Web Conference with Dr. Joseph Wu

Modeling cardiovascular diseases with stem cells.

Ebert AD, Liang P, Wu JC. 2012. Induced Pluripotent Stem Cells as a Disease Modeling and Drug Screening Platform. J Cardiovasc Pharmacol.

-Burridge PW, Keller G, Gold JD, Wu JC. 2012. Production of de novo cardiomyocytes: human pluripotent stem cell differentiation and direct reprogramming. Cell Stem Cell. 10(1):16-28.

-Plews JR, Gu M, Longaker MT, Wu JC. 2012. Large Animal Induced Pluripotent Stem Cells as Pre-Clinical Models For Studying Human Disease. J Cell Mol Med. 1582-4934.

-Kamileh Narsinh, Kazim H. Narsinh, and Joseph C. Wu. 2011. Derivation of Induced Pluripotent Stem Cells for Human Disease Modeling. Circ Res. 2011 April 29; 108(9): 1146–1156.


00:00 Doctor Rue, first of all, thank you so much for your time to join my class, which is seated behind me, this…

– Good, hello.

– Hi, everyone. This is a class largely of Smith College students, but we do have some other students from Amherst college that are part of this class as well and, so and, thank you very much for your time, as you know, we’ve gone through several of your papers and three of my students just got done, sort of, presenting the contents from that and we’ve had a nice discussion about some of the implications from the work that you’ve been talking about.

– So, to get it started, you know, I was hoping that you might tell us a little bit about yourself, your career path over the years and, sort of, what led you to begin working as on cardiovascular developments and the use of Stem Cells to treat diseases related to it.

– Yeah, so I’m just making sure the, I guess, the video’s on. So , it is the.

– One more check, I’m sorry. I’m just gonna, say something.

– Yeah hello? Can you hear me?

– Kay, good, one sec. let me check the volume.

– Okay.

– Okay, good.

– So can you hear me?

– Yup, is that better or a little louder? Like, good. Okay we’re good.

01:22 Okay, so it is a pleasure for me to be here. I guess talking with you guys over Skype, and I look forward to having a very interactive discussion, over the next hour or so. I’m an Associate Professor in the Department of Medicine and Radiology at Sanford. I see patience with heart disease problems and I also do basic science research related to cardiovascular diseases. I went to medical school at Yale University from 1993 to 1997. And this was followed by a medicine residency and cardiology fellowship at UCLA. And during my clinical training I decided to focus my career in exercise research. And so I decided to do a PHD. I received my PHD degree from the Department of Molecular Pharmacology at UCLA. My thesis at that time was mostly on molecular imaging of cardiac gene therapy and cardiac stem cell therapy. My PHD advisor was Professor Dr. Sam Gambhir. Who is a premiere physician scientist with the expertise in molecular imaging and nano technology. And he has had a very positive influence on my career in terms of mentorship I was. So around 2004 I came to Standford and had been involved in various aspects of cardiac stem cells biology research assistant.

– Okay. Any one have any questions for Dr. Ru about his career?

– How did you come to find, to select, where you actually ended up during your PHD? Just curious as to, you know your choice making those, that kind of direction.

03:11 Yeah, so for my PHD I chose molecular pharmacology, because at the time I was finishing up my cariology scholarship. And stem cell research, as far as gene therapy research, was quote unquote, quite hot at the time. And I was actually interested in figuring out the pharmacokinetics and pharmacodynamics of cell therapy and gene therapy. And at UCLA there was a core group of investigators, who was working on molecular imaging to track gene and cell therapy for cancer applications. So I decided to talk to some of the professors there and say that I basically want to learn your technology and apply it to cardiovascular field. And that’s how I got started. That’s how I chose my lab at that time.

04:01 What affect would the immature HPSC have on modeling cardiac disease? You speculate in your article about the effect on drug testing and regenerative mater. But you don’t really say much about how them being immature can act to model diseases. Would there be obstacles disease to model diseases that aren’t early onset cardiac diseases?

04:24 So, I guess the question is, what effect would the immature IPS cell derive cardio cells that we have on modeling cardiovascular disease? Well it happens that this is a very important question, in fact the question that we get asked all the time, by grant reviewers and by mediscript reviewers. So I’m very happy that you’re asking this question. I think the answer is that this is one of the main limitations for IPS cell technology for modeling cardiovascular diseases and other diseases as well. And this is because the IPS cell derived cardiac cells that we currently obtain is still quite immature. And resemble a fetal cardiomyocyte phenotype. And as far as I know, this is also a major problem for other fields of investigation, such as differentiation IPS cells into pancreatic other cells, or into hepatic cells, or into moana cells. So it would be nice if we can improve the differentiation process so that we can obtain more mature IPS cell derived cardiomyocytes that better resemble the real cardiomyocytes from patients. I think many investigators are currently working on this area. For example, use electrical and mechanical stimulation, or try switching the cells with different genes to promote the maturity. Now keep in mind that regardless of how well we can improve the differentiation process in the petri dish. It will never be the same as cardiac cells from a patient. Remember cardiac cells from a typical 70 year old patient has been exposed to decades of normal hemodynamic mechanics. You know, boom boom. So a heart beat at rest, during exercise, and when they’re sleep. Like wise the cardiac cells of a typical 70 year old patient had been exposed to other potential stresses such as diabetes, hypertension, and high cholesterol, to name a few. So these are the kind of non genetic factors that we can not yet simulate, using our in beats for a differentiation process. And so this is also another reason why, in general, most experts in the field believe that it is easier to use IPS cells to model and deviate inherited monogenetic disease. Such as a mutation in one gene that gets passed on from say a grandfather, to a mother, to a daughter. And it is much more difficult to use the IPS cells to model polygenic disease, which gets multiple genes involved. For example, coronary artery disease, diabetes and hypertension. But you question is a very good one. At present that is one of the limitations of using IPS cells.

