Fully Automated Workflows for High Plex Multi-omic Profiling with GeoMx® Digital Spatial Profiler and BOND RX/ BOND RX^m Research Stainer

Webinar Transcript 

It's a pleasure to be here today to be representing NanoString and to be able to present on some of our automated workflows that have been Co-developed between both NanoString and Leica Biosystems, combining the High Plex multi-omic profiling ability of the GeoMx digital spatial profiler with the automated sample prep system of Leica Biosystems on their BOND RX and BOND RXm. For those who aren't familiar with NanoString, we launched our first platform back in 2008, which was our encounter system for doing bulk gene expression in clinical grade FFPE tissue. And we entered the spatial world in 2019 with the launch of our GeoMx digital spatial profiler, which is our very first spatial platform. And since then, we've made it our mission to map the universe of biology spatially. We believe that the next great round of diagnostics as well as the next great round of therapeutics will be discovered utilizing platforms like GeoMx spatially. And that is because biology is extremely spatially heterogeneous. Just take something as complex as the tumor microenvironment, in this case, what we're highlighting here is a publication out of Doctor David Rimm's lab. He's a professor in pathology at Yale School of Medicine. And he's trying to understand in non-small cell lung cancer cohort, potential responders or predictors biomarkers of therapeutic response to checkpoint inhibitor therapy. There's obviously a lot of promise now in immuno-oncology with checkpoint inhibitors. We've made great strides over the last 20 plus years now with these therapeutics, but still, it's still not entirely clear which patients will respond positively to these treatments, and which will not respond positively. So doctor Rimm, understanding that heterogeneity of something as complex as the tumor microenvironment, understands that spatial tools become a bit of a requirement to piece apart and understand these predictive biomarkers, not at a bulk level such as bulk RNA sequel give you or mass spec, but to understand these biomarkers within their spatial context and how that spatial context relates to response to treatment. So, in this case being able to use a platform like the GeoMx digital spatial profiler helped Doctor Rimm and his lab accelerate their discovery of potential predictive biomarkers that they can then focus on and bring into the clinic in the future. These spatial tools allow you to have all the power of your high Plex profiling of a bulk RNA or mass spec but now allows you to do that with once again conserving that spatial context. And so. These tools have really helped you advance our understanding of spatial heterogeneity and the role that it plays within a lot of these different disease states or therapeutic states. And as I mentioned, NanoString has made it its mission to understand and resolve spatial heterogeneity using our different spatial systems. GeoMx is what I will focus on today. That's the system that we launched back in 2019, and it is the platform that allows you now to look at multicellular functional units of the tissue, but profile them with unbiased plex all the way up to the power of the entire transcriptome. So spatial biology actually forces you to understand not only the XY and Z spatial resolution of your system or tissue of interest, but not only where cells are located, who their neighbors are, but also what those cells are doing, and how they interact with each other. Which pathways are being activated. So, both resolution as well as plex become a requirement for understanding spatial heterogeneity and piercing apart things as complex as say the tumor microenvironment. And so, as I mentioned GeoMx allows you to do that on a multicellular, extremely high plex level, and then more recently, we just recently launched our CosMx Special Molecular Imager. And what CosMx allows you to do with the power of imaging is to do extremely high plex analysis of your tissue. But to have that resolution all the way down to both single cell and subcellular resolution. So, while GeoMx allows you to do high plex profiling very quickly with the power of the whole transcriptome, in these multicellular functional units, CosMx allows you to understand which cells are present at both single cell and subcellular level using high Plex molecular imaging. So both systems working hand in hand really give you a full understanding of your biological system of interest. This is just really summarizing our spatial portfolio as it stands today, both our GeoMx digital spatial profiler, as I mentioned for doing the whole transcriptome and a higher throughput multicellular profiling. And this is important when you want to understand the difference in expression of spatial or functional structures across a large cohort of samples. And then that's complemented by our cosmic spatial molecular imager, which allows you to look at up to about 1000 Plex currently, but at both single cell and subcellular resolution and therefore giving you a finer granularity of resolution for understanding not only the cells that are present, but also the expression level of different targets within those individual cells. Both systems are automated. They both have FFPE capabilities because we know how important it is to work with clinical grade tissue specimens and what most of our customers have access to in the lab today. They are also both multi-omics. So, they can look at both RNA and protein together and that's also important