Hello, everyone. Good morning or good afternoon, depending on which part of the world you’ve logged in from. Welcome to this webinar hosted by Apple instruments and thank you for joining us. My name is Aditya, but I am the VP of technology at Apple and I’m based out of Northern Virginia in the US. I’m also delighted to be joined by Adam Bergen, who is with the National Institute of bioprocessing research and training, or, as it is commonly called, has ninebot in Ireland. Welcome to the webinar. Adam, it’s great to be presenting with you. Moreover, we’ll also have Professor Mike butler, who will join us for the Q&A session right after our presentation. As most of you would know, Mike has years of experience with dielectric spectroscopy. So be sure to send in all of your questions for Mike. Together, we’ll be addressing the topic of dielectric spectroscopy and how frequently scanning could provide even more information about your sales in bioprocess applications. Before we begin, though, just a few reminders. The presentation should last about 30 minutes and we’ll keep the final 10 minutes for your questions. And that’s when Mike will join us. The platform is meant to automatically mute listeners, but just as a precaution, please unmute yourselves anyway. We hope to get a lot of interesting questions after the presentation, so I’d certainly encourage you to go ahead and type them in the box provided on the platform. We’ll try and address as many questions as possible, but if your question is not addressed because of time constraints, we’ll be sure to respond to your question via email. And the only thing left to do really and the most important thing is for you to sit back and enjoy the webinar, right? So you have the overview, the structure for the presentation for today. I’ll open the presentation with a brief introduction to our instruments. We then go through the technology how it actually works, which can be quite interesting for those of you who are unfamiliar with the concept of capacitance measurement. I also want to quickly demonstrate how some of our customers have used capacitance technology in various applications thus far. Following this, I’ll introduce the concept of frequency scanning as well and how one can derive additional parameters from the scan. I’ll describe some applications where frequency scanning has been used successfully, and then I’ll hand over the presentation to Adam, who will introduce the ninebot introduce neighbor and the team working there. He’ll then continue to talk about the benefits of frequency scanning Adam will talk about different cell death pathways. More specifically, it focuses on autophagy and how frequency scanning could relate to this as well Adam will then move on to discuss the frequency sweep in a little more detail and its application, along with the use of media analysis with Apple technology. And finally, Adam will conclude the presentation and bring it all together. And when our presentation is complete, as I mentioned earlier, we move to the Q&A session with Adam and mike, right? So let’s get started with a very quick introduction to ABA. The Apple headquarters are based in Wales in the United Kingdom. The company is situated in a town called aberystwyth, and in fact, that’s where the name of our instruments comes from. And this is where the technology was actually invented back in 1988. In 2017, we opened up a subsidiary in the u.s., which is where I am based in Alexandria, Virginia, on the East Coast Abbott instruments, as I mentioned on the last slide, invented the use of dielectric spectroscopy to measure biomass in bio processes and the technologies sometimes commonly simply referred to as capacitance measurement. And this has been our primary focus for the past 30 years or so. We have customers in over 140 countries, and most of them most of the major biopharmaceutical companies today and all of the major brewers, interestingly, use our biomass solutions. Now, for any technology, the transition from R&D and process development through to manufacturing is quite an important one. As we speak, capacitance is considered as a mature technology, with weather solutions being used in manufacturing and production environments globally. Because of the wide range of solutions from aba, the equipment can be used in small glass bioreactors through to large stainless steel vessels and are single use capacitance solutions are also available via our collaborations with sartorius and thermo. That’s making it a truly scalable technology. Right? so let’s jump to the interesting bit. How does the technology actually work and what is the principle behind the measurement now? Simply put, above measures two parameters we measure. Capacitance and conductivity capacitance in very simple terms is the ability of any substance to hold charge around itself, whereas conductivity is the ability of any substance to give a weight charge. Now why these parameters are relevant when measuring cell density. The answer lies in the property of a typical cell membrane. The cell membrane, as most of you would know, is impermeable with respect to charge. This means that it doesn’t allow the free flow of ions across it and therefore under the influence of an electric field. And bear in mind, this is a tiny, infinitesimal feel that the ab approach generates. The cells with intact membranes can hold charge it on themselves and hence can act as capacitors. So if you were to imagine a cell suspension across which an electric field is introduced using a positive and negative electrode, there’ll be a build up of charge around live cells and hence, each tiny cell would act as a capacitor. Therefore, the more the concentration