– Thank you.

07:27 I have a quick follow up Dr. Rue. I find it actually interesting that, your end response there was essentially the impression that IPSCs are almost not good to, or the best model to try to get at pleiotropic diseases. You know diseases in which maybe there are multiple factors that we don’t even know what all the factors are. But in fact aren’t IPSCs essentially, if we want to stay human with things, the only way to do it?

08:00 Yes, So I think this and quote unquote, element of time here, and element of time meaning that at present it’s much easier to study monogenic disease using IPS cells. Because the noise that you get from making the IPS cells and differentiating these cells into certain phenotype can be, for example, I’m just pointing out my fingers, could be this big. The monogenic disease in which it is inherited, from one generation to another generation, the noise could be this big. So it’s easier study, where polygenic disease, the noise could be this big, and make it very hard to study. I think in the future once we understand more about these IPS cells and can reproducibly produce these cells to differentiate these cells. It is possible then to tackle polygenic disease reliably. But for now I would say that if you wanna study IPS cells disease, the model some kind of a disease mechanism that’s much easier to study a monogenic disease.

– Okay.

09:08 So my first question for you is, assuming that sometime in the future, it’s a cell line stride from clinging IPSCs, or and or, ESCs are approved for human transportation, and assuming that IKSC and EFC cardiomyocyte, and hundreds of over different SC derived tissue types are available. Do you think it might be possible to extend life expectancy far beyond the age Like being able to replace old or damaged parts of the body indefinitely? And would you personally approve of this kind of use of stem cells?

09:44 This question has to do with, I guess question number three, which is, using IPS cells to possibly replace damaged parts of the body, or even to replace whole parts of the body, so that we get to live forever. I think the first part of the question is whether in the future can we replace damaged parts of our body with ESL so with IPS cells, infact these efforts are already being investigated by companies such as advance cell therapy or ACT, using human ESL derived retinal pigment epithelium cells, or RPE cells. These cells are used to treat patients with congenital immaculate dystrophy and with aids immaculate dystrophy, which can cause blindness in patients. And actually about a month a go, there was a paper published on acid where by investigators from UCLA reported that injection of these human ES cells derived retinal pigment epithelium cells, seems to improve visual acuity of two patients, with aids related immaculate dystrophy. Now granted this is only a preliminary study, I think the results are actually quite encouraging in terms of answering, is it possible to replace damaged parts of our body? For cardiac stem cell therapy, which is my area of research, focus, I think in the future it’s certainly possible to inject patients with a heart attack using ES cells or IPS cells derived Cardiomyocytes. Have these cells grafted to the heart and improve the cardio function. I think in fact this is something, that has already been actively pursued in many clinical trials today. By using adult stem cells, adult stem cells that’s just a bone marrow stem cells, knees and common stem cells, and cardiac progenitor cells, for patients with a heart attack. But for ES cells and IPS cells I think it would take much longer period of time to validate the safety, the feasibility, and efficacy, and many due to issues such as a tumor tenacity and immunogenicity. Before the FDA, which is the Food and Drug Administration, can approve a set trail. I personally believe within the next decade or so, there will be a clinical trial, focus on using allogeneic ES cells derived cardiac cells. Of the treatment of heart attack patients. Allogeneic means that person A receiving cells from person B, or from another person. However, I personally think that for autologous IPS cell based therapy, which is what you are referring to, autologous meaning person A is receiving cells from person A himself. This would be much more difficult, because of additional regulatory hurdles, and financial reasons. For financial reason, I’m referring to the fact that for most of the RND, research and development of drugs, that we give to patients. These are typically sponsored by pharmaceutical companies, and as you know, pharmaceutical companies are in the business of making money. So they would prefer to validate one drug, one cell type and sell it to as many people as possible. They would not prefer a business model in which they make one drug or one IPS line. And sell it to one patient, with heart attack, for the purpose of autologous therapy that you’re referring too. I think the second question, that you’re referring to, whether we can use IPS cells, ES cells, to quote unquote replace old part of our body. I think this is a difficult question to answer, simply because we’re not there yet. Meaning that we don’t even know how efficacious, and how safe these cardiac stem cell therapies are, as of now. So I think within my life time at least, is it possible to replace all my old cardiac cells with new cardiac IPS cells to prolong my life, as you’re asking. Probably not, I think in the future, lets say 100 to 200 years from now, is it possible to completely replace someones old cardiac cells or old heart with new IPS cell derived cells or new IPS derived organ? I dunno. I would not be surprised given how fast that technology evolves with time. I think the analogy. That if you could say for example, have somebody from a couple hundred years ago, you know somebody in the Victorian era, and ask him or her is it possible to do heart transplantation, most people would shake there heads and say no, right. But this is something that we now routinely do to patients with heart attacks and chronic heart failure. So I guess my message is, never underestimate the element of time and how fast technology could evolve with time.

