Exploring Multicomponent COVID-19 Mutant Variant Analysis of Immunity Indicators in Human Serum

0:00:00.0 Speaker 1: Welcome everyone to our presentation today entitled Exploring Multicomponent COVID-19 Mutant Variant Analysis of Immunity Indicators in Human Serum. My name is Peter Fung, and I'm the Senior Product Marketing Manager at Carterra, and I'll be your moderator for today's joint presentation between Bio-Techne and Carterra. Now, for a little background on our talk today, as most of you are aware, the COVID-19 pandemic and the global efforts around vaccine development has really heightened the focus on immunity resulting from both natural infections, as well as vaccination. As a result, the need for more sophisticated techniques to characterize immunity are urgently needed to guide, not only basic research, but drive effective and efficient public health policies. Using serum from COVID-19 vaccinees, our speakers will share how the combination of Bio-Techne's ACE2 protein and their large panels of high quality spike receptor-binding domain variants are examined using multiplexed high-throughput SPR on the Carterra LSA to maximize data that's gleaned from serum samples.

0:01:19.5 Speaker 1: The assay design allows for straightforward interchanging of variance by use of a conserved capture approach, making it easily adapted to evolving research needs, both within COVID-19 and beyond. Collectively, the assay enables a snapshot of several indicators of immunity, totaling hundreds of data points from just micro liters of serum and adds to the tools available for researchers to drive towards a more sophisticated understanding of immunity. So, to share with you a little about our speakers today, Anthony Person is the Senior Director of Protein Business Unit at Bio-Techne. Anthony was trained as a cell biologist and has done academic work at both the University of Arizona and the University of Minnesota, with focus on elucidating the molecular mechanisms of Went Biology in the context of cardiac development and human genetic diseases.

About Anthony

0:02:26.4 S1: Anthony worked in the stem cell industry for multiple years, developing methods for reprogramming somatic cells to play out potency and trans-differentiation of cells from one type into another with messenger RNA over-expression approaches. Anthony has been in the biotech industry for over 14 years with experience running the Bioassay, molecular, and cell biology portions of the Protein Business Unit at R&D Systems prior to taking over the entire department in 2018. He's focused on new product development and expanding R&D Systems' portfolio proteins for immune oncology and regenerative medicine research applications.

0:03:12.8 S1: Now, due to the COVID-19 pandemic, the Protein Business Unit at R&D Systems has prioritized the development of COVID-19 related products with over 200 recombinant proteins available to the biotech and pharmaceutical industries with critical tools to combat COVID-19. Our next speaker is Noah Ditto, who drives the development of transformative high-throughput surface plasmon resonance hardware and software solutions as a Technical Product Manager at Carterra. Prior to joining Carterra in 2014, Noah supported drug discovery and early clinical development for nearly a decade at Bristol Myers in Princeton, New Jersey with the focus on biophysical characterization of proteins and peptide-based biomarkers, drug targets and therapeutics. Noah earned his MS from Pennsylvania State University, developing chromatography-based fractionation techniques to isolate disease-specific serum biomarkers in Dengue infection, and more recently earned an MBA from Westchester University focusing on business analytics.

Today’s Topic

0:04:32.1 S1: So, as a brief outline for today's talk, what we'll cover today is Anthony will give you an overview of Bio-Techne's offerings on the SARS-CoV-2 RBD variant designs, their expression and purification as they're making these reagents. And then Noah will share on immunity monitoring on the Carterra LSA. He'll give you an assay overview and the development work that was done, and then share some pilot studies using the serum samples. We'll end with the takeaways from the presentation and a Q&A. Now, last bits as a bit of housekeeping for this presentation, I just wanna mention that we'll have a live Q&A at the end of the presentation, so if you have any questions, please feel free to enter your questions in the Ask a Question box shown on the dashboard or screen, and we'll do our best to answer these questions in the allotted time. If we can't get to your questions, we'll pass them on to our speakers, and they'll respond to you shortly after the presentation. Also, please check the resource box for downloadable documents from Bio-Techne and Carterra. And with that, I'll turn things over to our first speaker, Dr. Anthony Person.

