The concept of tissue engineering with part cell and part synthetic materials was proposed nearly 20 years ago. Bhatia explains why one would choose to use an engineered tissue and the challenges of tissue engineering that reproduces the micro-architecture of tissue in vivo. She goes on to describe how the ingredients and the methods of fabrication of a hybrid tissue are chosen.
In the second part of her talk, Bhatia tells us about research from her lab and others to develop an implantable, engineered liver. She explains the challenges of co-culturing hepatocytes and the supporting cells necessary to keep the hepatocytes functional. Her lab has successfully engineered microscale human livers that are showing promise as a mechanism to identify drugs that are toxic to the liver and to study viruses and parasites that attack the liver.
00:00:00.23 Hi, my name is Sangeeta Bhatia. Today I'm going to tell you about tissue engineering.
00:00:05.12 In particular I'll be telling you about engineering tissue replacements for patients.
00:00:09.28 In this seminal paper in 1993 Bob Langer and Dave Vacanti described the concept of tissue engineering.
00:00:19.22 The idea was that one would take a degradable biomaterial, basically a plastic,
00:00:25.25 and mix cells with this degradable biomaterial, culture them together,
00:00:31.00 and create an engineered tissue.
00:00:34.05 These hybrid tissues that are part cell and part material could then be implanted
00:00:40.14 for patients with tissue failure of various kinds.
00:00:43.20 In this same paper they examined the number of procedures per year
00:00:49.20 that were done for patients that had tissue failure of some sort.
00:00:54.24 And you can see that many, many millions of procedures are done
00:00:58.14 for all kinds of medical conditions every year.
00:01:01.07 So, how has this gone?
00:01:04.03 Well, first one of the challenges.
00:01:07.06 If one looks at how tissues are organized in the human body,
00:01:11.04 we see that they're hierarchically organized.
00:01:14.09 So for example, here is an image of the liver and if you zoom in on the liver
00:01:19.14 you can see this fundamental unit of the liver which is known as the liver acinus.
00:01:23.20 This unit is about 1 millimeter in diameter and there are many millions of these unit in the liver.
00:01:30.19 If one zooms in on these further, one can see that within the acinur unit
00:01:36.21 are hepatic cords and these hepatic cords are rows of hepatocytes (liver cells)
00:01:43.05 that are organized in a way to process the blood efficiently as it's flowing through the liver.
00:01:48.25 Finally, if one zooms in all the way to the single cell level,
00:01:53.08 one can see that these hepatocytes, these liver cells, are surrounded by at least 5 other cell types
00:01:59.28 in their home, in their what we call micro-environment.
00:02:03.18 So, these are the challenges of the tissue engineer: to recreate
00:02:08.16 the microenvironment of these cells
00:02:10.24 in a way that there could be many, many of them in healthy microenvironments
00:02:16.07 in tissues that could replace the tissue function of the damaged tissue.
00:02:20.20 So, first you might ask why one would create a bioengineered tissue that's part plastic and part cells?
00:02:30.02 This is a very famous image of one of the first tissue engineered pieces of cartilage
00:02:35.21 in the shape of a human ear in the back of an immune compromised mouse.
00:02:40.19 So why would one do this?
00:02:42.25 Well, tissues engineers would typically do this only if
00:02:46.26 one needed to improve upon one of the following existing medical therapies:
00:02:51.14 if no simple medical device could improve the tissue function.
00:02:57.17 For example, an implantable hip is a piece of metal or a piece of plastic
00:03:02.10 and if that could achieve the tissue function that you were interested in,
00:03:06.08 you wouldn't create a hybrid device.
00:03:08.07 A drug: if one could administer a drug to patient, again, one wouldn't need a hybrid device.
00:03:14.20 A cell that was minimally manipulated: this is the term that
00:03:20.27 has been invented by the United State Federal Drug Administration
00:03:24.05 and it refers to cells that have come out of the body and are re-infused with very minimal manipulation.
00:03:30.14 So if one could minimally manipulate a cell,
00:03:32.28 one wouldn't need to engineer a tissue to improve the tissue function.
00:03:36.27 A surgical reconstruction: if one could simply borrow another part of the patients body
00:03:44.02 for replacement of the tissue that's damaged, surgically, again,
00:03:48.19 one wouldn't need a complex functional replacement.
00:03:52.14 And finally, a transplant: if one could use a cadaveric or donor organ,
00:03:57.09 a whole organ, and replace the tissue function of interest
00:04:00.23 one wouldn't need a bioengineered tissue.
00:04:02.22 But importantly, these criteria are often still not met
00:04:07.17 and one does need to, in fact, turn to tissue engineering.
00:04:10.02 So, there's actually a spectrum of tissues that can be engineered
00:04:15.02 and all of these fall in that category.
00:04:18.19 So engineered tissues include biomaterials and cells in various combinations.
00:04:24.16 And on the first slide I described to you biomaterials that housed cells.
00:04:29.14 So these are plastic, typically, and cells together in a hybrid form.
00:04:34.24 But there are also various incarnations of these engineered tissues.
00:04:40.15 So, for example, there are complex biomaterials that can be implanted and recruit cells.
00:04:46.11 So, upon implantation they don't carry any cells but they can recruit host cells
00:04:51.26 and become hybrid in situ.
00:04:54.11 These are also in the category of what one would call engineered tissues.
00:04:59.04 And finally there are materials that are completely derived from cells.
00:05:04.13 So we know that cells themselves can make scaffolding which is called extracellular matrix.
00:05:09.15 And if one cultures cells in the laboratory, and creates structures that are part cell and part extracellular matrix,
00:05:18.00 those also can count as engineered tissues but they have no synthetic materials within them.
00:05:23.22 So all three of these categories of engineered tissues have been
00:05:29.12 widely experimented on with varying degrees of success
00:05:32.01 and what I'd like to do now is to show you some examples of each of these.
00:05:35.20 So we are going to start with acellular biomaterials.
00:05:38.28 This is an acellular biomaterial which is a very effective engineered tissue.
00:05:46.13 It's called small intestine submucosa.
00:05:50.12 It has no cells whatsoever.
00:05:52.26 So if we look in this bottom image here, what we see if a cross section
00:05:58.04 of the intestine, in this case of a pig.
00:06:01.00 So we see these vila structures of the intestinal lining
00:06:05.28 and that's how the intestine absorbs nutrients,
00:06:08.19 and underneath that structure are these muscular layers.
00:06:12.23 And lying between these two layers is what's called the submucosa.
00:06:17.01 The submucosa looks like this on an electron micrograph.
00:06:22.26 This is a very strong extracellular matrix structure
00:06:26.20 made up primarily of collagen and mixed with lots of other extracellular matrix types of molecules.
00:06:33.17 So what was done originally by Steve Badylak and coworkers and many other groups since,
00:06:38.27 is that one can take porcine small intestine and
00:06:42.25 strip away the cells with various detergent type procedures,
00:06:46.20 and come up with large scaffolds that are acellular, that are very strong mechanically.
00:06:52.25 It turns out that when one implants these at sites where regeneration is required,
00:06:58.02 so for example in tendon and ligament repair,
00:07:01.10 these recruit host cells by inducing proliferation and differentiation of those cells
00:07:07.22 and promote wound healing very effectively.
00:07:10.11 These types of materials have been in millions and millions of patients.
00:07:14.06 Another more complex example is a hybrid example.
00:07:20.13 So I told you before that once can combine biomaterials and cells
00:07:25.11 to make hybrid engineered tissues and this is one of the best successful examples of having done this.
00:07:31.17 So here we see engineered skin.
00:07:34.20 So, on this side we see a cross-sectional micrograph of what human skin actually looks like
00:07:43.02 and we see that there's a dermal layer here and an epidermal layer here.
00:07:47.20 The dermis is composed of connective tissue cells like dermal fibroblasts
00:07:52.20 and the epidermal layer is composed of epithelial cells
00:07:56.18 which undergo what we call stratification as they mature and form this multilayer surface.
00:08:02.12 And way out here is where the skin interfaces with air.
00:08:06.14 So this structure is what's damaged in the case of burns, for example.
00:08:11.07 And tissue engineers sought to create tissue engineered constructs to replace this skin.
00:08:17.20 So, on this side what you see is an engineered version of skin
00:08:21.17 and in this material what's been done is a biomaterial, a collagen gel, an extracellular matrix gel,
00:08:29.25 that has the consistency sort of like jello, has been mixed with dermal fibroblasts.
00:08:35.12 These dermal fibroblasts remodel the collagen and they form very tight mess-work.