as we try to understand that kind of central dogma of molecular biology and whether the presence of the RNA transcript translates over to the functional unit of the tissue or cell which is the protein. And by utilizing both platforms together, as I mentioned, you really get a clearer understanding and picture of the heterogeneity within your biological sample. I mentioned previously the tumor microenvironment, in this case what we're showing you here is actually a cross-section of a mouse brain and in this case, if we're interested in looking at the hippocampus, for example, which is circled in the center, there we can do kind of multicellular whole transcriptome entire pathway profiling utilizing the GeoMx system on the left, but then also understand what cells are present and what their cellular state is by using high Plex molecular imaging on our CosMx system on the right. So, with both platforms understanding which cells are present, but also what those cells are doing, and both require these two unique capabilities. As I mentioned, I'll focus mostly on GeoMx today because our automated workflows using the BOND RX and RX m have been pre validated and tested currently on the GeoMx system. So, we have a nice integrated workflow between both the Leica system and the NanoString GeoMx. So, I'll focus on those today. But first I'll take you through some of the similarities between both platforms as well. One of the main similarities, as I mentioned, is both systems have been designed to give our customers ultimate plexibility with what samples they can profile or image using their NanoString system. Both have a wide area of scanning, and that gives investigators plexibility to work with things such as core needle biopsies, whole mount tissue sections. Obviously, both platforms have been FFPE validated, so they both work with FFPE, but they also will work, and we do have pre validated protocols for fresh frozen tissue as well. And then finally, because they both have these wide scan areas, they allow our customers to also work with things like tissue microarrays where you can put multiple samples on a single tissue section or on a single slide to utilize High plex or higher throughput of the numbers of samples that you can view at a single time. For the GeoMx platform to do our profiling, rather than attaching fluorescent probes to antibodies for protein detection or in C2 probes for mRNA detection, what we've actually done is come up with is this DNA oligo barcoding system, which allows us to Multiplex with all of the power of alumina next Gen sequencing with the numbers of targets that we can look at a single time. We're no longer limited by the fluorescent spectrum of our microscopes for being able to do spatial multiplexing. We now can utilize the power of next Gen sequencing and as I mentioned, we have a partnership with Illumina to do this where we can utilize their platforms now for doing high plex multi-omic detection using our GeoMx system. As mentioned, for protein, it's all antibody based. These are IHC tested and validated antibodies. We use standard immunohistochemistry protocols for doing all our staining and that's what allows us now to also utilize the BOND system for putting that automated staining protocol on that platform. The unique capability here is not only the DNA bar codes that we're using for the multiplexing, but also these UV photocleavable linkers that are attaching those oligo barcodes to the probes. What's unique about those photocleavable linkers is they allow us now to stain up an entire tissue section with almost an unlimited number of these targets and then utilize the UV power of the GeoMx system to cleave off those photo linkers, photocleavable linkers and release those DNA barcodes into a little liquid bath that's sitting over our tissue for doing aspirating and then downstream analysis of those oligos on next Gen sequencing. As I mentioned, multi-omic utilizing antibodies for protein detection and C2 probes for MRNA detection. And we're just using standard immunohistochemistry protocols. As I mentioned, we are combining those algo tagged probes with up to 4 fluorescently labeled probes and the reason we do this all on the same tissue section is that those 4 fluorescently labeled probes allow us to visualize the tissue section on the GeoMx system. At its core, the GeoMx platform is a fluorescent microscope. It allows you to visualize up to 4 fluorescent channels on that tissue section, but layered on top of that fluorescent image are also these oligo tag probes and barcodes. We combine all of those, both imaging reagents and profiling reagents onto a single tissue section, all using a single antigen retrieval and one single stain. And then once those are combined, then that slide is ready to go onto the DSP system. That antigen retrieval and standing has been automated, as I mentioned, on the Leica Biosystems BOND RX, so it allows us to take a workflow which can take a very long amount of time, few 100 minutes of hands-on time for doing all the dewaxing and deparaffinization and antigen retrieval and scanning. And it combines it down now to less than 30 minutes of hands-on time utilizing the power of that BOND system. So great automated workflow and once again kind of integrates right into nicely standard pathology labs and workflows. This is that workflow once the once that slide goes on to the GeoMx system, what's happening on that platform? As I mentioned, the system is going to go through and do a scan of the four-color fluorescence, those imaging reagents that were included with the cocktail