SARS-CoV-2 RBD variant designs

0:05:53.7 Speaker 2: Thank you, Peter, for that kind introduction, and I will spend some time going over the reagents to use in the Carterra LSA device that were made at R&D Systems, part of the Bio-Techne corporation. At Bio-Techne, we've spent a significant amount of time and effort into developing Coronavirus proteins over the last two years. It's actually amazing to think that it's been two years since the pandemic really started in December of 2019. In the Protein Business Unit, which I manage, we've brought to market over 235 Coronavirus-related protein products, focusing on spike, ECDs, receptor-binding domain for portions of the spike S1, S2, portions of the spike proteins, and of course, ACE2, the receptor in our cell surfaces. We have also made various other proteins, including nucleocapsid, proteases, NSPs, ORFs, and the membrane protein. We've also brought to market various biotinylated proteins, both amine biotinylated and Abby biotinylated proteins, and recently, we've also developed fluorescent-labeled spike proteins for testing or blocking antibodies in flow cytometry workflows.

0:07:04.8 Speaker 2: In addition to proteins, of course, R&D Systems, part of the Bio-Techne Corporation, has developed many antibodies for Coronavirus research, including over 200 primary antibodies that can be used to detect Coronavirus, match pairs for ELISAs, flow cytometry antibodies, very nice neutralizing antibodies that block the spike binding to ACE2, as it'll be articulated in this presentation, as well as Western and IHC antibodies.

0:07:36.5 S2: Over the last two years, the spike protein and the ACE2 receptor on our cell surfaces have probably been the two most researched proteins on the planet, as we all know. The vaccine approaches, both mRNA as well as the J&J vaccine, focus on overexpressing the trimeric spike protein in our bodies, and it serves as an antigen to produce blocking antibodies that block the spike ACE2 interaction. And as this virus mutates and evolves in our bodies, the question that always comes up is, how protective are these vaccines that use the canonical spike protein, how effective are they in protecting us from variants of the Coronavirus as it evolves? And our reagents developed at Bio-Techne along with Carterra LSA really provides a high-throughput platform for us to screen patient samples, vaccinated patient samples, even monoclonal antibodies, to look at how well the blocking antibodies in these vaccinated samples protect against the various spike variants in reductionist assays, where we're using ACE2 as well as spike protein variants.

0:08:58.5 S2: A little bit of anatomy on the spike protein, it's a longer protein, over 1200 amino acids. There's an N-terminal domain, there's the receptor-binding domain, which we'll really dig into during this talk. This is the actual domain that articulates and binds to ACE2, and this is where the... The majority of the blocking antibodies that are protective after vaccination target this receptor-binding domain portion of the spike. There's a fusion peptide portion of the protein which is involved in fusion after cleavage of S1 and S2 portions of the spike. And of course, there's the heptad repeat 1 and 2, which are involved in trimerization. There's also the green portion that's the transmembrane domain in the C-terminus. The major focus of this talk is on the receptor-binding domain. As you can see here, as we've seen variants of concern evolve over time, these variants show mutations in the receptor-binding domain portion of the spike protein. This can lead to a higher virulence, more severe infections, and it can escape vaccination protection due to these mutations, where the antibodies that are raised against the canonical spike may not recognize these mutated versions of spike as this virus evolves, and that's the focus of our talk today.

32 Different RBD Proteins

0:10:21.9 S2: At Bio-Techne, we have made over 32 different RBD proteins that encompass amino acids 319 through 541. They all have a C-terminal His-tag. We've used transient expression and our HEK 293 platform to make all of these proteins presented in this study today. In addition, we use pretty straightforward nickel column and sizing column purification strategies to purify all of these proteins in a similar fashion. One thing to note, when we made the recombinant receptor-binding domain portion of spike proteins, we tested and looked into glycosylation, using various host systems to produce these proteins. The spike proteins have 22 N and O-glycosylation sites on them, and as you can see in the upper right, published by Zhao et al and Sel in 2020, the white portion of the protein is the spike, but all the color blue, and purple, and green portions are actually sugars on the surface of the protein. I believe strongly that the vaccination in humans is successful partially because you're putting mRNA, for example, with a Pfizer and Moderna vaccines into the human body and you're getting humanized glycosylation patterns on the immunogen, if you will, resulting in relevant antibodies that block the interaction with ACE2.

