Cell cycle mapping in tumors with advanced multiplexing and spatial profiling

Video transcript

• 0:00 Welcome and good day everyone.
• 0:02 Thank you for joining Part 2 of our Human Relevant Models and Spatial Profiling series.
• 0:08 For those of you who didn't catch our last program, we dove into advancing Human Relevant model research and that program is available on demand.
• 0:17 A link has been put into the chat, so feel free to click on that.
• 0:20 After this program.
• 0:22 Today we'll focus on multiplexing and spatial profiling and how they can help you gain critical insights.
• 0:29 We're very excited about this topic and I have an amazing guest speaker, Doctor Wayne Stallard from the University of Pittsburgh.
• 0:36 Wayne's going to share his research into cell cycle mapping and tumors.
• 0:41 So before we jump in, just a bit of quick housekeeping.
• 0:45 This session will run for about an hour and we'll kick off with a brief introduction to Danaher and the significance of multiplexing and spatial profiling.
• 0:54 This will be followed by a presentation from our guest speaker and then we'll wrap up with the live Q&A throughout the webinar.
• 1:02 We encourage you to drop your questions into the comments.
• 1:06 We'll have product, product, subject matter experts joining us and our guest speaker during the Q&A session.
• 1:14 So my name is Thomas and I'll be your moderator for today's event.
• 1:18 I'm here on behalf of the life sciences companies of Danaher and I'd like to take a few minutes to provide a quick introduction to Danaher in today's topic.
• 1:27 Danaher is a leading global life sciences and diagnostics innovator.
• 1:31 As evidenced by our portfolio of world class companies, we're committed to using the power of science and technology to improve human health.
• 1:40 Our mission is simple.
• 1:42 We help scientists like you bridge the gap between research and real world applications.
• 1:50 For anybody who's new to spatial profiling and Multiplex imaging, I'd like to quickly highlight how researchers worldwide are using these techniques.
• 1:59 Spatial profiling helps elucidate how cells or biomolecules interact with their environment, and Multiplex imaging is key to enhancing spatial profiling because it allows multiple related components and processes to be observed in parallel.
• 2:17 These approaches allow you to uncover cellular heterogeneity, function, interaction, architecture, and even sub cellular structures.
• 2:29 By understanding location, you can better understand biological mechanisms.
• 2:38 So while these techniques can transform the insight you can gain from your samples, it's important to plan your approach properly.
• 2:46 Asking yourself some important questions upfront can help you determine the best approach for your sample types and experimental questions.
• 2:56 Your sample and the biomarkers you're targeting, along with the imaging platform and antibodies selected influence the potential outcomes.
• 3:06 Avoiding risk like poor imaging resolution, non reproducible results and wasted resources should always be the goal.
• 3:15 Because of property, planned spatial profiling experiment has the ability to deliver high quality data including the identification of new biomarkers.
• 3:29 Since the selection of fully characterized and validated antibodies is so important to multiplexing, I want to highlight how ADCAM is helping researchers in this process.
• 3:42 So Abcam has a rigorous quality management system outlined in the cartoon, which is applied to its ever expanding portfolio of knockout validated antibodies.
• 3:54 Biophysical characterization is an important part of this quality management system.
• 4:00 Abcam leverages several approaches to monitor product and critical quality attributes like sequence identity, integrity, aggregation, and purity.
• 4:12 Their antibodies are even tested for their specific applications and within control models like knockout cell lines to ensure they meet your specificity and performance needs.
• 4:24 And they are constantly expanding the variety of validation to include what is most relevant for you to be able to select the right antibody quicker.
• 4:38 Selecting the right imaging platform is equally as important as choosing the right antibodies like a Microsystems cell dive, Multiplex imaging solution, and app cams.
• 4:49 Validated antibodies make it easier to conduct your spatial profiling experiments.
• 4:55 The Cell Dive can even be integrated with an automated platform to help with your sample preparation needs.
• 5:01 And when it comes to analysis, Avia is their AI powered software that makes it easier to draw conclusions and biological contexts from your samples.
• 5:18 So you've planned your experiment, you've selected the right tools.
• 5:22 Now what's next?
• 5:23 Right?
• 5:23 Well, now it's just a matter of acquiring the data that will help you answer your research questions.
• 5:30 We've highlighted 2 research areas where multiplexing and spatial profiling are highly utilized.
• 5:39 First in study utilize multiplexing and spatial profiling on the cell dye to uncover the cellular relationships influence the ideology of Alzheimer's disease.
• 5:51 The second case study highlights aspects of the CAC tumor micro environment that could be targeted for treatment.
• 5:59 However, these approaches are also being applied in developmental biology, infectious disease research, cardiovascular research, and many other fields where cell models and tissue samples are being used.
• 6:17 So now I'd like to invite our guest speaker to the stage so we can learn how he's leveraging Abcam’s antibodies and like those imaging platforms to further his labs research goals.
• 6:29 Doctor Wayne Stoller is an assistant professor in the Department of Computational and Systems Biology at the University of Pittsburgh and a member of the Cancer Biology Research Program at UPMC Hillman Cancer Center.
• 6:44 He holds a PhD in Biochemistry from the University Of Montreal and performed his postdoctoral studies at the Max Planck Institute for Molecular Physiology in Dortmund, Germany, and at UNC Chapel Hill.
• 6:58 Doctor Stalich's research focuses on the intricate relationship between cells and their environment, particularly the plasticity of the cell cycle and its adaptation to changes in the environment or genome.
• 7:11 His lab employs quantitative single cell microscopy, machine learning, and advanced computational approaches to study how the cell cycle changes during tumor genesis, metastasis, and drug treatment.
• 7:26 A key interest is understanding how the tumor micro environment controls cancer cell proliferation.
• 7:34 Ultimately, his goal is to predict disease outcomes and therapeutic strategies by directly looking at the cell cycle program driving the growth of a given tumor.
• 7:47 So Wayne stages all yours.
• 7:52 Thank you very much, Thomas.
• 7:53 And I would just like to in general, thank Danaher for giving me the chance to share my lab's research with you today.
• 8:01 As Thomas said, we use single cell imaging to explore this idea of cell psychoplasticity in cancer and in other physiological and disease context.