of probes. Once we scan those four-color fluorescent probes, the system will give you a fluorescent image. The end user then can go through and using that fluorescent image, select key structures or compartments or areas of the tissue that they want to profile. You know, and they can go through in the software. They just kind of click and drag the areas of the tissue that they want to profile and then they hit run and the GeoMx will go through, and it will move the objective field, the view of the GeoMx over the first region of interest. One of those first circles that you see there, it'll turn the UV light on right through the objective of the GeoMx system and that will release those photocleavable linkers and release those DNA barcodes, as I mentioned, into a little bath that's sitting over top of the tissue. A micro capillary system will just come by. Once those oligos are released and aspirate them up so the little microcapillary system aspirates up those oligos that were released from the tissue and it will dispense them into a single well on a 96 well microtiter plate where each well on that plate has been indexed back to the XY area on the tissue of the region that was selected, so we are actually connecting based off of the well of the plate. The XY area of the tissue that we're profiling. That's how we're combining now. The multiplexing with the spatial information. The system will go through each region that was selected one at a time, until either you fill up the 96 well plate or you've exhausted the number of regions that you've selected on your tissue. That plate then comes off. It will do a very simple standard library prep on the plate and then those oligos are pooled and they can be read out on our encounter system, which will allow for targeted readout of up to 96 targets at a single time. Or on alumina next Gen sequencing, where once again we can go all the way up to right now currently about 22,000 probes at a single time using the power of next Gen sequencing. Once that data has been read out, it gets ported back into the GeoMx and that expression level data gets mapped back to the XY area on the tissue that the regions were selected on. There's just an example of a tonsil on the bottom right-hand side that you see and you can see germinal centers. You can see the T-cell beds and the aphelia layer in green and then those regions, those circles were selected across that tonsil. And what we can do is compare the digital expression of a handful of these protein targets, in this case, back to the regions that they were selected on the tissue. And we can see that they match up nicely. And because we're using UV light to release those oligos from the tissue, we basically have the flexibility of how fine or controlled we can control that UV light to select different regions across the tissue. We have tons of different capabilities and different questions that we can answer utilizing this platform. Anything that doesn't receive UV light will not actually be profiled. You'll have the fluorescent image of the entire section of the entire slide, but only for regions that you selected for doing your profiling will you have that multiplexed expression level data. And if we're just selecting geometric circles or squares, for example, we can ask questions about XY, you know 1 area of the tissue, comparing expression versus another. For example, if we just want to drop a circle over, you know, say the TME, the tumor microenvironment, we can do that and just profile everything within that circle. But say for example we want to look at the expression of potential biomarkers within specific compartments of the tissue. Say we want to look at expression of potential biomarkers within the CD 45 positive department so lymphocytes or say we only want to look at the tumor compartment. We have this ability to control our UV light to only profile specific cell types or compartments based off that combination of those fluorescent markers. In this example, we're showing you here for segmentation. That yellow area is a pan sided keratin stain. We can profile only the pan sited keratin positive cells. And then it's binary, so can you pick the inverse of it as well and profile everything around the pan CK positive cells. This allows us to have finer cellular resolution and granularity to really understand where the expression of these targets is coming from. And then take that one step further and say there is a definition of a specific cell type that we want to profile. I mentioned CD 45 and Pan CK, but maybe you only want to look at macrophages for example, and so we can include say at CD68 stain to only profile macrophages or we want to look at CD3 positive cells and so any combination of those 4 fluorescent markers can be used to create a gating strategy to look at expression level profiling and data of those cell types. If we want to look at expression gradients so like the GeoMx we want to take it one step further and look at different layers of expression as we go from point A to point B in space. We can do that by drawing these cool unique expression gradient layers and look at expression as we go from one area tissue to another and just to see how that expression changes as you go in the X in the X&Y. Finally, if you have no spatial question in mind other than just you want to look at overall expression and understand you know across this large area what the expression of these targets are, we can do gridded profiling as well, and the system allows you to just kind of set your gridding criteria. So how many boxes do you want and