– Thank you.

15:12 I just wanted to, this is the fourth question In the production, in the paper, sorry. Paper production of de novo cardiomyocytes, you imported human stem cell differentiation in direct providing pure cardiomyocytes, following, attempted differentiation, the derived IPSCs were unchanged, if you’re getting great a mixed progresionary cells after great big addition and different impures Does that mean that the exact proportions, and times and factors, to generate certain types of cardiomyocytes haven’t beenyet, also do you have any inkling as to what might be missing from the differentiations process of IPSCs, and ESCs to cardiomyocytes, such that they are not maturing? Are we missing an entire step that’s one chunk of factors.

16:07 Yeah, so I think my Skype is set but it’s in and out. So I’ll just repeat the question, just for the video purpose. I think your question is How do we improve the differentiation process of IPS cells to cardiomyocytes and how do we improve the maturity? And are we missing an entire step, a chunk of factors, I like that you put a quote unquote chunk of factors. I think this question, the answer to this question, relates over to my answer for the previous question. Question number 2. I think so far we do not have perfect ways for differentiating cardiac cells from ES cells, or IPS cells. Infact I would say that we do not have perfect ways of differentiating different cell types. As far as I’ve searched, hormonal cells, hepatic cells, or pancreatic or iris cells. And this has to do with what I just mentioned earlier, which is quote unquote the element of time. So if you think about, human ESLs was only discovered about 14 years ago in 1998 by Dr. James Thomason. And human IPS cells was only discovered about 5 years ago in 2007 by Dr. Shinya Yamanaka. If you add in the restrictions human ES cell research over the past decade, due to political pressures, realistically scientist have only begun actively work on human cardiac cell differentiation protocols over the past five to ten years. So within the past five to ten years I do think we have made significant headway. And we are now able to add different growth factors into the petri dish, and after about three weeks or so we can make a human ES cells and IPS cells in that petri dish to become a beating cardiomyocytes. How ever we can not make 100 percent pure cardiac cells and one of the reasons that you’re referring to is that we do not know a lot of the answers to these questions. Such as, how long do we expose the IPS cells to various growth factors, at what time point should we expose these IPS cells to growth factors, at what dosage, in what particular order and then what combination. We do know that there are several major signaling pathways that are involved in the early cardiac differentiation process of the IPS cells and these include, for example, the BMP pathway, the TGF activin nodal pathway, WIT pathway, and FTF pathway and certainly there are other pathways that are involved. And many investigators including our group are trying to figure out using, either mouse models, or human fetal heart samples to model that. So these new pathways could be what you are referring to, as quote unquote, missing entire chunks of factors, that we’re not aware of. And hopefully some of you in the audience, will take an interest in doing, academic research, and work on these, figuring out what these, missing entire chunks of factors are, yeah.

– Thank you.

– Anyone we have a follow question? You have one?

19:40 I do, yeah. Okay we were discussing this earlier, before we Skyped with you. In one of your papers you have different, it’s a chart of three different blank. Here and there’s two in this one where you’re doing suspension embryonic body, one where your doing forced aggregation, and one where you’re culturing the cells in a model there. And in those, there aren’t many different researchers, who are doing, adding completely different factors, at different times and they’re getting cardiomyocytes at the end. How is, what they’re all added in completely different things. Do you know which chart I’m talking about?

– Yeah, yeah.

– Sorry I’m just throwing that out at you.

– No that’s fine. I think you’re, you’re Samantha right? So

– Yeah but that’s not a keeper I’m just asking.

– So you’re referring to question number one. That is.

– Oh I guess.