0:11:54.4 S2: We found in our own internal studies that when we probe with fluorescent sugar, sialic acid and fucose, along with our glycobiology-related enzymes, we see that the RBD His proteins made in HEK 293 cells are the most mature, they're more humanized, they have terminal sialylated N-glycans and sialylated core 2 O-glycans, whereas RBD His proteins made in show or insect cells are slightly more immature and may not mimic what the true spike looks like in a human. So we believe that the human system is the preferred system for making spike proteins due to their more complex glycans, as I said and just demonstrated in this application note that we have on our website. Please inquire if you want specific information about that, and I can send it to you.

0:12:45.6 S2: Now, on to the evolution of the spike protein. Of course, the vaccinations that have been put into everyone's arms across the globe are the canonical spike protein, but over time, as this virus jump from animals to humans, it's evolving rapidly, as expected, and we've had four main variants of concern that have really been newsworthy. It all started with the UK variant that notably had the N501Y mutation in the RBD portion of the protein. This has been shown to have increased binding affinity to the ACE2 protein, and we've seen similar things in our internal studies at R&D Systems. Shortly after this, the South African as well as the Brazilian variant came out, and these also both had the N501Y mutation in the RBD portion of the spike, but the South African, we see the K417N and the E484K mutation that may involve increased binding affinity to ACE2 and evasion of blocking antibodies elicited by vaccination.

Brazilian Variant

0:13:55.5 S2: When we look at the Brazilian variant, it's very similar to the South African variant, but the K417N in the South African variant is replaced by the K417T in the Brazilian. Again, more antibody evasion, and slightly a better binding affinity to ACE2. Then came along the India, or Delta variant, that is now 99% of the genotype cases in the United States. This clearly has increased fitness over the other variants of concern, and in the receptor-binding domain portion of this protein, we see the T478K and the L452R mutations that are hallmarks of this India Delta variant. These, again, are involved in greater virulence, slightly different conformations of the spike, as well as antibody evasion, and more breakthrough cases, as we're seeing in the news, and so people that are vaccinated with the canonical vaccine, which we all have been, are not seeing the same level of protection that they'd see against the canonical virus comparative to the Delta.

New Variant Omicron

0:15:04.6 S2: I wanted to bring up one more variant that is definitely newsworthy. Over the Thanksgiving holiday in the United States, there's many news stories on the Omicron virus, the variant of this coronavirus, and we are working hard at Bio-Techne to make the RBD, the ECD, the S1, S2 versions of the spike protein as quickly as possible, but as you can see from the top of the slide, there are over 15 mutations alone in the RBD portion of the molecule and there's 30 mutations and seven deletion insertions in the longer spike extracellular domain portion of the protein. So what's going on here? There has been stories in the news suggesting that there is potentially a recombination event in patients that are infected with COVID-19, and the common cold coronavirus 229E, this came out... This is a not peer-reviewed article out of entrance, out of Cambridge, Massachusetts, where there is this hallmark insertion, three amino acid insertion, this EPE insertion, that's a hallmark of the spike in 229E common cold coronavirus, and it's thought that potentially, this is arising from a recombination during replication in a person that was infected with both viruses. Interesting to speculate about. Clearly, there's desire for this reagent, and we're working hard to make it.

0:16:39.3 S2: In some of these articles, they're suggesting that there could be greater transmissibility, but lower virulence with this new Omicron spike and variant, but we're still... The jury is still out on this, and much investigation is in place right now to try to come to the mechanisms, and the virulence, and the severity of the Omicron spike. At Bio-Techne, if you're interested in the Omicron spike protein variants, these are all of the variants that we're making currently, you can go to our website, put your name and information into our website, and we'll notify you as soon as we have these ready to ship to you, so please do that if you're interested in Omicron spike variant and nucleocapsid variant proteins.

0:17:26.9 S2: In the study, these are the proteins that we made at Bio-Techne. These are all receptor binding, the main portions of spike. On top, you can see the canonical sequence, the UK variants, we have the E484K variant, found in Alpha, Beta, Gamma variants, we have the Beta variant, we have the Gamma variant, we have the Delta Variant, Delta-plus, Kappa, Lambda, Mu, all of the major variants that were newsworthy. We have receptor-binding domain proteins that encode those variants. We also have many single-point mutations in different combinations of point mutations. In addition, at the bottom of the list, we have the original SARS RBD spike, we have MERS RBD spike, we have three common cold RBD proteins, as negative controls, and we even have a version of a coronavirus that infects bats and the spike RBD for that as another control. Also note that we have custom protein libraries available upon request. Please contact me, and we can put together whatever proteins you want in a plate format, or in a tube format, whatever fits your needs going forward.

