Scientist Stories: David Mooney, Viscoelasticity From Stem Cell Fate to New Medical Devices

Started. First of all, thank you for accepting our invitation, Professor Mooney. Welcome to another Terasaki Seminar Series. Today we have Professor David Mooney from Harvard. He is the Pincus Family Professor of Bioengineering in the Harvard School of Engineering and Applied Sciences. He is also the founding and core faculty member of the Wyss Institute. He earned his bachelor's from University of Wisconsin-Madison and his PhD in chemical engineering at MIT. His laboratory designs biomaterials to promote tissue regeneration immunotherapy, and he has made significant advances in tissue engineering, drug delivery, and mechanotransduction. He has an h-index of 158 and with over 100,000 citations. He's won many awards, just to cite a few: Clemson Award from the SFB, Merit Award from the NIH, Distinguished Scientist Award from IADR, 5 Beta Kappa Prizes for excellence in undergrad teaching, Everett Mendelson Excellence in Mentoring Award from Harvard College. He's a member of many national academies including National Academy of Engineering, National Academy of Medicine, and the National Academy of Inventors. In 2019, he was named one of the top 10 translational researchers in biotech by Nature Biotechnology. His inventions have been licensed by over 18 companies leading to commercialized products and founding of companies and he's active on industrial scientific advisory boards. The floor is yours, Professor Mooney.

Well, thank you for the opportunity to come and talk to the group today. And, um, there it goes. I'm just going to share my screen here. Um, you see the slides okay?

Yes.