• 8:12 I'll be sharing some of this with you this morning.
• 8:16 So, you know, our main reach of interest is really this idea of cell psychoplasticity.
• 8:21 And it stems from this, it's, it's really this, this new area of research in cell cycle biology that builds upon, you know, decades of a very classical and, and, and, and, and rigorous research into the molecular mechanisms that drive the cell cycle.
• 8:38 We gave out Nobel Prizes back in 2001 for figuring out the series of molecular events that govern passage to the cell cycle, really culminate in this, in a single generalized model of the cell cycle that we use to understand cell proliferation across a variety of contexts.
• 8:57 A lot of this work really went into dissecting the molecular mechanisms that that regulated this progression.
• 9:05 You know, these mechanisms are really thought of as a sort of this fixed series of molecular events and switches that allowed a cell to assess whether it was ready to engage in these very big events in life history, to replicate their DNA or divide it into two daughter cells.
• 9:23 And so the series of this fixed series of events sort of functioned like a checklist that a surgeon or a pilot might use before they begin their activities.
• 9:34 But really in the in this era of single cell biology, with the advent of new approaches that allowed us to watch individual cells as they progress from the cell cycle, we've come to realize that cells don't always progress along a single fixed series of molecular events.
• 9:51 In fact, a lot of work has been done really looking at the molecular events that occur immediately following cell division as cells reenter, daughter cells reentry, a new cell cycle.
• 10:02 But we've really come to realize that at this point, cells can use different sets of effectors to perform key regulatory events in different orders, really letting us know that there are alternative paths that cells can take through the cell cycle.
• 10:16 We've also done some work not just in the proliferative cell cycle, but in the arrested cell cycle a couple of years ago, really identifying that that the arrested cell cycle is not a single monolithic state, that cells can exit at different parts of the proliferative cell cycle into different states of cell cycle arrest.
• 10:37 All of this is really demonstrating to us that the cell cycle is not simply a fixed series of molecular events.
• 10:45 And now this has really opened up a lot of doors in in the cell Cycle World and new questions to explore really centered around this idea of, you know, what is the extent and then the role of this plasticity in cell cycle progression across human health and disease.
• 11:05 Now these are cool questions, of course, but they're very challenging questions to address.
• 11:10 You know, the cell cycle, as I've alluded to, it's, it's, it's really regulated by this, this complex interconnected network of, of, of effectors that, that regulate the progression of cells through the cell cycle in space and in time.
• 11:23 And so in a given physiological context, for example for looking for plasticity in the cell cycle, how do we choose where in this huge signaling network to zoom in and look for these differences?
• 11:40 Well, we decided that maybe the easiest way to do this would not be zooming in on a specific locus of signaling, but actually to zoom out and try to observe the cell cycle as a whole.
• 11:51 This idea being that what about if we could get this really overhead view, this bird's eye view of the cell cycle and all the paths that cells can take through the cell cycle, including any detours they might take if they encountered problems, as well as any sojourns into different arrest states that they can take through the cell cycle.
• 12:12 And in thinking about this, we really developed over a little while this methodology that we call cell cycle mapping that aims to do just that.
• 12:22 Now in the cell cycle mapping approach, we start with either 2D cell cultures that are that are fixed in place or we can use FFP tissue sections or tumor tissue microarrays.
• 12:38 And here we perform multiplexed amino fluorescence to measure, you know, 50 to 100, sometimes more than 100 individual biomarkers in those same, in those samples in the in the same exact cells.
• 12:53 Now of course to do this one must assemble a large and validated antibody library to get good robust readouts of these things that we're interested in.
• 13:05 And we were very closely with AB Chem to identify antibodies that sensitively read out the biomarkers that were interested, both interested in both in the cell cycle, but also in the you know tissue and tumor microenvironments for projects where those things may be important for the questions that we're we're trying to answer.
• 13:28 And so through AB Cam, there are a number of pre conjugated antibodies that are pre conjugated to Alexa Fluor dyes that are compatible with our imaging instrument.
• 13:38 And for the antibodies that don't come pre-conjugated, AB Cam also often provides carrier-free antibodies that are then amenable to labeling in-house, which we also do when required.
• 13:51 And as you can see here, here's a list of 56 antibodies that we are currently part of our antibody library that come from Abcam.
• 13:58 In particular, there are a number of the core cell cycle effectors here in this library.
• 14:05 As you can see all of basically the the, the canonical cyclins are here as well as the canonical cyclin dependic kinase, the CDKS that we typically measure in our approach.
• 14:19 As I said, there are also number of of markers of cell types or states, extracellular matrix protein signaling proteins in the tissue and tumor microenvironment that we also have here in our antibody library.
• 14:34 Now to measure these antibodies in our sample, we need to use a specialized imaging system.
• 14:42 And for this we've chosen to use like a cell dive imager.
• 14:46 This is a Microsoft that's dedicated to automated hyperplex or Multiplex amino fluorescence that that really automates the both image acquisition and then all the post processing and alignment steps, which are really if some of you have have tried to do multiplexing immunorescence are really the some of the most difficult steps there that are the really the bottlenecks in the process.
• 15:11 That was one of the reasons that was really that this technology is really attractive to us when choosing our multiplexing technology that we we would use.
• 15:20 The other reason why we chose to go with the like a cell dive for our experiments is that the availability of a, a liquid handling robot that, that it can be set up directly in series with the cell dive imager that then now can automate the sample preparation, automating the staining and the bleaching of, of the of the labeled antibodies.
• 15:49 So now we can actually automate the entirety of our process.
• 15:54 And in doing so, this also increased the scale at which we could acquire data.
• 15:59 So when we're running at really Max capacity and we're using 15 tumor microarrays of let's say 60 cores each, our throughput here is around 900 tumors with 6060 biomarkers in around 15 days, which is an incredible amount of data.
• 16:18 And the ability to do this actually open the door to entirely new questions that we wouldn't have been able to ask otherwise.
• 16:27 So from here, we acquire all of these images, a stack of perfectly aligned images where we've measured all of our antibodies of interest.
• 16:33 We then use a deep learning based cell segmentation method to identify individual cells across the entire tumor microarray and quantify every single biomarker in every cell across all of these samples.
