Nerem Lab Meeting 01-11-2002 Transcribed by Jim Davies Jon: So, I guess what I'm-- my plan is to just kind of go over what I'm doing, not only for me but for other people who may not know what I'm doing, and what I've done, what I plan on doing.. Kind of a overview. So is this... okay, here we go... Yeah, so I'm going to just kind of present, you know, what I, where my research area I think is going to be and my-- the problem that, or the question that I want to answer, and uh, how I right now think I'm going to go about that, and then, uh, I've got some preliminary data that I want to pass on, and uh, based on that data, what I should do, kinda where I should go from there, and then you guys can give me some feedback. That'd be great. So, uh, just to kind of bring us to the beginning. Um, in the clinical realm, there's uh, the main problem that I'm looking at is the fact that we have all these people that have, uh, require heart valve replacement procedures, either through some sort of required problem or uh, a congenital one, uh, and, uh, the number of people requiring these procedures is increasing every year somewhat significant around 89 thousand hospital discharges, and uh, the year 2000, uh, something that's perhaps more disturbing is uh, from a tissue engineering perspective is the fact that we have some 40 thousand infants born every year with a congenital heart defect that oftentimes will affect the heart valve. Um, and the aortic valve is the most prevalent injured valve because it is the interface between the aortic system and the heart system. So what clinitians do is um, right now to replace these valves is to replace them with mechanical or biological valves. And, uh, the mechanical valves you can see here are some examples. And uh, the biological valves, for the most part anything that is biologially-- well, uh, most of the biologically derived valves will be fixed somewhat with the gluteralterhyde or some other photooxidized method or something. Um, the only valves that are living that are used are something that is done like an autograph which is called a ross procedure. Basically they take another valve which would be your pulmonary valve and put it position and then they just try and fix the pulmonary valve. The idea being is that the pulmonary circulation is much lower pressure much lower stress environments so maybe we could deal with that one more easily from an engineering perspective. So um those are things that they do and so far, um, the results they have are, to date, as you can see in this graph um, is that these valves basically last somewhere between 10 and 15 years. Now, different valves will show higher or longer success rates, um, depending because you can kind of select your patient group and such and some valves might be optimized for a particular patient situation but this, ah, this study was done like a randomized study between a mechanical valve and a biological valve. And what you find is that the different valve types fail differently um, the mechanical valves fail through what seems like biological means and the biological valves fail because of the mechanical environment, it's just that there's more stress and you can see examples of what happens in those situations. So the tissue engineering world looks at that situation and there's several advantages that we can use or maybe, ah, to try to address these problems. So the construction of a living heart valve substitute would be great because it would be able to grow and adapt with the body which would be useful in the case of these infants with these congenital heart defects because now the valve doesn't need to be replaced every so often. Uh, the procedure is very difficult. To only have to go in once would be good. Um, and uh, it it may turn out to be a depending on how the technology develops you can get the resources could be readily available if we could harvest the appropriate cell type and uh, the there's several objectives or challenges to this to this applying tissue engineering to heartvalve similar to every other system that tissue engineering used: mechanical stability, tissue repopulation, long-term effectiveness, construction time of the immune response. All these things are critical to the success of tissue engineered, ah, substitute and likewise it would apply to the heartvalve. Um, what people have done so far in basically trying to push forward into these tissue engineering using different biomaterials. All right here you can see that there's a selection of all the different things that have been tried, um, in this area, and some of the research that's been going on and uh, there is, uh, I guess, treatment methods that people have done to try to use those biomaterials um, so, I guess I don't know if we're all familiar with some of these but if not I guess you could as me, but so far, um, what I want to point out, I guess, at least, from where I'm going, Is the cell sources we're using in um, creating these constructs are all cells that are not from the heart valve but they're not indigineous populations. Cells that are a lot more easier to isolate and grow and uh, have been characterized in the literature to a much further extent. Um, and uh, anyway, the uh, current progress I guess as the way I see it with these valves is that these the-- the valve type that has gotten the most amount of exposure in the last five or ten years has used a polymeric scaffold where they seed endothelial cells, or smooth muscle cells or other kinds of things. So what um, people have been doing is trying to mimic the, ah, the shape of the valve and try to get it to open and close and they're uh, running into a problem because you can get the shape but then you can't get the strenth or you can get the strength but not the shape, and you know it's basically an optimization problem. Um, and uh, some other people have had a success with acellurized tissue in reducing immune responses and uh, also other people have had success in trying to recede endothelial cells onto like a fixed valve so you can kind of have this living monolayer of cells and now that showed that has some lower regurgitation rates. But none of these have really gone far enough to any kind of clinical trial where we can try to see any actual improved success rate and uh, time will bear that out. So, um, I guess just to get into more of what I want to do, just a little background, so, the aortic valve is the one in the middle. It's got these three leaflets, um, that are very thin, um, almost transparent, yet very strong. You can, like when I dissect the uh, the aortic-- when I dissect the aorta, the aortic valve you can actually pick up the entire valve by holding up one of the leaflets, which is quite substantial, uh, strength-wise. And uh, the-- each leaflet is made up of three layers-- actually kind of two layers with kind of sort of an interface which is kind of a loose collection of proteoglycants, fibroso layer. You've got this bunginosa and tricularus and they have different amounts of collagen and elastin that are uh, the proportions of which their orientation are designed to suit their function and in that particular thickness of the leaflet. Um, the thing that's uh, interesting about these leaflets is that they