultimately mapping to. Do they map to the correct chromosome and set as a chromosome band? ACMG has a lot of guidelines that they recommend in terms of, or recommendations, I should say, about sort of how to go through with metaphase-based studies. So again, in terms of confirming the localization, they always say use male cells. And again, that makes sense. Again, you wanna make sure that the Y chromosome, too, is just a map to rule in or rule out in terms of any cross-hybridization. You wanna make sure the signal appears only on those chromosomes and appears sort of at the right intensity and chromosome band you'd expect. It's recommended to do that over about five metaphases. And then it's also recommended to confirm the specificity and the sensitivity of the probe. So again, does the signal only hybridize to those particular chromosomes? As it should. Is there any background? And again, that's typically recommended to do 40 chromosomes in terms of the initial evaluation. Again, at this point in time, we're just talking about metaphases. Now, from a cancer perspective, we really wanna get into ultimately using this probe for the purpose of saying this patient either has or does not have, let's say, a deletion, like a deletion of 20q. The first, one of the first steps is establishing a normal cutoff. And so there's a few different ways you can do this. Typically what happens is that you'll take a collection of individuals, something like 200 individuals, 200 nuclei. This is what ACMG recommends in terms of the number of nuclei per individual. And ultimately just say, what's the hybridization pattern? We'd expect in a normal disease-free individual, you're always going to see two signals. But why don't you sometimes? Why sometimes does that not work? Pardon me? Yeah, it could be something as simple as, it's either hybridization just didn't work as well as you'd hoped. It's also, it's a 3D object in 2D space. And sometimes, again, it just so happens that again, the way you're looking at the chromosomes, two signals end up being on top of each other. You just can't visually appreciate them. There's a number of different situations that can lead to sort of slight variations in those patterns. And so it's quantifying that as much as possible. And then using the numbers you've established to basically, there's a few different methods, and I definitely don't want to get into any of these here, but a few different methods have been proposed to basically take that data and mathematically model it to say, this is what the variation in terms of the probe signal is, and this is what I expect it to be. This is what I expect a normal to be greater than at that point in time for it to be a true signal. Okay. And the last couple, two minutes here, we'll end on a few different points in terms of just things thinking. So we've gone through a probe. We've calculated this cutoff. We've determined sort of what normal range would be at this point in time. When I say carefully evaluate the probe characteristic, now a lot of this comes down to specifics of individual probes. So if we talk about, by example, like a break apart probe, if I see small gaps, again, how much of a gap do I see between those break apart probes ends up being critical in certain situations? And that's because, again, depending on what I'm looking for an example of this would be in renal cell carcinoma, there's a subgroup of this tumor type that has TFE3 rearrangements. So the gene TFE3 on the X chromosome gets rearranged. Fine. Break apart probe, fine. Should be identified, no problem. However, it was recently identified that also inversions happen. And so why would inversions potentially, paracentric inversions happen specifically, but why would inversions potentially be a problem? Should it be a problem or not? For a break apart probe. So it's not actually that it's broken. To some extent, that's what you want. It has to do with how far it moves. And so with paracentric inversions, inversions in general, if it's an, well, it's an intrachromosomal event necessarily, that means that you might only have a slight separation of the two probes. And so depending on how much normal variation you have in that break apart probe, it could mean that you'll never detect it. Your probe may not have been set up for that in general. And so it's really doing a deep dive any time you're launching a probe. Do I understand what the characteristics are? What am I trying to find? Is there something else that I may want to detect that I haven't? And just think about it in the back of your mind before you go full deep into a validation effort. All right, last two points I'll end with. It's just thinking points and both of these are points we'll come back to in the last lecture at least. Fixed material. So we talked a lot about FISH. We didn't really get into what we were looking at. Most of what, certainly for hematological latency, we're often looking at whole nuclei in those instances. It's fresh material, drop them, drop them, spot them, count them. That's FISH at its best. Having said that, again, with fixed material, and again, this is often under the currency what we use for solid tumors when we evaluate the presence or absence of rearrangement or enumerations. Again, we're typically using five micron sections. And that means, by necessity, we're actually cutting through the individual nuclei. And that can lead to these truncation artifacts. Now, that may not be a problem at all. But if you're looking for deletions, necessarily, that probably, not always, but probably means that it's gonna cause, you know, your cutoffs to be higher. And so in those situations, it's possible that with fixed material, by example, that your ability to detect deletions, certainly if there's a significant admixture between normal and tumor, can go down quite a bit. And so your possibility of a false negative increases in those instances. So just being aware of what you're looking at. Not, I should say, this is likely not gonna be an issue for amplifications, none, nor for rearrangements. But really, deletions end up being a critical point as it pertains to fixed material. Okay, last point I'm gonna skip because I think we're gonna bring it up on the solid tumors. So I think I'll end there. And if there's any questions, I'm happy to take them. Thanks for your time. Thank you.

Adrian Jubeck

fish

fluorescence in situ hybridization

hematologic malignancies

DNA sequences

probe types

interpretation of results

fixed material