• 16:48 So we arrived at a, a, a single cell data set like maybe some of you are are familiar with more in single cell transcriptomics where we have, you know, thousands, 10s or hundreds of thousands of cells each described with this high dimensional feature data set.
• 17:05 Now, unlike transcriptomics where we obtain one measure per gene of its, you know, gene expression, another advantage of using a an imaging base, you know, spatial approach to acquiring single cell measurements is that we can exploit the spatial nature of it to to measure more than a single outcome for each antibody.
• 17:31 So we can measure the sub cellular localization of a given protein, its intensity in the nucleus or the cytosol or at the plasma membrane.
• 17:39 We can also obtain measures of some sort of distribution.
• 17:43 Is this protein kind of equally distributed in nucleus or does it form puncta?
• 17:48 As these could be very important differences in the biology of that protein.
• 17:54 We can of course look at the the ratio of cytosol to to nucleus of a given protein is is cytoplasmic to nuclear shuttling can be also important for biology.
• 18:04 We can take measures of morphology, nuclear cell size or shape, etcetera.
• 18:10 And then of course we can capitalize on the maybe more straightforward advantage of doing of the spatial nature of these approaches.
• 18:19 We can look at the the immediate surroundings of the cell and ask who are their, who are its neighbors and what are they doing?
• 18:26 So from here we can, even if we measure 100 antibodies, we can still obtain a feature set, a feature signature, our case a cell cycled signature of sometimes thousands of features derived from a a, you know, 100 biomarker panel.
• 18:41 From here, we use a dimensionality reduction approach or manifold learning approach that then projects these cell cycle signatures down into A2 dimensional cell cycle map.
• 18:51 Basically, we're trying to find the trajectory in this high dimensional space that represents progression through the cell cycle and then, you know, capture those differences and project them down into A2 dimensional representation that we can actually interact with.
• 19:05 We also have some fun ways of assigning cell cycle face to each of these cells.
• 19:10 We then refine these maps through feature selection.
• 19:13 So something I won't be going into detail today, but I'm happy to discuss with anybody that's interested.
• 19:17 There are a number of ways to try to do this to arrive at our final cell cycle map in which we can very clearly see this is a a map that we derived from a a cell line of pancreatic ductal adino carcinoma where we can actually see two distinct routes through the proliferative cell cycle.
• 19:35 We see this main route here.
• 19:37 I hope you can see my cursor, but if you if not, it's this main route along the bottom of the structure where most cells are going through G1SG2 and then looping out through M phase.
• 19:47 But we also captured this other alternative cell cycle program as I called.
• 19:52 I've labeled it P2 on this left most diagram here, where a minority of cells are taking this this completely distinct route through G1S and G2 before converging with the other P1 proliferative cell cycle at the G2M transition.
• 20:06 And we also detect a, a route through, into and through cell cycle arrest.
• 20:12 Now once we identify multiple routes, multiple cell cycle programs, of course, then we want to understand how do these programs differ in a mechanistic and a molecular sense and how, why is that important to the biology of interest here?
• 20:25 So there are a number of ways we can play with this data computationally to try to identify these mechanistic differences and and explore them a bit further.
• 20:35 So in our first proof of principle project to to sort of demonstrate the capacity of this approach to map the cell cycle, we generated a completely data-driven map of the human cell cycle, including both proliferative and arrest trajectories.
• 20:49 This was in a in in retinal pigment epithelial cells.
• 20:52 This is the sort of darling of the of cell cycle biologists as it represents a a relatively normal H turt transformed cell line.
• 21:00 So it has a, you know, as normal of a cell cycle as, as we can hope for in a cell culture model.
• 21:06 But as you see, we get that we, we detect both a, a cyclical proliferative trajectory and then this arrest trajectory that branches shortly after cell division.
• 21:15 And then we followed oh, and, and once we found these trajectories, we can then, you know, perform trajectory inference and, and quantify how these cell cycle effectors or other things are changing along these trajectories.
• 21:29 And we're able to, you know, report of decades worth of, of cell cycle biology in this single experiment.
• 21:38 So, for example, we see the DNA content exactly double S over the course of S phase.
• 21:43 We capture this classic U-shaped dynamic of cyclone D1 expression.
• 21:49 We see that cyclin A expression only takes off the cells transition from G1 into S We're able to capture, you know, rich cell cycle biology with this approach.
• 22:00 We then followed this up with a a, a a paper that I think really demonstrates the the promise of this approach is a, is a comparative technique to show how the cell cycle changes in different contexts.
• 22:11 So in this case, we took retinal pigment epithelial cells again, but we subjected them to different cell cycle stresses to see how the cell cycle would respond.
• 22:19 And to make a very long story very short, we found that as I alluded to earlier that cells can exit from the proliferative cell cycle at at different phases into into molecularly distinct states of arrest, but in response to different stresses.
• 22:37 But I think the most important thing that we observed in this in this project is that cell cycle arrest isn't this stable state that cells continue to change in a molecular functional and a phenotypic way after exiting the proliferative cell cycle.
• 22:55 So for example, we see that we detected this trajectory here in response to sustained replication stress where cells would exit from G2 and they perform something called mitotic skipping where they after replicating their genome, they switch from AG2 like molecular state to AG one like molecular state and sit in this this region here of of cell cyclores that looks like senescent cells.
• 23:21 It's beta gal positive by all measures in a molecular way looks like a senescent state.
• 23:27 However, we detected that at a a relatively low frequency, but you know, it's happening, cells would exit from the senescent state, re enter the cell cycle.
• 23:37 But since they've done this mitotic skipping to this more G1 like state, they would then re enter the cell cycle at G1S after having already replicated their genome.
• 23:49 So they do S phase once again and undergo endo reduplication to polyploidyce to do whole genome doubling.
• 23:57 And so this is really the path that we detected through which sustained replication stress could lead to whole genome duplication.
• 24:03 And this is something that we continue to explore in various models of cancer in my lab today.
• 24:10 OK.
• 24:10 So with the rest of my time, I'd like to give you little snapshots of of four projects that we are currently working on that I think really demonstrate how we can use this approach to uncover novel biology both in cancer.