Okay, well, thank you for the opportunity to come in and talk to the group today. Um, I'm looking forward to sharing some of our work and getting some feedback on some of the things that we're doing. And I'm going to describe today just one aspect of the research in the lab today and that relates to our exploration of material viscoelasticity, where we started by exploring its role in stem cell fate, and that has led us, as I'll describe, to work designing new types of medical devices. Before I get started, I need to mention that the funding for the research I'll describe today has come from the NIDCR, NIBIB, the NCI Physical Sciences Oncology Network, as well as Novartis. And commercial engagements that are relevant today include Novartis, and then Amend Surgical has licensed some of the IP related to some of the work that I'll describe today. Now, as I mentioned a moment ago, you know, our starting point in the line of research I'm going to talk about today relates to tissue regeneration mediated by stem cells. And so we've had a long-standing interest in taking stem cells from the body such as those present in the bone marrow, and using those to create new tissues and organs in the body, such as this engineered bony organ you're seeing here with the hematopoietic compartment that we've been using to reconstitute T-cell immunity. So we've had this longstanding interest in designing materials that enable stem cells to create new functional tissues in the body, and while this was the goal of our work, Along the way, some very talented postdoctoral fellows and PhD students noticed that some of the basic biological principles that we were starting to understand might have relevance in a number of other spaces. So where the talk is going to go by the end is to talk about medical devices, such as this medical adhesive you're seeing here, that's been adhered to a beating porcine heart, demonstrating really outstanding and unprecedented adhesion energy, really based on some of these fundamental principles of material-cell interactions. So that's where we'll end up going by the end. So to start with, let's go back to regenerative medicine. When we're interested in using stem cells therapeutically, we very frequently go to the normal stem cell niche for inspiration for the types of cues or information that we could provide these cells to control their behavior. And in the lab in particular, we're interested and using biomaterials as a means of organizing and presenting some of these natural cues to control cell fate. Now, there's many cues in the stem cell niche. I'm not going to try to talk about all of them today. Instead, I'm just going to focus on one type of interaction that will form the basis for all the things that I will describe throughout my talk. And that relates to the active mechanical interaction between cells and their surrounding matrix. And one nice way to visualize this is shown here, which is a movie of a human cell crawling around inside a 3-dimensional collagen matrix. And what you can appreciate if you look carefully at these collagen fibers is that the adhesion of the cell is not a passive chemical interaction where a cell binds and unbinds, but instead it's also an active mechanical interaction where the cell reaches out, grabs, pulls, pushes, and manipulates the extracellular matrix as it engages with it. Now, a couple of decades ago, researchers first recognized the mechanical dynamics of this interplay, and this led to a very important question, which is: if cells actively and mechanically probe their environment, does the resistance that they feel to their own forces feedback and alter their behavior. So, that became a really important question in cell biology for well over a decade, and there was a variety of studies that went on to demonstrate that indeed this was quite important. One of the historically first of these was some work by Yuli Wang's group, where he demonstrated that the migratory ability of cells is regulated by the stiffness of the substrate, and thus the resistance the cells feel to their own cell-generated traction forces. Jun Kong, now down at the University of Illinois, when he was in the lab, demonstrated that the proliferation and apoptosis of cells is regulated by the stiffness of the substrate. And Dennis Discher's lab showed a number of years ago as well that the fate of merostromal cells or mesenchymal stem cells, in terms of their differentiation, was regulated by the stiffness of the substrate. So, after about a decade of research, the answer was pretty clear that yes, the stiffness, or the resistance the cells felt to their own traction forces, was an important cue that fed back and regulated many aspects of cell biology. However, all of these studies were done in two-dimensional cell culture, and clearly our body is more of a 3D kind of world. So, what we've been very interested in for about the last decade is whether or not these ideas translated to 3D, and whether we could exploit these ideas to impact tissue regeneration back in the body. So to explore this, we've been using a particular model system that I'll describe a little bit here, and it is a polysaccharide called alginate that is a block copolymer comprised mainly, or comprised of, two types of sugars, what I'll abbreviate as G sugars or M sugars, and these are organized into blocks. And the relevance of this block structure is that only the G-blocks can actually participate in cross-linking or formation of a hydrogel from these polymers. And when these polymer chains cooperatively share divalent cations such as calcium between the G-blocks of adjacent polymer chains, this initiates gelation, induces gelation. The M-blocks don't participate in this process. Now, the system allows one to very easily and safely encapsulate cells simply by adding divalent cations to a solution, and the cells become trapped in a 3-dimensional matrix, and we can vary the stiffness of the substrate with usual, with the usual tools of polymer science, controlling the polymer molecular weight, the cross-linking concentration, or the composition of the polymer, the M to G ratio. As you can see here, we can get families of hydrogels in which we can very readily manipulate the stiffness, as indicated here by the elastic modulus. And actually, the background image that you're seeing here is an AFM image of one of these alginate hydrogels. So you can see the individual polymer strands. They kind of look like pieces of spaghetti. And then you have the spaces in between, which are in the nanoscale. The mesh size in these gels is about 3 to 5 nanometers. In size, and so the cells are immobilized and encapsulated within this gel. Now, cells don't have a normal means of engaging with these, with this particular polymer, which is typically derived from seaweed. So to mediate a specific and defined mechanism for cells to engage via their integrin receptors, one can simply covalently conjugate small peptides taken from sequences found in molecules like fibronectin, and the RGD sequence is obviously a very ubiquitous sequence that is very frequently used to mediate adhesion of cells to synthetic materials. And this takes a hydrogel that is otherwise non-adhesive to cells, as you can see in the top image, to one that instead mammalian cells can readily attach, spread, and function on these materials. Now, Nate Hipp, shown here in the lower left, who's now running his own lab at Washington University, was the first to demonstrate that in 3D, one could indeed regulate the fate of stem cells simply by varying the mechanical properties of the gel. So here he took MSCs, he put them in gels that were either very soft, as indicated by an elastic modulus of 2, 2.5, and these cells actually differentiated and became fat-forming cells or adipocytes, In gels that were a little bit stiffer, the cells became osteoblasts, or bone-forming cells, and in the really rigid gels, they actually didn't differentiate at all. And all the, the rest of the data here is just some molecular characterization and confirmation of the fate of the cells being regulated by the mechanics in concert with the density of the adhesion ligands. So this definitively said that we could indeed regulate 3D, or cell fate in 3D, with mechanics. And then Nate went and followed that up with a demonstration that this could allow us to regulate the amount of bone regeneration that would occur using these cells in a transplant model. In this case, in a calvarial defect model, he took a small number of cells, transplanted them in gels that were either suboptimal for osteogenesis, either too soft or too stiff, or one that was otherwise chemically identical but had actually what we consider to be the optimal modulus. And what he found was the amount of bone formation was increased at that optimal modulus that we knew was optimal for osteogenesis in vitro. So the first demonstration in the body that we could regulate the ability of stem cells to mediate regeneration simply by controlling this mechanical property. Now, this was all a nice story. However, it ignores some fundamental biology of the body. And so I've been talking about controlling cell fate by controlling the stiffness or the elastic modulus of these hydrogels as a mimic of the extracellular matrix. And so in essence, we're kind of pretending that the matrix is a purely elastic material like a rubber band, and that we can characterize it by a single measurement, a modulus. However, the tissues in our body reconstituted extracellular matrices as you can see on the graph on the left, when we look at the loss modulus and storage modulus using rheological analysis, are not actually purely elastic. Instead, they have properties that are intermediate between being fluids and being solids. They're viscoelastic as being pure— as versus being purely elastic. Another way one can actually monitor viscoelastic properties is shown