20:34 That is, why is it more effective if the culture, human IPS cells in the model there? It would seem intuitive that a differentiation protocol would be more affective if it could mimic in Vivo environmental factors. And a question to do with the forced aggregation, whether it produced less differentiated cells with each treatment right? So think, you know this is a very important question, and I was emailing Mike that you guys ask so many tough questions, that most of the time we don’t have good answers for. I think it relates to what is the optimal method for differentiating human IPS cells and human ES cells and the cardiac cells. The present the differentiation process is actually a very delicate. And there’s still a considerable variability in the yield of these psychotic stem cells. I think that’s when we highlighted in the cells stem cells review article that there are three major techniques. One, just to give you a background, One of the techniques is using suspension embryo bodies. Which is often referred to as 3d technique because it’s a three dimensional embryo body. The second technique is using a model layer technique throughout the whole differentiation process. Which is also referred as to 2d technique, or experts in the field would call it the Geran protocol. And number three is using a monolayer followed by forced aggregation into a smaller embryo body later on, which is a hybrid of both the 2d and the 3d technique. At this moment there is no single technique that is better than the other. And I would say that even within the same technique there is still considerable variability from one cell line to another cell line. From one lab to another lab, and even from one post-op to another post-op within the same lab. So I certainly think more studies that will be needed in the future to optimize the cardiac differentiation protocol. I think the second part of your question relating to mimicking the in vivo, the in vivo environmental factors is a very good one. In fact that is what we are already doing. So most of the growth factors that we currently use in the differentiation protocol is based on what we know from cardiac development to biology in terms of what genes or what proteins get turned on at different stages of the cardiac development. This is the reason why, for example we add, as you can see in that figure activin A and FTF and BMP 4 and other kinds of growth factors, to mimic the cardiac differentiation processing rebuild. Actually I just got offline, lets see if I can get back. The question about differentiating the cells to mimic the in vivo environment is a very good one. So one of the concepts is, is it possible to differentiate these cells, for example, using some kind of three dimensional tissue engineering construct, that better mimic in vivo environment. This is something that a lot of people have thought about in the past, for example, one question that we’ve tossed around is that is it possible to inject IPS cells into beating heart, and allow the local environment to promote the cardiac differentiation? The other question is, is it possible to mechanically and electrically pace the IPS cells and to promote the electrical mechanical maturation? Some of these techniques have been tried. Without much significant success. And other techniques are still being tried. And that’s why as of today the three major techniques for differentiation remain the 2d, the 3d, and forced hybrid aggregation technique that we mentioned in the review article.

24:51 I have a real quick follow up. So everything that you’ve been talking about right now has been, a lot on how to get these cardiomyocytes to differentiate. You just mentioned this kind of, interesting idea of injecting IPS cells into a beating heart. I’d love to hear you comment very briefly on Doris Taylors work, right. Where she is eliminating all cells of a particular heart trait to provide just the left over extracellular matrix scaffold, and then reseeding that with a progenitor cell. As far as I can tell those cells are in fact differentiating quite nicely after that scaffold, based on the, what ever the local cues are that are residing in that extracellular matrix.

– Yeah.

– So have you ever tried to think about what those cues are, and utilize those, or take down approach.

25:42 No, I think the tissue engineerio codes, in my opinion, is much easier for other organs, For example, if you wanna create a trachea, if you want to create a bladder, if you want to create a, other tissues. I think it’s pretty hard, it’s actually much tougher, because the heart is a beating organ. There’s a very very, in a nice element of the number cardiomyocytes, the number of fiberblast, the number of epithelial cells that all need to align together to form a beating heart. And after that one, it’s almost like gods of the science. I think it’s very very difficult to do that, as far as Dores work is involved, yes she basically perfused a heart, digest the heart, and leave the heart basically within extracellular Matrix, seed them with a cardiac progenitor cells, I think inorder that experts in the field, will be interested in persuading this area, and also in trying to figure out, what are the cues that tells the cardiovascular progenitor cells to say I want to become a cardiac cells. And this to me there’s a fiberblast next to me there’s a blood vessel and these are the questions that still need the answer in terms of her set up. But she’s probably a better person to ask than me

– Because it relates to differentiation and trying to identify

– Yes, yes.

– Hi.

– Hello.

27:17 My question was, would the application of de novo cardiomyocytes be plausible in infants who were born in, with the congenital heart disease, for example hypoplastic plus left heart syndrome, where the left side of the heart does not develop completely?

27:33 Yeah. I like this question because, it’s something that’s dear to my heart. I’m a, as I told you early an adult cardiologist, and I see patients with congenital heart disease. And that means I do see patients with the hypoplastic left heart syndrome, with left ventricle noncompaction syndrome, with a tetralogy fallot, with ventricular septal defect, just to name a few. Certainly this is something that we have thought about, That is could it be possible that you could take, a skin biopsy of an infant, reprogram the fiberblast IPS cells, do genetic correction to replace the mutated gene with normal gene, and then differentiate these IPS cells into cardiac cells, and inject into the infant heart to treat the problem. I think this type of targeted gene correction with mutation and IPS cells has been demonstrated in mouse models before. For example, there was a paper in Nature 2001 showing that genetically modified mouse IP cells can be used to treat alpha 1 antitypes and deficiency. Which is a monogenic disease involving mutation in the Serum protease inhibitor that can cause liver failure. So I think your idea is an exciting one, certainly more research would be needed, to validate this concept in the future.

29:15 My second question was, in the review article, in production of de novo cardiomyocytes you acquire potent stem cell differentiation and direct reprogram, it states that, a minor cortical infarction could lead to needed replacement of a billion cells, how would such a large number of cells be reproduced in such a short time for the person to survive?