of live cells, the more the capacitance measured across the suspension. Now, in stark contrast, dead cells, which more often than not have leaky membranes with respect to judge these are not they do not have the ability to act as capacitors or in other words, they cannot hold charge around themselves and hence these are not measured by the average technology. For the same reason, solid media particles gas bubbles are not measured directly since they cannot contribute to capacitance. Likewise, the technology also works quite well with micro carrier cultures because carriers, much like dead cells, do not store charge around themselves and hence cannot contribute to this capacitance. It’s the same story with fixed bed bioreactors, and that’s why you see capacitance being used quite regularly in bioreactors such as the body cells series. So what you end up measuring is purely the viable by volume of the suspension. Now, the challenges associated with offline measurements are quite well known, but with a technology like ab, you are measuring your biomass in real time in situ. You don’t need to take a sample for the measurements. You’re not really exposing your bioreactor to contamination risks. The other measurement can be performed once every few seconds. So there’s a lot of detail in the capacitance profile, which will be evident with some of the graphs that I’ll share with you later on. This allows for timely intervention when required. So essentially, not only can you monitor the cell density over time, but the technology will also allow you to control something critical during the process. Based on the real time profile. But the bottom line is that overall you end up with a complete and detailed knowledge of your critical process parameter. So just to recap very quickly, what we’ve said is that AP technology only measures like cells since they can hold charge it on themselves, and hence they measured that cells with leaky membranes are not measured. And because we measure the buildup of charge around cell membranes, the capacitance is directly proportional to the total membrane bound volume of the cells. So it correlates not just with the number of cells, but also the size of the cells. And this is quite an important concept to remember going forward, and I’ll just show you an example for that. There’s a lot of applications whenever capacitance correlates very well with cell density, especially for processes where cell size changes are not significant. But here I’m going to show you an example where the ability of abour to measure by volume is utilized for the benefit of understanding a process better. So here you see a case where was used to monitor sf9 cells insect cells infected with a particular virus. So the black continuous strand is the relative permeability or capacitance. The trend with the circles is the offline cell density, whereas the one with the diamonds is the offline cell size. So if you look closely capacitance and off-line cell density correlated quite well with each other as the insect cells grew early on here. But up until the infection point, though, because right after the infection, you can clearly see that the offline cell density plateaued, whereas capacitance kept increasing sharply at this time. You see, that cell size was seen to increase as well. Now, this is clearly a case where capacitance was responding to cell size changes, along with changes to the electrical properties of the cell membrane. Now, this is quite valuable because the capacitance profile and its increased post-infection can be used as a parameter to determine successful infection. And as you keep moving along the capacitance change in real time and because of the detail in the profile, you can also determine peak viral infection point and when the cells start rising upon release of the protein. In most cases, you simply plot about capacitance at a fixed frequency of measurement to get the growth profile of your culture. That’s that’s what the research group working on this process did as well. So overall, in terms of the applications, I often see that the most basic use of the technologies to monitor your cell concentration over time, which then gives you a detailed fingerprint of the process in real time as well. Now, this fingerprint has been used for troubleshooting purposes. When the growth profile of your culture or indeed a fermentation process deviates from, let’s say, the golden fingerprint of the process. The golden capacitance fingerprint, if you will put capacitance measurement, brings additional value when you can identify and control something critical automatically in the process. Now, this could be controlling the cell concentration at a specific value for process intensification or continuous perfusion processes, or it could be controlling the nutrient feed based on capacitance as well. For example, Biogen published a paper where they’ve demonstrated how they control the complex nutrient feed in real time based on raw capacitance. They don’t really convert capacitance to off-line accounts for this. In fact, Biogen has published about four people over the past four or five years, and it’s all about using capacitance in manufacturing for applications like C transfer troubleshooting and for controlling the feed of their complex nutrients. I also mentioned about how capacitance works quite well with micro carrier cultures and fixed bed bioreactors. It has also been used to determine success of a scale up process since it can be used as a key performance indicator across different skills. So before we move to frequency scanning, I’d like you. I’d like to take you through some classic examples of using capacitance at a