Automated, Multiplexed Co-detection of RNA and Protein

Welcome to our life science webinar, the symposium series. Here we are at Cambridge with the CR UK team. My name's Fiona Smith. I'm the marketing manager for UK and Ireland. I'm here together with my colleague Gareth.

Hi, so I'm Gareth and I cover the UK and Ireland life science business for Leica Biosystems. And I would like to introduce you to the CR UK team from Cambridge.

Hi, my name's Jo Arnold. I'm the head of the Histopathology and In-Situ Hybridization Core Facility here at the CRUK Cambridge Institute. Our department is divided into four service areas, routine histopathology, immunohistochemistry, in-situ hybridization and scanning and image analysis. And I have two of my colleagues that head up two of those service areas who will introduce themselves now.

I'm Julia Jones and I run the in-situ hybridization service for the facility, mostly using RNAscope from ACD and other methods such as DNA FISH and microRNA in situ hybridization. Hi, I'm Jodie Miller. I run the IHC arm of our core facility where we run as our bread-and-butter routine validated antibodies. Also, we do method development and new antibody validations.

What are the drivers and changes that have brought about CRUK being so well established here at Cambridge?

I think one of the main successes of being based in Cambridge on the biomedical campus is being so close to Addenbrooke's Hospital. That helps with the translational nature of the research that's carried out here, going from clinic and the patient to the research bench and back. I also think that we have twelve strong core facilities and expertise within those to support the research groups within the building, of which there's 22. They all have specialist cancer research focuses.

We are part of the University of Cambridge, and the clinical school specifically, so we attract some  good quality scientists within the building. The quality of the research and the quality of the scientists is of high standard. And I think all those factors together, as well as now the biomedical campus expanding and having people like AstraZeneca and Abcam also on campus, as well as the rest of the university, allows us to have large collaborations and get the best out of the science on the campus.

Thank you very much, Jo. That's interesting to hear about how you work with Adam Brooks and Abcam and AstraZeneca. I'm interested to know how you choose which products and how you decide on which areas to focus on for your research.

There's 22 research groups within the building and they all have their own individual research focus, whether that be a specific cancer type or just a mechanism. We're very driven by the research that's carried out within those 22 research groups. We can always keep an eye on all the new technology out there that's concerned with histopathology and present that to the research groups. And then if there is a need for that within their research area, then we can work that up with them. But it's very much driven by the 22 research groups that are within the building.

Thanks, Jo. It's interesting to hear how you're focusing on the different techniques and when they do become available, it helps to spark those areas of interest into new research projects. Today we're talking about multiplexing, but I’d just be interested to know how multiplexing adds value to your research and to the goals of your organization here at CRUK.

There's a big drive for multiplexing now. The more targets that we can look at in one tissue section helps the researchers within the building to understand more about the tumor microenvironment and the interactions between immune cells, for example. The more data we can get out of very precious human tissues, the better. It is the focus of all the research in histopathology now.

I'd like to hand you over now to Julia Jones, who's going to do a presentation this afternoon on automated multiplexed co-detection of RNA and protein.

I'd like to thank Leica for the invitation to talk about our work today. I wanted to say up front that this work has been carried out by myself and Jodie Miller as a joint collaboration of the service areas of our team.

An overview of what I'm going to be talking about. First, I'm just going to give a summary of CRUK, Cambridge Institute, and a bit more of a rundown of our histopathology and in-situ hybridization core facility. Then I'm going to go into some detail about how the RNAscope and integrated co-detection workflow assays work. And then I've got a couple of case studies that I'd like to present, and finish up with our future plans.

We are the histopathology and in-situ hybridization core facility. We're run by Jo Arnold, and we have four arms for our facility. We've got routine histopathology where we carry out paraffin processing, embedding and microtomy. Frozen processing and cryotomy, as well as H&E stains and other special stains. This area is run by Jo Heffer. We've got immunohistochemistry, which is run by Jodie Miller, in situ hybridization run by me, and scanning and image analysis run by Cara Brody.

We have a fantastic team of six research assistants who rotate through those areas to help with all our user requests. We are a core facility and there are twelve core facilities within the building, all with their own expertise. We've got genomics, microscopy, proteomics, to name a few. And we all serve 22 research groups that have their own interests in different cancer types and cancer mechanisms, including breast, ovarian, prostate, pancreatic, intestinal, and just to name a few.