00:08:41.19 On top of that mess-work, one can seed those epithelial cells, the epidermal cells
00:08:47.28 and they will stratify and form a multi-layer structure on top of that dermal layer.
00:08:54.25 And this engineered skin can be implanted in the setting of burn or other injured skin
00:09:02.02 and promote wound healing and to great benefit.
00:09:06.11 Another example of a hybrid tissue engineered system is this one.
00:09:11.25 This one was developed originally by Toy Atala and coworkers
00:09:15.00 and in this system instead of using a natural scaffold, like collagen,
00:09:19.19 we're using a synthetic scaffold called polylactic co-glycolic acid.
00:09:25.15 So this is a polymer which I'll describe later in a moment,
00:09:29.03 but for now you can think of it like a plastic
00:09:31.09 and this plastic was cast in the shape of a balloon structure that you see here.
00:09:36.11 On the inside of this balloon, we see that lining cells from the bladder were seeded
00:09:45.03 and on the outside of the balloon muscular cells that are important for bladder function,
00:09:50.04 for the contraction of bladder, were seeded.
00:09:53.03 Together these were cultured in vitro and then implanted in animal experiments and now in humans.
00:09:59.07 And these hybrid tissues that are composed of two cells types and a polymer scaffold
00:10:04.13 together promote better bladder function, better compliance in this setting.
00:10:10.01 Interestingly, in this classic paper, where they first did the experiments in dogs,
00:10:15.23 these hybrid bladders were also re-enervated.
00:10:19.28 So they promoted nerve regeneration in the host
00:10:23.29 and these dogs regained voluntary bladder function.
00:10:27.00 This was unexpected and one of those exciting areas of research that is still unexplained.
00:10:33.16 So I've told you about acellular grafts and hybrid grafts.
00:10:38.25 Now let's talk about completely cellular derived grafts
00:10:42.11 that have no synthetic or natural biomaterials in them.
00:10:45.21 So here's an example of a biomaterial-free engineered tissue.
00:10:50.13 This is blood vessel that's made by Nicola L'Heureux and colleagues
00:10:55.03 and in this system they do what they call sheet-based tissue engineering.
00:10:59.15 The vision is that one would take a sample from a patient, a little biopsy sample,
00:11:04.23 from which one could harvest endothelial cells which line blood vessels
00:11:09.05 and fibroblasts, those same dermal cells that I described earlier, and culture them in flasks.
00:11:15.16 And if one cultures them in flasks...if one cultures the fibroblasts in flasks,
00:11:20.02 one can see that the fibroblasts deposit their own extracellular matrix.
00:11:25.05 They deposit the same types of collagen that you saw in the small intestinal submucosa earlier,
00:11:31.08 but they deposit it on the bottom layer of the flask.
00:11:34.01 And one can manipulate the culture conditions to promote
00:11:37.00 lots and lots of extracellular matrix deposition.
00:11:40.11 Then one can detach the system and it creates a sheet
00:11:44.14 that can be manipulated and rolled on a mandril.
00:11:47.15 This mandril then, now is a living cinnamon roll of cells and extracellular matrix
00:11:56.13 and it's cultured in this living format so that
00:11:59.19 the layers of the cells and the extracellular matrix will fuse into a very strong wall.
00:12:04.28 One can then seed the inside of these tubes
00:12:08.19 with the endothelial cells that were originally obtained
00:12:11.27 and now you have a construct that has endothelial cells on the inside
00:12:15.26 and this nice, strong surrounding structure made of fibroblasts and their extracellular matrix.
00:12:23.23 And these can be implanted in patients.
00:12:27.01 And these are what these tissue engineered blood vessels look like.
00:12:31.02 These blood vessels have no scaffolding material that was introduced from the outside.
00:12:40.15 OK, so these are the different flavors of engineered tissues that exist.
00:12:44.12 If one were thinking about designing a bio-engineered tissue to replace a particular organ function,
00:12:50.17 one important thing to note is that, as tissue engineers,
00:12:53.23 we can frequently...we are unable to replace all of the functions of interest.
00:13:00.03 So, for example, if one thinks about the functions of skin that are are being replaced,
00:13:04.03 it's simply the barrier function in the example I showed you.
00:13:07.16 Those pieces of skin have no hair follicles, no sweat glands, no immune cells,
00:13:13.19 and that's a deliberate choice on the part of the tissue engineer
00:13:17.13 to go after the function that's most critical for life.
00:13:20.09 Similarly, if one thinks about pancreatic tissue engineering,
00:13:24.10 one often sees that folks are interested in the implantation of beta cells
00:13:28.24 that will produce insulin in a glucose responsive way for patients with diabetes.
00:13:33.10 There are many other functions of the pancreas
00:13:36.02 that are often excluded there from tissue engineering
00:13:38.28 and again this is a deliberate decision on the part of the tissue engineer.
00:13:42.21 So it's important to keep in mind that one can't do everything, one has to have priorities.
00:13:48.00 OK, well then, once you've made your decision about what functions you're most interested in,
00:13:54.02 what would one then to go further and actually engineer a tissue from scratch.
00:13:58.22 So the next thing to do would be picking some raw ingredients.
00:14:02.22 The raw ingredients that one has to work with are cells.
00:14:06.09 There's actually a variety of cell sources that one can consider.
00:14:10.06 So there are somatic cells and stem cells.
00:14:13.24 Somatic cells are cells of the body that are not germ cells, that is, not eggs or sperm.
00:14:21.08 So they are part of the adult organism.
00:14:23.18 And they can come from the person that you plan to treat. Those are called autologous.
00:14:29.22 They could come from another person. That would be allogeneic, of the same species.
00:14:36.10 Or they could come from an organism that is another species.
00:14:40.14 and those would be called xenogeneic.
00:14:42.11 And in tissue engineering one actually sees examples of all of these types of cell sourcing.
00:14:47.18 So, for example, the skin that I showed you earlier was allogeneic.
00:14:52.05 The dermal fibroblasts and the epidermal cells
00:14:55.12 were from a non-self human to create those skin tissues.
00:15:00.21 Another source of cells, which is becoming more and more exciting, is stem cells.
00:15:06.02 One can think about using both adult stem cells or pluripotent stem cells
00:15:10.13 that have been derived from various methods that I'll describe in a moment.
00:15:13.20 So one certainly needs raw ingredients of cells.
00:15:16.29 The other thing one needs is biomaterials.
00:15:20.07 And finally one needs to cultures these together in a way
00:15:23.25 that the cells have sufficient nutrients to remodel those biomaterials and create a tissue.
00:15:30.00 Once one has the right raw ingredients, we have to think about fabrication.
00:15:35.13 How to assemble these structures, how to process these, how to create culture environments
00:15:41.19 how to preserve them so they can get to the clinic
00:15:44.28 and finally how they can then integrate with the host at that site.
00:15:49.03 So in the next few slides I'm going to talk to you about some of the issues
00:15:52.21 that are related to both the raw ingredients and the fabrication of engineered tissues.
00:15:56.27 First, let's talk about cells.
00:15:59.07 So, if one is thinking about somatic cells, culturing primary cells,
00:16:05.12 they can be derived in various ways;
00:16:07.24 from an adult organism, from an embryo, or from an egg.
00:16:11.28 Regardless of the source, what's typically done is these are mechanically dissected
00:16:16.14 in the laboratory and then they can be further processed in one of three ways.
00:16:20.14 If they are just slightly processed and they retain their three-dimensional structure
00:16:27.20 from the in vivo environment, we call that organ culture.
00:16:31.18 These organ cultures are often done at air-water interfaces
00:16:35.04 that mimic the environment of particular tissues in vivo.
00:16:39.17 For example, the skin that I showed you earlier is actually cultured at the air-liquid interface,
00:16:45.03 just like the skin on your body is at an air-liquid interface.
00:16:48.05 One could further dissect these tissues, finely chop them,
00:16:51.28 and culture them in a flask and allow cells to grow out of these structures.
00:16:57.19 That's called explant culture.
00:16:59.11 And that's actually the way that the bladder cells were first derived that I showed you earlier.
00:17:03.26 Finally, and more commonly now, these tissues can be enzymatically digested
00:17:09.22 to create a single cell suspension, a suspension of cells in a liquid.
00:17:14.18 And those can be cultured in flasks and if they are a proliferative cell, if they can grow,
00:17:19.20 once they grow to what we call confluence, once they fill up that flask,
00:17:24.13 one can remove them, put them in a new flask and passage them.
00:17:28.16 That's called passaging--grow them, expand them, over and over again.