0:18:46.6 S2: In addition to the 32 RBD His proteins that are used in the study, here are some other reagents from Bio-Techne that were used in the study. We used an ACE2-Fc protein, that was actually cleaved with Factor Xa to remove the Fc portion of the protein to have an untagged ACE2, and that is used in the study for looking at binding to the various RBD variants. We also provided an anti-His antibody to immobilize the RBD His-tag proteins. And finally, we also used and Noah will present this blocking lambda body antibody that beautifully blocks spike proteins from binding to ACE2 and some data below shows flow cytometry assays showing the beautiful blocking of this reagent in the context of human embryonic kidney cells expressing full-length ACE2. On the top right, you'd also can see the ACE2 used in the study binds very nicely to spike RBD proteins in an ELISA assay, which is showcased on that product insert. That was my last slide. I guess we can move to the final slide regarding questions. If you have any questions for me, please go ahead and ask them, and then I guess we can turn it over to Noah.

Introduction of Noah Ditto

0:20:16.1 Speaker 3: Great, thank you Anthony for that terrific introduction to the proteins used in these studies. So hi everybody, my name is Noah Ditto. I'm gonna be talking about the LSA assay that we developed using these terrific Bio-Techne proteins, first kind of starting off with how the assay was designed and some observations from the assay, and then moving on to the application of the assay itself using a set of sera. So at a high level, the SARS-CoV-2 RBD variant assay looks like kind of on the left-hand panel here, an array of 32 variants that Tony highlighted a second ago, RBD, His proteins captured on a surface in triplicate. So we effectively put down 96 proteins simultaneously onto the chip surface, these are captured on to an anti-His antibody surface, which I'll detail in a second.

Introduction of Serum Across the Variant Array

0:21:12.3 Speaker 3: The next step in the cycle is to introduce serum across the variant array, and we measure a binding response from that serum. In the same cycle, we immediately then come in with the ACE2 receptor to assess the potential for blockade in this format. And then finally finish up with anti-IgM, A and G isotyping injections to get an understanding of where their need response is coming from from an immunoglobulin perspective. A few interesting observations from setting up and working with the reagents as we're developing the assay, one overall, is that all these RBD variants provided by Bio-Techne, all were very active in our hands, very well behaved, really encouraging because oftentimes in assay development, there's a lot of upfront work to find reagents that do work and work well on the assay and behave similarly. In this case, I would say that all these very much were plug and play. We ran 32 variants in this particular assay format that I'll showcase, but in reality, Bio-Techne had much more than that, so we had the luxury of picking and choosing maybe the most interesting variants and mutants to put into the assay.

0:22:28.1 S3: And it was really just remarkable how well they all performed in terms of activity. This particular slide highlights ACE2 binding. So in the process of assay development, we were obviously arraying the variants and then looking how ACE2 binded, in this case, since the LSA measures real-time binding interactions, we were able to quickly screen for an off-rate kind of profile from all the variants. You can see in the left-hand of the chart that we had no binding to many of our controls, and then when we look across, we have interestingly some enhancements and binding or off-rate as it's measured for a number of the mutations. Which was very interesting in our hands, but overall, from our perspective, the big picture is, we wanna make sure that we're getting good uniform activity across the array, which we did. Then kind of looking at the variants at another angle, we use Bio-Techne protein antibody CR3022 to look for just reactivity, just to get a sense of whether we have strong antibody binding is in conjunction with the ACE2 binding, and in fact, we do. We see that across the board, the measured off-rates are actually at the limit of detection. They're very high affinity, which is documented for this antibody, one E to the minus five, you can see here is the sort of maximum off-rate that we're measuring in this assay.