do have an indigineous population of cells that, at least, uh, as far as the research that's been done so far have somewhat of a different characteristic from other aortics cells or even other endothelial cells. So the valvular endothelial cells people have found they have particular requirements about fibernactin and a particular inability to synthesize it. Um, and people have also done some research to show that, um, the adhesion properties are uh, are slightly different than um, aortic endothelial cells and um, another interesting cell type is the interstitial cells. So basically you can think of the leaflet as being a sandwich with two endothelial bread slices and interstitial cells in the middle. And what people have done in understanding these interstitial cells is they've kind of found that there's a.. It's kind of a hybrid between fiberblast and smooth muscle cells, um, there's two different kinds of forms like a spindley shape and a cobblestone form and some people think that the spindle form is more myofiberblast and the cobblestone form is part of a fiberblast cell. Um, and uh, people have done studies with uh, responding to stimuli with vasoactive agents and stuff like that and uh, gotten some results with that. So we see that these cells behave at least initially, they behave somewhat different. If we want to adhere to this, you know, structure controls function thing and we control mechanical environments affect this cellular phenotype then one would think that these particular cells may have a particular advantage in this area that we could exploit for a tissue engineering perspective. So that kind of brings me to my research question, at least right now that's what I think what I want to kind of hone in on. And uh, so, the two-- the two ways that I'm thinking I want to look at this now is um, how do they, how do the valvular cells respond to the mechanical stress and my monolayers three dimentional constructs um and then likewise the different cells, how do they interact in the environments. So there's kind of two different cell types of interest kind of two different environments. Uh, so the way I'm thinking about going about this is uh, first I've got to isolate the cells and characterize them you know, what do they look like, how do they grow, that kind of thing. How do I keep them alive, how do I freeze them and stuff. Um, and the literature has been halfway decent in that not that many people go into the techniques and that kind of thing so that's kind of a challenge but I think I'm finally there and uh, I've had some great help from a guy by the name of Steve Hilbert who's at the FDA and he is going to do some microscope work and stuff like that for me for these cells. The idea is I want to compare the vascular endothelial cells and their interstitial cells to other known cell types that are in the same area, which are, namely, the aortic endothelial cells and the smooth muscle cells. Those are much better characterized and uh, you know, I'm hoping I can, uh, kind of make a graph or some sort of span to see where these cells fall in the middle kind of thing. And then next, uh, I want to run shear stress experiments using patented flow plate parallel technology, um, and uh, the things that I'm thinking of looking at right now are adhesion parameters, which I'm kind of grouping as a big topic to report which would include like more parametric things and certain adhesion molecules. Another thing I'm interested in getting into is RNA expression of the gene array technology. And I want to look at both monolayers and 3D constructs. And then the other thing is uh, so I would use the shear stress to evaluate the endothelial cells. They're the ones that are exposed to the shear. And then, um, a device that we're getting from the university of California San Diego um, is this biaxial strain device that we can use to strain cell cultures, um, I want to use that to do the same kind of gene array stuff. but I also want to see if I can modify that somewhat to be able to analyze three dimensional constructs. Or try to make it so that I can grow cells three dimensionally and strain. So that's kind of what I'm thinking of doing, and hopefully I can get some feedback. So this is getting into the data that I've got so far, uh. These are all passage four pictures of these cultures. And you can see that there are some differences between the two cell types. It's interesting to note that the, ah, the valular endothelial cells, at least when I $$ appear much flatter, and uh, their borders have these little kinks in them that would imply, like, these gap junctions. I'm not all up on my knowledge of that specific biology's that great. I'm sure that with Steve's help we can figure out exactly what these differences are as opposed to that just being some sort of actual thing. Um, you can see, actually I have these arrows here to show there are-- at least when I capture these interstitial cells I too find these kind of different groups of morphology. They seem to grow, um, you know, they seem to grow well with each other. There's not one that will dominate over the other one. Um, so there's that. And uh, this-- some initial attempts to try to characterize the interstitial cells um, with, or, with um, two different dyes that are endothelial cell markers.Um, So this is kind of preliminary because I only have an n equals 2, and the fact that um, like for instance my i selectance thing I'm not quite sure whether or not the smooth muscle cells should press it.. um, I'm not sure it says that after a while it can get macrophageous and take everything out. So I'm not sure if that's the case but I do more. But nevertheless the interstitial cells at least appeared to take up both dyes. Um, and then this is, uh, again, a, uh, a similar thing but with flow cytometry to try to see intensity differences. And, um, you can see that there are some differences between the interstitial cells and the other cells that, um, that we already have better characterized. Jan: What does isolectin account for? Jon: Isolectin is an endothelial marker. There's some surface $$ that uh, it attaches to. : $$ is the same. Jon: Yeah. It's a $$ protein. So that's the smooth muscle alpha. So, these are, this is kind of what I've been doing for my shear stress experiments. I subjected my monolayers um, to 12 dynes, or 30 dynes for 48 hours and I used the passage 4 cells. Um, and this is kind of my hypothetical protocol for these constructs that I want to make. It's basically using the similar protocol when making the vascular grafts. When I uh-- when we open them up to $$ and uh, entrap the cells I'm going to use my $$ as opposed to the rat smooth muscle or aortic endothelial. So what I'll end up with is kind of like a half-leaflet model where we have an interstitial uh, kind of medial layer and a uh, endothelial surface uh flat, and do an approximation that resembles the geometry of a leaflet surface. So I haven't done any of that yet but we'll see how that works out. Uh. And this is kind of in description of the biaxial strain protocol that I want to try to do um, so these heart valves in addition to opening the leaflets actually stretch up to ten percent. So I wanted to use the biaxial strain because that's