00:17:32.00 And that's how you would create the cells for your engineered tissues.
00:17:35.26 One limitation of using somatic cells was first described by a scientist named Hayflick in 1965.
00:17:45.26 And that's known as the Hayflick limit.
00:17:48.02 So, what he observed in his experiments was the following.
00:17:52.04 If one takes fibroblasts, those same fibroblasts that I was describing to you earlier,
00:17:56.11 from patients of different ages (so, fetal fibroblasts all the way up into your 80s)
00:18:03.24 and cultures them and passages them, in the flasks as I showed you and counts doubling time,
00:18:10.04 one sees that the number of cell doublings that one can achieve is less as the patient source is older.
00:18:19.21 So as patients age, their cells are able to double a fewer number of times.
00:18:25.23 And what that implies is that cells have a finite lifetime in culture,
00:18:31.05 and, in fact, in vivo.
00:18:33.00 And we know now that the molecular basis of this is that
00:18:36.00 the ends of chromosomes have something called telomeres, the so called molecular clock,
00:18:40.29 and every time the cell divides, those telomeres get shorter until the cell reach senescence.
00:18:46.18 So, this observation has led to much of the excitement in the field of tissue engineering
00:18:53.03 about stem cells because in fact stem cells don't have this property.
00:18:57.11 They are what we call immortal.
00:18:59.06 They elongate their telomeres every time they divide
00:19:02.21 by expressing an enzyme known as telomerase.
00:19:05.06 They actually are not susceptible to the Hayflick limit
00:19:08.26 and they can grow in perpetuity in culture.
00:19:12.06 So as a cell source they are very exciting resource.
00:19:15.18 Furthermore, stem cells have been identified that are pluripotent.
00:19:23.27 So here's a diagram describing a number of ways that one can make
00:19:27.29 stem cells that can differentiate, that can be encouraged
00:19:33.23 to grow into the various cell types of interest for any tissue of interest.
00:19:38.15 The idea here is that you could have a single cell source.
00:19:41.20 It could be immortal. It could be expanded.
00:19:44.14 And then, it could be used to engineer any tissue of interest.
00:19:47.10 A further idea that's been very exciting in the field
00:19:50.19 is the idea that one could make individualized pluripotent stem cells.
00:19:55.18 This is the idea of patient specific, or personalized medicine.
00:19:59.14 So let me walk you through how this idea goes.
00:20:02.05 There are a variety of ways to make these cells.
00:20:05.03 So, in this system what we see is that you take, again, a fibroblast, a skin fibroblast
00:20:09.19 just like the ones I showed you earlier
00:20:11.17 and they have two copies of DNA. So these are somatic cells.
00:20:18.00 One can take out the nucleus of that cell and put it in an enucleated egg.
00:20:23.21 Those cells then will start to grow into a blastocyst structure shown here.
00:20:29.29 And the blastocyst, which is the early embryo, has a pocket of cells known as the inner cell mass.
00:20:36.14 If one cultures cells from the inner cell mass,
00:20:39.22 one can create so called pluripotent stem cells or embryonic stem cells.
00:20:44.25 Pluripotent stem cells, the embryonic stem cells, will have the same genetic identity
00:20:49.28 as the original patient.
00:20:51.15 Another way to create pluripotent stem cells from that patient
00:20:55.15 would be to take the same somatic cell that we saw earlier
00:20:58.15 and fuse it with an embryonic stem cell.
00:21:01.13 And that process of fusion also generates patient specific pluripotent stem cells.
00:21:07.12 Finally, recently, in a very exciting set of discoveries,
00:21:11.03 Yamanaka and coworkers have defined a way to reprogram somatic cells
00:21:16.26 without going through an ES cell intermediate or an egg intermediate.
00:21:21.29 What they did, in their classic experiments, was show that
00:21:25.18 a certain number of molecules known as transcription factors
00:21:29.06 could be expressed in a somatic cell and that transduced cell would be reprogrammed
00:21:36.09 into a pluripotent stem cell.
00:21:38.02 So, this now offers the potential to create differentiated cells
00:21:43.19 that are genetically matched to patients.
00:21:46.27 It offers the further possibility to make pluripotent stem cells
00:21:50.27 from patients to fix a genetic defect in the laboratory
00:21:55.14 and then to create a tissue out of those genetically altered cells
00:22:00.07 to give back to the patient.
00:22:01.29 So, given the Hayflick limit,
00:22:05.20 the promise of stem cells has been very exciting in the field of tissue engineering.
00:22:10.21 OK, so we've talked about cellular ingredients. Now let's turn to biomaterials.
00:22:17.00 So, one classical tenant of biomaterials in the tissue engineering field
00:22:23.13 is that one wants to match the degradation of the biomaterial
00:22:29.17 with the synthesis of extracellular matrix scaffolding in the host.
00:22:34.12 So, one would not want the biomaterial to degrade too quickly.
00:22:39.01 Otherwise, it wouldn't perform its scaffolding function.
00:22:42.15 On the other hand, one typically would not want the biomaterial to be persistent indefinitely.
00:22:48.21 Because one might incur what's called to foreign body response.
00:22:52.02 So, where does this concept come from?
00:22:54.26 This concept was originally defined by Yannas and coworkers
00:22:58.27 in these classic experiments that were done in the skin.
00:23:03.01 In these experiments, what they did was they made tissue engineered skin
00:23:07.27 that had a variety of degradation rates.
00:23:10.25 And what they looked at was the wound healing time of a full thickness wound,
00:23:16.07 a cut through both the epidermal and dermal layers of the skin in an animal model.
00:23:22.09 And what they found was that if one didn't put any scaffolding in that wound,
00:23:29.19 it would contract and close very quickly and form scar.
00:23:35.02 That was what we were hoping to prevent with the use of the tissue engineered scaffold.
00:23:40.18 If one then implanted tissue engineered scaffold that has a very rapid degradation time,
00:23:46.09 those wound proceed to contract, just as in the control conditions,
00:23:53.02 as if there was no scaffold in the system at all.
00:23:55.04 However, if one further cross-links the material,
00:23:58.24 one can see that one can delay that contraction time as seen here.
00:24:05.18 And then after that, if one makes the material even less degradable,
00:24:09.11 there's really no benefit to that and the wound actually just persists the same amount of time.
00:24:13.25 So these experiments led to the idea that there's an optimal degradation rate
00:24:19.07 for tissue engineered scaffolds
00:24:22.06 and one can actually see that in the SYS experiments, in this bottom graph here,
00:24:27.13 they conceptualized this as follows.
00:24:29.21 So, the degradation time of the material is decreasing
00:24:33.06 and the production of extracellular matrix is increasing.
00:24:38.26 And the SYS actually bridges the gap in those two time scales.
00:24:43.20 So, it's important to match the degradation with the host synthesis.
00:24:47.25 So that's a material property that's important. What are the materials actually?
00:24:53.18 So there are a variety of synthetic scaffolds that one has to choose from
00:24:58.06 and if one looks in the community we see new chemistries coming out on a daily basis.
00:25:03.17 I've just listed for you here some of the classical chemistries
00:25:07.22 that are used for tissue engineering.
00:25:10.07 Many of them are what we call polymers. So, for example, PLGA is polylactic co-glycolic acid.
00:25:17.26 These are macromolecules that have repeating structures.
00:25:22.03 In this case, a lactic acid and glycolic acid.
00:25:25.16 And when they're implanted and they degrade in the body by hydrolysis,
00:25:30.13 by the interaction with water, they degrade into units of lactic acid and glycolic acid,
00:25:36.02 which already occur in your body.
00:25:38.10 So these are actually originally what sutures were made of, biodegradeable stitches.
00:25:45.01 So there are many, many novel, polymeric materials that have been developed.
00:25:50.09 Another very popular one which has the property of being a hydrogel...
00:25:54.14 Again, a hydrogel is a material that absorbs a great amount of water and
00:26:02.02 has a sort of intuitive feel of jello, gelatin.
00:26:04.27 Here we have a polyethylene glycol which is a very popular biomaterial these days
00:26:11.15 and you can see that it looks a lot like a soft contact lens.
00:26:14.05 What's interesting about some of these hydrogel materials
00:26:17.08 is that rather than being pieces of plastic, where cells can be attached to the plastic and they can remodel the plastic,
00:26:26.07 these are materials where cells are imbedded within the plastic.
00:26:29.28 And this material in particular is very interesting because it can be crosslinked with light.
00:26:35.04 So the way the experiment works is
00:26:37.10 you take polyethylene glycol macromers (so long chains of polyethylene glycol).