• 24:26 But also I will wrap up and and look at how we're using it in the context of of immunology.
• 24:31 And of course, my lab has been trying to bring this both together and look at immuno oncology here and how these interactions influence cancer cell proliferation and tumor growth.
• 24:43 So the first project I'll talk about today, we're really going to be exploring this big idea that we have that we're really, you know, kind of just starting out with now with this idea that, you know, asking whether we can classify tumors, in this case breast tumors, by the cell cycles that are driving their growth.
• 24:59 And if we in doing so, can we then predict clinical outcomes by looking at these cell cycle programs?
• 25:06 And so in a, in a proof of principle experiment that we performed in in collaboration with Adrian Lee and and Steffi Usterak here at the University of Pittsburgh, we took a tumor microarray of 35 estrogen receptor positive breast tumors.
• 25:22 I'm using cell dive and our antibody library perform Multiplex immunofluorescence of around 120 biomarkers.
• 25:30 These are about half and half phenotypic and and biomarkers and markers of the of the tumor microenvironment which I will get to in a moment.
• 25:39 If we zoom in on on one such tumor, here you can see some of the imaging that we performed where we've, you know, in this first panel here we're looking at cyclins or P27, an important CDK inhibitor.
• 25:55 In the second panel we're measuring gamma HJAXA marker of DNA damage, cyclin B1 and CDK four.
• 26:01 And the third you can see we're starting to look at cell types.
• 26:04 Of course, we need to distinguish cancer cells from other cells in these tumors.
• 26:08 We're looking at a pancettokeratin momentum to identify stromal cells and then KICC 7 like the classic clinical biomarker of active proliferation.
• 26:19 And finally in the fourth panel here we also measure just important, you know oncogene or tumor suppressors, other molecular changes that might be contributing or converging down on cell cycle to regulate cell cycle progression.
• 26:32 So here we have measures at the single cell level P53 of CMIC beta catenin.
• 26:38 Even just looking at these data right now, it's very clear to see a lot of, of single cell heterogeneity and expression of cell cycle effectors and, and you know, and other markers that could be influencing their, their progression through the cell cycle.
• 26:54 So again, to make a, a very long story very short for our our purposes today, we performed cell cycle map and we extracted single cell cell cycle signatures from every cell across this entire tumor microarray and generated A unified cell cycle map where every dot here is a cell is one of the cells from the microarray.
• 27:14 And as you can see, we detected 4 distinct proliferative programs across these breast tumors and a single cell cycle arrest state that is characterized by a high P 27.
• 27:29 Now how do these programs differ from one another?
• 27:31 If you look up at the top here, we see that the cell cycle programs really differed in the combinations of cyclin CDKS that were driving their, their, their progression through the cell cycles.
• 27:41 For example, in this top cell cycle program #2, we see that this cell cycle program has very low expression of cyclin D1, but let's say very high expression of of of ACDK 2.
• 27:55 If we look at cell cycle program for the bottom, we see that on the other hand, this program program has very high expression of cycle in D1, very low expression on CDK 2, and all of them can be.
• 28:05 All these 4 cell cycle programs can be defined by distinct combinations.
• 28:09 Here it.
• 28:11 So not only can we detect these distinct cell cycle programs across the tumors, now that we can assign individual cells to a given program, we can now describe tumors by the repertoire of cell cycle programs that we detect within them.
• 28:26 And so if we now we have these bars that represent individual tumors across the TMA and we're coloring these bars by the proportion of cells that we find in that in each cell cycle program.
• 28:37 So we see there's incredible inter tumor or heterogeneity in the programs that these breast tumors are using to proliferate.
• 28:45 But also we we detect that some tumors are using multiple cell cycle programs to proliferate.
• 28:52 So within the tumor we're detecting multiple cell cycle programs.
• 28:55 This of course may result from distinct genetic genomic clones within that tumor subclones.
• 29:01 But another possibility is that these cells are growing in distinct micro environments and that the local environment is influencing or selecting or inducing specific cell cycle programs.
• 29:13 And so of course, if this is the case, we're very interested in how the tumor microenvironment might influence the cancer cell cycle.
• 29:20 So taking this same data set, as I said, we also measured around 60 biomarkers of the the tumor microenvironment including markers of cell state, cell types of extra cellular matrix components, other signaling molecule cytokines, etcetera in the tumor micro environment.
• 29:38 And in doing so, now we can derive A tumor micro environment signature for every cancer cell.
• 29:44 So within some radius, you know what cells are there, what's you know ECM components are there etcetera.
• 29:50 We can derive a ATME signature for every cell now only that we can then we can then cluster cancer cells by these TMB signatures by by you know sort of differences in their local environments.
• 30:02 And now we can ask is there enrichment of cell cycle programs within tumor microenvironments?
• 30:10 And we see this is just preliminary data, but we see very nice promising signs of this may indeed be the case.
• 30:16 So for example, in TME signature 3, this is a TME signature that we found that is immune infiltrated.
• 30:24 There's an enrichment of CD 8 positive T cells.
• 30:27 We find that there's this enrichment of cell cycle program #2 Now cell cycle program 2 is defined by a very high expression of CDK four.
• 30:34 And indeed, if we look at a tumor that is using sound psychoprogram to to to proliferate, we indeed see that this tumor has a high abundance of CDA positive T cells.
• 30:48 And around those CDA positive T cells, we see that the cancer cells are expressing higher levels of CDK 4.
• 30:56 We also find that this other TME signature that is immune excluded, this is a ATME that consists of entirely of estrogen receptor positive, progesterone receptor negative tumor cells.
• 31:08 So we don't see any immune cells here.
• 31:11 Indeed, in this tumor microenvironment, we see enrichment of cell cycle program one that has low expression of CDK four.
• 31:18 And again, if we zoom into a tumor that is using the cell cycle program, we don't see any CD 8 positive T cells and we see very low expression of CDK 4.
• 31:30 Next, I'd like to move on to another application of this approach in cancer and that's to, to look at how the cell cycle adapts to targeted cancer therapies, particularly those that are that are directed at the cell cycle, at stopping the cancer cell cycle.