the most controlled strain environment that we can apply. I think it's more important to see how the cells respond to the most controlled environment as opposed to something uniaxial which would be in fact biaxial but would be very difficult to characterize. But one of the things that I'll have to do if I want to use them for three dimensional constructs is I'll have to try to characterize the tree dimensional strain field. Now, I don't think that means entirely three dimensional I just want to see how well the strain transfers from the bottom of the silicone membrane to the top of my collagen construct. I want to make sure that the top strain $$ what like the bottom so that way either $$ is receiving some degree of a similar strain thing. And I'll have to use some sort of opaque beads and cameras and so forth. Uh but this is my-- this is what I think so far how I want to try to get these in a three dimensional construct kind of like the way Steph does the migration phase where she's blocking-- she makes a little agar frame $$ with collagen. That way I'll focus in on a smaller region inside on top of the membrane that will be, maybe have some thickness. $$ cells. So this is the data that I have from the shear stress experiments that I've run so far. And uh, the top one is the static control. The middle one is 12 dynes and the bottom is 30. And something that I found that's interesting so far even though It's kind of preliminary, don't seem to respond as strongly to shear stress as the endothelial cells. 12 dynes at 48 hours is more than sufficient to cause alignment changes and the cells to enlongate into the direction of flow for aortic endothelial cells but that doesn't seem to be the case for these valvular endothelial cells. I do see a change in the way that the cytoskeletal is arranged with the 12 dynes. We just don't see the change in alignment and uh, orientation. So you can kind of see, looking from the establishment control to the 12 dynes you can kind of see that the microfilaments you see this is an acron stain back there so you can see that they start to, um, you start to get the bands across the middle of the cell $$ peripherally oriented actin. What you don't see is that they're all aligned in the direction of flow. Now there are studies that show in vivo that the preferential orientation valvular endothelial cells actually perpendicular flow. What that, I think, suggests is that these cells are more in tune with the strain environment as opposed to the shear environment because they're strained perpendicular flow. Obviously I can't verify that with this but that's just something that I might be suggesting. But when I move it up to 30 dynes, you can clearly see that the cells really align quite nicely to the flow and they elongate. And something else that's interesting is it seems as though these arrows-- oops. Sorry about that. Almost. Ok. What these arrows kind of show is that the preferential-- flow is left to right. So you got these real dense bands on the left sides of the cells, which seems to suggest a anchor, you know, in particular ways. I'd like to look at other adhesion molecules, to see whether or not they're present preferentially in these particular regions. I mean I don't know if that's necessarily useful. No one else has done it, but I don't know if I have to do that. But anyway, I think that that's-- it's something that happens in aortic endothelial cells and so that's been demonstrated here. As well. Um. And this is a comparison of the, um, endothelial cell morphometric parameters. You can see that the increase in shear, you have, uh, significant decreases in the shape index which implies that the cells are becoming more elongated. You have a significant increase in the angle of orientation which implies that the cells are becoming more aligned in the direction of flow. Excuse me. $$ flow. And then, uh, something else that's really interesting is that, uh, with the increase in shear you have a decrease in the number of I guess the contacted neighbors.. It's not what I want to say.. $$ these cells are interacting with. So, uh, in static culture you have a lot of cells, they're all grouped together and they talk to each other, and somehow in shear, when they elongate, they seem to only like to interact with a couple neighbors. Um, I'm not so sure why that is, it's just something I wanted to point out. Nerem: That's number. That's similar. Number of cells contacting an indivitual cell? Jon: Yeah. Yeah. So what I did was we got a picture of a cell, and I look at the cell, and I look to see, you know, the contact at least to be determined by actin filaments. So it'd be like some sort of cell-cell contact and would just $$ them around Nerem: You know, I think the result is interesting but at the same time it only goes from five to $$ Jon: Right, I know that. I understand that. But you know, with, uh, Yeah. I understand that. The significance with the number of samples I guess makes it significant. But when you go back and look, I don't know. To me it just seemed like something to bring up. I don't know what that means. But uh, this is kind of a trend that I wanted to show. So with the increase in shear it seems like we have-- it's just really confirming what I've said before, but with the increase in shear you got more alignments and elongation uh, so, you know, I guess I $$ with more data points along that we could probably continue to move down in that-- down that tradjectory, which is similar to what happens with aortic endothelial cells. But I think the interesting thing being- because I don't know whether or not there's some sort of transition shear where like the cells don't-- are not really affected by shaer until you get to a certain point and then they start, you know, almost it seems like you can force them to align and orient $$. So that's something that I don't know if I should persue. Whether or not that's interesting. and this is just kind of a histogram of the ah, the number of neighboring cells, as I call it, because I kind of wanted to show that there is a trend to these fewer numbers of contacted neighbors with shear. And that's-- that's been shown to some degree with the aortic endothelial cell. So my next steps as I see it is that uh, soon as Dr. Hilbert gives me some of his data I can take a look at that what some of the things that are different. Different cell types at least $$ with microscopes and of course the question I have is whether it's useful to repeat some of this flow type cytometry experiments I was looking at those different markers and pick some other ones, you know, if that's something that won't really give me as much information for my time and maybe I won't. I'm looking at more shear stress data. I could try to look at it in immediate shear stress between the $$ the thirty. I'd like to get some more data on the 12 and the 30 because what I have right now is preliminary. And then the other issue is the adhesion molecules as thinking of looking at that as venicular $$ because they've been shown to be associated with the actin filaments and their adhesion bands. And then maybe looking at a shorter time frame for my shear right now $$ 48 hours. So I was looking at maybe six hours which would be some sort of $$ type thing. Um, of course if you start getting into that I could have this whole, you know, time and shear and that could be really involved. So I guess I'm kind of interested in what's important here. Um, and then, I will start to take these valvular constructs. Uh, so, soon as, uh, I guess I think all of our equipment for the flow loops and making these constructs are in so I can start making these, growing these cells and embedding them and seeing how they-- whether or not I can make a construct out of them. What they look like. Maybe do an initial histology. I have some histology slides of actual leaflets $$ prepare. And then, um, one thing that's probably going to be pretty big for both me and Tiffany is learning this gene array techniques, um, down at morehouse. Um, one thing that we've been or at least that she's been doing a lot of it, with poking around is we found that this new method of RNA amplification $$ linear RNA application to be a lot more promising than the, uh, PCR techniques because it has the ability to maintain uh, the correct proportions of RNA between the samples. So, uh, we're exploring that. Uh, we've got a lot of information on that. so that's something we think might be better way to amplify the RNA. People I guess use-- develop these techniques specifically to get, to make, uh, programs of RNA out of $$ from the sample which would be very useful for us because our monolayers are so small. And we get such small amounts of, uh, RNA out of that. So, hopefully that will be positive. Um, and, I think that's it. So I'd appreciate feedback, comments, direction. Ann: I have a thought for you. Jon: mm hmm Ann: When you're making your constructs ## but do you want to make them into tubes or do you want to make them into disks? ## Jon: I think that I want to make them in the tubes and then open them up like Ann: Like we do now? Jon: I think I do because even if there is some sort of like when they compact in the tubes and start to get some sort of a preferential arrangement, I still think that that preferential arrangement, if it's circumferential, would be similar to the approximation in vivo because like I said the cells are aligned perfectly with the flow. So I think that-- I think that wouldn't be a bad thing. Nerem: You'd actually want some preferential alignment ## mechanical preconditioning. And that's when Gore got really aligned collagen fibers and uh, smooth muscle cells. Jon: Yeah, I don't know if that's-- Yeah, I don't know if that's something that I would need to go that far at least I want to make sure I can get these cells to live together in harmony first but I thought that that would be the easiest way to do that. Seems like the disk-- Nerem: I should give you a copy of the $$ paper which is in press because um, she showed, um, in the absence of flow-- Jon: mmm hmm Nerem: --distinct differences with constructs, depending on whether you preconditioned them and got the $$ or didn't precondition them. Jon: ok. : Are you talking about preconditioning the cells before making the construct? Nerem: Preconditioning the construct prior to doing endothelial experiments. : Oh, the seeding. : The culturing end. : Yeah. ok. Joe: I have a question too. Looking between the different cell types and what the differences are between them, have you also other factors such as, like, the phenotype shifts passaging, or, depending on what sort of matrix they grow on, how that will affect it? Jon: Ah, so I've done some studies with the different matrices, where the matrix is grown on-- I'm kind of limited there because I'm $$ endothelial cells really only grow on fibrinectin. Like, I've tried collagen gelatin and tissue culture classic and they're just $$ and don't grow at all, so, um Joe: So when you get to the three dimensional are you going to have to have some sort of fibrinectin-- Jon: See that's-- Joe: concentration? Jon: Yeah.That's something that I'll have to, to look into. I don't-- what I'm thinking I'll probably be able to do. The interstitial cells grow great on anything. So what I'm thinking of doing is, once I trap these interstitial cells and open up my, uh, tube, I can maybe coat the top of it with fibrinectin and then see the cells. And see how that works. Ann: I wonder how much cell contact $$ interstitial cell contact you would get. ## you know what I mean, you might want to like, slice through it. Jon: Yeah. Ann: to see if they are even touching each other. Nerem: Now when you compare porcine and endothelial cells with porcine aortic valve endothelial cells.. Same passage? Both on fibrinactin? Jon: Yes, uh, so what I haven't done yet is I haven't looked at any of the aortic cells, or any of those because I've just haven't had any time. I've just-- all I've done is the valvular endothelial cells to see what they do and I've just been comparing them with literature, which would be, you know right know it could be bovine, whatever, ambilical valve. Nerem: I guess when you begin to make conclusions about, uh, the valvular cells, not, uh, aligning at 12 dynes per square centimeter. Jon: Yeah, I know it's very important. I have those same, you know, I compare it with the porcine-- Nerem: Because different species respond to a $$ point of view very differently $$ shear stress. Joe: Are there also other endothelial cells like vascular endothelial cells which are far away from-- I mean I imagine that the way that the valvular and the aortic would be similar enough that to design $$ construct you would have isolate valvular. You could you know, maybe somehow switch the aortic to respond like the $$? Jon: mmm hmm Joe: So to sort of see how far away some other cell type might be as well. Jon: Yeah. I did actually find a paper once $$ some sort of a microvascular entity endothelial cell that uh, appeared to be more like, at least the pictures of microvascular cells $$ the valvular endothelial cells. Uh, I can't say much about anything else because no one's really tested the valvular ones. So, I mean.. : Yeah. Ann: I have a question about, I don't know, is there another way to characterize your interstitial cell because I think an initial thought might be to have a join population and you'll have solely interstitial cells, you've just accidentally gotten some endothelial cells in there with the varied morphologies Jon: Uh, yeah, that was like a big fight for a couple years ago, but ah, with all the different staining $$ what I've been doing and stuff $$ like these cells actually-- they don't posess the endothelial markers and stuff like that. They have distinct characteristics that make them interstitial, except they have, you know, there's like a spectrum of like two markers like smooth muscle actin or something else that aren't endothelial at all these cells just have it to a lesser degree. So I guess it seems like that kind of a question is an answer to. Ann: I think that, I mean I think that's a natural question that a first audience ## Jon: Yeah. So one of the concerns I have is, like if I'm doing all these experiments for four different cell types, you know, different time points, different sheer stresses, whatever, like you get this pretty huge matrix. So, ah, I don't know, I guess from experience if yo guys know of any easier way to either reduce