on the right. Where if one takes a material and places it under a constant strain, one can measure the amount of stress that's required to maintain that deformation. If one has an elastic material like a rubber band, the actual stress will stay constant with time, so the line along the top here. In contrast, if one looks at tissues in our body, they actually dissipate the stresses over time So the amount of stress actually is diminished. So these materials in our body, the normal extracellular matrices, are actually viscoelastic. And so that led to the question starting a few years ago, is do cells then actually read and care about the viscoelastic properties, or is it really just the stiffness that matters? Now, to give you a little bit of a visual sense of kind of what viscoelastic and elastic materials are and how they differ, Here's actually just some videos of a purely elastic material being deformed on the top, a polyacrylamide gel, and a tissue, brain tissue in this case, being deformed on the bottom. And what you can appreciate from the material on the top, that as we deform it, the material, as we basically apply a load, the material deforms, and then immediately upon release of the load, the material elastically and instantaneously goes back to its original dimensions. In contrast, The viscoelastic material here on the bottom, you'll appreciate, is after we remove the load, some significant deformation remains in the material. So it's actually exhibited some, some flow. It's dissipated some of the energy that we put into it. Now, throughout the rest of the talk today, you know, materials are not typically purely elastic or purely viscoelastic, so I will tend to talk about materials that are being either more elastic or more viscous, or elastic versus viscous. And one way that we oftentimes look at this is by looking at the rate at which they dissipate the energy, and the more viscous materials will dissipate the energy more rapidly, as shown in the curves here to your right. Okay, so we want to examine now a number of variables and understand their impact on cell behavior, and we need to actually have a system where we can independently vary each of these. The stiffness, the viscoelasticity. We want to independently vary the mesh size and the ligand density and type for cell adhesion. Now, this will lead us to actually re-examine for a minute why we use these alginate gels. So the reason we don't use materials like Matrigel or collagen or fibrin as they are traditionally used in the biomaterials field is that while one can readily make stiffer or less stiff gels, out of these materials, this is typically done just by adding more and more material. This makes the gels more stiff, but also increases actually the number of adhesion sites. It actually decreases the spacing between the gels or the mesh size, which can lead to changes in diffusional transport of things like macromolecules. So all the variables end up getting basically dependent in the setting. Now we can take a classic synthetic polymer approach to make hydrogels instead, Like the alginate system, if the cells don't mediate adhesion themselves, we can decouple the ligand density and type by covalently conjugating it to the polymer chains after we form them into gels that are either more or less stiff. But again, as one varies the crosslink density in a traditional polymer system, one varies the mesh size for diffusional transport. Now the reason we use these alginate gels is because of this cross-linking structure, this cross-linking between blocks of these sugars that I previously described. And so when it's an alginate gel, if one cross-links it to a low level to make a relatively soft gel, as indicated by these few red dots, basically bridging adjacent polymer strands where we have these G-block structures, Actually, all the, the polymer chains end up falling into a specific alignment, and then as we add more and more cross-links, we in essence just occupy more and more of these locations that are already aligned adjacent to each other. So we can make the more, the material more stiff without changing the nanoscale architecture, so we don't change the mesh size or diffusional transport. Further, we can independently vary the viscoelasticity as versus the stiffness by controlling the molecular weight of the polymer chains, which allows us to alter their relative mobility, and thus their ability to actually dissipate stresses independent of stiffness and these other variables. So here's just an example of a group of hydrogels that Ovi Luo on the right, who's now running his own lab at Stanford, and Luo Gu, who's running his lab now at Johns Hopkins, demonstrated a few years ago, where you can see from the top left that these 4 gels all have very similar initial elastic moduli. But on the right, you can see that they have varying rates of stress relaxation, which we've characterized by the half-time for stress relaxation on the bottom. And you can see that we get a couple orders of magnitude variation in the stress relaxation behavior while we keep the moduli similar or constant. Now, if we go ahead and start taking cells and exposing them to these materials, we find that the stress relaxation has a dramatic impact on cell behavior. So, here you're looking at individual cells that we visualized from the left to the right, as the gel is otherwise the same, but simply relaxes to a faster and faster rate. So, they're more viscoelastic to the right, and you can appreciate the ability of cells to spread, is dictated by the viscoelasticity. This is dependent upon the ability of the cells to mechanically engage with the material, because if we take away the cell adhesion ligands, the cells are not able to spread even in the very rapidly relaxing gels. And furthermore, we see the proliferation of the cells is similarly regulated, as well as is the migration, by controlling the rate of stress relaxation in these materials. Now, the fate of stem cells is also regulated by the relaxation behavior. So here, if we take MSCs and we monitor their ability to become osteoblasts, as indicated by alkaline phosphatase staining, on the left you can see that if we have too soft of a material, we get very little osteogenic differentiation. If we make a permissive gel in terms of its modulus, we see very low level of osteogenesis. If the gel is too elastic, but as it becomes more and more viscoelastic, we get greater and greater osteogenesis., as indicated here by ALP staining. And actually, we see the opposite behavior for adipogenesis, that basically adipogenesis is favored by more elastic materials as versus viscoelastic materials. Now, one of the questions that may arise is whether, you know, this really has any physiological relevance, or whether it's just, you know, some parlor trick that we've developed with these materials. So, Max Darnell, a very talented PhD student who's now at the University of Washington doing a postdoctoral fellowship, collaborated with Edie Lippens in Georg Duda's lab in Berlin to look at environments in which bone regeneration normally occurs in the body and to see if they are also viscoelastic. And so, if we look at either coagulated bone marrow on the left or fracture hematomas that are removed from human patients, coming into the Charité Hospital for revision surgeries. On the right, you can see if we do a stress relaxation test, these materials are viscoelastic and actually exhibit stress relaxation behavior that quantitatively is very similar to that which we found empirically to be optimal for osteogenesis, suggesting that what we're really doing is rediscovering the microenvironment that our body normally uses to drive osteogenesis during bone regeneration. Now, to confirm that this actually could be a driving variable in the body, Max did a really simple experiment. He went back to those rat keratovarial defects, transplanted a small number of cells in gels that were otherwise matched in all features except in the rate of stress relaxation, and he found that the rate of, or the amount of bone new formation was dramatically increased in the Fax-relaxing gels as versus the slow-relaxing gels, in agreement with our observations of the differentiation of the MSCs in vitro. Indicating that actually stress relaxation is a variable that seems to have a profound impact on not just cells outside the body, but also on the ability of stem cells to promote regeneration in vivo. Now, we find that this is a phenomenon that is relevant to many, if not all cells, actually all cells we've examined to this point in time, the relative importance of these different variables of the matrix can be distinct. Here, for example, if we look at human iPSC-derived neural progenitors, and we do an experiment where we vary the rate of stress relaxation, the stiffness, or the adhesion ligand density, and then do sequencing, and look at the number of differentially expressed genes for each of these variables, what we find is that altering the adhesion ligand density, or soft versus stiff, will impact, you know, let's say hundreds of genes, but the rate of stress relaxation will actually impact the expression of thousands of genes. So, it seems to be far and away the dominant variable with this particular type of cell. So we believe that stress relaxation is something, and viscoelasticity is something that's going to be very relevant in many, if not all, aspects of biology. Now, we've recently become interested in not just the relevance of viscoelasticity in regeneration, but also its potential relevance in disease, because one of the common features of many diseases is fibrosis, where we get additional extracellular matrix deposition and cross-linking that changes the mechanical properties. And so Kyle Vining, who's just finishing up in the laboratory with his PhD studies, became interested in whether the mechanical changes of the ECM in fibrosis