CR3022

0:23:56.3 S3: So across the board, that largely holds true and interestingly we find exceptions to this where certain mutations do seem to be decreasing the off-rate or effectively reducing the strength of binding for some of these reagents. And again, as a control, we see no binding to certain variants which the CR3022 is not expected to be reactive towards. Kind of looking at the overall ability the assay to measure circulating levels of immune response, we went ahead into the spike recovery using the CR0322 antibody, and it looks like at about down to 50 nanograms per mil is where the limit of detection works in this assay. And this is looking at it as the antibody spiked into serum, you can see kind of on the left-hand most panel our dilute serum at 1-100, and we're about two-fold above that signal, at 50 nanograms per mil. So that appears to be where we're feeling confident in the assay, at least using this particular antibody as a surrogate where we could judge limit of detection to be.

0:25:11.5 S3: And then also, as Anthony alluded to, we use this antibody LMAB10541 as a particularly potent inhibitor of ACE2 to assess antibody blocking, and in that case, it appears that we get nearly 100% blocking somewhere in the one microgram per mil range. So in the first slide, I just mentioned that we typically would say that we can detect antibodies down to 50 nanogram per mil, and it looks like around one microgram per mil, maybe slightly below that we would expect to see near complete inhibition in this particular assay format. So kind of moving on to the exercise of actually putting serum into this assay and understanding actual real world responses using vaccinee serum. So a little bit about the subjects used in this, there were six subjects in total. They were all dosed two times with the mRNA vaccine, the Moderna vaccine, 1273. The first dose occurred at zero weeks and the second dose occurred at four weeks, and there was serum drawn at zero, two, six, 12 and 16 weeks total. So for each of the six individuals we have five time points, some details on the individuals shown here, nothing unusual about these individuals, at least in terms of reported comorbidities, and overall their BMIs fall under a similar range, generally, age-wise have a fairly broad range from 26 up to 58 years of age.

MAB050 Antibody

0:26:47.4 S3: And I'll go on to the next slide here, which is our experimental details, so in the assay set up, we ran the experiments at 25 degrees, our chip surface was Carterra's HC30M linear polycarboxylate chip surface, which we then attached the amine coupling and anti-His antibody to. In this case, the anti-His antibody used was MAB050, which was that 50 microgram per mil, so fairly standard conditions that we use in the LSA to make a capture surface. And on the side of immunity monitoring, we then changed over our assay buffer from a low pH coupling buffer to a neutral pH physiological buffer, HPSTE, for the ACE2 binding studies as Anthony highlighted, we used the ACE2 with the Fc cleaved at 30 nanomolar final, and then we used isotyping sera to assess the isotype profiles across.

0:27:50.4 S3: And at the end of each one of these binding cycles, the surface was regenerated with 10 millimolar glycine pH 2.0. And this regeneration effectively removed everything down to the capture antibody, which is MAB050. So in each cycle of this assay, when we're exposing it to a new subject serum or new subject time point of serum, we have a fresh array of RBD His-type proteins on the surface, because we do regenerate them and then rebuild the array in each cycle. So this avoids some of the complications of trying to find optimal regeneration conditions for each of the individual variants since we simply strip the entire surface and put down fresh protein in each cycle. So this is just a quick sensogram view of the variants being captured on the surface, they were only prepared at one microgram per mil, we did about a four-minute contact time, and typically the ranges during the assay were about 200 to 500 RU on this chip surface, so we tended to keep the capture fairly low because that just generally improves data quality, obviously conserves reagent and overall helps out with some of the downstream steps including the competition portions of this assay where we're looking for ACE2 in the patient.

ACE2 and anti-IgM

0:29:12.2 S3: So if we kind of step back, this is, you can imagine looking at one individual's time course of serum across the entire set of experiments. For every individual, we initially ran a buffer sort of series where instead of injecting serum we just injected buffer in place of serum, and followed all the injection routines of ACE2 and anti-IgM, IgA and IgG. So we did initial upfront buffer injection, then we started with serum at zero, two, 16 or six, 12 and 16 weeks. And even if you kind of step back and view this data, it's pretty clear that as the weeks progress here we see clear increases in signals, and I'll go into detail in a second on what those particular signals are, but this is the heart of what we're tracking for each one of these zero time points, there's 96 unique locations on the surface that have captured RBD His protein, and we're looking for reactivity of these various reagents against those surfaces over time. And drilling down on the particular assay cycle, so we start off with a 1 to 50 injection of serum across the surface, so we see that binding onto the surface, then we come in with ACE2 to see whether or not this immune response can change the binding profile of ACE2.