then just drop them right over your tissue and do a kind of high Plex high region profiling using this kind of unbiased breeding approach. Lots of flexibility on the system and the reason we can do this and the unique capability of the GeoMx platform that gives us this ability to select these specific genotypes or cell types of interest are these little digital micromirror devices that are built right into Microscope. These DMD digital micromirror devices allow us to control UV light in little one Micron increments, so we can set this gating criteria based off those fluorescent markers. As I mentioned, those imaging reagents that we're including in our Cocktail, and then once we set this gating criteria, those little DMD's will reconfigure based off the criteria that you've set in the software. In this case if we want to look at you know the tumor cells, those will be paying side of carotene positive, we'll set that gaining criteria those DMD's will reconfigure to only shine UV light on anything that's NCK positive. And then say we want to understand the immune microenvironment around the tumor. If we've got a CD45 stain, which we typically use in our imaging reagents. We can set that gating criteria for CD45 positive Pan 2K negative and we can profile all the immune cells that are adjacent to or around the tumor. What this allows us to do is once again just have finer granularity into these unique expression level profiles of these different compartments. So where bulk RNA seek is going to just give you kind of overall expression and you'll be impacted by averaging effects going across that large area or large tissue with this capability. Now we can not only conserve the spatial XY, but we can also select specific compartment to understand that the signal that we're seeing can specifically be contributed to those compartments or cell types. We’ve launched in 2019, we've had a lot of early success with the GeoMx system just being out there in customers hands and where we've really found a really strong value for the system is in these really large cohort analyses where we want to not only conserve the spatial information, but we also want to understand all of the underlying mechanisms and all the different pathways that are being activated within the biological system. And so this is now between the GeoMx being able to do kind of larger cohorts these and combined with the automated ability and high throughput of the BOND RX and RX m system and then the high multiplex and pooling ability of next-gen sequencing. We can now start to enable high-cohort spatial precision medicine and really start to understand how these potential biomarkers are impacting patient outcome and how different patients are responding to different drug therapies or treatments. In order to do this type of large cohort analysis, obviously you have to have a lot of samples and you need to be able to not only get through all of those samples in a reasonable amount of time, but you also need the reproducibility and quality of data that the once again the BOND RX plus the genomics allow you to have because it's cutting down on a lot of the variability that can happen with a lot of the different hands on time between multiple steps. By automating this process and having a great throughput on all these platforms, we really can now start to enable this type of precision medicine work. This just showing you a little bit of a comparison on the three different workflows that are enabled using the GeoMx system. The first one is manual; it's about 3 1/2 hours of hands-on time to do a sample prep. So anyone who's done, you know, kind of tissue work in the past, immunohistochemistry, you understand that you’ve got to do your baking, your dewaxing, antigen retrieval. There is an overnight hybridization step for both protein and mRNA staining. So we do allow those probes to sit on a tissue overnight, and then the next morning you come in, you wash the excess probe away, you input your imaging reagents or what we call morphology markers and then you can put it on the GeoMx system. It’s about 3 1/2 hours of hands-on time. So, it is not incredibly difficult, but still lots of steps and lots of room for variability within those steps. What's nice about the BOND system? Once again you can automate. We both have a semi-automated which does all the baking, dewaxing antigen retrieval, protease K treatment but leaves manual staining for the overnight high and then coming in and washing the next morning. You will need to manually stain using semi-automated, but all kinds of pre tissue work and treatment is automated on the BOND RX, which is great because that is labor intensive, but what's fantastic is this new fully automated protocol that we launched. Like I mentioned a few months back, that takes all of that, the dewaxing the engine retrieval as well as the overnight staining and the washing and does all of that directly on the BOND RX system. It takes hands-on time all the way down to about 30 minutes. And what that does is it opens up your morning. So instead of coming in in the morning with both the manual and semi-automated and then doing your wash and then your morphology marker. Which can take some time in the morning and not really getting on your GeoMx system till the afternoon with the fully automated now you can actually show up first thing in the morning and your GeoMx is ready to run. You're ready to take those samples right off the BOND, put them on the GeoMx, and that just opens up more hours in the day for you to run your GeoMx system. So really nice three options for