00:26:43.19 You mix in cells and a chemical that's sensitive to light
00:26:47.22 and when you shine light you get a network of polymer
00:26:51.17 that looks like this contact lens structure with cells imbedded within it.
00:26:55.09 So these are examples of synthetic scaffolds
00:26:58.27 and one can tune these to have different material properties.
00:27:02.08 They can be inert, like polyethylene glycol.
00:27:05.05 You can modify their chemistry to interact specifically with cells.
00:27:09.17 For example, one can put in peptides that can bind to cellular receptors like integrins.
00:27:15.08 They can have different mechanical properties
00:27:17.08 which we know is important for some cells types.
00:27:19.19 And then they can have different mechanisms of degradation.
00:27:22.04 So they don't all degrade just by interactions with water.
00:27:26.00 Some of them can degrade in very specific ways;
00:27:28.07 for example, by interacting with enzymes that the cells are making.
00:27:31.18 Another example of a kind of scaffold that one would use are the so called natural scaffolds.
00:27:38.11 So I already mentioned collagen earlier.
00:27:40.26 Collagen is one of the most popular natural scaffolds.
00:27:43.29 Hyaluronic acid is another natural scaffold.
00:27:46.13 The SYS material that I showed you is a more complex natural scaffold
00:27:51.00 because it contains all the extracellular matrix molecules that were in the submucosa
00:27:55.16 in addition to the growth factors
00:27:59.03 and other bioactive molecules that would be in that natural material.
00:28:03.25 An example of such complex natural scaffolds that is becoming quite popular these days
00:28:11.23 is an idea that was first described by Doris Taylor's group a couple of years ago
00:28:15.27 and these are whole organs that have been de-cellularized
00:28:19.23 in much the same way that the small intestinal submucosa was de-cellularized.
00:28:23.26 So here what you're looking at is a picture of a heart that's being perfused with the detergent.
00:28:30.12 And you can literally see the cells melting away
00:28:33.13 so that you have what they call a ghost heart at the end.
00:28:37.01 This preserves all of the microarchitecture of the organ that I described to you earlier.
00:28:43.10 You can see here under the microscope that it has a fibrolar collagen organization
00:28:49.14 and what she did in these experiments was re-cellularized this de-cellularized heart
00:28:55.29 using that as a scaffold for tissue engineering.
00:28:59.23 So various groups have now gone on to do this in other organ systems including the liver.
00:29:06.07 And what's exciting about this potentially is one can think about reusing organs.
00:29:12.03 Instead of having to just transplant a living organ from a donor into a patient,
00:29:17.11 one can think about using cadaveric organs as complex scaffolding
00:29:22.03 for cells from some of the cell sources that I described earlier.
00:29:26.00 OK, so I've described the raw ingredients; cells and biomaterials.
00:29:31.01 I won't say much about nutrients except for to say that
00:29:33.24 there is a whole science about how to feed cells in culture.
00:29:39.10 And that's pretty well understood at this point.
00:29:42.19 The next thing I'd like to just touch on briefly is how to fabricate these organs;
00:29:47.15 how to assemble them, how to process them before implantation and
00:29:51.17 importantly, how to preserve them so that one can get them to the clinic at the time of need.
00:29:57.11 OK, so I mentioned earlier that tissues are organized hierarchically.
00:30:03.07 One can see here in this diagram that when one starts at the level of the liver,
00:30:07.23 we have this multi-scale organization where the liver acini are 1 millimeter in dimension,
00:30:14.17 the liver cords here are 500 microns long,
00:30:17.22 and the cells here are about 20 microns in diameter.
00:30:21.07 So it's this hierarchy which tissues engineers have to try and grapple with
00:30:26.07 in the sense that cells respond, actually, only locally to their microenvironment.
00:30:31.07 So even though you're trying to build a large structure,
00:30:34.12 if you care about cells performing the functions of interest
00:30:38.22 and here are some functions that one might care about,
00:30:41.04 proliferation, differentiation, apoptosis, migration, or metabolism,
00:30:46.23 If these are the functions of interest for having a healthy tissue,
00:30:49.29 that we know that the cells are actually responding to their local microenvironment
00:30:54.23 to these kinds of cues; soluble factors, interactions with other cells,
00:30:59.17 extracellular matrix interactions, physical forces, and cell shape.
00:31:04.24 So the challenge for the tissue engineer is to maintain the cellular microenvironment
00:31:09.25 that will impact the cell behavior on the sort of 10 to 100 micron length scale
00:31:15.12 but simultaneously be scaling up the structure
00:31:18.06 to be a large enough body of cells to implant to achieve a therapeutic outcome.
00:31:23.17 So a good example of how cells have local microenvironmental interactions
00:31:30.12 that one calls 3-dimensional in the field
00:31:33.09 is cartilage.
00:31:34.09 We see this is cross section of articular cartilage
00:31:37.17 and if one looks at cartilage one sees that the chondrocytes
00:31:40.23 that make the extracellular matrix that makes up cartilage,
00:31:43.26 have different organizational structures.
00:31:46.09 So here we can see that they are in these columnar structures.
00:31:50.06 Whereas, in the superficial layer of cartilage that sits right next to the joint space
00:31:54.25 they actually are flat, elongated and in clusters.
00:31:58.24 So if you were to look down at the cartilage, you would see that they were in clusters.
00:32:02.18 So, it's been known for quite some time now that chondrocytes in culture,
00:32:09.07 when one puts them on plastic of any type for tissue engineering purposes
00:32:12.28 they will actually spread out and lose many of these functions
00:32:17.14 They will start proliferating and stop producing the extracellular matrix
00:32:20.25 that one is trying to recapitulate in a tissue engineered construct.
00:32:24.23 And so, in fact, if one embeds them in a 3-dimensional microenvrionment,
00:32:28.24 where they have this round shape and more favorable mechanical properties,
00:32:32.23 one stimulates them as they are stimulated mechanically in vivo,
00:32:36.18 one can actually improve the quality of the extracellular matrix that they produce
00:32:41.23 and therefore the mechanical strength of the graft.
00:32:44.26 So this is just one example of a tissue where
00:32:47.11 the 3-dimensional organization of the microenvironment,
00:32:50.14 in addition to the chemical composition of the microenvironment is very important.
00:32:54.19 And one will hear a lot about 3D tissue engineering in years to come.
00:32:58.09 One can then think about how to assemble these microenvironments
00:33:03.18 where cells have very precise, local chemical and mechanical cues into larger structures.
00:33:09.12 There's a set of technologies emerging as 3D fabrication technologies
00:33:14.05 that have come out of the rapid prototyping field
00:33:17.23 that are starting to be borrowed in tissue engineering.
00:33:20.10 Some people call this organ printing.
00:33:22.12 This is an example of a couple of parts that were made in a layer by layer fashion
00:33:27.17 known as stereolithography
00:33:29.28 and stereolithography is one example of one of these rapid prototyping technologies.
00:33:34.23 So, in this technology, what's done is
00:33:37.28 a 3-dimensional drawing is drawn on the computer and sent to this robot.
00:33:43.14 And the robot has a movable stage that sits in a vat of light sensitive material.
00:33:50.16 And in every layer of the material, light is shone in a pattern.
00:33:55.21 It polymerizes and the stage drops and the next layer is built and so on and so forth.
00:34:00.28 And using this, one can create complex structures; some that are actually biomaterial structures.
00:34:06.06 Some folks envision that one could make personalized tissue engineered parts
00:34:10.26 using a technology like this because one can envision
00:34:13.09 going from a 3-dimensional medical imaging data like CAT scan data, to tissue engineered parts.
00:34:24.02 Another thing that's required is to culture these tissues before they're implanted
00:34:28.18 and these are a couple of examples of the kinds of bioreactors that are designed and used in the field.
00:34:34.06 At the top what you see is a quite well known bioreactor developed by NASA.
00:34:39.06 This is a circulating bioreactor and it never allows the cells to touch the walls of the bioreactor
00:34:47.02 so they're forced to aggregate in 3-dimensional clusters.
00:34:49.29 And this has been shown to promote tissue formation in both cartilage and heart cells.
00:34:55.01 Furthermore, techniques from the computer chip industry
00:35:00.13 are now being borrowed to create micro-reactors.
00:35:03.22 So here what you're looking at is a coverslip and some inlet ports from a microfluidic device.
00:35:09.18 And if one looks in the microscope, inside that coverslip,
00:35:12.19 one can see these little tissue units that have microstructure,
00:35:16.24 that have been fabricated in the computer lab at 100 micron length scale.