0:30:33.6 S3: And how we do this determination of ACE2 inhibition is because in the previous slide I had mentioned, we inject buffer first kind of as a control, so that gives us our baseline to understand how much ACE2 would bind without the presence of serum, and we use that to determine percent inhibition which I'll show in a few more slides. We then come in with our isotyping reagents, so IgM, A and G, and in this particular cycle that I'm highlighting here, you can see a strong anti-IgG response indicating that this individual had lots of proteins, immunoglobulins that targeted RBD proteins and were IgG in nature. And the data I'll show subsequently was characterized by taking the response at the end of each injection kind of shown here in the blue arrows and using that to make our inferences. So we'll just dive right into the data and get down to looking at each individual variant here, so this is subject one serum reactivity, so we're effectively looking when this individual serum is injected on the surface, what change in signal do we see against these RBD variants and controls? 

SPR Response Units

0:31:46.2 S3: So we can see kind of in the bottom here, and I should mention too before I start, the scale here is response unit, so this is the traditional output in SPR response units, and we start off in this lower range where these blue dots are indicating that all the variants at time zero or week zero, excuse me, effectively have no reactivity with the serum, so this individual appears naive to RBD variants derived from SARS-CoV2 and even the controls used as well, common cold, there's minimal reactivity across all this as well for example, but over time, we can see that the signals do increase and interestingly, this particular individual, subject one, at week two reached their maximal signal for all RBD variants, so across the board, they show reactivity against each variant. There's slight differences here that are somewhat indicative of the starting levels, probably the variants on the surface, but proportionally, we see the same shift across the board and past the week two time point, which is the maximal time point here of reactivity, we see a decrease over time, owing kind of to the changing nature of the immune response, but definitely this individual had a robust signal, peaked early and dropped off.

0:33:03.1 S3: And the great part to see is several of the controls that Anthony highlighted earlier on are showing largely unchanged behaviors here, showing that this is a very specific response as measured in the assay. In contrast though, I highlight here subject two serum reactivity, so this individual, although given the same vaccine and collected at very similar time points, fails to show much reactivity against RBD variants in the assay which is very curious, but clearly across the board I'm using the same scale here I had in the previous slide, you can see just that there's really minimal change over the background, the week zero to week 16, even there's really nothing signal-wise, which occurs here, so this individual, we will generally classify as having little to no RBD variant reactivity, and despite getting two doses of the vaccine, so there's obviously the question of whether or not maybe the first dose was somehow mis-administered or something but clearly, even with the second dose that failed to elicit a robust immune response against these RBD variants.

0:34:15.1 S3: And if we tend to take... For the sake of today, I don't have time to go through every individual's detailed responses but if we just kind of look at the overall reactivity trends, so this is kind of condensing down the RBD variant responses into an average signal, obviously excluding the negative controls from this averaging exercise, we can see that clearly the responses rise for many of the individuals from that first dose, particularly subject six and subject one, we have an increase, strong increase while others are fairly flat in their response and they take longer to reach the maximal serum reactivity. Ultimately, many of them end up around a similar range as the immune response starts to wane at week 16, but fascinating how these individuals have differences and we're monitoring this both in this very high level scale of overall average signals and like I shown in the previous slides, we can individually confirm that each variant is seeing a similar level of reactivity. So switching gears a bit and then moving on to the next portion of the cycle or assay, is the ACE2 blockade exercise. So in this case, the surface of RBD variants has already been exposed to the individual serum, and now we're looking for whether an injection of ACE2 going across that surface is able to still bind relative to the control situation where there was no serum.

0:35:44.4 S3: And in this case, this is subject one, we had seen previously that subject one had what we would call a robust response in terms of serum reactivity and it corresponds that this individual had strong blockade, getting near 100% for many of the variants, at least about 60 in some cases. So there's some interesting differences here that we don't always see complete inhibition but we do uniformly approach it, and it seems to be robust starting at week two and 16 onward, it sort of maintains a high level of inhibition across the board, so very interesting here that this individual both has a strong response and in fact ACE2 binding to the RBD variants is largely disrupted. In contrast, we have our subject two who had a very kind of lackluster immune response that we can classify it as, we do see some inhibition, but it's definitely on the order of maybe 40% inhibition or something, so definitely not the true level of inhibition, so overall, corresponding to the serum reactivity, we can infer that this individual did have some level of response to the vaccine, but it was not very robust and overall the levels here probably suggest that they don't have as high of circulating immunoglobulins to inhibit ACE2 binding as subject one did for example.