• 31:47 So for instance, in breast cancer, the use of CDK 46 inhibitors like Pablo siclib have really revolutionized treatment of the disease.
• 31:58 And of course, with targeted therapy there comes this other side, which is the development of acquired resistance, which you know is, is is seems to be inevitable even with the most effective therapies.
• 32:11 Of course, there's been a lot of work now trying to understand how tumors are becoming resistant to the CDK 46 inhibitors.
• 32:17 And there have been a number of of genetic or genomic changes that have been associated with resistance, including amplification of key cell cycle genes like cyclin E, cyclin D, CDK 4 or loss of the retinoblastoma protein.
• 32:33 Now as I hadn't said, these are genomic changes.
• 32:35 These are really changes that are going to take place over, you know, weeks to months of of treatment with the inhibitor.
• 32:42 Now the cell, the type of cell cycle adaptation that we're interested in here is actually a more rapid adaptation, the kind of cell cycle adaptation that is occurring in hours to days with the idea that cells can find a way to sort of bypass a, a, a pharmacological blockade of the canonical cell cycle.
• 33:04 Like if we were to block CDK 46 activity, we're looking at these drug resistant sort of cell cycle detours.
• 33:11 If you want to use our, our, our mapping metaphor.
• 33:15 Now, why we think that this is important is that, you know, these genomic changes that lead to more stable solutions to drug resistance.
• 33:25 These really require that cells resume DNA replication and chromosomal division, right?
• 33:33 A cell that's actively proliferating and and playing with its DNA and dividing two daughter cells is much more likely to make these kinds of errors that will lead to the genomic changes that could confer this stable drug resistance.
• 33:48 Of course you know these things can happen in the absence of proliferation, but the cells high much more probable to acquire these mutations if it is actively proliferating.
• 33:57 And so we reason that if we can understand how the cell cycle is adapting in the short term to drug treatment, perhaps we could we could target that that fast non genomic adaptation to to prevent the acquisition of of more stable solutions to drug resistance.
• 34:14 And so in breast cancer, there are a number of CDK inhibitors beyond CDK 4/6 that are being investigated clinically, either in clinical trials, preclinical trials right now they're really targeting different cell cycle effectors that are that are acting at different phases of the cell cycle.
• 34:34 And so to get a sort of a global idea of the contribution of cell cycle adaptation to of the breast cancer cell cycle, we performed a, a massive cell cycle mapping experiment where we took a panel of eight breast cancer cell lines exposed to seven different CDK inhibitors or cell cycle inhibitors.
• 34:57 These are breast cancer cell lines that really span the, the molecular and histological subtypes of the disease.
• 35:04 And we performed our Multiplex imaging of around 50 biomarkers.
• 35:07 Remember in this case, these are cell cell lines.
• 35:10 So we don't need to measure the the microenvironment.
• 35:13 We're just measuring signaling and, and cell cycle effectors to obtain our single cell proteomic cell cycle signatures and the performer cell cycle mapping.
• 35:24 Now here is a sort of grand total of that entire experiment that I'll show you right now.
• 35:29 As you can see, these cell cycle maps which consist of for these are each of the 8 cell lines.
• 35:35 What we're showing here is a map that shows all of the conditions, both the control cell cycle and all of the drug resistant cell cycles we detect across all 7 inhibitors.
• 35:45 As you can probably as you, as you see, you know, these, these, these maps are far more complex than any of the maps that I've shown you earlier in this presentation, except for this one cell line, MCF 10A.
• 35:59 And this happens to be the, the normal breast epithelial cells that we're using as a kind of comparison to these, these other breast cancer cell lines.
• 36:09 These normal epithelial cells have the least complex cell cycle map, really hinting at this idea that that these cancer cell lines actually might have a higher capacity for adaptation.
• 36:20 They're able to find more cell cycle programs than normal cells to adapt and respond to inhibitors that are trying to block their their passage to the cell cycle.
• 36:32 Now, instead of sending these things, these very complex maps, we found that the the the more practical solution to understanding what's going on in terms of drug resistance is to focus on specific inhibitors.
• 36:43 So for example, if we can make the map of this triple negative breast cancer cell line, we can just show the cells that in control or cells that were treated with Pablo siclib, the CDK 46 inhibitor.
• 36:54 And in doing so, we see a very clearly distinct cell cycle program cell cycle.
• 37:00 The cells are taken in the presence of the CDK 46 inhibitor that differs from the normal path to the cell cycle.
• 37:06 Now of course, we want to just understand what is different about this path.
• 37:09 So we can take a look at our cell cycle signatures across all of the cell cycle phases from G1 out to M phase.
• 37:19 And you can see what is up regulated in our drug treaty condition relative to control.
• 37:23 And we very clearly see that these cells are seeing this massive up regulation of cyclin D1 across all cell cycle phases and also the other isoform cyclin D3 to some extent.
• 37:35 Cyclin D of course is the cyclin that binds directly to CDK 4/6.
• 37:40 And of course then you know this, this makes sense as a potential mechanism of drug resistance.
• 37:45 If we overlay cyclin D1 expression over onto our cell cycle maps, we really clearly see that along the CDK 46 resistant trajectory, the cyclin D1 is massively overexpressed relative to control cells throughout the cell cycle.
• 38:02 You see in in control cells it's mostly limited to early G1.
• 38:07 If we look across our cell lines, we see that this up regulation of cyclin D1 in the presence of palbociclib is we find it across all of our estrogen receptor positive lobular carcinoma cell lines.
• 38:21 We find it in our her two and our triple negative cell lines.
• 38:24 We don't find that to be the case in our normal MCF 10A cell line or in our in our two ER positive ductal cell line.
• 38:32 So there even seems to be some kind of subtype specificity to this mechanism of drug resistance.
• 38:38 So of course we want to know how is this cyclin DF regulation overcoming CDK 46 inhibition given that both the inhibitor and cyclin D both bind to CDK 4/6.
• 38:49 We you know, maybe the most obvious answer would be that this, that this, this massive upregulation of cyclin D is simply out competing the drug and rescuing CDK 46 activity.
• 38:59 And to to address that question, we turn to another technique that we often use in the lab, which is single cell time lapse imaging of cell cycle biosensors.