single or dual frequency over time to get the growth profile of different cultures of fermentation processes. So on the left, you have an example of batch fingerprinting. You see four identical real time strands of capacitance for the same streptomycin fermentation run one after the other. Now this enables you to develop your process fingerprint, and as I mentioned earlier, you can use this fingerprint in different ways. For instance, if your current process changes seem to deviate from your golden capacitance fingerprint, this would be a call to troubleshoot the process to either stop it or recover it based on the cause of the deviation. And you remember that you get this information in real time as well, which significantly improves opportunities to troubleshoot the process. The drop in the middle right here is an example of a Fed beach culture with a zero cells where Abu was used to track their growth. The dark blue line is raw capacitance in measured pico fighters centimeters. And that’s the unit of measurement for capacitance. By the way, the light blue line is capacitance calibrated to read in cells per milliliter. Using a simple calibration factor or multiplication factor, if you will, and the pink squares represent the off-line typekit blue zones. Now, this is a typical example of how a simple frequency capacitance measurement can use to can be used to track the entire growth profile of your profile of your culture. The device that you see at regular intervals, of course, represent the dilutions because of nutrient conditions, which is quite correctly picked up by capacitance. The graph on the very right shows an example where above capacitance was used to determine food timing. Now this application has been implemented at cultivar previously known as dow-, based solely on the real time capacitance profile yet again. So the optimum time to feed the cells was established based on the real time capacitance profile. And when this was done, the new food strategy impacted the growth of cells favorably, as you can see with the red capacitance chain, and it also increased the product titer by 15% I just do a couple more examples really on the left, you have a graph from Biogen from a Biogen publication, in fact, where scalability with the other sensors was demonstrated from five liters to 15,000 liters in the middle, you have about capacitance being used to monitor the growth of baby hamster kidney cells and macro micro carriers. As we discussed during the principle of the technology, Abbott isn’t. The Abbott technology can be an ideal candidate to measure processes with micro carriers because most micro carriers do not have the ability to hold charge around themselves and on the right. We have a perfusion application where Abbott capacitance was used as a part of an automated feed and bleed strategy to control the cell concentration at a fixed value for a prolonged period of time. And the final example that I want to show you was of the other single use new sensor that is pre-installed in the thermal subs. We recently launched this sensor. Just a few months ago, Biogen actually tested this solution in the 50 liter recipes for monitoring and controlling a perfusion culture. And this study was performed to compare Apple’s reusable and single use offerings. The cells were both in the growth phase for the first nine days, and then a cell bleed was enabled to maintain different, viable cell volume targets. You’ll see the reusable Abbott probe was used to control weather, whereas the single use meal sensor was used for monitoring the culture. And as you can see quite clearly, there’s a near-perfect alignment of the signals that we’ve seen for both the reusable and single use sensors for the duration of the culture. Right? so all the examples that I showed you just now, they were examples where the capacitance was measured at a single frequency or in some cases, at two different frequencies. But we now look at the concept of frequency scanning using other technology. Now, the other system is capable of measuring capacitance at 25 different frequencies in the radio frequency spectrum, so a 50 kilohertz, 20 megahertz. So if one were to blot capacitance against frequency when measuring cell culture, you’d see a graph much like this one. The frequency scanning curve is also called as a beta dispersion curve, by the way. So the typical shape of the curve is such that you see high capacitance plateau at low frequencies and the low capacitance plateau at high frequencies. And these two plateaus are separated by a region where a capacitance gradually decreases as frequency increases. Now this curve, the beta dispersion curve is described by the cool cool equation, which again will present in a minute the adult frequency scanning software, which is called futura. Escada can fit the measured scan onto a cool, cool model following which full additional parameters can be determined. So here’s the four parameters. We’ve got delta capacitance, critical frequency, cool, cool alpha and capacitance infinity delta capacitance position. Essentially, the difference between the high and low capacitance plateaus C infinity represents the capacitance of cell free medium FC is the characteristic frequency. And is essentially the frequency at which the dispersion is half complete. And alpha is nothing but the level of steepness of the curve, if you will, at FC. Now all of these parameters, especially if see an alpha, can potentially be correlated with other critical events in the process. And moreover, these four parameters can be used in conjunction with other dielectric spectroscopy equations to estimate even more information about yourself. And just to