0:37:09.9 S3: And an interesting contrast possibly is subject four, where we have a strong ACE2 blockade, which is similar in magnitude or percentage to the subject one, but it doesn't come on as fast as subject one, so here we're... I'm highlighting here, the week two time point, which this is after two weeks post-vaccination with the first dose, this individual does really have no measurable change in the ACE2 blockade, but it comes on later, and by weeks six, 12 and 16, we do see a stronger response. So just curiously how the immune response in this particular case was lagging versus subject one for example. And kind of doing that same exercise of condensing down the RBD variant binding across the board into some major trends, we see that there are differences in overall ACE2 blockade amongst the individuals as we go from weeks two, six, 12 and 16. We see that early on for example subject four, like I just highlighted, was quite lackluster in the ability to block ACE2 but then immune response did come on strong in the subsequent later time points and actually was one of the maximum responses seen in the assay.

0:38:32.0 S3: While there's others, for instance this Subject 5, which were low in the beginning, that really did not have substantial increase in responsiveness, so again, overall just a fascinating behavior that we're monitoring both at a high level and also down to the variant level, understanding all the collective behaviors with these individuals. And if we plot this in terms of reactivity, serum reactivity, which is response units versus ACE2 inhibition, which is a percentage of inhibition over the control, we see that there is a distribution, so each one of these individual spots is a RBD variant for that individual. So we're looking at the overall trends but we do have the granularity of each variant plotted out here. And what's really interesting, if we just take two of these individuals, Subject 6 and Subject 1, is that the binding response, the magnitude of it is fairly comparable in terms of our response units in the assay.

0:39:35.3 S3: But when we look at the ACE2 level of inhibition, it's quite different between these two individuals, where for Subject 6, for example, we're really not getting much above 50% inhibition across the assay despite there being an apparently high serum response. While on the other hand, Subject 1 who's relatively comparable to Subject 6 in terms of the serum binding response, actually has a much higher percent inhibition across the board against all the variants. So it goes to show that it's simply not just the level of serum and activity but there's other factors at play that seem to be disruptive. So it's not an absolute mass at the surface but really how that mass, which is ultimately an immunoglobulin targeting epitope, plays out that really translates into what we would perceive is a therapeutically relevant outcome being ACE2 blockade in this particular case.

Isotyping Responses

0:40:30.7 S3: And now, I'll take some time to go through the isotyping responses, the last half of the assay, which is looking at anti-IgM, IgA and IgG binding to the surface to understand what RBD immunoglobulins do we see. So this is IgM shown here. It's just Subject 1. Generally, the responses here were quite low. Overall, Subject 1 was the only individual where there was something, we would say, about 2X above background. In this case, this came on at about Week 2 as we would expect earlier with post-vaccination, and it seemed to be present for most of the variants. Interestingly, some of them seemed to be a little less robust, so I think there's some maybe interesting trends here that probably require some additional statistical power behind them to make really confident inferences, but we do see differences among the variants, which is somewhat curious and I would say, requires some follow-up to determine more. But major takeaway is really, IgM responses were not that high in this assay across the board; just generally, Subject 1 being the only one that really stood out.

Anti-IgA

0:41:43.9 S3: Anti-IgA, we see that there is really minimal responses. This maybe isn't expected since IgA is sometimes less prominent of an immunoglobulin to detect post-vaccination. And for example here, Subject 1, we just see a slight increase at Week 2 in the same time frame as the IgM was detected but very low and really, really just barely above the background. So for most statistical perspective, it would be difficult to infer a lot at this point from this individual. Interestingly, Subject 6 did have a fairly strong response above background showing up at Week 2 over that. So there is some value in measuring A, it just does not seem to be a highly robust marker, but I think this is well-characterized in the literature that it's not typically a primary marker in many cases, but easy enough in this assay to incorporate a test for it.

