The Three Pillars of Organs-On-a-Chip Technologies (Full Title in Description)
United States Department of Agriculture. Hi, thank you for the introduction, and thank you, everyone, for attending. Alright great, so let's get started. So today, I wanted to talk about organs-on-a-chip technology, and particularly the 3 pillars of organs-on-a-chip technology and where we can innovate. And so again, my name is Vadim Jucaud,
I'm a professor at the Terasaki Institute in Los Angeles, where we do biomedical innovation and research. So, the first step, or the first thing that I want to mention is a little refresher on what are organs-on-a-chip. So really, those are microfluidic culture devices that recapitulate the complex structures and functions of living human organs. So basically, really, it's a microfluidic chip that contains different microchambers and microchannels that allow us to deposit different cells within them and be able to recapitulate living organs. So again, I'm going to go over how we can fabricate those
little chips, what are the different materials, what are different fabrication processes, and I'm also going to discuss about what are the types of cells and tissues that we can integrate in those models, and also, I'm going to talk also about the control system and monitoring those organs to really be able to do some testing to further healthcare and so much. So, let's jump in. So, in the first part, I want to discuss really why this technology is really important technology in the future, because it's technology that can recapitulate the complex biological mechanism of organ physiology and disease states in vitro, and this it can do it a little better than conventional 2D models that don't really recapitulate all the components of physiology of organs, and also as compared to animal models, because there is always mismatches between species, different animal species, and human, particularly when you want to look at efficacy and human responses. Then also, it's important because, given the cost-effectiveness of this type of technology, it can really accelerate drug discovery and development pipelines.
And I think this is really one of the main goals of this technology which is to, again, try to facilitate and reduce the time for a drug to move from discovery to translation into the clinic. And another part also is that this technology is amenable for personalized medicine applications; what that means is that we can use patient-specific cells that we can put in those organs-on-the-chip to be able to look at different drug responses or diseases in the personalized manner that recapitulate all the genetic mutations of a specific individual. And so then, of course, the other reason why this technology is really important is because it can be a practical replacement method to improve laboratory animal welfare while maintaining or improving scientific results through accurately representing the human body and generating clinically translatable data. So, I think this is very important because definitely, in organs-on-a-chip, we can create human organs, which again, when we're going to do some testing in vitro and then translate it into the clinic, we might have a better chance at selecting the right candidate drugs. So
now, I'm going to discuss a little bit about the different organs that we can model, and of course, we can model virtually any organ in the body. But here are just examples of different tissues that are mostly recapitulated in vitro using this organ-on-a-chip technology. So, we have the brain, heart, kidney, bone, blood vessels, gut, liver, lungs, but also other organs such as like the eye, lymph nodes, spleen, skin, or pancreas. And here I want to also highlight that all those organs-on-the-chip models are made to really recapitulate the specific function of an organ. So, for example, for a liver, it will recapitulate like metabolism or metabolite secretion. For the heart, usually, we try to recapitulate contractility and conduction.