00:35:22.06 So they have 3-dimensional microenvironments
00:35:25.01 that have been specified in a scalable fashion for building larger pieces of tissue.
00:35:30.17 So, one can assemble tissues, one can culture these to mature them in the laboratory,
00:35:37.08 and then finally one needs to think about how to preserve them the deliver them to the patients.
00:35:41.16 So, there's a whole set of technologies that have grown up in the field
00:35:47.06 about cryopreservation. I should say that not all tissues are frozen.
00:35:51.18 Many are actually just refrigerated to 4 degrees C
00:35:55.06 and in fact, the skin tissue engineered product that I showed you earlier
00:35:58.02 is sent that way to the clinic.
00:36:00.13 However, in the field of cryopreservation,
00:36:03.15 which most of us agree will be important to really make this technology have it's highest impact.
00:36:11.03 There has been a lot of research on how best to make cells, especially in 3-dimensional structures,
00:36:17.10 survive the insults of cryopreservation.
00:36:19.18 So, the insults of cryopreservation have now been categorized
00:36:23.21 into two classes of insults and those are depicted here.
00:36:27.03 So I need to walk you through what a cell feels as it gets frozen
00:36:31.14 so you can see what the two classes of insults are.
00:36:34.25 So if you have a cell here and it's been cooled just under the seeding point of ice,
00:36:40.03 and one artificially seeds external ice in the environment, two things can happen.
00:36:45.15 If one slowly cools the solution further, you can see that water has time
00:36:52.03 to cross the cell membrane and leave the cell and form ice in the extracellular space.
00:36:57.25 And this causes what people call dehydration injury to that cell.
00:37:03.07 You can see that its shape has shrunken. There is a concentration of the intracellular solutes.
00:37:09.19 And those cells when they are thawed again, exhibit damage from this dehydration process.
00:37:15.13 If instead one takes this solution and rapidly cools it,
00:37:20.03 one actually finds that you can have intracellular ice crystal formation.
00:37:25.02 The formation of intracellular ice crystals, in and of itself, can cause damage to cells.
00:37:30.10 So the process by which cells are susceptible to slow cooling or fast cooling
00:37:35.11 turns out to be cell type dependent and dependent on various additives in the system
00:37:40.21 as well as the 3-dimensional architecture of the tissue.
00:37:44.01 And so this is being worked out now for many, many cell types in the field.
00:37:47.18 OK, so to conclude, I've told you that engineered tissue replacements can
00:37:52.15 be made from combinations of cells and biomaterials to replace a subset of tissue functions.
00:37:58.17 I've told you that the cells can be derived from somatic cells or stem cells.
00:38:03.06 I've told you that the biomaterials in these tissue replacements can be natural or synthetic.
00:38:09.02 I've also told you that tissue structure in the body is hierarchical
00:38:13.18 and therefore the tissue engineer has the challenge of maintaining the cellular microenvironment
00:38:18.27 in a way that can be scaled up to build a tissue that's large enough for therapeutic impact.
00:38:25.03 And finally, I hope you've noticed that the convergence of
00:38:28.20 cell biology, medicine and engineering is really advancing this field.
00:38:33.11 In part two of my seminar, I"ll be telling you about our work on liver tissue engineering
00:38:39.05 both in the progress we're making towards therapeutic goals like the ones I've described,
00:38:44.18 as well as progress that we're making in using these engineered liver tissues
00:38:48.20 to advance scientific discovery.
- Describe your favorite bioengineered tissue replacement mentioned in the first 12 minutes of the video. Given the opportunity, what tissue replacement would you engineer and why? What category would it fall in: acellular, hybrid, or hybrid with cell-derived ECM?
- What is the major limitation in culturing primary cells and how do stem cells bypass this?
- Why is it crucial to match the degradation of the biomaterial with synthesis of the ECM?
- What are the two types of insults imposed on tissues by cryopreservation? Can you think of other biological substances in which this might also be an issue?
00:00:00.00 Hi, my name is Sangeeta Bhatia
00:00:02.08 and today I'm going to be telling you about tissue engineering of the liver.
00:00:06.21 To orient you, this is a schematic of the liver.
00:00:12.24 The liver has over 100 billion cells and performs over 500 functions in your body.
00:00:19.04 300 million people worldwide suffer from liver disease
00:00:23.29 which one can acquire from viral insults like hepatitis
00:00:27.13 or a drug insult like chronic alcohol exposure
00:00:32.05 or even an over the counter drug in high doses.
00:00:34.15 What happens when the liver gets damaged is it forms scar tissue
00:00:38.27 seen here in the middle panel.
00:00:40.23 The scar tissue evolves into this nodular pattern known as cirrhosis.
00:00:46.09 A cirrhotic liver is prone to developing cancer over time.
00:00:51.26 So patients who have liver disease often progress to liver failure and liver cancer.
00:00:58.11 The gold standard for treatment of these patients is whole organ transplantation.
00:01:03.24 However, there are not enough donor organs available for transplanting those in need.
00:01:08.21 Because of this and because the liver provides so many functions that are vital for life
00:01:14.25 many groups have thought about supporting liver function with cell based therapies
00:01:20.24 wherein the cells would be hepatocytes,
00:01:23.24 the cell of liver that performs most of those 500 functions.
00:01:28.06 The idea is that living cells would perform liver function in a variety of hybrid formats.
00:01:35.05 So, here I'll just show you two examples.
00:01:38.03 These are extra-corporeal devices.
00:01:41.00 They are akin to kidney dialysis machines.
00:01:44.12 The blood would run outside the body through this cartridge
00:01:48.06 and the cartridge would house living hepatocytes that would process the patient's blood.
00:01:53.06 The idea is that this device would bridge the patients to transplantation
00:01:57.15 or even support the patient while the liver regenerates
00:02:01.12 because, interestingly, the liver is one of the few tissues that actually can regenerate
00:02:06.03 in certain kinds of injury.
00:02:07.25 Another idea which our group has been pursuing as well as several others
00:02:14.17 is this idea of an implantable, tissue engineered liver.
00:02:17.24 This would one day, we hope, serve as an adjunct or an alternative to liver transplantation.
00:02:24.24 The problem in the field has been that liver cells in these devices
00:02:30.13 do not perform the 500 some odd liver functions that are necessary for patient support.
00:02:36.07 We know from a variety of work in the field that this is because the liver microenvironment,
00:02:42.21 depicted here, is disrupted when the cells are isolated and
00:02:46.24 put in contact with the plastic of the dialysis cartridge or the polymer scaffolds
00:02:52.15 that are going to be implanted in people.
00:02:54.27 So the liver microenvironment, as you can see, is quite complex.
00:02:58.29 Here we see a repeating unit of the liver.
00:03:01.10 This is called the liver acinus and each of these units is about 1mm in diameter.
00:03:07.21 The hepatocytes, the cell of interest that performs most of those 500 functions,
00:03:13.05 are aligned in these cord-like structures.
00:03:16.13 These are actually 3-dimensional structures known as hepatoplates
00:03:19.27 and you can see that along each one of these cords lies a blood vessel.
00:03:24.21 So each hepatocyte is only one or two cell widths from the blood flow
00:03:30.10 and that provides them the ability to very efficiently process the blood;
00:03:36.18 both to metabolize drugs and hormones in the body
00:03:41.26 as well as to produce secreted products for the blood stream.
00:03:46.15 For example, your liver makes most of the clotting proteins
00:03:49.27 important for clotting blood after you cut your finger.
00:03:53.18 In addition, it makes albumin which helps you keep your fluid inside your blood vessels.
00:03:59.18 It metabolizes all the drugs that you take from your doctor.
00:04:05.04 And in addition it regulates energy metabolism.
00:04:10.23 So all of these functions are important and they reside in these hepatocytes
00:04:15.17 that are dependent on this very complex microenvironment to function.
00:04:20.07 When we remove hepatocytes
00:04:22.20 in order to incorporate them into one of these cell based therapies
00:04:25.25 and put them in culture this is what happens to their function.
00:04:29.23 So here I'm showing you data for just one liver specific function
00:04:33.22 and that's the production of albumin
00:04:35.18 which is important for oncotic pressure, the fluid pressure in your blood stream.
00:04:40.27 And you can see that these are hepatocytes here.
00:04:43.12 They've been cultured on collagen coated tissue culture plastic.
00:04:47.01 And after a couple of weeks in culture they're spread out, and they're dying,
00:04:51.20 and all of their liver specific functions dramatically decline.