29:37 Yeah, so these are very good questions as well. So I think you might say that the math doesn’t add up right, because when a patient gets a heart attack, as many as, one billion cells are killed. So how do we replace one billion cells in such a short period of time. The answer is that we don’t, we don’t need to replace all the cells. For clinical trails that are currently one the way, using bone marrow cells, and mesenchymal stem cells, also cardiac progenitor cells, the cardiologist actually typically, inject any where from about five million to one hundred million cells. Which is significantly less than one billion cells right, and the logic is that these are stem cells that we inject, can release pool survival, pool antigenic, and anti-apoptotic growth factors. These growth factors then prevent the native cardiac cells that are ischemic and dying. Specially those native cardiomyocytes that are near the border of the zone, These growth factors can also serve as a stimulus for recruiting peripheral bone marrow cells to hone to the heart and repair, and finally, these growth factors can also activate or stimulate endogenous cardiac stem cells in the heart to proliferate. This is what we call the quote unquote, paragon hypothesis of cardiac stem cells therapy. That is the stem cells that we inject are actually secreting some kind of pritikin fectors to prevent cardiac cell death, to recruit peripheral stem cells and to stemulate endogenous stem cells. In fact the stem cells that we inject, keep in mind, most likely do not survive for a long period of time. By all indications especially some of the molecular emission studies that have been done in our lab. These stem cells most likely do not replace the death cells in large numbers and these system cells are most likely also do not need significant mytocardio regeneration de novo. Another more practical reason, why we can not inject the one billion cells in to the heart is because it would simply plug up the coronary artery, and cause a micro embolisim. Basically just clog up the arties and causes small infarction and lead to further heart attacks. But I think your question of how to get so many cells in such a period of time also brings up what I had to discuss earlier regarding the practicality of using embryonic stem cells or IPS cells in a clinic. To generate such a large number of cells immediately, when the person gets a heart attack, say one million, a hundred million cells, it would be easier for a company to differentiate well validated ES cells or well validated IPS cell, the right cardiac cells and then send them to the hospital for use as allogeneic cells transplantation. I think in contrast it would be much more difficult, for a company to differentiate, well validate lots of IPS cells and send them to the hospital for use as autologous transplantation, because the whole process of making IPS cells and differentiating through cardiac cells would take at least five to six months at least.

33:10 So my question was on the paper, production of the And it basically states that, in order to reprogram cardiac hyperverse into cardiomyocytes cells that you began, slice up 14 key genes related to cardiac development. So wondering how did you get these 14 out of the key genes? And what you have encountered to find these and what do you define as a key gene?

33:36 Yeah, so I think you’re referring to the paper by Aiega Hurrow which was published by Deepak Shrivastava a group in cell in 2010. So basically these genes were selected using micro array analysis that compare the transcription factors that are present within mouse cardiomyocytes , where it says mouse fiberblast, Deepak group chose 14 factors that were highly express in cardiac cells, and the they transfected the cardiac fiberblast with the pool of these 14 factors and found that they can convert the cardiac fiberblast into cardiac cells. And then they started removing` each factor one by one to see which factors were not essential and finally using this algorithm. They were able to narrow down the combination to three cardiac developmental transcription factors. Which are gata 4, meth 2c, and TBX 5 , published in the paper. I think basically the authors took the same story book from Yamanaka’s 2006 cell paper on making mouse IPS cells, from mouse fiberblast. Remember in that parper the first author in that paper, I believe was Takayashi. He selected 24 genes that are highly expressed in mouse Es cells. Transcipted these 24 genes and mouse fiberblast and shared that these 24 genes were able to cause appearance of mouse ES cell white colonies. He then narrowed down the list of 24 genes basically down to the now quote un quote classic Yamanaka four factors of R4, SOX2, Semic and KLF4.

35:30 Hi, so my question is, Why do you think certain cell types such as marrow stem cells can be more easily reprogrammed than others? Does this difference in reprogramming ability in any way effect the viability of certain cells in disease modeling?