reiterate, with the above software future, SCADA estimates all of these parameters for you automatically in real time as well. Now, several groups have utilized the concept of frequency scanning and the additional parameters that it estimates. A for various applications, and I’ve sort of enumerated just a few of them here very quickly. For instance, Biogen published a paper a couple of years ago where they used the additional frequency scanning parameters to detect apoptosis in their tumor cell culture significantly earlier than their offline method. Not only that, they were also demonstrated how they could reverse the apoptosis based on this early information from the ab probe, whereas they could not achieve reversal with access to just that offline data alone. A research group at North Carolina State University utilized FC and for the distinguished 3D tissue constructs made up of different cell types. Another group was able to distinguish between the cell growth and lipid production phases in instrumentations using FC and delta capacitance. Pleasant, cool, cool modeling have been utilized as well to adjust capacitance, sometimes to match offline cell density, especially in the lead stages of the culture when it is absolutely necessary to do so and in some cases, using the dielectric spectroscopy equations and making an assumption, one can also estimate cell size in real time. So I’d be happy to talk more about these applications if anyone is interested to learn more in the Q&A session. And with that, I’ll handle the presentation over to Adam, who will expand more on frequency scanning and its potential use over to you, Adam. Thanks, addition. And so hello, everybody, and my name is Adam Bergen, and I am a PhD student in the cell technology group here in numbers for anybody who isn’t familiar with neighbors. We are primarily a Training Institute for the pharmaceutical industry and we’re based in Dublin, Ireland. But we also have a number of research groups in house in nivas, also with a variety of different research topics to research into CTG group in particular, is focused on monitoring of cell cultures and also global engineering. For my PhD in particular, is focused on the identification of changes in cellular health prior to the typical loss of membrane integrity, which is usually studied, want to understood to be an indicator of viability in moderate bio processes. So we’re under the direction of Professor Michael butler, who will be joining us for the Q&A after. So like adisa said, be sure to send in any interesting questions that you may have at throughout the webinar. So to get an understanding of why we would be interested in looking at frequency suites, so the mechanisms of cell that are very well understood and there’s a variety of different methods you can use to analyze these cell death pathways and apoptosis would be the primary cell death pathway that’s analyzed. However, there’s another mechanism more so related to cell survival. Both is very interesting as well, because it’s usually a precursor to cell death is autophagy. Apoptosis is usually a response to cell stress in the cells. However, autophagy is an act of cannibalism of the cells where they try to scavenge nutrients due to a lack of nutrient availability in the culture. So why are we talking about this as an interesting topic for our impedance measurements? Well, let’s think about the dielectric signals that we’re generating. If we have changes in the cells due to membrane changes or particularly intracellular conductivity changes, we may be able to detect those changes using frequency sweep data from the likes of an probe. And this is all represented by a change in the cell distribution over time. And that’s why, particularly when we move towards phases where the health of the cell is beginning to be lost, can we identify any earlier indicators around these changes in the cellular processes prior to when we use membrane integrity? And that’s really a focus of the work that we’re trying to do in our lab. So as part of this webinar, I’m hoping that we’ll be able to deliver the potential in these systems for analyzing these parameters. So as Ed introduced here is the beta dispersion graph that was shown earlier and why I just want to look at this again is I want you to understand what it is about the cells that changes a particular point in this graph. So I start low frequency plateau. What we get is we get a full depolarization of the membrane and you get a build up of charge and that’s typically related to either cell size. And then as you move through the sweep, you start to learn information on membrane properties of the cell. However, when you move to very high frequencies, what happens is we don’t get any depolarization of the membrane. And what we end up interrogating is more interested or aspects of the cell, including the cytoplasm. And as I said, if we get the likes of leaky membranes and the cell. And we get a change in that content in the cell, we may be able to detect that using our frequency suites. So let’s have a look at a typical growth curve that we produce in our cell line that we have in our lab, in particular, media also. So what we find is that we get very good correlation between our single frequency readings with the ever and triple blue counts up to about our 60. However, beyond this point, we get a loss in that relationship, so that becomes an area of interest for us in the lab. We want to try and understand, is there a change in the cell properties that’s causing disaffection in our single frequency readings to occur? And these readings are taking at 0.58 megahertz, which is a typical frequency you would use for analyzing mammalian cells