could actually be not just a result, but also a player in driving further inflammation. And he decided to look at this particularly in the context of myelofibrosis, which is a disease of the bone marrow where one sees increased deposition and cross-linking extracellular matrix is shown here in these histology images. Now, myelofibrosis is actually one of a group of neoplasms. It is the one that actually has the worst prognosis because a very aggressive form of these types of cancers. And so there's clearly an unmet need to really understand this disease better and to develop then better or newer targets for therapeutic intervention. Now, the first approach that— or first step that Kyle took was, first of all, to confirm that indeed in myelofibrosis, we would be getting changes in stiffness and viscoelasticity. We can't do this very readily in human patients, so he turned to an animal model of this disease in which hematopoietic stem cells that are genetically modified to have a very similar mutation to that which is known to cause many human disease of this type are adoptively transferred into another set of animals. What is found is they will over time come to occupy the bone marrow and will lead to the types of changes in the extracellular matrix of the bone marrow of these animals that we typically see in human patients. If we characterize the mechanical properties of the bone marrow over time, you see there's an increased stiffness With these transplants, the red bar, there is a diminishment in stress relaxation. They're basically becoming more elastic. And actually, if we look at the tan delta, which is an indication of viscoelasticity, you can see as the disease grade increases, the disease gets worse, we get more and more stiffening of the bone marrow in these animals. So then Kyle addressed hypothesis. That these changes in mechanical properties could directly regulate inflammation in this setting. To do this, he stepped back to an in vitro model, and as collagen is the major extracellular matrix molecule that we know is involved in fibrosis, I won't go in detail because he actually published the system, but he developed an approach to take— create interpenetrating networks of collagen and alginate, and by this he could create materials that would have varying rates of viscoelasticity, as indicated by varying rates of stress relaxation, without changing any of the other features of the system, such as the collagen fibril formation, the mesh size or porosity of the gels, or other features. Now, when he took human primary monocytes and placed them within these gels, in which we would vary the stiffness and the viscoelasticity, What he found is the differentiation of these cells was dramatically impacted, and specifically, their differentiation into dendritic cells, as first indicated by expression of a whole panel of relevant cytokines, was dramatically upregulated in the stiff elastic gel specifically, as versus any of the viscoelastic gels or the soft elastic gels. If you compare then just the stiff viscous and stiff elastic, he found that the cell surface expression of markers indicative of dendritic cell differentiation were dramatically upregulated, as seen in these FACS plots on the top right. And then if one looked at a variety of other markers of dendritic cell differentiation, one can see that those are upregulated with the stiff elastic gel specifically. So the viscoelasticity regulates the monocyte differentiation of dendritic cells, when these cells are taken from primary or from healthy human donors. If you actually look at basically cells in human patients of myelofibrosis and compared those cells' gene expression in early versus late stages of the disease and compare them to the viscous versus elastic gene expression pattern we see in vitro, one sees a very strong correlation between the upregulation of a number of these genes in vitro as the disease progresses, and in vitro as we make the transition to a stiff elastic substrate mimicking the matrix properties that would be seen in late stages of the disease. We actually see very similar patterns of gene expressions in vitro and in these patients. Furthermore, if we look at other types of fibrotic diseases, for example here, lung fibrosis, One sees that very similar changes in expression of a number of genes can be found in these patients as versus what we see in our in vitro model system. So since the model system has been able to identify these changes and show good correlation to what we see in patients, the question became, can we use it as an approach to identify new therapeutic targets then that may transfer then to human patients down the road. To explore this, Kyle started looking at what the mechanism by which the cells were impacted by the viscoelasticity, and he actually screened a great number of potential mechanotransduction targets. The one that ended up being most fruitful was PI3K gamma, which is a myeloid-specific form of PI3 kinase. And this is a molecule shown in this video that's taken from a publication from the literature. That is highly expressed as cells migrate and is very involved in cells assembling the extracellular matrix apparatus they need to push out against the extracellular matrix as they migrate and move around. And what Kyle was able to show is when he inhibited PI3K in these cells using a myeloid-specific inhibitor, that he was able to inhibit the actin assembly that is specifically seen in these stiff elastic gels. And when he inhibited this, he was able to demonstrate that the upregulation of dendritic cell differentiation from these cells that we see with the stiff elastic substrates is significantly dampened with inhibition of this particular molecular target. He then went on, though I don't have time to show, to demonstrate in an animal model of this disease that inhibiting this target in vivo also diminished the severity of disease. So what he was able to, in sum, to show is that the viscoelastic properties of ECM regulate not just stem cell differentiation, but actually regulate monocyte differentiation through this one particular mechanical checkpoint that we think could be very important and very relevant in at least this disease context. So this gives you a little bit of flavor of kind of the biological side of what we've been doing. Now I'd like to transition to talk for the last few minutes about how we've been exploring this, or trying to exploit this, in the context of medical devices. So while a lot of the initial mechanotransduction studies were underway in the lab, Shani Zhao, who's now running his own laboratory at MIT, was also in the lab. And he comes from a mechanics background. And one of the observations that he made of these hydrogels we were using in lots of our studies is that, you know, they're basically very weak materials, fracture energy of about 10 joules per meter squared, and it's not surprising for this class of materials that they'd be very weak. Well, as he was looking at Nate Hipps and others doing these mechanotransduction studies and beginning to recognize the importance of the mechanics of the materials, he recognized that the ability of extracellular matrix to dissipate stresses could be actually an important principle, not just for cell interactions, but more broadly in terms of toughening materials. And so what he and actually Sun, a visiting PhD student in the labs of two of my collaborators at Harvard, Suo and the Vlasic labs, what they were able to demonstrate together, or propose initially, was that having a combination of basically dissipative So stress-relaxing bonds within a material combined with covalent bonds could lead to significant toughening. The basic idea is if one has a typical covalently cross-linked hydrogel, like a polycrylamide gel, as one starts to stress it, one puts a significant stress on individual bonds, the stress becomes so great those bonds fail, and then you get stress propagation throughout the material. Now, in contrast, with something like an alginate, that is reversibly cross-linked and can dissipate energy because it's viscoelastic, that some of that energy would be taken up by decross-linking. Now, the gels, though, because they're formed from relatively weak bonds, don't absorb a lot of energy. But what they propose is if they combine a covalent network that could transfer energy over significant differences with this ionically cross-linked network that could dissipate the energy, this now creates a large zone that would be absorbing energy, dissipating energy, and really toughening the material. They were able to demonstrate this idea pretty spectacularly, and here's a nice video demonstrating the idea, where this is a 1-inch steel ball that was dropped from 6 feet onto a 1-millimeter-thick layer of hydrogel. And you can appreciate that the gel not only doesn't break, but it actually now is able to return a lot of the energy back to that ball and actually will send it flying back up again, although it will have dissipated quite a bit of the stress. So this was actually a nice demonstration of some fundamental material properties, but we didn't really know, you know, what to do with this. And then a few years later, Janyu Li came into the laboratory, and he was very interested in medical adhesives. And the challenge that he was interested in addressing is the fact that adhesion on wet and dynamic tissue surfaces is a big challenge, but that actually unfortunately describes most of the inside of our body. He started looking around for where nature has already solved this general problem, and he came across the slug that you see the picture of here. Other scientists had already characterized that these slugs secrete a mucus that they normally use to move around, but when they're threatened, for example, by predator, they change the composition of their mucus, and this allows them to adhere very strongly to whatever