• 39:08 In this case, we use like as Thunder imager to perform these time lapse experiments.
• 39:13 So we in these experiments, we take our breast cancer cell lines and through lechipiral infection, we put in a number of cell cycle biosensors.
• 39:21 We put in the PCMPCN AM turquoise biosensor which basically reads out the G1S and the SG two transitions where in S phase PCNA pharmacy's punctate structure.
• 39:33 So we can see cells go from G1, whereas some more diffuse expression, they then get punctated S phase and we can mark that transition and then the puncted disappear as cells transition back into G2.
• 39:44 So it lets us annotate cell cycle phase progression with a single sensor.
• 39:49 We then added in kinase translocation reporters that read out both CDK 2 activity and CDK 4 activity.
• 39:56 So these kinase translocation reporters work in that when the CDKS are active, they phosphorylate the sensor and Kick It Out of the nucleus.
• 40:05 So we can then measure CDK activity by measuring the cytoplasmic to nuclear ratio for each of these sensors.
• 40:12 Here's a quick movie demonstrating the CDK 4 sensor in the presence of our CDK 46 inhibitor.
• 40:19 And you see as the movie runs, we see this gradual accumulation of this sensor in the nucleus of basically all of the cells, indicating that this inhibitor is indeed inhibiting CDK.
• 40:30 And the bio sensor is reading this out by increasing the the nuclear component of this in this measurement.
• 40:39 OK.
• 40:39 So from here we can extract single cell tracks and you can see how each cell at single cell level respond to treatment.
• 40:48 And in the presence of Pablo Ciclib, if we were to look at CDK 4 activity, we see that compared to control up above, here we see that that CDK 4 activity remains inhibited in the presence of CDK four of the CDK 46 inhibitor across the duration of the experiment.
• 41:08 So in this case, it's really suggesting that that this this adaptation isn't at the level of CDK 4 and that the drug continues to work by inhibiting CDK 4 activity.
• 41:18 And in fact when we look at instead when we look at CDK 2 activity, we see that cells that are destined for mitosis indeed find a way to up regulate CDK 2 activity in the absence of CDK 4 activity.
• 41:31 So even when Pablo Ciclip has blocked the activity of CDK 46 early in the cell cycle, this up regulation of Cyclin D1 is allowing CDK 2 activity to increase.
• 41:44 And so we're currently exploring the hypothesis that that's this up regulation of cyclin D may actually allow cyclin D to bind directly to CDK 2 to increase its activity, sort of bypass the requirement for CDK 4/6.
• 41:59 Then of course, you know, we want to know, OK, well, this is the, if this is the drug resistant trajectory, can we block that trajectory?
• 42:04 So what about if we then block CDK 2 at the same time that we're blocking CDK 46?
• 42:09 So we also perform cell cycle mapping using an FDA approved inhibitor of CDK 2, four and six, this PF compound.
• 42:17 But indeed we still find that in the presence of this compound, a proportion of cells that find their way into the proliferative cell cycle and detect cells in all of the phases of the of the proliferative cell cycle G1S and G2.
• 42:33 If we look at the cell cycle signatures of these of these seemingly proliferative cells, they have all the markers of an actively proliferative cell.
• 42:40 They have high phosphor B, they express a lot of CDK 2 itself and their positive for KS 67 are clinical marker of active cell cycle progression.
• 42:50 If we look at the cell cycle signatures in these proliferative cells, in these resistant cells as opposed to what we saw with CDK 46 where it was a very specific signal in CD, in Cyclin D1, we block all of these CDKS.
• 43:03 We see a more broad up regulation of cell cycle proteins suggesting that maybe these cells are just, you know, they're broadly up regulating these proteins, just trying to eke out the CDK activity necessary to get to push them through these checkpoints.
• 43:18 The other thing that we see is in the presence of the CDK 246 inhibitor, we see this massive enrichment of cells in S phase and, and and MG2, which is telling us something about what's going on here.
• 43:29 So once again, we turn to our time lapse imaging experiments and treated cells with this CDK 246 inhibitor, the PF compound of the bottom here.
• 43:38 And if you look at these single cell traces compared to Pablo Siclib, we did see that these cells are able to get sort of partial activation of a bit CDK 4, but definitely CDK 2.
• 43:52 They're not able to get to the same levels as in Control or even Pablo Siclib for the CDK 2 activity.
• 43:57 But this is consistent with what we saw, you know, in our in the in the cell cycle signature measurements where so it's kind of brought up regulate brought up regulation cells just trying to eke out enough CDK activity.
• 44:09 And I think we that's what we're seeing here in these experiments, we're able to get this partial activation of CDK 2.
• 44:15 But the one thing that we don't see in these experiments is a single mitotic event.
• 44:18 So these cells are finally their end of the cell cycle, but they're not able to go into mitosis and complete mitosis.
• 44:26 And in fact, if we go back to our Multiplex data, our our snapshot data and we look at phospho H3 positive cells, this is our mitotic marker.
• 44:34 Indeed, we detect very, very few mitotic cells in the presence of this inhibitor.
• 44:40 Now in terms of evaluating this as a drug, this would seem to be a good drug, right?
• 44:44 If we are, if we're inhibiting mitosis, then there are a fair few, few cells that are dividing.
• 44:51 This is what we want in a in a cell cycle targeting agent to treat cancer.
• 44:55 However, if you look at all these other measures of the cells cycle, these cells seem to still be in a proliferative state.
• 45:02 They have, they have high KICKIC C7 and they're expressing a lot of these proliferative markers.
• 45:08 So it seems like this drug is inducing cell cycle arrest, but it's not inducing cell cycle exit.
• 45:14 These cells remain primed ready to resume the cell cycle.
• 45:19 And we're currently performing experiments where we then wash out the drug.
• 45:21 And what we entirely expect to find is these cells simply reenter the cell cycle and continue proliferating, which is probably not ideal in the context of treating a tumor
• 45:33 So with that, I'm going to end up having to skip my final project here looking at T cell activation.
• 45:39 But you know, you all came here to see what we're doing in tumors anyway.
• 45:43 So I will then skip to my acknowledgement slide here.