00:04:55.11 So in our group, what we've been intersted in is
00:04:58.19 seeing whether we could recreate a happy microenvironment
00:05:02.29 for these hepatocytes outside of their native microenvironment
00:05:07.03 in such a way that it could be useful for incorporation into a therapeutic device.
00:05:13.09 Because the environment of the cells in vivo is specified at the 10-100 micron length scale,
00:05:21.05 we have been drawn to technologies that allow us to manipulate 10-100 micron length scale.
00:05:27.08 And in fact, if one looks at the computer revolution,
00:05:31.02 the computer revolution was driven by miniturizartion technologies that have exactly this property.
00:05:37.12 So what I'm showing you here is what we achieved as a community, over 50 years of progress
00:05:43.28 from a single transistor to 100 million transistors
00:05:47.21 through a very simple process known as photolithography.
00:05:51.19 So this is a process by which computer chips are made.
00:05:56.07 And what's done is one takes crystalline silicon and coats it with a light sensitive material,
00:06:01.26 shines light through that light sensitive material and creates a pattern on the surface.
00:06:06.29 Now if you were making a computer chip, one could use that pattern then to make
00:06:11.14 integrated circuits, resistors, transistors and so on.
00:06:15.11 But we and others have been thinking about using these technologies
00:06:19.05 to create cellular microenvironments for the applications of tissue engineering that I mentioned.
00:06:25.08 In this talk, what I'll be telling you about is our work
00:06:29.13 in constructing human tissue microenvironments for the liver,
00:06:34.03 our interest in interrogating these microenvironments in high throughput
00:06:39.25 for, for example, drug development,
00:06:42.05 and furthermore, our interest in using these liver tissues
00:06:46.21 for studying the interaction with pathogens that only normally interact with the human host.
00:06:52.16 So I'm going to start by telling you how we are interested in
00:06:55.28 exploiting these tiny technologies for constructing engineered liver tissues.
00:07:01.14 So the first step that I need to tell you about is
00:07:05.13 how we adapted that same process I mentioned earlier which is called photolithography,
00:07:10.08 where normally one patterns metals on silicon,
00:07:14.17 here to pattern cells on glass.
00:07:18.00 And what we did was we took the same liver cells that I showed you before,
00:07:21.11 now they're color red, and we surrounded them with another cell type
00:07:26.00 So we knew, in the liver, in its native microenvironment,
00:07:28.27 that there's at lest four other cell types that are close to the hepatocytes
00:07:32.24 and so here we've just added one other cell type back
00:07:35.08 and this is a randomly organized co-culture of two different cell types.
00:07:40.18 What we did with photolithographic processing
00:07:43.21 was create what we called a micropatterned co-culture and that's shown here.
00:07:48.10 In this system what we've done is use the same light-based systems
00:07:51.27 to pattern collagen extracellular matrix on a glass coverslip.
00:07:56.06 The hepatocytes (the liver cells of interest) stick on those spots
00:08:00.20 and then we can fill in the spaces with the supportive cells.
00:08:03.23 And in this way we can control the architecture of the tissue in a two-dimensional plane.
00:08:09.23 And we found over the years is that the geometry of this architecture,
00:08:15.17 the control over which cellular neighbors are right next to the hepatocyte,
00:08:20.06 whether they're self (homotypic) or non-self (heterotypic),
00:08:24.22 determines the level of function that one gets.
00:08:27.04 And in particular, for human tissues that one can rescue the phenotype of the cells,
00:08:32.24 those 500 functions of interest, for about 4-6 weeks.
00:08:36.16 In addition, we've been interested in using another version of microfabrication technology known as MEMs.
00:08:43.05 In this process, instead of just making a flat, planar patterned structure,
00:08:47.29 what groups have done over the years is develop etching techniques
00:08:52.08 to etch crystalline silicon like this part that you see here.
00:08:55.13 So this part is first patterned with light
00:08:58.13 and then deep-reactive ion etching is used to etch out the structure that you see there.
00:09:03.23 Using this part, we were able to manipulate our cells at the microscale.
00:09:09.09 So here what we've done...You're looking at two interlocking combs.
00:09:12.17 You seed one cell population on one part and the supportive population on the other part.
00:09:18.01 And what we're able to do is move them together and apart
00:09:22.03 and ask questions about whether they need to signal each other continuously or reciprocally.
00:09:28.09 So, for example, in this experiment, in this movie, what's being done
00:09:33.13 is the cells are being brought together
00:09:35.26 using a simple micropipetman manipulation in a biosafety hood.
00:09:41.15 And you can see that they've been brought together in what we call gap mode,
00:09:45.02 where they're separated by an 80 micron gap but not allowed to touch.
00:09:49.25 Now they've been brought into contact mode where they're allowed to touch.
00:09:55.23 And now they're separated completely.
00:09:57.13 This allows us to study the dynamics of cell interaction using microfabrication.
00:10:02.29 What we've learned as an ensemble in these sets of experiments, is the following.
00:10:08.23 That this big fat cell here is supposed to represent the hepatocyte
00:10:12.15 and this cell is the supportive, stromal cell. In this case it's a fibroblast.
00:10:18.01 What we've learned is that the cells need to touch each other for about a day
00:10:22.16 in order to get the rescue that I described earlier.
00:10:25.29 After the first day, however, they can produce soluble factors that can diffuse over that 80 micron moat.
00:10:35.20 And this soluble factors are completely sufficient to support the hepatocyte.
00:10:39.29 So what's exciting to us about this finding as engineers
00:10:43.15 is it might suggest that after an initial priming period one could support hepatocytes
00:10:49.11 in an implantable tissue or in an artificial liver device
00:10:52.12 without needing to have another living cell taking up space and nutrients.
00:10:57.19 So I've told you about how we used microfabrication
00:11:02.15 to learn about how to stabilize the liver cells
00:11:05.17 but one can also use these light based patterning techniques
00:11:09.22 to build three-dimensional, implantable parts.
00:11:12.06 And in doing this, we were inspired by this rapid prototyping technique known as stereolithography.
00:11:19.04 So stereolithography is a technique whereby one will draw a three-dimensional part
00:11:25.00 using CAD software in the computer and then use it to drive this robot.
00:11:29.20 And the robot works as follows. It drives this stage.
00:11:33.20 The stage can move up or down in a vat of light-sensitive polymer.
00:11:38.19 So when the stage is at the top and one shines a light in a pattern it crosslinks the polymer
00:11:44.03 and then we can make it drop a level and shine a different pattern.
00:11:48.26 And so on and so forth.
00:11:50.07 At the end you build very complicated three-dimensional parts.
00:11:54.06 And in fact, there are programs now to be able to drive these
00:11:58.06 with three-dimensional, patient-specific anatomic data.
00:12:02.04 So one can imagine then taking CAT scan data and driving a robot like this
00:12:08.14 to make a three-dimensional living part.
00:12:10.03 So in order to make a living tissue using a system like this
00:12:14.08 we need to create cells within this three-dimensional light-sensitive system.
00:12:21.29 In order to do that we borrowed a chemistry
00:12:25.17 that was first developed by a scientist named Jeff Hubble, which is described here.
00:12:30.27 This is a polyethylene glycol polymer. So it's a long chain of polyethylene glycol molecules.
00:12:37.13 It's a long chain polymer with reactive end groups.
00:12:41.02 And what one does is simply mix in photo-initiator (a light-sensitive chemical)
00:12:46.29 with cells and when one does this mixing and shines light,
00:12:51.17 you get a photo-crosslinked network with cells imbedded in the polymer structure.
00:12:55.26 So this is like the cells are fruit and the polymer is the jello.
00:13:00.09 So the cells are the fruit in the jello.
00:13:02.20 And the nice thing about this is that one can then change the pattern of light
00:13:07.16 that one shines and create different structures.
00:13:10.22 So in order to get hepatocytes (the liver cells) living in this system,
00:13:15.06 as I mentioned, they're very sensitive to their microenvironment,
00:13:17.23 a graduate student in our group spent about five years discovering how to do that.
00:13:23.06 And in this slide, I'll summarize her work.
00:13:26.19 So what she showed was that if you put liver cells just in this material system,
00:13:31.07 and tracked the albumin secretion, which is one of those liver-specific functions I mentioned earlier,
00:13:36.15 they survive but they don't function like liver cells.
00:13:40.14 If instead you add those supportive cells that I showed earlier, the cells do a little bit better.
00:13:46.19 If one further adds a peptide into this inert polymer, this is an RGD peptide,
00:13:55.00 which mimics the binding site of fibronectin, an extracellular matrix molecule in the liver,
00:14:00.15 as it interacts with the integrins, the receptors on the hepatocytes,
00:14:04.14 one gets even more function.