that, or how to high points or which, you know. I thought that I should look at the valves cells first because we don't know anything of what they do. And if there's something interesting maybe that will compare with the aortic ones. Nerem: Well I think you have to do the aortic as a control. If nothing else I would probably focus on the endothelial cells first. Jon: okay Nerem: In terms of the shear stress. Jon: Okay. Nerem: As opposed to doing all four at the same time. Jon: Yeah, I don't think that I would-- I don't think I'm going to look at all at the uh.. Like I think I'm going to keep the shear to the endothelial and the $$ to the smooth muscle and the. That's like my smooth muscle being the control. But uh... Nerem: What's the status of the strain device? Jon: They said that they should be shipping it here on the twentieth. Nerem: Of January. Jon: That's what they said. Nerem: That's a sunday. Jon: Exactly. I don't know. : Personal courier. [laughter] Jon: So after that, you know, then give me a call, and we'll kind of go from there. Nerem: We're getting close. Jon: mmm hmm. Nerem: Good. Jon: Yeah. But uh, I guess that's... Nerem: Well, thanks Jonathan, we should probably move on to Kara. Ann: Kara thanks for the uh, food. Kara: Are you eating them? Ann: I'm getting ready to. Are there different ones, like for-- Kara: I have no idea. They're called like petite fours. I imagine yes. It doesn't say anything, though. : There's a little chart. ## Kara: They don't give any information. Jon: You want to switch? Cause I don't want to Kara: Yeah. Okay. : I want another cookie. ## Kara: So as most of you know I work with embrionic stem cells and the goal right now is to either 1: well, go to the next slide... Press the up arrow? Jon: Uh, down. ## : or the space bar. Nerem: Just press something. [laughter] Kara: See what happens. Ok. Jon: Is that in focus, for all you guys? I really wasn't looking at the screen when I.. [murmurs of agreement.] Kara: Ok, so one of the goals is to actually derive endothelial cells from embrionic cells. Um, and one way we thought that might happen is by exposing these embrionic stem cells to shear stress and I've done that and you'll see what happened. And then the other thing then is to derive these endothelial cells and to see if embrionic stem cell-derived endothelial cells respond to shear stress like mature endothelial cells that we're used to. And then also I want to repeat some of this work once I find like kind of a something I consider like successful repeat in primate and human cells. So I'm skipping all background because in August I focused mostly on the background of these cells. I'll move right into some of these experiments. So these are what my cells look like in culture. This is an alkaline alphatase assay. Alkaline alphatase is an enzyme which is considered a marker for embrionic stem cells, so undifferentiated stem cells and if you look here you see these cells grow in colonies. Um, let me go to the next slide. I have a blow up. Here we go. We can see these cells grow in colonies, and the darker purple indicates the most undifferentiated colonies, so they are probably like the best colonies, so like they're the best. And then in this, in this culture dish I'm growing them on $$ cells so they are like the cells that grow in between the purple colonies. Nerem: The white background is the $$ layer? Kara: Right. The white background. So you see the negative for the alkaline alphatase activity $$ should be. : There are actually cells ## Kara: THose are cells. Nerem: THose are Feeder layer are what-- Kara: Mouse fiberblasts. So when I grow these cells I plate them down, the embrionic stem cells are plated down on a monolayer of mouse fiberblasts. So I'm going to go backwards now... So you see the bottom picture that I took actually has so undiff-- um, some undifferentiated cell population because they're not growing in nice, tight colonies, they're more spread out and they're not as darkly stained. So that's just to show you some contrast. Ok,so moving on, Nerem: By the way,let me make a comment about the $$ layer. Beacause it kind of $$ the human embrionic stem cells, virtually all the human, well virtuall all the available human embrionic stem cells are available are growing on a mouse which doesn't seem to be a problem as long as you are doing research, but if you were to ever move to anything clinical, you would represent ##. So that's why ## people trying to move away from using animal cells to Feeder layers. Kara: Mmm hmm. And I think-- is it geron corporation that came out in august saying they've isolated and grown human embrionic stem cells actually without the feeder cell? Nerem: Yeah. There's one ## KAra: Yeah, one ## Josie(the newbie): So, is the feeder layer just a nutrient layer? What's the feeder layer? Kara: $$ it's a surface but even more so, it produces cytokines which are believed to help cells remain in their undifferentiated form. Joe: If the cells continue to proliferate, would they eventually form a monolayer? Kara: Um, if they start differentiating, into different cell types, yes, if I'm able to maintain them undifferentiated, which is usually the goal $$ culturing the cells, they stay in the colonies. Joe: So they're- I guess will never grow to the size that they Kara: If I let them keep going, and they start touching each other, and again they're still 3d, they will keep growing and start reaching out and touching each other but I don't let that happen because that could induce differentiation. So as soon as they start-- the colonies get big enough so that they're like touch one another they you know, Joe: Harvest the cells. Kara: Harvest the cells. Ok, so I've done a lot of different experiments, and I wanted to show all of them, so in order to keep them all straight in my head I have these dates, so you can see how many hours, or how many days did I do this. So first what I want to do is I want to see-- I'll do the embrionic stem cells, undifferentiated response to shear stress. So I see them on gelatin coated, microslides and then as a control I seeded these endothelial cells. This endothelial cell line that I purchased from ATTC. And then I set up my flows, and then I stained them with just the oxolin eeosin just to $$ the morphology. And I also have a few different types of media that I'm using. One's for the endothelial cells that's a $$ just DMEM or regular media with serum. ## And then my embrionic stem cell media which is special, and a lot of people have it figured out now I have this knockout DMEM. Like at least two people have taken my bottle. [laughs] Accidentally. And then I used-- Nerem: What does it knock out? Kara: You know, I don't know. I'm sure it says but I didn't bother to read it. The knockout serum replacement is actually considered a serum replacement, so if I use all knockout serum replacement I could potentially use that in place of FBS which is the, that's the embryonic stem cell qualified serum but that's basically FBS that's being, um, screened. Knockout serum replacement, I think the goal is to actually get serum out of the media, so you don't need serum when you're culturing the cells. And then, you know, some amino acids, and beta ## I think the goal with beta mercaptinawa