With the spleen or lymph node, we try to look at immune responses. With blood vessels, angiogenesis and remodeling. So, after all, this is really to highlight that organs-on-a-chip, although the ultimate goal is to recapitulate all the functions of a specific organ, so, for example, the liver may have more than 400 different functions, as of now, it is quite complicated to recapitulate all the functions, and really organs-on-the-chip are designed, as of now, to recapitulate a specific function that you want to study or understand better, or a function that when you add a drug, then it can act on that specific function. So, for example, for the heart, if you want to regulate the beating of the heart, maybe we can only look at contractility. So, organs-on-the-chip technology is made for a specific application. The goal is really to look at the full function of the organ so that we can see things in a more bigger picture. And so, all
those organs-on-the-chip also definitely are used to study different diseases. So, including neurogenerative diseases, cardiac diseases, liver diseases, inflammatory diseases, infectious diseases, and cancer. And so, this is particularly important because those diseases, especially inflammatory or infectious diseases, when happening in humans, might not really be the same as what would happen in animal models, because definitely, there is a huge mismatch which I can say with the immune system, is really where the biggest mismatch is seen, and responses, immune responses, in animals really might not be the same at all as what happens in humans. And so again, those types of models are really going to be helpful in understanding how human responses are before applying them to the clinic. And so, now I want to jump into the place where I
call it, where to innovate, because this technology was started around 2010, so it's a relatively new field, but it's a very interesting field because it includes many different areas of expertise to really achieve the best organs-on-the-chip to recapitulate organ physiology. So, here I want to discuss really what I call the three pillars of organ-on-the-chip, which consist of the microfluidic chips, the cells and tissues that are grown inside of the chips, and the control system that allows us to use this chip for some duration and really to be able to monitor and control the models. So, in the first part of the talk, I'm going to discuss the design, fabrication, architecture, and different features of microfluidic chips, and also give you a practical example of some chips that we've developed within my lab at the Terasaki Institute. So then, the next part I'm going to discuss is about the cells and tissue where I'm going to, again, discuss about the different cell types, functions, and the complexity of different tissues that we can integrate into those microfluidic chips, and we'll be also looking at the advantage of stem cells in patient-derived cells. And then
in the last part, I'm going to discuss the control system, looking at automation, microenvironment control, and real-time biomonitoring, because this is also an important and integral part of organs-on-the-chip, because without any good control systems, then it's hard to culture those organs-on-the-chip for a long duration. So again, here is really to highlight that this field is a really multidisciplinary field that includes the engineers, biologists, and many other fields, and also clinicians, because I think clinicians are very important in understanding what are the types of outcomes that they want to see in an individual model to really be convinced that this would be a good step towards drug discovery, development, and eventually improving healthcare. So, let's jump first into the microfluidic chips, and here I want to really tell you about why microfluidic chips are important in this field. So really, those chips are designed to manipulate a small volume of fluid with different channel valves and pumps. And
this really is a good approach, because definitely, we can reduce the amount of samples and liquids and reagents to be used as compared to conventional 2D models. And so those chips with the design can create a dynamic microenvironment that closely resembles the fluid condition of human organs. And also, it allows us to have really good control over the experiment, and which allows us to mimic organ-level functionalities. So here I want to show you different examples
of different chip designs. So of course, chip design is really limited by the imagination, but at the same time keeping in mind that we have to look at the organ physiology, structure, and function. So here on the left side, I want to show you different simple chip designs, where you can have single channels, where we can perfuse different liquids, and you can code the channel with different cells. After, here you have double channels where pretty much you can have two different cell types in two different compartments, separated by your porous membrane.
And this is the same one, your double channel, but in a sandwich mode, and then we can do it also with multiple channels as well. And so here are different examples. So, this one, those two microfluidic chips here are called really a bioreactor. So, it's a sort of a generic organs-on-the-chip device in which we can culture in those chambers, which are less than 1 centimeter wide, we can culture different tissue, like heart tissue, liver tissue, and so on. And so, this really will be more reflective of those single-channel types of microfluidic chips. Then here, we have another one, those two here which those are really three or multi-channel chips, which, in this case, on the right one, are used to create different vasculatures or microvasculatures of organs. And this other one here is liver-on-the-chip, actually, where we can put
liver organoids and then recapitulate the functionality of those cells. And then here we have two other chips that are blood-brain barrier chips, where we can see that we can integrate also sensors and other functionalities. And all this really comes into the design, because before you want to add any other features to this technology, you really have to think about the design of it.