Within our core facility, most of our focus is 80% mouse work and 20% human. So just to give you an idea of some of our throughput, we process around 1000 samples a month and cut 7,000 sections and stain around 1,200 H&E slides. We run about 1,500 slides for IHC, 85 slides for in-situ hybridization, and we scan around 3,000 slides a month.

A little bit more detail about the immunohistochemistry service. This is run by Jodie Miller, just as a reminder. We're like a heavy lab and all our immunohistochemistry services are run on the BOND platform. A standard IHC is a single plex or duplex offering and so far, we have over 750 antibodies tested, and of those, 400 plus are validated. There's also a multiplex IF offering, so immunofluorescence. This is also fully automated on the BOND RX, and it's typically been using Alexa Fluor conjugated secondary antibodies, but we're now also moving into using TSA opal dyes from Akoya as well.

Jodie has worked up a fully alternated tunnel protocol. This is detection of apoptosis using Promega's dead end colorimetric tunnel system. And she does this on the BOND RX. And there's also a new antibody validation service. We get more than thirty requests for antibody validations every year. And ideally, these are carried out on control tissue. We'll have siRNA knockdowns and knockouts, as well as treated tissue controls. And Jodie also always carries out no primary antibody and isotype controls to determine antibody specificity.

The in-situ hybridization service is a bit of a mixture of automated and manual methods. Most of what we do is using RNAscope assays from Advanced Cell Diagnostics, which is part of the Bio-Techne brand. And we're offering all their assays. We can do single and duplex chromogenic and then three and four plex fluorescent assays as well as their base scope. This is also performed on the BOND RX, but we can also offer manual stainings for different sample formats like cells and plates and whole mounts.

We also perform DNA FISH for whole chromosomes and genes using commercially available probes. One of our most popular is human versus mouse cells in xenograft. We're determining the ratio of these xenograft tissue using pan-centromeric FISH. This is a manual method as well as the following two. Micro RNA in situ is performed using double digoxygenin labelled locked nucleic acid probes and tyramide signal amplification. And then we also have padlock in situ hybridization, which is for the detection of single point mutations in RNA using padlock probes and rolling circle amplification.

Also scanning and image analysis, this is a very important end point of our facility. For all our brightfield scanning, we use a Leica Aperio AT2 scanner which gives us a 400-slide capacity, and we use eSlide Manager for user viewing.

For our fluorescence scanning we've got two different scanners. Firstly, the Zeiss Axia Scan and our recent acquisition is the Akoya Polaris. We can perform fluorescence and bright field on this scanner but it's revolutionized our multiplexing capability due to this multi-spectral nature and its ability to exclude autofluorescence from our resulting images, which makes our image analysis far easier and more accurate.

For image analysis, we have Halo from Indica Labs, and we've got a wide array of different modules for brightfield and fluorescent images. We can perform cytonuclear and area quantification in both brightfield and fluorescence. And we've also got modules for high-plex, which is very like multiplex slides. We also have ISH and FISH modules for spot counting.

As well as Halo, we also have some image analysis software within the Aperio eSlide Manager for brightfield only, looking at positive pixel and nuclear quantification. We use SurDown PathXL software for image sharing with external collaborators, where they can annotate images and we can share more easily.

On to the RNAScope assay. RNAScope is from Advanced Cell Diagnostics, and this was designed to be suitable particularly for formalin-fixed paraffin embedded tissue due to the partially degraded nature of the RNA. And the way that it works is that you have these pairs of Z-shaped probes. So here, along the base of the Z, we have these 18 to 25 base sequences that hybridizes in tandem with its partner along the length of your target RNA. We then have this linker in this upright and then along the top is a 14-base sequence which can hybridize a 28 base pre-amplifier strand.

When these two hybridize in tandem, you can get this amplification. We have around twenty pairs, typically twenty pairs on an RNA target. And you can see when they hybridize in tandem as they should, you can then hybridize this pre-amplifier strand. And when both probes within the pair aren't bound, this pre-amplifier is unstable and so it can't bind and reduces background signal.