is to basically to change the acidity, the ph of the media and LIF which is Lukemia inhibitory factor key cytokind that helps maintain the cells undifferentiated. And then flow media is the same but I don't use lif because at this point because it's very expensive. And I don't really want to use that much $$ take a lot of media. And then I have one more which is of all the differentiation media I take basically it a very similar so basically I take lif out and allow the cells to spontaneously differentiate. So those are the different media I'm using, flow loop, you've all seen. OK, so these are what my cells look like on the left, these are embryonic stem cells. Um, the small cells in between the colonies are residual fiberblasts and I plated them straight down on gelatin those fiberblasts came with the culture. They've been inactivated so they don't grow, you know, they're still present and alive. And you can see what I've $$ them just to four dynes, um, I'm basically losing all the cells. And the pictures you're looking at are of course the sections of the slides that had the most number of cells. [laugh] [laughter] Kara: Because otherwise you'd be looking at a blank picture. [laughter] Kara: So basically all my slides, all my cells, I'm losing them, so obviously they don't adhere strongly to gelatin, at least. And then-- Nerem: Um, I just wanted to comment on that but $$ if you want. Kara: Ok. Nerem: Well, you know, our seminar speaker yesterday. Kara: I know she said the very same thing. Nerem: And I'm talking to her afterwards. I think you probably need to go. Stephanie $$ control has 0.5 dynes per square centimeter. And you may want to go down to 0.5. Try it. ## Nerem: The speaker yesterday ## flow experiments. Kara: But she $$ culturing them-- Nerem: $$ takes very very little. Kara: That's what I found as well. That's what I found as well. Ok so contrast that with-- Nerem: And then that was only a couple of days ago when we were meeting with $$ but I've changed my mind. Kara: ## ok. Ok so this is my, these are my, uh, endothelial cells, my cell line, and so obviously there's a morphology change $$ four dynes, I believe that was 24 hours. And so looking more closely to cells, you observe the static controls, and they've very different. So if that's the extent of my $$ that's find. [laugh] But they remain adherant and all that, so. Ok, so moving on, now these are just pictures of what the cells look like growing them. So they're not stained, you know, so they're not purple now. You can see they grow in colonies. Each one of those round circles is like twenty cells, probably? And it's really difficult to see the individual cell boundaries, which is characteristic of embryonic stem cells they're growing on. Nerem: I need to ask a question, Kara. Back when you were showing your embryonic stem cells being almost totally removed by shear stress. Those are still on the feeder layer? Or not. Kara: No. So $$ I roll them on the feeder layer then plated them on the gelatin coated glass slide, some feeders. Nerem: I wonder if it would be worthwhile trying to do a flow loop experiment with a mono feeder layer... Kara: They would adhere more strongly? Nerem: Yeah. Kara: Yeah, I could try that. I could try that and seed down the feeders. Nerem: I don't know but it just seems like it would be worth a try. See what happens. And in fact if you put normal endothelial cells on the surface doesn't mean when you talk about what ES cells you have. Kara: Yeah, I could try that and see, because I would imagine they would probably adhere strongly to the fiberblasts. But I don't know. Ok. Joe: How three dimensional are they? Are the colonies? Are they-- Kara: How three dimensional? They're like little domes. you know, like that. Nerem: So what size are the domes? Kara: It depends on the number of days I've let them grow. They start with like one or two, three cells and if like they keep growing, they'll keep growing. I think-- I don't think they'll keep piling up more than a few cell layers and then they'll start spreading out a littl bit. Joe: The reason I ask is because the shear level they experience is a function of how high the-- Kara: I know. Joe: $$ up there-- Kara: I know, and I looked at that. But I have, I mean I have ten to fifteen cells for.. You know how, there's that spacer? 250 microns. So I have cells, and they can pile up for.. that's a problem. Joe: Ok but with one forth of that the shear is a lot different than ##. : It's like cubed, or... Tiffany: And that changes the flow after. Kara: It will just change the whole flow geometry. Really. Nerem: I mean think about it in the context of just, a chemical engineering background, flow over a spherical... Kara: Right. Nerem: And you've got half a sphere, so, so you've got that kind of a shear stress-- Kara; [laughs] $$ flow over a sphere. : Well, just try to make your height a much larger than you would ever experience cells and really increase the RPM of the flow loop. So that way, you know, you'll have a large band of shear but the height difference of your cells won't vary too much relative to that space. Kara: Oh, you mean just make a larger space? : Yeah, so make your spacer, you know, 500 microns or something. Joe: The six cell layers is negligible-- : Right. Joe: $$ the space between. Kara: Okay. Josette: $$ much easier. Nerem: But even so $$ the distribution. Kara: I still.. Oh, yeah. Okay. Ok, so then I took pictures of what the cells look like when they're starting to differentiate. These are still embryonic stem cells, you know I'm still trying to keep them undifferentiated so you can see some of them are differentiating because I have these monolayers that have these distinct boundaries, basically in both pictures. : I thought the ## was supposed to keep them from-- Kara: It is. So THESE are like, you know, : They're bad! Kara: These are bad cells! you know depending on like how you can culture you can get rid of these because the embryonic stem cells grow so much faster than the other cell types you can actually clean up your colony just by, you know, continuing with the culture. But at some point you just have to throw them away sometimes. Ok so now these have no $$ and I'm letting them spontaneously differentiate so you can see they're not colonies anymore, spread out, form these monolayers. And the bottom, left hand corner it's very, very confluent section. Nerem: And are they very $$ endothelial cells? Kara: [laughing] Who knows what they've differentiated into! Nerem: I'm figuring you're going to know. Kara: That's my goal. Josette: you're just looking for certain markers that are characteristic of endothelial cells? Kara: Yes. Josette: And if they don't turn out to be endothelial cells you said oops I messed up or do you try to find out what they became? Kara: Well, so far when people are trying to differentiate these cells, I have not heard of anyone being able to differentiate all the cells into what ever cell type they want. So you end up with a mixed bag, and then you should really have to go in and sort the cells. Now, Ideally, I'd like to be able to-- Nerem: Now I had the impression from the speaker yesterday that if she used $$ that she got a very high percentage of endothelial cells. Kara: Yeah. I don't think so. She, I mean, she did say if she used ## that she got endothelial cells but all we saw were like pictures of endothelial cells which could have been like one percent of the cell population. She never actually said how many were, you know, endothelial cells. Nerem: But I think if you use, you know, cytokine ## that are endothelial, in a sense, directed, Kara: Right. Nerem: You are probably going to get a higher percentage of endothelial cells. Kara: Right. Oh yeah. ## efforts, or something. Definitely high percentage but these people are trying to make neural cells, they are really hard, to try to make neural cells. They make like, 3%. And that's considered very successful. 