So then, after, the next part I want to talk about different materials and device fabrication. So, there are really many different materials that can be used to create those microfluidic devices. But the most common ones are PDMS glass and thermoplastic. So those
different materials really have different pros and cons of using them and depending on the type of organ that you're trying to mimic and the different functions of your organs on the chip, then you would select the different types of material, and also the selection of those materials will also influence the fabrication process. So, the main fabrication processes are soft-lithography and injection molding. There are other creative methods of hot embossing and 3D printing, and really, those fabrication methods are mostly to create different molds so that we can actually generate those different microfluidic chips with really intricate structures. So, the really most promising, or the one that has been used a lot lately is the 3D printing, because it's really much faster to create molds and, if we want to do prototyping and iterative process to optimize the chip design, this has really been extremely useful for us to create complex microfluidic system. And so here, this is where I want to describe how we can create those complex fluidic networks. And again, with those complex fluidic networks, the goal is to have a fluid flow control, make maintenance of physiological gradients, and establish a realistic communication between organ compartments when we want to create more complex organs on a chip design that includes different organ modules. And again, this type of design really
can help us study organ interaction. This highlights as well that even though this is designed, and more of an engineering aspect of the field, of course, it has to always keep in mind how biology works, so that we can make a real connection and really understand what physiology of the different organ. So here I want to highlight with those images that, on paper, looks somewhat simple, even though this is kind of a blueprint that is still complicated with a lot of connection, but once we actually translate it into the lab, into a platform, then you can see that things can become a little bit more complex, with lots of chips, lots of micro microfluidic tubing, and lots of connection, and this really is one place where innovation can happen to really facilitate the fabrication connection of all those organs together, as well as other physical sensors, or bio-electrochemical sensors. And so here, with that, I want to show you
a little example of a model that we made where we again, thought about the design and wanted to use a new approach to create a blood-brain barrier on a chip. So here, in this work, one thing that motivated us to use or to create this new chip was the goal of democratizing this technology. So as of now, this technology is limited to a few tabs in the world, and really, this is because you need a lot of different equipment, you need some special rooms which are pretty expensive. And so here, you know, again, when we develop an organ-on-a-chip, I mean, one of my goals is definitely to try to disseminate that organ-on-a-chip and be able to have it validated by other groups. And so, one way to do that
is what we call it, to democratize it, so, using materials that are relatively cheap. So, in this case, we're we're using a thermoplastic called PMMA, which can be cut using a laser cutter, which is a relatively cheap piece of equipment. And then in this chip, we were able to integrate also sensors or electrodes that we applied using screen printing techniques, and those are carbon electrodes that allow us to measure barrier properties or the resistance of the blood-brain barrier. So, in this chip, we culture a brain endothelial cell on one side and really try to see, while applying shear stress or flow in the chip, to see how the blood-brain barrier was established.
What we did, so this is just a schematic, we seed the cells, we apply low flow at the beginning, to be able to remove all unhealthy cells per se, and then, after we apply high shear stress or high flow, so which will allow the state to rearrange and form that strong barrier. And so here we can see the barrier, and we showed that the barrier was forming by using an immunofluorescence approach. But those techniques are destructive which, again, once you do the fluorescence, you cannot use that chip anymore. So, there is a need to actually have some
sensor that can non-invasively assess the barrier of the blood or the integrity of the barrier, of the blood-brain barrier. And this again, also this highlighting, that even though here I'm talking about the chip design, it really encompasses also what type of cells you put in it, and also how you're going to be able to integrate sensors. And so, this is what we did. We measured the resistance after 4 and actually we went up to 6 days to see how the barrier was improving. And so, this is what we see in this graph here, where you can look at the resistance and as we cultured the cells, and when we started applying high flow, because again, endothelial cells that line blood vessels are always exposed to shear stress with the blood circulating within them. So
those cells really like that mechanical stimuli and using the stimuli, we can actually increase the integrity of the barrier. So here we have some staining, showing that again the tight junctions were forming in this model, and as well as being a really impermeable model, that we didn't allow the diffusion of a certain molecule through the model when the cells were present. So, with that, now I want to talk more about the cells and the tissues, because definitely, this is a very important component of the organs-on-the-chips. So, there are 3 different
types of cells that are mainly used, which are cell lines, primary cells, and iPSCs. So, both have pros and cons, and again, cell lines are good because they're highly reproducible, they're easy to use, they're cheap to buy, and they're usually well characterized. But the problem is that they lack physical relevance, they can have genetic drift, and usually, they're from cancerous origin, so definitely to try to study the healthy state of an organ, this might not be the best type of cell to choose. But in that case, you can use primary