in it by a process. So now let’s move on to the multi frequency scams that we can generate with the ever what you can see here is a scan of 25 different frequencies. We get data points captured every 30 seconds in these graphs. So what you end up with is a huge bulk of data that we then need to analyze. But just so you can get familiar with what these frequency sweeps look like. Let’s have a look at the changing shape in these curves as we move through to frequency range. So very low frequencies, we can see that we get different plateau and peaks of their curves relative to where we move to higher frequencies. So that becomes of interest to us into that because we want to understand what the change of shape in these curves reflect when it comes to parameters of the cell. So to better understand, it’s very hard to see the changes in the shape of the curve at high frequencies because the magnitude capacitance is quite low. So what we can do is we can just normalize the capacitance to then understand the change in the shape of the curves. And what we find is we get a good distribution of the different frequencies up to that, our 60 20 points. But after that, we can see that there’s a dramatic change in the shape of these curves. Particularly ask those very high frequencies where we can see different points at which we get an inflection of the curve. But not only that, when we move to those high frequencies, we can see that we don’t get any plateau in the curves until hours 120 to our 140 four, which is where we would start seeing cell death. So this is all indicating that. OK, so if we measure one frequency, we might get a certain amount of information. But if we take that information that’s been generated by the Abbott probe, I relate that to our knowledge of how the changing frequencies can affect cell health properties weakened and draw up a more detailed picture. Now, using all this information, I don’t worry, there’s no exam at the end of this webinar, so don’t worry about the complex nature of this equation, but this is the cold call modeling equation that a detailed alluded to earlier in the presentation. And thankfully with the future SCADA software, it can take all the data points that I’ve just mentioned and fit this to debate a dispersion curve. So we can generate those for critical properties, as indeed you mentioned earlier. So this is really nice because we can then generate complex information in real time and use this to potentially interact with our culture to make changes in an already turning point relative to our offline accounts. And this has been shown in literature to be possible, particularly with the likes of the code called alpha value and critical frequency. I’ll talk about this more later in the presentation. So let’s look at our beta dispersion curves, and these are take taken at data points to the culture, and what you can see is we get a change in the shape of our beta dispersion curve with time. So for the first two days of the culture, we get profiles that are quite similar. However, when we move beyond these time points, we can see that change and debate a dispersion curve is evident. Now you could also represent this information as a heat map on the right. It depends on what way you want to view the information, but we can then generate these beta dispersion curves every 30 seconds and identify exact points at which dede’s curves are changing. And like adeed had mentioned also, we’re very keen on generating what we call fingerprints. That is a typical signal that might suggest to if we move beyond that fingerprint that we may need to introduce a change. So this is introducing the concept of what it is about these beta dispersion curves and what it is about the changing shape that might be affected by the properties of the cell. Now we can only take so much information by analyzing those curves by Ali. So that’s why, particularly with the onset of big data and industry and the discussions around industry 4.0 and analyzing these large data sets. Ideally, what you would want to do is take all of this complex information that we’ve generated and feed it through some sort of computational method of breaking down that variation. That’s more understandable. So let’s have a look at principal component analysis. I won’t go through all the details of this. I would encourage everybody to have a read up on principal component analysis to understand what the base is behind the analysis is. But put simply, we want to take all of our different frequencies and break them down into one or two components of variation that accounts for all the variation in the system that we’re seeing. So when I run the frequency sweep data of a typical batch true principal component analysis, you can see that we identify two principal components. Principal component one is accounting for, say, 90 87 of the variation in this case of the batch, and it’s really related to changes in cell volume. As we already know from the Jesus discussion, cell volume impacts the frequencies across the frequency range. So that’s why we see the effect of volume is having a large impact on one of the principal components. However, what really becomes interesting is principal component two, because the loading of this principal component would suggest that we get a change in the culture from low to high frequencies. If we were to analyze this component. So that’s why we get what looks like a map on the left of where we put these two principal components against each other. And what we can do is follow is related to the duration of the culture by mapping these principal components against time. So that’s what I’ve just done in this graph here on the right. So