they're basically on and avoid predation. And when one looked at the composition of this mucus, it was a hybrid network comprised of ionically cross-linked polysaccharides and covalently cross-linked proteins. And Jane, you realize that this sure looked a lot like these gels that Sean He had been making, that were able to dissipate so much energy. And so he proposed that one could create new types of viscoelastic medical adhesives that would provide ultra-strong adhesion by combining these, what we call, tough gels, these dissipative matrix, with an interfacial material that could form strong bonds between the underlying tissue and the gel. And they specifically have proposed to use chitosan that could form entanglements, ionic interactions, and also covalent linkages between the two. And the basic idea is we'd have our gel, we would coat it with a chitosan solution, place it in contact with the tissue for a short period of time, allow adhesion to occur, and now have a material that would conform to the underlying tissue, be capable of dissipating significant stresses as it moved and flexed in response to loading from the tissue itself. Just to give you a sense of what adhesion looks like with the system, here's a simple experiment where we've now adhered one of these materials to porcine skin, and we did it in the exposure to blood to mimic a bloody surgical field, and you can appreciate we get very strong adhesion. What the numbers on the right are showing you is that the tough adhesive, or TA, dramatically outperforms cyanoacrylate in the dry state, and cyanoacrylate is pretty much the state-of-the-art adhesive in the medical field right now. And really importantly, that in the context of a bloody wet environment, cyanoacrylate performance is dramatically reduced while the performance of this material is not. Now, they were able to— JNU was able to demonstrate that this material has performance that actually is better than any of the materials that were out there in the field at that point in time, including the fibrin-based materials, the PEG-based materials, the cyanoacrylates, and really provided some really quite appealing set of properties that potentially could be useful in many contexts in medicine. Just to give you a sense of what these— how this material is used, here's some videos that Ben Friedman, postdoctoral fellow in the lab right now, has demonstrated showing how one uses these materials, simply takes them, coats them, adheres them onto the tissue, and here you can see the beating heart image that I— or video that I showed you earlier, where these have been applied to a beating heart, and here you can show the adhesion is strong enough to actually hold the entire mass of the heart tissue. So we think that these could be very appealing in a number of settings, and we're currently exploring that. One of the newest areas of exploration for us is actually the idea of using these as an advanced drug delivery system. And we're specifically interested in looking at this in the context of tendon with our partners at Novartis. And the idea is that if we could adhere a drug delivery system to a tissue of interest, and it could be robust mechanically and allow for high loading, we could then have a very extended release locally of a drug agent of interest while minimizing systemic circulation. And I'm not going to talk about the stimuli-responsive part, but we can also make these stimuli-responsive. So to pursue this, Ben— and here's just an image of Ben— has been looking with their colleagues at Novartis led by Eckhard Weber. He's been placing these. For example, here you see one placed in a patellar tendon above a defect site. They've demonstrated that not only can these place these on really demanding sites like a tendon, but they will actually stay in place for several weeks as indicated by the imaging you're seeing in the bottom. Now we can load these with drugs, and unlike existing hydrogels that tend to enable only a very low drug loading, here we can actually load up to 500 mg per mL of drug into these patches. So at this point in time, half of the mass actually of the system is drug, about 5 to 10% of the mass is polymer. So we actually have a lot more drug than we have gel there, or at least polymer in the gel. This does not alter their really favorable mechanical properties, as you can see by their ability to be stretched, They adhere very well, and they allow for sustained release in vitro on the top of drug molecules of interest and sustained exposure in the body to these same drugs in the context of both injured or uninjured hosts in some preclinical animal studies in the bottom. So we're really interested in the use of these as adhesives, as drug delivery systems, and this work that's been going on and Ben's been doing exploring these viscoelastic adhesives Then in my last short story here, inspired another PhD student in the lab who was working in a very different space. And this student was working in the space of multi-electrode arrays that, as many of you know, are used to record and stimulate electrically active tissues throughout the body. And here we just see some examples of these different types of devices in the schematic. Now, existing multi-electrode materials are very different in mechanical properties from neural tissue or the cardiac tissue upon which they're typically used. These materials are made of metals or hard plastics that have, for example, moduli that are several orders of magnitude greater than the soft tissues upon which they're placed, and they also tend to be purely elastic materials as versus the viscoelastic tissues upon which they are placed. Now, this has some real repercussions So the mismatch in mechanical properties can lead to damage of the tissues over time. And here are some images taken from an in vivo experiment showing that even relatively soft for these kinds of materials can lead to significant deformation, in this case the spinal cord, over time after long-term placement. Furthermore, because these materials are stiff and elastic, they don't conform very well to the underlying tissue geometry, And if we think about recording from the surface of the brain, for example, the brain architecture is very complex, and about half the brain topology is simply inaccessible with the current materials. So what Christina Tzirngides in the laboratory proposed then was a new type of multi-electrode array that would be based on hydrogels, would be totally viscoelastic to match precisely the mechanical properties of the underlying tissue. So the electrodes themselves would be viscoelastic, the encapsulation material would be viscoelastic, the whole thing would be viscoelastic, and we would exploit the ability of conductive fillers, things like conductive flakes and particles and rods, to provide the electrical tracks. These viscoelastic arrays actually conform to underlying brain architecture very well. So here we're looking at an agar model of a sheep brain, and if we use EcoFlex, a typical material, it doesn't conform very well to the underlying brain tissue. When you remove it, you can see it has undergone no deformation. When we use one of these alginate materials instead, it will actually plastically deform or flow to precisely match the underlying geometry of the brain tissue, as one can see when it's removed after some period of time. As I mentioned, we want to make the electronics component soft as well, and viscoelastic, so we're utilizing graphene flakes and carbon nanotubes to accomplish this, and to minimize the quantity of these materials, which could otherwise actually dramatically stiffen the gels, we actually use a technique to create microporous materials to concentrate the particles just in a fraction of the entire volume, And this allows us to get percolation, electrical, electrical passage at very, very low solids content, where we don't have any significant impact on the bulk mechanical properties of the layers. This allows us to get completely viscoelastic substrates, as indicated by this purple sphere over here, which that's the range of properties that we've been able to demonstrate to date that overlap very nicely, for example, with the properties of the heart and the brain. In contrast, these other colors and shapes over here represent the other materials that have been described in the literature, which do not match the physical properties of these soft tissues. Christine has been able to show in a variety of settings that these materials will function in vivo very well. In this particular example, this is just a simple EKG recording, but one where we've taken the array and wrapped it completely around the heart. So done basically a 360-degree deformation of the material, and we're still able to get an excellent recording with the material. So I'm going to stop here, and hopefully what I've been able to give you a sense of through this kind of whirlwind tour is how we started out by being interested in mechanobiology, and this led us to interest in mechanoregeneration, how we could design materials, for example, to regulate the ability of cells that we want to use in the body to differentiate, to proliferate, to migrate, and to regulate overall tissue formation. We also have a very active interest in mechanically active systems, unlike the ones I described today that we're using for regeneration, although I didn't have a chance to talk about those today. But a result of these basic studies, we're really trying to understand and manipulate how cells engage in field forces, led us to appreciate the importance of viscoelasticity in the design of medical devices, and we're designing a variety of different systems today that we think may have some beneficial properties down the road when used in people in a variety of different settings. So I'll wrap up there, and I'd be happy to take questions in the time that's remaining.