cells which have high visually relevance you can have diverse phenotypes, and of course, you can generate patient-specific models with them. But the cons are that those cells really have limited lifespan, they could be variable depending on how you isolate them and culture them, and also they are difficult to obtain and maintain, because to obtain those types of cells you going to have to get a biopsy or something like that. And then the last one, which is the most promising cell type for this technology, would be iPSC cells. And why? Because they're pretty important. You can generate many different cell types, so therefore, many different organs, you can, of course, make patient-specific models, and they proliferate, you know, not completely unlimited, but they have high proliferation capabilities. So essentially they would be a little bit
easier to use than the primary cells. But of course, there are other cons that you can have differentiation variability, it could be long and expensive to differentiate the specific cell type into the one that you want, and of course, they could have a genetic abnormality. But then those cells are really used, and this, I think, is what's very interesting with this field is they can, we can use a tissue engineering approach to be able to recapitulate the physiology of the organs in a better way. So, we can use scaffold materials, bioprinting techniques, or
self-assembly, which can improve the functionality of those tissue. And again, we can have different extracellular matrices, and we can organize cells in a specific way and generate specific functions to achieve biomimicry like contractility, barrier function as I showed you in the previous model, or metabolic. activity. And this here is what you can see in those images. So, this is a bioprinted heart, which I don't know if you can see in the video, but this is actually beating. And so those are integrated into an organ-on-the-chip device here. Same thing
with liver cells that are photo-patterned in the shape of a lobule per se, and they're encapsulated in a gel that protects them from sheer stress, which again allows them to grow well and nicely. Here, it's a picture of vascularization again, which we are doing within the chip to really have, a physically relevant model that includes blood vessels and capillaries. And here is a last figure that shows, like some organoids that can be included in those organs-on-the-chip, and this is a liver organoid that still retains pCYP functions of metabolism function. So, with that, again, I want to give you another example of how, what we generated in the lab, and in this case is a patient-derived vascularized liver on-a-chip model.
So, this is a brand new approach where we actually got a discarded human liver that was supposed to be transplanted, but because it was not in good quality, good shape, it was not transplanted and discarded, and because the Terasaki Institute has a long history in organ transplantation, we actually were able to get some of those organs from which we isolated primary cells and put them into the chip. So of course, here we can characterize the cells before putting them in the chip. But the goal here was to really recapitulate liver regeneration, and so we have a vascular model where we create a vasculature. Then the hepatocytes start to rearrange within the model, creating buds, and then those buds start to become more like globules, where they become vascularized. And so here is really how it looks in the chip. So, you can clearly see that after growing this chip for
7 weeks, which is actually a very long time, and I think this is also one advantage of organs-on-the-chip technology, we can culture tissue and models for, you know, up to 7 weeks, or even more, I would say. And so here we can clearly see that the top row is a live-dead staining. So here you can see that most of the cells are alive, and then you can see throughout time that there are clumps of cells that are forming like, sporadically after 7 weeks this big blue blob which really corresponds to a liver lobule. So here in the bottom one, we are looking at a CD31 staining, which is an endothelial cell marker, and we can see at the beginning, there is a little bit everywhere, but they start rearranging in some network. So then there is more rearrangement in the second week of those networks, and then in the third week, that's where we start seeing in green, the hepatocyte, or the liver cells arranged in two buds. After the fourth week here,
we can really see that there are more hepatocyte buds that are formed and attached to the vasculature. And then, after 7 weeks, when you see those buds that are white here, buds that are not positive for hepatocyte, but positive for endothelial cells. So here we really have the bud or organoid that is vascularized. So again, this creates really a very functional model. Where again, the vasculature is an integral part of the organ, in this case, a liver. And so also what's very interesting with this type of model, so hopefully, you can see with this video. But so, we can also perfuse
immune cells, and so then we can really look at the interaction of those immune cells with the microvasculature, because, again, in the body immune cells usually travel within the blood. And so, this is particularly important where, with immune response, those types of models really, clearly have better recapitulation of what happens in humans as compared to animal models. And so now I want to discuss a little bit more about the control system, which again, also, is an integral part of this organ-on-a-chip technology, because it allows us precise and automated control over micro-environmental parameters, like oxygen, temperature, pressure, pH, or metabolites. And also, it allows us to manipulate fluid flow or apply mechanical forces and introduce other chemical stimuli to enable accurate replication of organ function. And here are different examples of sensors that we have generated in the lab to be able to monitor the organs-on-the-chips. So here we have an oxygen sensor that is using LED light where we can actually see changes in oxygen within the microfluidic chip in a real-time manner.