0:42:42.1 S3: And then maybe the IgG response, which we would expect to be quite robust, is in fact so. This is Subject 1 and we can see, similar to the serum reactivity in the ACE2 blockade patterns for this individual, strong IgG responses across the board that initiated around Week 2 and remained that way through Week 16 with the controls in this case, showing minimal to no changes in some cases across the board. Showing that there is a specific IgG response and not maybe just bulk IgG interacting with the surface, for example. And that was an example of a high anti-RBD interaction. This is an example here, in Subject 3, of a low anti-RBD IgG response. Just quickly looking at this chart, which has a similar scale to the previous slide, we see about a fifth of the signal seen for Subject 1 in this individual. So really, a minimal response that was reflected in the overall trends in serum reactivity previously that you just didn't see this individual produce much of a meaningful response. And through Week 16, despite getting two doses of the vaccine, they fail to elicit much of a change.

0:44:00.1 S3: So if we did, again, pool all of these individual data points together and look for an overall trend in the data, it seems to match the overall story we're telling, that certain individuals respond strongly. The great news for everyone getting these vaccines is that it looks like, across the board, they're getting good recognition of the variance and there's no variance, at least that we studied in our hands, that had a clear deficiency in reactivity versus the other, so that's very encouraging. And you can see overall that responses peak at about Week 6 for almost all these individuals and then decrease in terms of the IgG trends across the board. So just making some highlights out of the data here, this was just an initial pilot study to understand the assay. You can see that early on in the assay characterization, even in itself there, we found tremendous value in understanding off rates and the ability of a SARS-CoV antibody to interact with these RBD variants. So the LSA itself is well-suited to characterize the reagents and also execute the assays I've shown in the latter half of the presentation with serum. And really, a lot of this, the ease of this assay, was attributable to the high quality of the variants used in this assay. As I said, they were all quite active, all captured really well on the surface.

0:45:23.7 S3: What's also great is in the course of developing this assay, because they do capture very seamlessly via the His-tag that they have, you can easily swap them in and out of the assay, and that's what we did. We, in fact, started with a bigger pool and ultimately had to shrink it so that we could get triplicate measures within a 96-block for a particular assay that we designed. And really, the assay is portable to other infectious diseases. We're showcasing it here for SARS-CoV-2 and there's just a wealth of reagents but really, there's nothing fundamentally about this approach that couldn't be translated to another infectious disease, especially if there's the availability of such high quality proteins like shown here.

0:46:07.3 S3: The assay described here was done against 32 variants and triplicates of 96 total measurements but in reality, the LSA has the ability to screen up to 384. The array capacity is 384, so it definitely can do more than 32 and triplicate, and that's an option as well to really maximize the amount of data per serum injection. And that's really emphasized. The replicates and the controls really added confidence to the data, particularly the negative controls used in this captured equivalently to the active RBD variants but we could see clear delineation where the serum reactivity was targeted towards SARS-CoV-2 and not SARS-CoV-1 or the common cold RBD variants or RBD proteins used in this assay. So I think that's a tremendous point to highlight as well, that having these controls and be able to put them into the same assay really boosts the understanding of the response and doesn't leave questions lingering of whether or not it was specific or whether or not the response may have been broad and covered other variants, for example.

0:47:15.3 S3: And lastly, it's really worth noting that we're just using microliters of serum in this assay, but there's 480 measurements that come back and in real time, so it's a huge amount of data that really I think is tremendously powerful and gives a lot more information than some other assays do in terms of what we're measuring from the serum in this case. And I didn't even get into it for the sake of time today but since the LSA is a real-time measurement device, you have the opportunity to measure the off rate of serum as well. There's some interesting manuscripts that have emerged recently this past year showing that looking at the off rate, for example, of serum reactivity as strength of binding, as it's called, can be one way to understand the immune response. And that's a whole another facet maybe not explored here but maybe explaining some of the underlying trends of reactivity we see where we're looking for growth binding but possibly there's something more to it in terms of maybe the changes, the affinity of the polyclonal response against the RBD also is a factor and can provide more detail as well.

0:48:22.2 S3: And again, that's all put into this because it's real time in nature, so I just wanted to highlight that as one more piece that comes out of this type of workflow. So with that, I definitely want to thank everyone for attending.