the workflows. As I mentioned, kind of advantages for all three here. A little bit more on the advantages. This is really breaking down those times in specific numbers. So you can see how that hands on time breaks down between the manual, the semi-automated and the fully automated. Like I said, kind of 30 minutes of really just for the fully automated, about 30 minutes to setting up the equipment and then once you come in the next day, after the hybridization, you take those samples off and you put them right on your GeoMx. So just opens more hours in the day. It cuts down on your on your full-time employee time so you know it allows you to be more efficient and it also has some really nice advantages for things like reproducibility and the quality of data as well. What's nice about the fully automated protocols we have come out with some pre-validated and pretested panels that work right out of the gate with both the manual and fully automated. And we've configured these to work with either workflow, but we do have special kits available for both manual and fully automated, and one of them is our whole transcriptome, Atlas. Our WTA whole transcriptome atlas, this is one of our biggest sellers, our best product and the reason being is it's it gives you that kind of unbiased ability to look across the entire transcriptome for both human and mouse. And so you can look it's a non-Poly pull down. These are probes that will hybridize to your target RNA. So as I mentioned by not utilizing the Poly-A it removes that Poly-A bias that we can see. For just targeting like the highest expressed genes, so it removes that that Poly-A pull down bias and allows you to use the hybridization ability of the probes to target and get good sensitivity across a large dynamic range of both high, medium and low expression genes. It's got complete coverage. So does the entire transcriptome. These are all of the protein encoding genes in the transcriptome. They can be combined with the visualization ability of the GeoMx, so including your morphology, scanning for doing ROI selection as well as cell type or compartment profiling. As I mentioned, there are streamlined workflows in that we have pre-configured kits that work with both manual and fully automated. So regardless of what options you have available to you in the lab, we do have preconfigured kits for both. They do allow for customization, so you can spike in specific probes that you're interested in. Say for example, you want to look at PCR probes or if you want to look at non protein encoding genes you can spike those in through custom through our custom capability. And because it's non Poly A, because you can select the biology of the tissue that you're really interested in, there are minimum sequencing requirements, and it is a nice cost effective option versus some of the other you know kind of like gridded Poly A based approaches that are out there. It is a nice cost-effective option as well for just focusing on the biology that you're interested in. And we do have the ability to do our whole transcriptome with protein goat code detection. Our morphology staining for visualization, we do have protein options for that. So that can be combined with our whole transcriptome, but we also have our high Plex profiling ability that can also be combined on a single tissue section for both High Plex protein and whole transcriptome profiling from that single tissue section. We call that workflow our proteogenomics workflow. There is a white paper with that workflow outlined on our website. Visit the reagents section of our GeoMx web page on our website. If you're interested in doing code detection for both RNA and protein together. Bringing it all together, this ability to combine whole transcriptome kind of unbiased pathway profile. With the spatial resolution that GeoMx provides at the cellular level, allows you to do some amazing things. This is just one example of a data set that is available on our website to download. We have what we call our spatial organ Atlas, which is a series of publicly available data sets. On our website, in case you're kind of curious to see what GeoMx data looks like or if you want to download the data just to get a sense of you know how we approached our spatial studies and what kind of pathways and targets are being expressed within these different unique regions of tissue. But this is a really great example. This is a pancreas tissue that we profiled using the GeoMx system, and we wanted to look at these unique little structures. They're called the eyelets of longer homes. They represent about 5% of the total pancreas. So once again, if you're doing kind of a gridded approach to a pancreas, you may or may not land on one of these, you'll be lucky if you do, and you're going to wind up spending a lot of money profiling things that you probably don't care about. But if you are interested in biology, you know, GeoMx allows you to go right in, identify the eyelets using that fluorescent staining, and then just profile them with the power of whole transcriptome. Not just profile the entire eyelid, obviously you can do that, but you can also piece apart the unique cells that are within that eyelet. In this case we're looking at these alpha and beta cells and we're profiling those alpha and beta cells with our whole transcriptome Atlas. And we're looking at the expression of all these different targets within these unique compartments to understand functionally and biologically how these two cell types within these eyelets are different. Obviously, if you had a lot of samples, you