00:14:06.23 And finally, if one adds a third cell type which exists in the liver,
00:14:10.18 the endothelial cells which are the natural neighbor of the hepatocytes and
00:14:15.14 normally live only 1 micron away from the hepatocyte.
00:14:18.14 If one adds those into the network,
00:14:20.13 now one can get a highly functional liver cell in this photo-encapsulated system.
00:14:25.14 So using this system we've created many, many different PEG-based tissues.
00:14:31.26 And this is just a smattering of the kinds of architectures we've been able to make over the years.
00:14:36.08 In this movie, what you'll see is in this latest version,
00:14:41.01 we're now able to make these microfluidically.
00:14:43.17 So this another type of microfabrication where one can make a microfluidic device out of polymers.
00:14:50.01 Here the polymer is called PDMS, poly-dimethylsiloxane.
00:14:54.14 And what you see entering into this microfluidic nozzle is a mixture of cells and the polymer suspension.
00:15:01.12 As they bud out into these droplets then they are ready to have light shone on them.
00:15:07.26 And then they can be polymerized in these microfluidic droplets.
00:15:11.19 And this an example of cells inside these microtissues.
00:15:16.19 So, we've made progress now on the fabrication of these hepatocyte-based tissues.
00:15:23.27 The next thing we wanted to do was see how they would do in vivo.
00:15:28.16 And that experiment is shown here.
00:15:32.05 So in this experiment what we've done is taken these hepatocytes, these liver cells,
00:15:37.03 and genetically modified them with a virus, a lentivirus,
00:15:41.15 that drives a gene known as the luciferase gene
00:15:46.11 when a particular promoter is active, the albumin promoter that I was mentioning earlier.
00:15:52.15 So if the hepatocytes are happy and functional, they will express luciferase.
00:15:58.21 And here what we've done is implant these tissues in the the interperitonial space,
00:16:04.14 the abdomen of the animal or in the subcutaneous space.
00:16:07.25 So we're looking at different site of implantation.
00:16:10.22 And then we look for light generation by this tissue as a marker of how it's doing over time.
00:16:16.10 You can see in these experiments these tissues survive in the interperitonial space
00:16:21.23 for up to three months by this measure.
00:16:24.26 And if we look in the blood of the animals and look for markers of the human proteins in the mice,
00:16:31.00 we see that the cells are making human albumin in the blood of the animal.
00:16:36.21 And this is a sign that the blood vessels in the mice
00:16:40.22 have been recruited to that construct and they're hooked up.
00:16:44.17 So the construct has been vascularized.
00:16:47.23 OK, so that's as far as we've gotten in terms of fabricating tissues
00:16:54.25 that we hope to implant in patients one day.
00:16:57.03 We have lots of work to do on the scale up of these tissues.
00:17:00.14 The tissues that I've mentioned to you only have about a million cells in them
00:17:04.18 and in order to get a therapeutic effect in patients we probably need a billion cells.
00:17:09.17 So we have lots of ideas now about who to scale up by three orders of magnitude.
00:17:14.23 Hopefully next time I talk to you I'll be telling you about that.
00:17:18.01 What I'd like to now mention in the last two parts of my talk
00:17:21.00 is how one could use these model systems not just for therapeutic applications
00:17:26.15 but for scientific discovery by interrogating them.
00:17:32.01 So, the application that we first became interested in
00:17:36.02 when we realized that we had human microtissues growing in the laboratory was this one.
00:17:41.13 This is the infamous drug discovery pipeline
00:17:45.24 and what you can see here is that typically it takes...
00:17:49.21 one can start with about 10000 chemicals
00:17:52.06 and over the course of about 15 years and close to a billion dollars,
00:17:56.11 one can get one FDA approved drug on the market.
00:18:00.02 What's problematic is that actually even after all of the pre-clinical studies that are done,
00:18:09.00 all of the in vitro screens and the required animal screens,
00:18:12.08 so these are typically rodent models and large animal models,
00:18:15.25 when one enters into Phase I trials, which is when the drug is first exposed to patients,
00:18:20.29 a third of the time they exhibit liver specific toxicity.
00:18:27.11 So they exhibit toxicity to the human liver
00:18:30.22 which was not predicted earlier in the development process.
00:18:34.09 So we were interested in this idea
00:18:38.13 which was whether we could bridge the gap between animal models and clinical trials
00:18:43.25 with some of these technologies we could make engineered human livers in vitro
00:18:48.27 and use hem to study this process.
00:18:51.08 So, what I'm going to tell you about now is a two-dimensional, in vitro human microliver.
00:18:56.06 And one can also do the implantable three-dimensional models that I mentioned earlier
00:19:01.14 to humanize a mouse model and again bridge the gap between mice and clinical trials.
00:19:07.06 So when one discovers new drugs, one often doesn't have very much of the compound.
00:19:14.17 So you'd like to develop a much higher-throughput way
00:19:17.20 of fabricating the tissues than the ways I mentioned in the past.
00:19:20.22 The photolithographic process that I described to you in the beginning of my talk
00:19:24.28 was one wherein we patterned coverslip by coverslip
00:19:28.04 and we created these microarchitectures that would stabilize the cells for about four to six weeks.
00:19:33.27 But that's not a practical way to screen thousands of compounds.
00:19:38.01 So one thing that our group worked on was miniaturizing that technology further.
00:19:42.29 We came up with this. This is a multi-well device.
00:19:46.02 And the idea is that this looks just like a 24-well or 96-well plate
00:19:51.01 This is sort of the workhorse of the field of biology.
00:19:54.15 And this can be fed with fluidic robots.
00:19:57.08 At the bottom of each of these wells is a micro-patterned co-culture
00:20:01.23 of the two cell types I mentioned earlier, human hepatocytes and the supportive fibroblasts.
00:20:07.06 And now they're organized in their optimal size and shape for human cells.
00:20:12.23 The way we do this is we create a soft polymer part
00:20:16.23 that has a stencil at the bottom of each well.
00:20:19.20 When you pour collagen through this stencil you get collagen spots on the underlying surface.
00:20:25.27 And now again, as before, you can seed human hepatocytes
00:20:28.29 and then surround them with those stromal neighbors.
00:20:31.20 And in this way one can reuse this part over and over again
00:20:35.06 and never have to go back to the microfabrication facility
00:20:37.22 and have a sort of high throughput format of the microliver tissues.
00:20:42.06 So if one now looks at the albumin function of these tissues,
00:20:46.08 and this is the curve I showed you earlier where the albumin function is lost very rapidly,
00:20:50.25 one can see that it's rescued and is stable for about four to six weeks.
00:20:55.27 So, the next question then is are these useful for drug screening.
00:21:02.01 So, one important aspect of liver tissue in vivo and in vitro, is the one I already mentioned
00:21:08.27 which is that animal livers and human livers differ in their drug metabolism.
00:21:13.20 There are a couple of reasons for that.
00:21:16.11 The first are a class of enzymes known as the P450 enzymes
00:21:20.13 and they have species specific differences.
00:21:22.25 Furthermore, there's a process known as induction.
00:21:26.23 Induction is the process whereby exposure to a chemical
00:21:31.02 upregulates the P450 metabolism and alters the metabolism of that drug.
00:21:37.27 So, the induction process also varies in a species-specific way.
00:21:43.02 So what we wanted to do is make micropatterned co-cultures out of rat cells and human cells
00:21:49.10 and expose them to inducers that are known to have species-specific differences
00:21:55.08 and ask whether in vitro we could predict these findings
00:21:59.00 that are normally reported at the level of the organism.
00:22:02.21 And you can see here with these two drugs, Omeprazole and beta-naptholflavone,
00:22:07.07 we were able to upregulate the P450 metabolism in the human system
00:22:12.29 and relatively less in the rat system and that recapitulates what we know about the clinical picture.
00:22:19.01 So this was encouraging for the utility for species-specific induction studies.
00:22:25.09 Furthermore, this process of induction
00:22:28.13 is actually the mechanism by which drug-drug interaction occurs.
00:22:33.01 So, many of you may have noticed when you go to the pharmacy and you get prescribed a medicine,
00:22:39.01 it will tell you sometimes not to take the medicine with another drug;
00:22:42.16 that the two drugs might interact.
00:22:44.09 And the drug-drug interaction often happens at the site of the liver.
00:22:48.11 So this is an example of Tylenol (acetaminophen) exposure to the liver.
00:22:54.27 And so what you can see here is that the acetaminophen exposure to the liver is not consequential.