After the pre-amplifier stage, you can then hybridize amplifier sequences. These are just oligos that are specific to the pre-amplifier strand. Then we have label probes. These contain enzymes, either alkaline phosphatase, which is used with fast red, and horseradish peroxidase which is used with DAB or the TSA opal dyes. We can then plex this by changing this top portion of the Z. If you change this sequence and the sequence of the amplifier, the pre-amplifier and amplifiers, you can then build up to four targets in one slide.

To show you what this looks like, this is a single Z pair. This is a single plex assay using the red kit. This is a duplex using, so we've got Red, that's alkaline phosphatase, and in blue, this is the fast green from ACD. Using the RNAscope multiplex kit, I can demonstrate here a three-plex assay. We've got three different target probes here in red, blue and green. Before ACD brought out their integrated co-detection workflow, we had to perform both pre-treatments for the RNAscope at the start. We would do our RNAscope protocol in full, followed by the IHC or the IF protocol.

I've got a couple of images to show you how this worked. In this slide you can see we've got some signal for RNA in white and in red. And then we've got our protein signal here in green. This is a lovely example. We've also got a single-plex red version. You can see here we've got the signal in the crypts. And then we've also got the protein signal in DAB here. These were some nice examples, but most of these projects have fallen because the protease treatment required for the RNAscope is just too harsh for most epitopes. We often ended up with suboptimal antibody signal.

I'm going to show you how the ICW, the integrated co-detection workflow, is different. So firstly, we perform the tissue pre-treatments. We split those. So firstly, it's heat-induced epitope retrieval with Tris-EDTA, followed by the antibody binding. Then in this step here, we fix with 10% neutral buffered formalin before then performing the protease step from ACD. And then we move on to the RNAscope protocol as we would have done previously. We hybridize the probes, perform all of the amplifications and detect each probe in sequence.

Then we add the secondary antibody and detect that, and then we visualize by scanning them and perform the image analysis. The antibodies are diluted in the proprietary co-detection antibody diluent from Bio-Techne. And we've got a couple of case studies now to show you how this has been working for us.

The first study that we performed was for Celia Martinez. She was a researcher here at CRUK Cambridge Institute and is now a group leader at Helmholtz Pioneer Campus. And she and her team wanted to detect 2 RNAs and one protein and also another single RNA with one protein. And they wanted a fluorescent output for the project. We used the RNAscope LS multiplex kit in combination with the co-detection antibody diluent and TSA fluorophores. Unfortunately, this did mean that we ended up with some redundant channels within the multiplex kit because this is for three RNA targets and we're only using one and two separately. We also scanned the resulting slides on the Zeiss AxioScan.

We thought very hard before we started this project about which controls we would need to include and decided that we needed an IHC control on test tissue. We wanted to test positive control RNA probes. These are housekeepers with a primary antibody, and then also those positive control RNA probes with no primary antibody, and then negative controls, here this DapB, with and without the primary antibody.

We also included a run control, which is a set of positive control probes run on mouse intestine tissue, so we use that run control to assess variability from run to run and ensure that everything was okay with the reagents on the day. I'll go through the controls that we ran. We've got this IHC on the test tissue. This is liver and they had chosen this protein, it's a membrane marker. We can delineate the cells, and you'll see that there are a few binucleate cells. They wanted to be sure when they did the RNAscope in fluorescence that they can determine when a cell is binucleate.

We've got our positive control probe here with the primary antibody. This is DAPI with PPIB in green and DAPI and POLR2A in orange. And then we've got DAPI alone with a protein and altogether they look like this. I'm not sure the green is particularly standing out, but it is here. You can see we've got some autofluorescence here from red blood cells. It's yellow.  Overall, we're quite pleased with that.

This is our primary antibody control. If I zoom in, we've got our signal for the PPIB and the POLR2A and you can again see the autofluorescence from those red blood cells, but we're not getting any signal from the no primary antibody, so we're quite satisfied here with this control. Moving on to the negative RNA control with a primary antibody. If I zoom in, we've not got any spots here for the PPIB or POLR2A. We've got a few autofluorescence again, but the antibody signal looks nice. You can see we've got a binucleate cell here. Then our double negative, if I zoom in, you can see we're only picking up the autofluorescence of those red blood cells.