can understand how the eyelets and beta cells interact across a large cohort as well. GeoMx gives you that ability. This is a little bit of an overview on that. As I mentioned those pre configured kits that we have for using the BOND RX and RXM systems, we have the human whole transcriptome Atlas as well as the mouse whole transcriptome Atlas. You can also include custom spike INS into both of those. And we have our cancer transcriptome Atlas as well, which is an 1800 gene panel. So obviously not as high Plex as our whole transcriptome, but 1800 genes of the key functional targets within cancer biology. So that has been a curated panel in case you're looking for something a little bit more targeted. And we have these sample prep kits as well as pre-configured morphology marker kits that go along with the BOND RX and RX m system as well. And so this is just the entire GeoMx assays and portfolio in case you're interested in utilizing the GeoMx system. As I mentioned, we've got a number of different protein assays. These different protein assays. They come in these little modules that can be added up to get to higher and higher plex. The most plexing we've done with protein currently, about 165 Plex. I'm still very high plex utilizing the GeoMx with this modular approach, but you can also mix and match the modules. So if you're interested in some of the targets on the immune cell profiling and you want to combine them with some of the targets from the neuro profiling panel and you can do that in, in the GeoMx, the protein assays give you that flexibility. With RNA, we have 3 current RNA panels, one of them is that whole transcriptome Atlas that I mentioned for both human and mouse. We also have our cancer transcriptome Atlas, which is at 1800 gene panel focusing on cancer biology. And then we have an 84 gene RNA immune pathways panel. These are some of the key immune targets 84 genes and that panel is read out on our encounter system. Finally, we have custom abilities for both spike in as well as standalone. If you want to do a custom protein panel or a custom MRNA panel, you can reach out to us and we can help you design that panel with our informatics team. Or if you are just interested in spiking into any of our current panels where obviously we can help you out with that as well. This shows you what's getting a little bit of what's being done on that BOND RX system and that is the baking and the dewaxing, the antigen retrieval, as well as the blocking and that overnight incubation. Just a great automated workflow and the real power of that is in the reproducibility as mentioned. Now of course, all our assays are extremely reproducible, and the GeoMx system shows amazing reproducibility both between user to user as well as instrument to instrument, but I think that what you see here is with the BOND RX and RXm. It takes it to the next level as well. So just amazing reproducibility scores here. This shows you kind of a standard tonsil tissue with 300 Micron geometric ROI selected. The amazing reproducibility between slides as well as between the RX and RXm systems as well for both protein and MRNA. And what's nice about the GeoMx system, especially when it's coupled with the like BOND RX and RXm, is that we have this large dynamic range. About 6 logs of dynamic range. That expression level profiling and when you combine that with like I said, the minimum hands-on time and reproducibility that the BOND RX provides. You get these amazing reproducibility scores. And as you try and, you know, strive towards, you know, larger and larger cohort studies or multi system studies or multi institutional studies, being able to automate all of this and cut down on that reproducibility and that learning curve for new users allows you to scale the numbers of the not only the number of studies, but also the number of samples that you run in a single study. This is a list of all the equipment and reagents that are required for running this protocol so you can see here you need both access to the GeoMx as well as the BOND RX system. And number of the reagents that are available through Leica Biosystems, and if you have any questions on what to order of course Leica Biosystems as well as NanoString would be happy to help you identify the right kits and right reagents. We have an entire section of our website dedicated to customer education. So if you're also interested in looking at the protocols and the workflows, we have this section of our website called NanoString University or Nano U and you can log in there. You can find all of our white papers, all of our workflows, as well as lots of information on how to get started utilizing both the BOND RX as well as the GeoMx platforms. And so we've seen just this influx of discovery coming out of these systems. These are obviously new capabilities and are really starting to expand our understanding of biology and of the role that spatial heterogeneity plays within different biological systems. And so we've seen many publications. We're pushing almost 200 to date so far and most of that has come in the immuno-oncology space. But obviously a lot of work is being done now in neurology as well as infectious disease, and many covers utilizing the GeoMx system have been granted because of just the amazing data and some of the amazing discoveries coming out of our customers lab. With that said I will just conclude here. I want to thank Leica Biosystems for having me and, yes, happy to take any questions that come up and looking forward to working with all of you once again to map that universe of biology.