00:23:00.13 These two drugs, one which is an anti-seizure medicine
00:23:04.05 and the other which is a heart medicine, are not consequential to the liver.
00:23:08.07 However, if you give acetaminophen together with the anti-seizure medicine,
00:23:12.10 or with the other one, then you see the emergence of drug-drug interaction.
00:23:16.17 And we know the mechanism of this has been well worked out,
00:23:19.07 that that's recapitulated in this system.
00:23:21.16 So this is promising for the ability to maybe study drug-drug interaction in humans
00:23:27.24 again, in vitro and in high-throughput.
00:23:30.00 And finally, one can then study the hepato-toxicity of drugs with these livers.
00:23:39.26 So, this is an experiment where one calculates the toxic concentration
00:23:44.27 at which 50% of the cells die. That's called the TC50.
00:23:50.18 And one can do this over 24 hours of exposure with a variety of clinic compounds
00:23:55.26 and this is what's typically done now for novel small molecules.
00:23:59.20 And one can rank order these chemicals
00:24:01.13 and we can see that they rank order in the way
00:24:03.16 that they should based on what we know about clinical exposure.
00:24:06.27 What we're more excited about
00:24:09.10 is the potential to do things actually differently than the way people do things now.
00:24:13.12 And that's to do chronic exposure.
00:24:15.18 We know that all of us take our medicines in low dose, multi-day format.
00:24:20.08 And if one looks at the emergence of hepatotoxicity clinically, in clinical trials,
00:24:25.13 it actually usually unfolds on the order of weeks.
00:24:28.24 So, in similar experiments, what we've done is shown that one can do
00:24:34.10 multi-day exposure and score the relative toxicity of compounds over a chronic, lifetime.
00:24:44.15 And that in fact the answers change if you do chronic exposure as compared to an acute exposure.
00:24:49.15 So we're excited about the potential to change the way people do toxicity testing with small molecules.
00:24:55.23 OK, so I've told you about building implantable livers,
00:25:01.15 about building arrays of tiny livers in vitro,
00:25:04.09 and the last thing I'd like to tell you about is how one might learn something
00:25:08.08 about pathogens that infect the normal human liver using a platform like this.
00:25:14.13 So, in fact I've already told you about one class of pathogens that infect the human liver--Hepatitis.
00:25:20.06 So there's Hepatitis A, B, C, D and many more.
00:25:24.00 In particular, we've been working on Hepatitis C for which there's
00:25:26.26 no vaccine and no good, effective drug therapy.
00:25:29.29 Another pathogen that's very interesting for which there are not good in vitro models
00:25:35.07 are the human malarias.
00:25:37.26 So, just to remind you how human malaria work
00:25:43.01 Plasmodium falciparum and Plasmodium vivax are two parasites
00:25:46.28 that are transmitted by a mosquito bite. So typically the mosquito will bite a human.
00:25:53.00 The sporozoite, the parasite, will travel through the bloodstream, infect the liver.
00:25:59.11 When it infects the liver, it will traverse several hepatocytes
00:26:03.01 and pick one in which to set up shop.
00:26:05.03 It sets up shop in that hepatocyte. It will then multiply and grow and undergo morphogenesis.
00:26:13.04 And then burst out into the bloodstream
00:26:15.25 where it can then infect the erythrocyte (the red blood cells).
00:26:19.19 And it's at this stage that one would get the clinical symptoms that one associates with fever.
00:26:24.26 What's interesting and exciting about the opportunity to get drugs to work on the liver stage
00:26:31.08 is that typically 30 sporozoites going into the liver
00:26:35.12 would be amplified to about 300,000 merozoites bursting into the blood.
00:26:40.26 So, if one could kill the parasite at the liver stage of infection and before there were any symptoms,
00:26:47.12 one could prevent the symptomotology and also this amplification in the patients.
00:26:52.11 So, the trouble with studying all of this is that the human malarias
00:26:58.06 don't infect animal models.
00:27:00.28 So for example, they don't infect mice or rats in the laboratory.
00:27:04.00 So, we've been interested in studying our human micro-livers in vitro
00:27:09.17 and seeing whether one can infect them with the human malarias
00:27:12.27 and recapitulate this liver stage of the malaria lifecycle
00:27:16.27 both for drug discovery of new anti-malarials and for vaccine applications.
00:27:23.00 So, in this movie what you see is a typical sporozoite that would be swimming through the bloodstream
00:27:27.23 and you can see what they do when they swim like this
00:27:32.08 is they undergo this gliding mechanism
00:27:35.05 and as I've mentioned, they traverse several hepatocytes on the way
00:27:38.08 before picking the one in which they set up shop.
00:27:40.27 The hepatocytes that they traverse ont he way are wounded.
00:27:46.11 So, the membranes are temporarily breached and they self seal if they survive.
00:27:52.04 So we've recently been infecting our micro-livers
00:27:56.19 those 500 micron colonies of hepatocytes that I mentioned earlier
00:28:00.11 with the human malarias.
00:28:02.03 So, here what we've done is add a fluorescent dextran, a red dye, to the media
00:28:08.23 and as the sporozoite traverses through the cells, it injures the membranes
00:28:15.03 and the cells take up the dye and if the membranes seal, then they stay lit up like this.
00:28:20.05 So, essentially, this is a molecular footprint of the sporozoite traveling
00:28:24.22 through our human micro-livers in culture.
00:28:27.21 And so, we've now done this with both Plasmodium falciparum and Plasmodium vivax
00:28:33.18 and we're very excited to get a glimpse of the elusive stage of Plasmodium vivax
00:28:39.13 that's referred to as the hypnazoite.
00:28:41.21 It's a dormant stage of the parasite that no one's ever seen before in vitro
00:28:45.29 and we're dying to see if we have it growing.
00:28:49.18 So just to summarize, I think that the tiny technologies
00:28:54.12 have been very powerful for the study tissue engineering and tissue microenvironments
00:28:59.18 and I've given you the example of the liver but I think this is broadly true because
00:29:04.16 the microarchitecture of all tissues actually have repeating units on these same length scales.
00:29:09.12 So here are just a list of technologies that are emerging in the bioengineering field
00:29:14.17 from the 100 micron length scale all the way down to the 100 nanometer length scale
00:29:18.15 that are really ripe to be borrowed in this field.
00:29:25.14 Just to summarize, then, I've told you how one can use microscale technologies
00:29:29.14 to fabricate liver tissues, to interrogate arrays of liver tissues,
00:29:33.08 and then finally to learn more about pathogens that infect the liver.
00:29:37.11 So that you for watching and I'be like to thank our group for all the hard work
00:29:42.13 and in particular, Megan Shan and Alice Chen who helped me work on this talk.
- Relevance: How does the therapeutic approach for liver diseases contrast with other organ systems? What is the microenvironment of hepatocytes? What is a major challenge that has hindered the advancement of cell based therapeutic strategies?
- Construct: Describe two examples of light-based microfabrication techniques used in hepatic tissue engineering. What features do these techniques control for and why is this important? What would you design using one of these techniques?
- Interrogate: What are some applications for the arrays of liver tissue in drug development? Can you envision a biological question that may be addressed via these micro-patterned co-cultures, be it with liver arrays or other organ arrays?
Paper for this Session’s Discussion
Stevens, K.R., Ungrin, M.D., Schwartz, R.E., Ng, S., Carvalho, B., Christine, K.S., Chaturvedi, R.R., Li, C.Y., Zandstra, P.W., Chen, C.S., et al. (2013). InVERT molding for scalable control of tissue microarchitecture. Nature Communications 4, 1847.
Discussion Questions for the Paper
- What are the steps for the InVERT molding technique? Briefly describe figure 1a (or 2b) and redraw the steps on the board.
- What experiments were done to further characterize the new InVERT techniqe and why were they important when trying to understand the relevance or versatility of this new technique? In the first part of the iBiology lecture, Sangeeta Bhatia talks about designing the right biomaterial. What aspects of the InVert technique allow for biomaterial design?
- Describe how interpenetrating, juxtaposed and paracrine conformations were patterned (figure 3). How do the organization of compartments (figure 3 and 4 a,b) relate to the importance of cell organization discussed in the first part of Sangeeta Bhatia’s talk?
- What experiments were done with the nude mice in figure 4? What other experiments would you have done?
Sangeeta Bhatia received her Sc.B. in biomedical engineering from Brown University, her M.D. from Harvard Medical School and her Ph.D. from MIT. She is currently Professor of Health Sciences and Technology and of Electrical and Computer Science at MIT and a Howard Hughes Medical Institute Investigator. Research in her lab is focused on using micro-… Continue Reading