Thank you so much for this excellent talk, Professor Muni. Several questions came in. One of them is If I understand correctly, you mentioned by increasing the concentration viscosity of the fibrin, the cell adhesion decreased. What is the logic beyond this?

Actually, sorry, could you repeat that question? I didn't quite get it.

Yeah, by increasing the concentration viscosity of the fibrin, the cell adhesion decreased. What is the logic behind it?

Are you saying the concentration of calcium with fibrin? I'm sorry, I think I'm missing something.

This is from the beginning of the talk.

Oh, yeah, I see. Okay, so actually, there must be a little— I'm sorry if I wasn't clear. Actually, none of the materials that I described today are using fibrin. So these are actually alginate-based materials. In some of the studies, we've added some other materials. For example, in Kyle's work, we have an interpenetrating network of alginate and collagen. But none of the materials I described today are we using fibrin. So I'm not sure what the context of that question is.

Okay, maybe we can pass on it. Second question, you can ask, is the fast relaxation hydrogel good for all kinds of applications? What kinds of applications need slow relaxation hydrogel condition?

Yeah, you know, so that's a great question. And so there— it depends, I think, very much whether a rapid stress relaxation is good or bad for you will depend on the context. So as I briefly mentioned, for example, adipogenesis actually is disfavored in rapidly relaxing matrices. We see actually much greater adipogenesis of MSCs in more slowly relaxing and more elastic materials. So if one wanted to, to favor adipogenesis, let's say in soft tissue reconstruction, one might be better served by having a more purely elastic material such as we typically use in medical devices. So it's going to really depend on the context and what you're trying to achieve, but I do think what our data suggests is that we should at least have some understanding of what the viscoelasticity is doing before you design your device.

Thank you. A recent study found a link between aged ECM and triggering of invasive cancer-related genes. Have you noticed any changes on ECM viscoelastic properties with aging?

Uh, the first part that was— sorry, what was the first part of that question?

Uh, there's a link between aged ECM and triggering of invasive cancer-related genes.

Yeah, okay, yep.

Yes. Notice any changes in ECM viscoelastic properties with aging?

Yeah, so we've done— there's actually, in general, in this space, you know, there's some characterization of the stiffness of tissues with age. As we know, generally tissue stiffening tends to happen with age, but even there the data sets are actually not all that overwhelming. With viscoelasticity, there's actually very little data on the effects of age. So I can't say anything universal, but we have done some measurements ourselves in the context of development and age, and we do see in general that one loses viscoelastic properties and things become more elastic and less viscous with age, probably because the cross-linking of the matrix tends to increase with age. So we do see in general, but it's not absolute, and A lot of characterization still needs to be done. So we do think that viscoelasticity is actually changing, and it may have some relation to, you know, the reference that the question is— that the individual is referring to here.

Thank you. Cell differentiation is better in viscoelastic substrates. The person is wondering how viscoelasticity is helping cell differentiation.