• 45:46 We're seeing very cool things in, in T cells, by the way, happy to discuss if anybody's interested.
• 45:51 And just say, you know, I hope that I, I was able to explain to you guys today about the different ways we're using spatial multiplexing to answer questions that are pertinent to, to tumor biology and, and the treatment of tumors.
• 46:04 With that, I would like to thank my amazing lab who performed all the experiments you saw today.
• 46:09 I do thank our collaborators at Pitts and elsewhere and our funding sources.
• 46:13 And with that, I'd be happy to take any questions you might have.
• 46:19 Yeah, I'm waiting, waiting.
• 46:21 Thank you so much.
• 46:22 I mean, that was some really great insight.
• 46:26 And it was really great to learn about your labs research, right, and how you're using these multiplexing and spatial profiling experiments to really map the cell cycle and those tumor samples that you were going through.
• 46:38 And we want to thank everybody again for listening and and participating in the webinar.
• 46:44 And if you have any questions for Wayne, please do feel free to drop them into the comments section and we'll have those answered live here.
• 46:55 So now what we'll do is transition over to the Q&A part of this program.
• 47:01 So feel free to continue adding your questions, like I said.
• 47:05 And what I'd like to do is bring Nadine from Abcam and Natasha from Lyca Microsystems to the stage as well.
• 47:13 So that way if you have any product specific questions, right, we could also get those answered.
• 47:17 We heard a lot about Abcam’s antibodies, we heard a lot about the imaging platforms from Lyca as well.
• 47:23 So if you have any questions for those, feel free to ask as well.
• 47:28 And while we're waiting on questions, actually I I'd like to kick things off myself, Wayne, and maybe ask you a question here.
• 47:36 Because in that last part, the last story that you were going through, you were talking about this time lapse imaging experiment that you were doing and then you transition back to making the connection with the Multiplex imaging.
• 47:50 So is it possible to combine the time lapse imaging with the Multiplex imaging as an endpoint measurement?
• 47:59 Yeah, great question.
• 48:00 And, and, and the you know, the short answer is, is yes, absolutely.
• 48:04 And in fact in our our first sort of proof of principle papers that we published for the cell cycle mapping we performed, we did this exactly.
• 48:13 We basically watched cells progress through the cell cycle with our cell cycle biosensors and then fixed the cells and performed cell performed our Multiplex imaging on those samples.
• 48:24 What this allowed us to do is really connect the cell cycle history of a, of a given cell with its subsequent molecular profile with its, you know, with this, this more high dimensional, more quantitative, more more nuanced measurement of, of its cell cycle.
• 48:40 And that we use these, these histories to then map onto our cell cycle map itself to understand and to sort of validate that the cells move through this map.
• 48:50 It was consistent with them moving through the cell cycle phases.
• 48:53 And the cell cycle age also increased.
• 48:56 Now in the context of looking at, let's say drug resistant, etcetera, you know, connecting these histories, you know, how long was the cell in G0 before it entered the cell cycle?
• 49:07 Or, you know, how does it's, it's it's molecular signature change the longer it's in in cell cycloresis, it's hard to move more towards a a senescent molecular signature.
• 49:19 These are all super cool questions and totally amenable in combining these these these types of things.
• 49:26 Pretty cool.
• 49:28 I Speaking of pretty cool.
• 49:29 I think the data that you presented, right, the way you visualize that was really great because it was really compelling and you could clearly see what you were trying to portray.
• 49:40 So I'm curious like regarding that data, like what software are you using to really analyze your imaging data?
• 49:47 Because it was really beautiful.
• 49:49 Yeah, thank you very much.
• 49:50 I mean, I can thank all of my, the folks in my lab for their hard work developing some of these things and and, you know, and visualizing the data.
• 49:59 The, the answer to your question is.
• 50:00 So, you know, one of the things my lab does, you know, we, we, we recruit people that are very good at this is to develop custom image analysis solutions both for the, the time lapse imaging part and for the, the, the multiplexing where we are stringing together often, you know, open source packages that are designed to do the individual kind of components of the, of the analysis there.
• 50:24 Now I recognize that that's, you know, not that, that, that that's not a sort of straightforward thing to do from for most labs.
• 50:31 And, but we've also worked with, with Leica and Danaher beta testing and, and providing some support for their Avia software, which honestly, if we, you know, if we, if we were already, you know, if we hadn't already devoted resources to solving these problems ourself, we would certainly be using that, that software to actually make things very, very nice and easy.
• 50:53 And I, I hope that we actually contributed some important updates to the, the software itself.
• 50:58 I've seen what I've seen.
• 51:02 Nice.
• 51:03 I, I bet Natasha might be able to answer that question if, if you have.
• 51:07 But yeah, great.
• 51:08 And maybe, Wayne, if I could ask one more question from, from you as well, 'cause I'd like to also get Natasha and Nadine in on the conversation as well.
• 51:18 But when you started off your presentation, right, I noticed that on one of the slides, you had a really nice chart of the biomarkers you were going after.
• 51:26 That included some of the antibodies that you were using to go after those biomarkers.
• 51:32 Like how did you identify those biomarkers and and how did you really decide on which antibodies to use?
• 51:41 Like I'm just curious how you built that library.
• 51:44 Yeah, yes.
• 51:53 Are we there, Wayne?
• 52:02 Wayne, I think, I think we're, we're having a little bit of difficulty hearing you.
• 52:08 Can you hear me now?
• 52:10 Yes, there we go.
• 52:10 Yeah.
• 52:11 OK.
• 52:12 Yeah.
• 52:12 So I think there are two parts to that question.
• 52:14 I think first is like choosing the choosing the biomarkers you want to measure the things that are important for your biology and the second is finding good antibodies for them.
• 52:23 And as we anybody that chooses antibodies use, anybody that's used antibodies knows that this is sometimes not a trivial task.
• 52:29 So in terms of choosing the biomarkers themselves, this really comes down to identifying that things are important for you.
• 52:37 The question that you're asking there, I mean, I don't think anybody can really tell you the things that you should be measuring there.
• 52:42 But certainly, you know, we as cell cycle biologists, we and you know, and the cell cycle has been studied for decades and decades.