35:45 Yeah, so you guys are asking all these tough and very important questions. We don’t exactly know why certain cell types are easier to reprogram. Most likely I believe it has to do with several factors. One major factor is the stemness of the cell, meaning that in general it is easier to reprogram, a multi purpose stem cell such as neuro stem cells, or mesenchymal stem cells than to reprogram a unique potent stem cell, Such as skin fibroblast. For example, you know, newer skin cells can give rise to nuance astocytes and dendrocytes, mesenchymal stem cells that can give rise to bone cells, to cartilage cells, fat cells. Where skin fibroblast is committed to become a skin fibroblast. Because these bulky, potent skin cells already express many of the transcripts that are factors. That are related to cell renewal and to full potency. In general it is easier to reprogram them in to prepotent IPS cells. I think after sampling all that we published a paper in PNS in 2009, showing that human adipose lipoma cells, which we can regularly obtain from patients undergoin lipo suction, is about 20 times more efficient to reprogram than embryonic human fiberblast. Another major reason has to do with delivery of the reprogramming factors. Either by using bioreactors or nonbioreactors or protiens or chemicals. So typically cells that are adherent in a dish, such as mesenchymal stem cells, or newer stem cells are easier to transplant than cells that are grown under any suspension culture, such as, hematopoietic stem cells. I think the last part of your question has to do with whether the difference in reprogramming efficiency can also effect the variability in disease modeling. Which is what we are interested in doing. I think it turns out that as we know, there’s a certain component that we call, epi genetic memory that’ll involve in these cells. Meaning that if you take bone marrow cells of a mouse and take fiberplast from a mouse and reprogram both cell types into IPS cells and the differentiate these IPS cells back into the blood cells. You will find it easier to get blood cells from IPS cells that were derived from bone marrow cells than from fiberblast. This was eligantly demonstrated by George Daly’s group, in a paper published in Nature 2001. So modeling cardiac disease, does it make better sense to make IPS cells that came from cartilage cells than from fiberblast? Truthefully I’d say yes, but keep in mind that some of these epi genetic memory, they do get you erased if they prolong past time IPS cells, and therefore one very interesting question that you may ask is, How good is the disease modeling using heart visis fiberblast derived IPS cells? I think another question is, How good is the disease modeling using early versus late pass the tide ES cell. I don’t think any one has looked into these questions yet. You know maybe these set of questions that you guys are interested in doing and doing a summer research rotations or something.

39:28 So I had a follow up question to Allys question on number eight. And if I understood correctly you were saying that possibly later passage, I can seed, or bond pheno seeds, might have structure that change drastically or lost to them characteriestics, would that make the cells more plyable in some way to differentiation, if those epi gene the original epigenetic structures have changed or are gone possibly?

39:57 Yeah, so typically if you look at it, lets say for example you take a, let me get you a real live example, you take an epithelial cells reprogram the IPS cell and differentiate this IPS cells back into the epithelial cells. And early passes you will get a higher yield, at the epithelial cells. Late passes the yield will come down. And so this is because some of this quote unquote epi genetic memory is being erased on a late passaging, and wiping out the epi genetic memory. There’s less resemblance in the epithelial cells. So after the late passes, you don’t see it anymore. And this is also finding that George Daly’s group had, with the mouse model of IPS cell that came from blood versus IPS cells that came from fiberblast. Yeah.

– Why thank you.

– Hi!

– Hello.

40:54 I just wanted to say thank you so much for being here today. So my question is number nine, You’ve mentioned many pros and cons for using large animals, and the IPSCs to model diseases with an over all message that there are more superior than most other forms of research subject. So basically my question is, if they make such better research models then how come no one else has strongly suggested this before and or has used them more commonly in their own research?

41:22 Yeah, so, this question has to do with, why not use large animals IPS cells? And it’s a very good question. That you are asking. And the answer has to do with level, Or the extra hassle of large animals, And budget issues. Now remember, most basic science investigators, are quite familiar with the mouse models. However not that many basic science investigators are familiar with large animal models such as pigs, dogs, and sheep, right? It is also much more difficult to work with large animal models because the extra scrutiny, and the documentation that is needed. Which can include regular business sites, side visits from the USDA or Us Department of Agriculture. And also the use of a large animal, because it can eat more can be costly than mice or rats. Just to give you an example, you can purchase a mouse from Jackson laboratory for about 30 to 50 dollars. And then house four mice in a cage for about 50 cents to a dollar a day. Now to purchase a pig from a commercial vendor will cost you any where from 800 to 1200 dollars. Now the daily housing fee could be any where from about 40 dollars to a hundred dollars if you add on the indirect cost, depending on which university. For example at Stanford, I either pay 60 dollars a day to house a pig in a small cage, which honestly I believe is more expensive than staying in some of the cheaper hotels, right? So for these three reasons; Comfort level, exhort hassle, and high cost, not that many investigators can work with large animal IPS cell models.

– Alright, thank you.

– Yeah.

– One clarification there, so why are you going through all the hassle to use them?

43:42 So I’m a, you know, I’m basically a physician scientist and at the end of the day, all the therapy that we’re interested in is pushing toward human application. Now I told my wife that my career would be disaster if I become a mouse cardiologist for the next 30 years. So obviously in the end you want to translate to humans. So to translate to humans, there are times the FDA or whatever ask you to do it in a mouse model, in a large animal model, before you can apply it to humans. That’s the reason why we work with large animal IPS cells, differentiate the cardiac cells, inject into large animals, and to see how they, how well they work.

– Well under that logic, why don’t you go straight to primates?

44:37 Primates, you can also go to primates, and it also depends on the comfort level of the individual PRAS as well as the institution, and also if you’re thinking about how much hassle I’d get working with pigs and dogs at Stanford, the primates level is gonna be a lot harder in terms of the extra hassle in documentation. And that’s why a lot of investigators don’t work with primates as well. Having said that, you’ll see that a lot of primate work has actually been done out side of the US especially in Asia, in which there is less regulations compared to us.

– Kay, thanks.

– Hello Dr. Wu.

– Hello.

45:26 I was just wondering what your current understanding of the complete epi genetic proposal, just different cells, different somatic cells, and typically also how human embryonic stem cells compare here epi genetic profile with square button stem cells.