Yeah, I think so. There's a variety of— so first of all, I got to say is that I, you know, I hope I didn't leave the impression along the lines of the first question that viscoelasticity is always what you need or always what you want. It's going to depend on the context, and so some cells will differentiate better, or some fates may be preferred in more viscoelastic versus less viscoelastic. So it will depend a lot on the specifics, so there's not an absolute there. The mechanisms that the cells use to interpret the viscoelasticity, I'm sure, are quite varied, and there's a number of them. In Kyle's work, for example, what we found is that what seems to be very important is the ability of the cells to actually push against— if you think about when a cell is, you know, reaches out, it attaches, it starts to push against its environment, If the material is purely elastic, as the cell pushes against it, tries to move against it, maybe deposits extracellular matrix, that actually all of that stress is going to be stored in the matrix and pushed back on the cell. In contrast, if it's viscoelastic, then those stresses get dissipated, and so the cell is not under the same level of stress or tension as it is in a viscoelastic substrate. So, some of the classic mechanotransduction pathways seem to be very important here. As I said, you know, Kyle's work, PI3 kinase, seems to be very important, and some other work in the lab, ARPP2III is quite important. Ovi Chowdhury, who I mentioned, did some of the early work in the lab in this space, has been running his own laboratory at Stanford. He's demonstrated that some of the stress-dependent ion channels are very important mediators of some of the effects of viscoelasticity. So there's a few different mechanisms that have been identified already, and I'm sure more that remain to be identified that mediate some of these effects.

Thank you. Can you encapsulate cells in the tough adhesive hydrogels?

Yeah, so it depends which one. So the, the original material was designed to be non-degradable, as we wanted it to actually last for a while, and we weren't worried about incorporating cells because the material wasn't going anywhere. So if one used the, let's say, the fabrication process described in that Science paper— no, the Nature paper in 2017 or '18, I would expect— I don't think we've even bothered to do it, but I would expect most of the cells would die. We have since then developed biodegradable versions of these in which one can encapsulate cells. Either cells can be directly encapsulated, or one can, for example, immobilize the cells within microbeads that are then placed within the gels or the adhesives as they are formed. So, there are a variety of ways of incorporating cells into these devices.

Thank you. Adopting your perspective by which the relaxation time is the main driver of biological response to the microenvironment, What would you reckon is happening on the intracellular side? In other words, if a particular time scale matters for, say, osteogenesis, there should be an intracellular mechanism for which that time scale is crucial as well.

Yeah. So, actually, I want— again, I'm going to kind of keep making the same qualification. Hopefully, I've not tried to give the message that I think viscoelasticity is the most important variable of the extracellular matrix. There is many important— variables, you know, if cells can't adhere, then, you know, many of these mechanotransduction pathways are not relevant. So, you know, the type of adhesion ligand, the specific integrin cells use to engage with the material is very important. The stiffness is important here. You may have noticed in a lot of the data presented that we see effects of viscoelasticity that are very strong. At certain stiffnesses and not as strong— and it's not as important at other stiffnesses. So all these are important variables. What I'm hopefully trying to do today is convince you that viscoelasticity does matter and is one of the things that we should be considering along with these others. So with that prelude, so I briefly talked in response to an earlier question about some of the mechanisms of mechanotransduction. That we've been finding and that we think underlie a number of the effects of interest here. The time scale question is a really interesting one. You know, when we first started going down this path, that I thought it was going to be relatively straightforward and there would be some biological processes that would be impacted by viscoelasticity because the time scales matched up really nicely, and others where it wouldn't be important because the time scales, you know, were too big or just too offset. I've since come to think that it's not quite that simple, and I'll use the example that I gave a moment ago. When a cell is encapsulated within a material and it starts to adhere, and let's say push against the material around it, the cell activities that are happening there are happening on really short time scales. However, when the material is purely elastic, As the cell, let's say, continues to push, as the cell starts to deposit its own extracellular matrix, that takes up a volume that actually is going to continue to basically strain the surrounding material. That's going to increase the amount of stress that's stored in that surrounding material and the amount of stress that then is basically pushing back at the cell. So it's not quite as simple as at least I was first thinking about it, that the biological process and the rate of stress relaxation, you know, had to be aligned, because I think some of these things happen over bigger time scales, recognizing that energy can be stored and dissipated, and how that happens over big time scales may have a really profound importance. Having said that, you know, I don't think, let's say, individual integrin binding events are probably all that dependent on viscoelasticity. I think that formation of focal adhesions, which is happening over timescales of minutes, is probably very dramatically impacted. We've seen actually that there's impact on the rate of stress relaxation because it's happening in that same kind of timescale. So, there's clearly some coupling between the timescales of the cellular processes and stress relaxation, but it's not absolute.

Thank you for the detailed answer. Your data shows that the stiffness can change the level of PD-L1. Is it possible to use this data to explain the efficacy of anti-PD-L1 PD-1 in cancer?

Yeah, so there certainly is some interplay there, though. This is— I would put it— so this is in the context of differentiation of monocytes to dendritic cells. So there's actually a number of repercussions there relative to— or potential repercussions there in relation to immunotherapy. So, I certainly think that there may be some mechanical elements and aspects to signaling that are important to immunotherapy and may give us some new targets to intervene and perhaps make patients who are less responsive to current immunotherapies more. But I don't think one— I can draw a direct relation between our effects of viscoelasticity on monocyte differentiation and let's say response to, you know, anti-PD-1 immunotherapy at this point.

Thank you. One person is curious, how do cells sense viscoelasticity? Which mechanosensing pathways are involved in sensing viscoelasticity? And do we know how viscoelasticity affects focal adhesion formation, maturation, or turnover in 3D culture?