• 52:48 So there's a very obvious set of of a few dozen proteins that we want to measure and then in the tumor microenvironment to figure out what we should be measuring there to sort of broadly assess as many cell types and cell states as we can.
• 53:02 You know, we just, we did a literature search and looked at other studies that were sitting through a micro environment and just trying to find antibodies for as many things as we can.
• 53:10 Obviously with, you know, the work that we do and how important this technique is, It's like the bread and butter of the lab.
• 53:15 We are constantly building this antibody library to make it as big and as versatile as we can.
• 53:20 And for every single project, we kind of take a subset of the antibodies library that are, you know, that are that are can address the question that we're interested in.
• 53:29 So that's the first part.
• 53:31 The second part is finding the antibodies, as I said, which is not a, not a trivial thing to do.
• 53:37 There are a couple tools online.
• 53:39 There's a tool called bench side that allows you to kind of search among all of the antibodies that are available.
• 53:45 And it'll find these antibodies that we will find, which is very helpful.
• 53:49 It'll give you figures where the antibody was used in a given paper or to assess, you know, that's really helpful in assessing whether an antibody is actually working in the real world.
• 54:00 The other thing that I would highly recommend is reaching out to the reps of these companies, like your Abcam Rep to talk to them about the biology that you're interested.
• 54:09 If you say, hey, we're in cell cycle biology, one of the things they're really great at is putting together a list of antibodies in a given field or subfield that you might be interested in.
• 54:20 The second is if you have a very specific target in mind is, is say, hey, do you have something for cyclin D2?
• 54:26 And what you'll find is that, you know, of course the websites have a lot of information.
• 54:31 You can find all the antibodies that are available there.
• 54:34 But sometimes as you know, you know, you go to find and for your target, you can't find an antibody that's been used in the specific application that you're trying to find.
• 54:42 In our case, we're looking often for, you know, immunofluorescence or or IHC, something that's working in paraffin.
• 54:50 And if it's not, if an antibody hasn't been, is not sort of approved for that application, it's either it could either be because it's been proven not to work there or it hasn't been tried there.
• 54:59 And getting that information from the, the Rep can be super helpful 'cause sometimes it's just that it hasn't been tried.
• 55:05 And, and they're willing to, you know, let us keep it a go and see if it works.
• 55:08 And, and you know, we have, we have luck with that sometimes.
• 55:12 So that's my advice.
• 55:14 And, you know, in general, I think reaching out to, to the reps at these companies is this is a, a very important resource that both on the imaging side at Leica and on the antibody side at AB camp, you know, we're causing communication with these folks to help push our science along.
• 55:31 Yeah, great advice.
• 55:33 And, and I think if I could draw anything from what you just said, the collaboration part of it is really needed.
• 55:41 And also the antibody itself, you know, the sensitivity and the specificity seems to be really paramount for what researchers are trying to do with their, you know, tissue samples or, or cell cycle cell samples, I should say.
• 55:55 So Nadine, actually I'd like to direct this question to you then, given that antibody sensitivity and specificity is paramount for a successful Multiplex interspatial biology experiment, like how does AB Chem help address this need for researchers?
• 56:18 Thank you for the question.
• 56:19 Thomas, I think you touched on or very robust quality management system in the beginning of the presentation where we do extensive validation of our antibodies not only at the application level but also at the molecular level.
• 56:34 When it comes to application testing.
• 56:36 As Wayne pointed out, antibodies that are used for IHC, for instance, are routinely validated on tissue microarrays including multi normal and multi human, multi normal and multi tumor Tmas for human and human activity.
• 56:51 And also we have TM maze as well for normal mouse and rats.
• 56:54 So we do extensive validation on the application level.
• 56:58 We also do knockout validation as best as possible to really make sure that the antibody is specific as needed.
• 57:06 And we also, as I mentioned, have what we do a molecular identity at the biophysical QC level.
• 57:13 I think apart from that, Wayne touched on the fact that the sensitive specificity work that we do on one end is actually crucial.
• 57:20 But also being able to build this library of validated antibodies for particular platforms like Cell Dive and working with collaborators like Wayne who are building those panels really helps to build that confidence that not only our validation works internally, but also externally, that the validation on the platforms helps in choosing the right antibodies in the clones for people's experiments.
• 57:43 Yeah, got it.
• 57:44 Really, really important.
• 57:45 And I think you, you hit it right on the head and, and I appreciate the response to, you know, because I think it researchers, whenever they're selecting an antibodies, they want to make sure that it's going to work for what they're trying to do.
• 57:56 So it's really important that they select the right ones.
• 57:59 And Natasha, this is a question for you because I think what we just saw was very heavily focused on proteomics, right?
• 58:07 And there was some hint from Wayne's presentation about being able to leverage these approaches for a multiomics type of analysis, right.
• 58:16 So is it possible to integrate multi omic approaches on the cell dive?
• 58:23 Thank you for that question, Thomas.
• 58:25 Yes.
• 58:25 So the answer is we are working towards developments to make changes on our automation platform so that we can fully integrate imaging RNA transcript as well.
• 58:37 Currently we do have users that are already imaging Tenex stadium slides on our platform as well.
• 58:43 Yes, OK, great.
• 58:46 Well, I know we're we're coming up on the top of the hour, so I want to be respectful of everyone's time.
• 58:51 I want to thank you again for joining today.
• 58:54 Nadine, Natasha Wayne, excellent stuff.
• 58:57 Thank you so much.
• 58:58 And to everyone that's on the event, we hope you found today's program on cell cycle mapping with advanced Multiplex and the spatial profiling informative and we'd love to hear from you and better understand your needs.
• 59:10 So there's going to be a simple survey posted in the comments and sent to you after the webinar.
• 59:15 We greatly appreciate if you could provide some feedback that's only going to help us better understand how to support your work.
• 59:23 And if you have any questions, you know, don't worry because you can always reach out to us directly through our web page or feel free to follow Wayne as well in terms of his research and ask him any questions you'd like as well.
• 59:37 So with that, we'll wrap up today's session.
• 59:40 And thank you so much for joining us.
• 59:41 We really appreciate you.
• 59:43 Thank you so much.
• 59:43 Thank you everyone.