45:50 I think my answer to this question, relates to the previous answer. I think the area of epi genetic memory still being highly debated. But in general it is safe to say that, yes there is epi genetic memory in the reprogramming process. Yes, most of the epi genetic memory will be wiped out over time, and also yes there will be still some minor residual foot print that remains over time. In our lab for example, we published a paper in General Clinical Investigation in 2001, showing that there’s also increase of heterogeneity of IPS cells, compared to ES cells on a single cell level, using a microfluidic cell platform, Single cell platform. This single cell platform allows us to analyze 96 genes at a time at a single cell level. And we saw that there’s significantly more hetergeneity in IPS cells on a single cell level which we thought probably explains why IPS cells typically proliferate slower in animals, and why IPS cells are more inconsistent in terms of differentiation of cardiac cells, and epithelial cells in our hands. So the million dollar question is then I think, which is what you’re asking is What is the biological consequence for the residual footprint of epi genetic memory? And how that will affect disease modeling, how that will affect drug screening, and how that will effect cell therapy. I think these questions will need to be further investigated in the future, obviously. However keep in mind that one can see the glass as half empty or half full. That means instead of keeping, instead of thinking of epi genetic memory as all bad maybe we can take advantage of it. For example, if you want to treat patients with a heart attack, maybe it is better to use IPS cells derived cardiomyocytes That would derive from heart biopsies, instead of skin biopsies, right. Like wise, if you wanted to use IPS cells to overwrite cardiomyocytes with drug screening of cardiac toxicity, maybe it is better to use IPS cells that were derived from heart biopsies instead of skin biopsies. So you know that’s my answer to your questions on this IPS cells. Now I think you also had a question about the ethical uneasiness in the public using Es cells and IPS cells. Well this is a tough question to answer. I think as scientist we are, should be aware of the ethical uneasiness in the public but we try to be as objective as we can. From a purly scientific standpoint, would it be scientifically better, to have access to many different biological samples or human Es cell lines. The answer is yes of course. But we also need to work within the confines of the laws, confines of politics and public perception as well. So will IPS cells ever reach the point of eliminating the need of human ES cells in the future? Which is what you are asking? You know again with the element of time, this is certainly possible in the future. Infact one would argue that IPS cells at present is probably more valuable than ES cells already, for drug screening, and for disease modeling. Simply because of the sheer number of patients whom we have access to. But I think for now, we will still need human ES cells a school standard for us to make careful comparisons of IPS cells and to human ES cells.

– Great, thank you very much.

– Hi.

– Hello.

50:20 My question is because MESP1 is a master regulator of cardiac message mm general specification, could you possibly just manipulate the amount of time and age the ESC or HIPS that you expose to this gene and see a different cardiac cells rise, and whether or not they act like a normal cardiomyocytes?

50:41 Yeah, so the question, again, I’m just repeating it for the video, is that whether MS1 as the master regulator of cardiac progenitor cells specification, could be used to improve the yield? I think this is a great idea for those of you who are not so familiar with MS1. MS1 was first identified following a screen of a transcription of factors that were enriched in the posterior region of the mouse embryo, and embryonic day seven, and initially was called mesoderm posterior 1 of MS1. There were several mouse lineage tracing experiments showing that Ms1 marks the earliest marker of cardiovascular progenitors and can give rise to derivatives of the primary and second heart fields. And there are also some studies as you’re eluding to that have shown that MS1 can promote cardiovasc, cardiac differentiation. For example there was one study that used a CMV promoter, to over express MS1 in embryonic stem cells, mouse ES cells. And then there was another study that used ES cells with stable doxycide inducer with MS1 expression. Both of these studies are reported about a five fold increase in cardiac differentiation. Well what’s interesting was that these studiers show that only transient over expression of MS1 promote cardiac differentiation where as continues over expression of MS1 could actually inhibit cardiac differentiation. I think the approach is very sound for basic research reasons, but for clinical translation we probably need to move away from genetic transcription of IPS cells and ES cells and use more growth factor base technique. And this is because it’s much more difficult to get FDA approval for cell product if they are of genetic modifications that are involved.

– Thank you.

– Hi I’m Connie.

– Hello.

– Thank you for your time and yourself for answering our questions tonight.

52:56 My question has to do with a review of the event of IPS cell generation and it seems that, so far, RNA delivery is the safest, most faithful to that judging expression of embryonic stem cells. And the most efficient method so far. Has there been any research where cardiomyocytes or other cardiovascular cells that have been derived from IPS cells using this method? And how, why, do you use this?

53:26 Yeah so this question has to do with, why not just use RNA delivery to improve the cardiac differentiation? Because RNA delivery has been used now to reprogram IPS cells. I think this is a great question and it is something that we actually thought about during our lab meeting. That is take IPS cells and modify RNAs that in code for cardiac transciption of factors and see if it can improve the cardiac differentiation deficiency. We haven’t done that yet and I’m not aware if there are any papers that have been published that have used this approach. One potential draw back I can see is that because of the short biological half life of the RNA you may need to repeatedly transfect the IPS cells with these cardiac specific RNAs. And the draw back is that the repeated transfection maybe actually quite toxic to the IPS cells. To repeat a transfection would also be a turn off for drug companies because it makes the production process of these cells very complicated and very labor intensive and very costly. I think having said that, your idea is a good one and is certainly worth a try to better understand the biology, better to understand the feasibility better.