Yeah, so in terms of the latter part of the question, I'd say that we know a little bit that qualitatively that it has an impact on focal adhesions. We, at least we have not done any significant quantitative studies to get at some of the specifics of that question, and I don't believe anyone else has. But I expect that there are indeed, because we see qualitatively some pretty significant impacts. The mechanisms are kind of along the lines of what I described before, that we know that, you know, you know, from Ovi's work that ion channels are part of the story. We know from some of our work and other people's work that it certainly is dependent on adhesion. If the cells can't adhere to the material, we don't see an effect. We see FAK phosphorylation is a key part in some of the contexts we're looking at this. As I mentioned, you know, we see ARPP2III plays a really important role, and the inhibition of ARPP2III inhibits the effects of viscoelasticity in some contexts. Of the biological processes that we're looking at. So, you know, a variety of players. You know, YAP is kind of an interesting one that most people think of as the kind of master regulator of mechanotransduction. Both in our work and in Ovi Choudhury's work, YAP is doing some different things here in relation to viscoelasticity than I think people classically think of it. So I'm confident it's still playing a role. It's maybe just not the same role that we think of it playing purely in the context of stiffness.

Thank you. How stable is the alginate calcium ion system in biological system? Will the calcium ion eventually leak out of the hydrogel in physiological condition?

Oh yeah, yep. So there will be— so certainly an equilibration between the calcium in the gel and the calcium in the surrounding tissue over time. You know, alginate gels do tend to be pretty stable if they're low molecular weight. That's actually one way that we control their dissolution over time or their degradation in the body. Then they tend to, over time in the body, continuously slowly chelate calcium. The high molecular weight ones, where there's more chain entanglement, you actually don't tend to see that. In vitro, what we see is that we'll have an equilibration that typically occurs in the first 24 hours where you'll see some calcium loss from the gels, and then it tends to be quite constant with time after that. And if there is, you know, most of the cell culture media that we use have physiologic quantities of calcium, so let's say about, you know, is it 2 millimolar calcium, and so you'll see fairly stable properties of the gels with time. Though, as over in vivo, as I said, over time courses of weeks and months, you know, then you will see a slow decline, and over, you know, a couple months, you will see some changes. You can also make these gels degrade much more rapidly. It's one of the goals that we have for our non-mechanistic studies, where we want them to exhibit certain properties for, let's say, a few days to a few weeks, and then to be gone. So you can also make these gels degrade actively, in which case then, you know, you'll also end up losing the ionic crosslinks as they degrade, let's say, via hydrolysis.

Thank you. What do you think is the most important problem of the current surgical bioadhesives? You have done great works on adhesive patches. Is it possible to achieve similar adhesion strength with injectable adhesives?

Yeah, so I think there's a variety of approaches that people have been exploring that look really promising. I've not seen anything that gets this level of adhesion energy from an injectable at this point in time, but you may not need it in many situations. It's going to be very situation-dependent what the requirements are. I think the first part of that question was, what do I consider to be something about the most important property? I don't think there is a single most important property. I think in a particular, uh, most important problem— almost important problem. Okay, yeah, that was it. So the most important problem will be dependent on the specific clinical situation. There's not a problem that applies. I mean, if you think about it, medical adhesives are used in many, many different contexts. So along those lines, I don't think there's going to be a medical adhesive that is going to be the answer for everything. I think there's room for actually lots of products that have different properties and properties that actually are really important and ideal for a particular setting. And, you know, there's just so many different ways that these are going to be used in the body or are used in the body that there's not, you know, a single magic adhesive. There's going to be many adhesives that will have utility.

Thank you. For the gel that is used as a drug delivery device for tendons, do you think there's a scope to print this gel into a fibrous scaffold Is there biomimetic tendon graft?

Yeah, so I think it's an interesting idea to have basically, you know, a kind of aligned fibrous network as part of the system, both for mechanical properties as well as potentially, you know, stimulating regeneration and controlling topology. So, yeah, I think one can very readily incorporate a fiber-based material and have composites with these adhesives, as well as other adhesives that other people have described. I think that's something that makes a lot of sense and can readily be done using a variety of techniques, whether it's pre— whether it's kind of traditional, kind of wet casting approaches for making fibers, or 3D printing, or other approaches. But I, yeah, absolutely, I see nice synergy between some of the fibers and using those as part of a composite material with these and other adhesives.

Thank you. Have you measured the mechanical properties of your hydrogels after in vivo studies? Are the mechanical properties of the hydrogels modified over time?

Yeah, so this kind of relates to the earlier question. So I think that they will certainly be related— they will certainly be modified to some level over time. And I'm confident of that, that there will be some changes. Even if we don't— if we make use of what we consider to be non-degradable, I think that they will have some changes with time. I'm trying to think. I don't think we've done— you know, when we typically either use the gel to induce regeneration, in which case we design the gel so it's gone after a couple weeks, or we, you know, like in this adhesive work I described, we've been using it to deliver something, have it be adhered, for example, for a few weeks. So in the former case, we don't go back and look at the mechanical properties because we designed the material to degrade away. In the latter, we've not. Kind of the closest we've come is probably just to look to see whether or not they stay adherent and stay in place. And we know for at least 3 to 4 weeks that these materials stay adherent and stay in place, but we haven't pulled them out and actually done a re-characterization of their mechanical properties.

Thank you. Great talk. It seems like a map of viscoelastic properties of each tissue might be helpful in guiding regenerative medicine. Does such a comprehensive tissue-specific index of mechanical properties exist?

No. And I think that's one of the— that is one of the limitations, is we don't have a very good understanding of what the viscoelasticity is. And actually, if you look into it, even things something as simple as the stiffness of tissues, you know, is actually not that well described outside of a few tissues where people have had a lot of interest. So I think a mapping out of the physical properties in general of tissues and pathologies is something that would be very, very useful.

Well, that's all the— you know, there's many other questions, but I just want to thank you for your time for this excellent talk, Professor Mooney.

Yeah, my pleasure. Have a nice afternoon, everybody. Very nice.