We're quite satisfied at this point that our controls are all working, so we went ahead with the staining with the target RNAs and the antibody, and this is the resulting image. We've got our two RNA targets first in green and orange. And you can see we've got mutual exclusivity here. Only either of those targets expressed in a cell. And secondly, then we have the single RNA target with the antibody. In this target, they were looking at differential expression of the target around the central and periportal areas of the liver. This work has now gone on to be published.

Okay, and on to our second case study. This was for Cristina Jauset, who's a member of the Hannon Lab at CLUK Cambridge. And Cristina wanted to look at three RNA targets and two proteins. We used the RNAScope multiplex kit again with the co-detection antibody diluent and opal fluorophores this time. And we were scanning on the Acoya Polaris.

I should note here that detection of two proteins is currently unvalidated by ACD. We've had to change the protocol slightly to suit our needs. As before, we've permeabilized using the heat retrieval alone, done the antibody incubation step, fixed with 10% NBF, and then done our protease step. We've then hybridized the RNA probes and performed all of the RNAscope amplifications and detection. We added our secondary antibody and developed this with an opal fluorophore. We then blocked this with HRP blocker and added the next secondary antibody followed by a different opal fluorophore and then carried on with the scanning and the image analysis. It's worth noting here that our antibodies were raised in different species. We've got a rabbit and a rat primary antibody, which did make our detection a little bit easier.

Back to our controls, it's a little bit more complicated than the previous study because we've added in that extra antibody. But we've got the same set of controls. We've got our IHC on test tissue, followed by the positive control RNA probes with both primary antibodies and without the primary antibodies. And then we've got our negative control probes with primary antibody one and then two and then neither. We've also added in here an autofluorescence control for the motif scanning on the Polaris.

And this is a slide that has run through the entire protocol, but it's not stained with any of the fluorophores. That just gives us a measure of the autofluorescence within the tissue that we can then remove with the Inform software on the Polaris. And then we've also added in our run control as we did last time.

This is the IHC on the test tissue. Antibody 1 here and antibody 2. We're getting quite nice strong signal, and this is what we're hoping to be able to achieve in our other run controls. We used two positive control RNA probes in these controls along with the two antibodies. In yellow and red we've got the RNAs. And here in green we've got antibody 1 and in orange is antibody 2. With our no primary antibody controls, you can see we've got very strong signal here from one of the positive control RNA probes in yellow. And the negative control RNA with both antibodies, we can see antibody one. We're not seeing so much of the second antibody in orange. Then with our negative control probes with antibody one, we're getting still nice signal with antibody one. The negative control RNA and antibody 2, we're not quite sure that this is representative of the IHC we saw originally. If I show you that again, it's not looking quite as strong here as it does. as it does there.

What we decided to do was to increase the concentration of both primary and secondary antibodies. And what we got was an increase in the background. We do think that there's this extra, we've got an increase in the signal from the antibody 2, but also an increase in the background. We're still trying to optimize the steps here. to get the second antibody looking good.

On to our future work. The next big thing for us is the BOND RX software upgrade. We're currently on version 5.2. We're limited by the number of steps that we can have now. We're always at the limit. We're looking forward to those 500 step protocols, and the ability to run 6 markers on a slide and access to the new Leica chromogens too.

We've got some more ICW projects in the pipeline. A chromogenic duplex RNAscope with the detection of 1 protein. And the second project that's on this list, the single-plex RNAscope with detection of 1 protein, we've already started on this one. It's going quite well, but we have had similar background issues to the previous case study for Cristina with our two antibodies. And what we've done to combat this so far is to add in a HRP block right at the end of the protocol. After the final secondary antibody and before the DapB counterstain, we've added in this HRP block, and it's helped. We're going to go back and test this again on the project for Cristina.

We've also started to test out using the single plex RNAscope kit with opal fluorophores so that we don't have to go using the LS multiplex kit when we've only got one target anymore. So that's been quite successful. I'd just like to acknowledge all my colleagues in the histopathology core facility and thank Celia and Kelvin at Helmholtz for letting us use their images and Cristina here at CI. Thanks to ACD and Leica and Indica for all of their help in getting these protocols up and running. And then finally, just to let you know that there's a placeholder for our in-situ hybridization symposium, which might be held in May 2022. Thank you.