as I alluded to, you can see that the principal component 1 over time matches quite closely to the cell growth profile and might even produce a more accurate representation of the cell volume of the culture. Purely because you’re taking all of the different frequencies into account and not just relying on the one frequency. However, when we look at principal component two, we can see it as a dramatic change in this component at the 60 hour mark, which was the point of interest that we were looking at earlier in our culture. And that’s telling us, OK, so we’re getting a change of the variation from low frequency to high frequency at 60 hours. Why is that and how can we relate that to the cells in our culture? Now, if you were looking at that as biologists, that might trigger you to say, OK, maybe the internal cell structures are changing, or perhaps you’re getting a change in the membrane channels, which does occur during cell death pathways that might be reflecting on this component. So that becomes really interesting for us to analyze further. And as part of the work we do in our lab here, and we then use other analytical methods as well to understand at these time points what’s changing by running like fluorescent assays or perhaps running imaging to them really build up the picture of what these different components mean, when we look at our cell cultures. So Ed had to do to add 2 to 1 or two to key parameters when we look at the beta dispersion curve fitting. That is the cold call alpha and critical frequency values. Why are you really like these parameters? Is that you’re not depending on this multivariate analysis, which you need to explore offline to analyze. These parameters are generated in real time using the Scottish software, so it gives us a telling point at which we can say, OK, this is how when we need to act, but we generate that information in real time, and that allows operators or researchers to understand that exact critical points at which we get these changes and approaches. So what we see is the critical frequency tends to be related to changes in cell size and then the whole alpha value tends to be more so related to change in the cell distribution. That is more so changes between high and low frequencies, like I’ve discussed earlier. There has been publications put out by the likes of Biogen that use this information to determine when you need to add feeds into cultures. So this evidence could be a strong indicator to become, in my opinion, a critical parameter for analysing cell cultures in the future. Because you can use this information to know when to introduce feeds in a smart manner, rather than relying on offline counting methods which can be unreliable or perhaps give information when it’s too late to feed the cells. So this all becomes very relevant to the topic of understanding changes. And so how earlier then membrane integrity, as does. So just as a final concept to introduce as part of this webinar, quickly to talk about the swan modelling equations, so we had mentioned that you can look at different parameters of the sale using the frequency sweep data outside of just looking at cell volume or cell density changes. So we’re a little bit of manipulation of the equations that you see on your screen. We can use offline data to understand how we can get changes in the cell membrane value and intracellular conductivity values of our bulk measurements and culture. So how do we do this? Well, when we look at the scale of software these values are just mentioned are generally set constants to aid into mathematical equations. However, as biology biologists, we realize that these can vary these values. So that’s why by changing the volume that the other estimates of the biomass and changing that with known values of offline data, we can then calculate these parameters to see how they change over the duration of our cultures. So what we see is that we see quite a steady membrane capacitance value over the first two days and our culture. However, this dropped significantly after that point. But interestingly as well, we see a constant decrease in the intracellular conductivity of our cells using these values also. So it’s really interesting value for us to then use against other methods that we have in the lab also today and understand how these values are changing relative to what the average is estimating. So it’s already valuable information that we can then collate as part of our data analysis. So just to wrap up, one of talked about today, Parsons measurements, in my opinion, hold potential to explore culture dynamics beyond just simple biomass estimation. Using the likes of multivariate data and analytical technologies and capacitance sweeps, these can be tied back to the changing dielectric properties of ourselves. A way that becomes particular importance is during the onset of cell death mechanisms. Ultimately, we want to try and identify critical points in the cell health that would indicate we need to make a change to a process instead of relying on the likes of tripe and blue staining, which is waiting until the sad cell is dead. We want to act nearly as doctors. We want to diagnose an issue prior to that point of reaching. So that, and also a parameters such as the critical frequency and cold called alpha value, may be key indicators of process changes in the future and could potentially be worked into future development processes and how we analyze our biopharmaceutical processes in the future. So with that, I’d like to thank you all for listening. And I’d also like to thank Ed for giving me the time to talk about the research that I’ve been conducting here in nivas. So thank you, everybody.