So then, underneath there is here a temperature and pH sensor that again, is just an electronic chip that we can integrate within the microfluidic chip where we grow the cells. Here in the middle one, this is really to show actuation, and how we can easily control and route liquid within different channels. So, this is again important to apply different forces or introduced chemical stimuli, and so on, and this we can do in an automated manner, using what we call quake valves that can block or open any microchannel. And then here
we have another chip with integrated sensors. So here is also a specific design that allows us to put the probe within the cell culture chamber to be able to monitor dopamine secretion from brain cells. So here again, we are able to really put the probe very, very close to the cells, because the dopamine is not going to travel so far out of the organs-on-the-chip. So, this is another approach again to be able to monitor responses within the chip. And then after, here's just a schematic of the oxygen sensor, and here is another microscope that we have integrated, but which is a little bit more complicated where it allows us to see three different color dyes. So again, having this integrated, you know, it's really much easier to study all those responses without having to place it into a confocal microscope or other heavy equipment. And so here again, I want to give you another
example, that of a model that we developed in the lab that includes sensors and real-time sensors. And so here we developed a model where we have a liver-on-a-chip model that can be treated with drugs or not. And then here we have a PC-TIR sensor which is a label-free optical-based sensor that allows us to detect different biomarkers. So, in this case, we look at biomarkers of liver injury to be able to assess drug-induced toxicity in the liver. And so here, what we did again, we created this model that is actually inspired by the architecture of the circulatory system in the liver, where there are two inlets, one from the portal, the vein, and one from the hepatic artery, that drains into a central vein. So, in the middle chamber, this is where we culture our hepatic cells that are encapsulated into a collagen gel, which is an extracellular matrix that is highly present in the liver. So, what we have shown is that,
using this type of model, drugs, nutrients, and oxygen can diffuse within the model easily, allowing us to culture this model for about 5 days. And here we can see that after 5 days, the cells are growing and again staying alive. Here we look at that also in growth, and how they are deforming sort of like spheroids within the chips. And so those cells remain alive for 5 days, and what we also can see in terms of function, they secrete albumin, which is one marker of function, and when we use the chip with a dynamic system, so with the flow, we can see that the cells actually secrete much more albumin. So again, showing that organs-on-the-chip technology is superior to static, conventional 2D models to look at the functionality of cells. And so here is what we did; we treated those cells with doxorubicin, which is a chemotherapeutic agent that can have little toxicity. And so, what we
see again, those are just the standard tests that usually are used to measure toxicity, which is live data, say, using confocal or ELISA, and those are usually done at like endpoint measurement, whereas here, we have this new sensor that allows us to really detect in real-time how many biomarkers are secreted by the hepatocytes or the liver cells here. And so here, what is nice is that we can confirm that using our sensor here, we don't really see any difference in albumin secretion in response to this drug, which we also see here with the ELISA, so suggesting that albumin might not be the right biomarker to monitor toxicity. But, on the other hand, when we look at GST-alpha, which is a liver injury biomarker, we can see that after 24 hours under ELISA, we can see a difference when it's treated with a 20 micromolar of doxorubicin. But in the case of our real-time sensor, we can detect those changes as early as 3 hours. So again, showcasing, that is much faster than conventional models to monitor toxicity with physical relevance. And so now the last part is I just want to
touch back on some of the challenges that remain, so again, with the microfluidic chip the main challenges are going to be scalability, reproducibility, and the standardization of fabrication process. For cells and tissue, again, is viability, functionality, and also scalability, and for the control system, there are still challenges regarding multi-scale control when we are looking at like high-flow rate or low-flow rate, and so on, and looking also at dynamic stimuli and also standardizing all those control systems. And so, with that, I want to just acknowledge all the team at the Terasaki Institute that contributed to this work and all our collaborators and funding agencies that allowed us to develop all those very interesting in vitro models. And then, with that, I would like to thank you, and I will take any questions. USDA is an equal opportunity provider, employer, and lender.
2024-11-06 12:29