3%. So. But you know some of these look like, you know, they could be endothelial cells. They're on collagen which is supposed to encourage, you know, $$ dermal lineages, of which endothelial cells is one. : Jan Tell you that morphology's not the whole picture. : Right. [laughter] Kara: Who would say that? : Jan. Kara: Jan. Nerem: Who's Jan? [laughter] Kara: Ok so then what I did now is now I'm staining the cells for an ethel k 1 um, antibody, using an ethel k 1 antibody. Ehtel k 1 is a receptor for fgf. And this receptor is, um, has been known to present itself as an endothelial progenitor $$ differentiating, it presents itself and then as they become more mature, um, the expression is lost. So, here are these cells, And you can see I have some positive expression. Eh, looks ok. And then what I did was I wanted to sort the cells that I have, with the flk 1 antibody expression. So I started with a million cells sorted for flk1 positive cells and then also ## which is another marker ## is a marker for the undifferentiated, um, the undifferentiated cells. so I ended up with a total of 13 thousand flk1 positive cells which is only 2.3% so, not very many. Again, I didn't give them veg-f, you know I didn't give them many $$ differentiation. So maybe that's not too bad. So I isolated these cells and tried to plate them on collagen. Four coded dishes and now I'm giving them them veg-f cytokine, which will hopefully induce further differentiation with these cells. So I waited 2 days with these non-attatched down. I thought something went wrong and we threw them all out, thinking that they were dead. So that's like part of that story. I'd like to continue that story a little bit because I've done some more work. So now we're looking at, I've looked at the $$ marker, $$ undifferentiated $$ stem cells. So I've plated cells on gelatin, and then I stained for Ecoherin, $$ ecoherin positive cells. So I got excite thinking oh, I could adhere $$ positive cells is actually going to clean up my cell population because those cells should be the most undifferentiated cells. So I did that and I got approximately two thirds of my cells were $$, which is good, and then I plated these cells and then you'll see what the pictures look like. This pre-sorting, this is what my cells look like before I sort. So I'm just staining, you can see the $$ positive, they're still growing in colonies. Those are nice images. This is post sorting of positive cells. And they're not even growing in colonies anymore. They just look bad. So in my opinion the shear stress from the sorting is just too much for these cells. And it just tears them up or induces differentiation of something. Or possibly the antibody binding to the ecotherin surface molecule might signal the cells to differentiate. One of the two, but it's obviously not a good mechanism to clean up my cell population. Okay, so now I'm moving on. Now what I want to do is my embryonic stem cells didn't want to stay here. Now I'm going to try to sort partially differentiated stem cells. So I'm allowing them to differentiate four to five days, on collagen 4 coded slides, and then expose them to shear stress. So what I did was I plated the cells on the slides, and waited four to five days. But in those petri dishes, those $$ petri dishes, with 20 mils of media, so I had way more media than they possibly need, not a lot of surface area for the cells to grow. So what happened is, I got the things that are called embroid bodies, which are basically floating aggregates of cells. and if you work with embryonic stem cell population people spend like a lot of time making these embryoid bodies and stuff. So it's a big deal, and they just spontaneously formed for me. So I'm hoping Athens is going to be interested in that. So anyway I got excited about that and I figured $$ and then isolated them using very rough means. And then uh, and then I also did the flow um, the flow experiments as well. So you'll see what they look like. Okay so this is what my cells look like on the glass slide, with and without lif. So you can kind of see the comparison. These on the right, they're still kind of maintaining their morphology [change discs] Nerem: -- fungus [laughs] Kara: Fungus! It is not! : Told you, that's why ## Kara: It is not a fungus! [laughter] Ann: Are the embryoid bodies like coming apart? Are the cells like crawling away? What's going on? Kara: Yeah, the embryoid, the cells are-- they're, they're growing out of the embryoid body on to the, on to the-- the gelatin. Josette: So the embryoid body is not breaking, then. Kara: The embryoi-- no, the embryoid body is still staying here and ## : I see stuff like that when I isolate like my interstitial cells, into like a blob, and I-- Kara: Oh, really? : --just come out of the tissue. Kara: Oh, yeah. : But it's-- they-- I mean those look more endothelial in the sense that they're polygonal and the cells don't do that at all. kara: Yeah. : You know? They don't-- they don't like come out of a tissue like that they'll either start in like a layer and come to a plate down and grow. They won't like-- Kara: oh. : --come out of this big ball. Usually cells that, like, one way you can tell is if.. Have you ever grown your cell cultures like say the differentiated ones, superconfluent and see what happens, see if they will re-join into nodules. Cause like, smooth muscle cells and my interstitial cells like once they get super confluent they might nodule up in these big, three dimensional packets. And you can take that and put it somewhere else and $$ again so-- Kara: Huh. : So I don't know if you could kind of reverse engineer that. Kara: That's interesting. Well, like, for example, a lot of other groups-- the reason why they make embryoid bodies is-- the idea is to try to mimic the three dimensional structure in the embryo. Because that's-- you know how these cells are used starting their differentiation process. Maybe it's necessary, maybe it's not, this is what a lot of researchers are using this strategy. I would, in the top right, lets-- it might look like neural cells, like little neural cells reaching out. But we'll see what the Athens people think because that's like what their scenario, the neural cells. So this is day three. Lot more cells. The top two ones I thought well maybe they're endothelial cells or at least misodermal lineage cells. I thought you know, I'm just going to isolate them and see what happens. So I washed off the embryoid body came off the places I washed. And then I tripsinized, um, those two ## says. I washed off the embryoid body I trisinized the endothelial cells the ones that I thought looked like cells and then plated them again on collagen with this media and then gave them $$ to see what would happen. And after three days they're still floating, like my other cells were floating $$ here down $$ christmas, so I just froze my cells, because it's time to go. [laughter] Kara: So they didn't get more than three days of growth. Okay so now I'm trying to isolate more of the flk cells because I thought I'll just do it again, I'll have more cells, I'll use more antibody, and maybe this time I'll be more successful. $$ the cells, I think they had four to six days? Six days for spontaneous differentiation and then I sorted based on this flk1 and the marker again, and I have one point for percent this time so still very low percentage of flk1 positive cells. And then I plated these cells again on collagen 4 with um, my differentiation $$ and they had four days $$ christmas. And after 4 days they were still floating. They looked like they might have actually been growing, though. It looked like the numbers had been increasing but all of them were floating. So I froze those as well. So we'll see. That's to be continued. [laughs] $$ those cells. Okay, so, conclusions. My cells are positive for ## and $$ which are both markers for undifferentiated stem cells so that's nice. Um, stem cells and partially differentiated stem cells not staying here under shear stress. I had to kind of figure that out but my endothelial cell line does so at least I know that I'm doing everything right calculating the flow, and the flow loops, $$ so that's good. And then also I've got these cells that are not attaching. And I've isolated them twice flk1 positive cells now and my crazy embryoid body technique $$ and none of these are $$ down and I've given them only four days and I think maybe they need more. I don't know. So what I did was I went back to the literature and I thought if I really am in fact $$ endothelial progenitor cells, maybe I am. Maybe this is what they're supposed to do. Um. So I went back to the literature and I looked at what people are doing when they isolate circulating endothelial cells cause I thought maybe these are $$ are circulating so maybe they're $$ cells anyway, right? So I thought well, maybe I'll just mimic what they did. So I looked at the literature and it looks like it does take a while for the endothelial cells to attach. Four days, up to four weeks on the culture. And then um, when you first isolate the cells. After you first isolate them and then they differentiate and then they mature into endothelial cells and then $$. and then I also found that they use fibernectin, which these researchers have found to be more successful in getting the cells to adhere. Attach. So maybe I'll try that. That's what I was thinking. Ok, that's all it says. Try a similar protocol to try to simulate circulating endothelial cell isolation technique. And the other thing I was thinking was I try this with endothelial cell medium and see if that maybe induces more differentiation of these, um, the stem cells, or if maybe it provides a better media to $$ cells to attach. But one of the-- maybe what I'm going to do is try and see what that media does. Because $$ is, supposedly idealized endothelial cell medium. Thought I'll try that one. Anyway, now I'm looking for ideas. What can I do to get the cells to adhere? Basically that's my main problem. Getting the cells to attach. So I thought maybe I'm not using enough antibody to isolate my cells. Maybe I'm-- I-- the sort is killing them. I don't know. Well they're not dead because I tried ## blue and the cells are alive, they're just not really growing much in the $$, and they're not attaching. Other substrates... Any ideas? Joe: Have you tried centrifuging them? Kara: Yep. Well yes, because I have to isolate them. Joe: No, to, um, you know you put them on a plate and then you centrifuge the plate for 6 hours. Kara: Oh, I didn't try that. Like a cytospin? Joe: Well, just, I mean, kind of, put the cells down to the bottom and keep them ##. Kara: Because gravity's ## Kara: Could try that. Joe: I don't know. I mean that's-- Andrew does that with his I guess retrovirus transfections. Because for some-- even though we're not exactly sure what the mechanism is it's not strong enough to pull the virus particles down, but for some reason that seems to increase the transfection efficiency but also.. Kara: I could try it. Joe: Even though ## Kara: Ok. Cells on the dish. ## ## Nerem: When are you presenting up in Athens? Kara: Tuesday. 9am. : You have plenty of time to come up with these before then. Kara: Yeah! Plenty of time. Oh and the cells that I froze down on I lost-- I mean I only had a few thousand to bein with. I've lost most of them. So ## anything. We'll see. Nerem: They didn't like the white christmas. Kara: They didn't like being frozen. Ann: ## embryonic stem cells similar to the way ## Kara: Yeah, very similar. They're a little bit more $$ as far as adding dmso and $$ time over like 5 minutes. A little bit more rigerous not to shock the cells. ## Basically the same. : So um this is flk1 positive cells that you're sorting out. Kara: Yes. : Is-- are those the adherent ones? Just $$ with? Or do you have any of like-- because you have some that are in this ## and some that are in this growing out of it, right? Kara: Well I isolated-- well I-- : Are these all in colonies? Kara: Those were-- I'll show you what they look like. When I isolated the flk1. When I let the cells spontaneously differentiate. Let's just go here.. They look like that. The left side. So they're not really in colonies anymore, they're kind of, you know, spread out. : So to try to figure out if it's actually your procedure that's disturbing it. Is there any way of $$ while they're there if they're adherent before then ## Kara: Right. You know I did that. I mean it looks like we have flk1 positive cells here. so I mean it looks like I have some flk1 positive cells. Not that many but here, yeah. : But they're adherent before you sort them so yeah, ## what's changing? Kara: Well, I don't know. ## Kara: It looks like they're adherent before I sort them : You're not staining for adherence, you're staining for the veg f ## Ann: Veg f would be an adherent cell or am I going too far in my assumption? [laugh] : ## can you stain for anything? Nerem: endothelial cells have ## one, right? Kara: Mature ones are not expected to express ## one. Although according to ## he surprised me. He stained the endothelial cell line ## one and he got positive cells. Maybe I'll try that. Some. I mean, I could try it. But it's supposed to be trans-- like it's ## in the early progenitor stages and then it loses that. Nerem: So there must be some other veg f receptor on endothelial cells besides flec one. Kara: There are two of them. There are two. Nerem: And they are? Kara: I think FLD is the other one. ## So there's another one and then, I also decided I'm going to try CD 34 which is a marker for hermaphlotic stem cells and endothelial cells. Nerem: Well we should see what the experts up in Athens think. Kara: I know, I'm interested in hearing what they have to say. So. Nerem: ok. Thanks Kara. Kara: We all done?