54:54 I have one question though. I dunno how the efficiency is measured or calculated in this thesis. But it seems like, other methods had .001 percent efficiency. But this method it said that you had 4.5 or 4.4 percent efficiency. How does that, how well is it tough if the method produces a toxicity?

55:16 Yeah, so this is actually a difficult question to answer. It turns out that in some of these papers. Investigators would report, different numbers and these different numbers maybe a particular way that the investigators report how they report these numbers. And maybe a particular prep, a particular set up. But for a lot of, basically there stuff happens trying to do this RNA reprogramming. We haven’t been able to get such a high efficiency of reprograming IPS cells.

– Alright, thanks.

– Thanks Dr. Rue.

56:00 I have a couple really basic questions that are actually for the purposes of the documentary, That’ll probably be good that I don’t think you’ve had a chance to really expound upon. So first off I was just curious if you could explain a little bit more kind of exactly, perhaps even give an example of how your lab is truly modeling cardiovascular disease with IPSCs or human embryonic stem cells? Which ever, and give you an opportunity to give us a hard example of that.

56:33 Yeah, so, we have a couple projects in works for example, I see a patient in the clinic and the patient has very heart and we basically get the blood samples of the patient, do a whole exome sequencing, figure out what the mutation is. In this particular case it’s a mutation in a TNT2. It’s a mutation that’s a pontine, which is a very common, protein in the heart, and the patient himself has started cardiomyopathy. It turns out the patients mother also has the same mutation, the patients brother has the mutation, the patients son also has the same mutation. So we basically ask the whole family show up. Gave them a biopsy, make IPS cells differentiate of them, into IPSL derived cardiomyocytes. We then take these cardiomyocytes and use a whole battery of acids, for example, oast camping, multi electrode arrays to make sure the EKG powders, single cell profiling to see what genes get cut on and what genes get turned off, and then challenging these cells with pharmaceutical agents. Just to give you an example we challenge patients, IPS cells derived cardiomyocytes from thiolated patients, with norepinephrine, norepinephrine is a drug you give to patients in the clinic if they’re not doing well, because it’s a better agonist and it cause the heart to beat much faster. And we can gradually improve the blood pressue, the contractility of the patient, but we don’t want to give this drug to patients on pure time. Because it is actually quite toxic to the patient. So you know, in our in vitro read out, we saw that, for example, that patients dethiolated cardiomyopathy IPS cells, when you get exposed to norepinephrine the heart rate will go up, but after four or five hours their heart rate will drop very fast to a point it stops beating, and when he do the needles thinning or when you do control commence oscopy electron microscopy, you see a lot of disintegration of the myofibril elements within the thiolated cell. That you’ll see much less in the normal cell. So this is a example of how we use model genetic disease, and show that the thiolated cells are much more susceptible to the current medications we’re giving the patient. You can then screen these cells to look for other medications, to see which one the medication work to improve cardiac function. Within our lab we have a mission or goal, which is to create 500 to 1,000 human cardiac IPS lines, and use these lines as a vial repository to give to other investigators. And use these lines as our repository to do drug screening. So when you think about drug screening for example, most pharaceutic companies would use their drugs and test them on animal models first and then test it on cho cells or HEX cells that over express herc chanel and then to see, use that to simulate what would happen interms of cardio toxicity. We’re thinking with this IPS cells once you create the spiral repository, you can then take the same drug and screen on a thousand patients in a petri dish, and get the information back to the pharmacutical company saying hey your drug screened fine on 500 females 500 males, your drug screened fine on ethnicity group caucastions, hispanics, asians, african americans, your drug does well on patients with normal phenotypes patients with hyper 12 genotype, how ever your drug can cause some type of toxicity in validated phenotype. And this is what we’re thinking in terms of using IPS cells for disease modeling and also for drug screening.

1:00:55 That’s perfect, fantastic, thank you for that. The example that you cited was a really good one in showing how the disease modeling actually in some respects mimics what you perhaps actually see in the patient. Can you give me a crazy example, if you have one of, where you did not see that? That the induced cardiomyocyte produced IPSCs from location and actually behaved a little differently or perhaps now it does actually match what the mutation would do but we don’t know that about the patient because we’ve not really been able to study it. Do you have any examples like that?

1:01:31 So that’s a very important question, right? So we haven’t done that and mainly because we tend to pick the low hanging fruit first. We pick the low hanging fruit which is what I refer to earlier as studying monogenetic disease. And we know family has a monogenic disease that’s a certain mutation in TNNT2. We haven’t dived into polygenic disease, which could be several genes adding up together causing increased risk to cordial artery disease, or increased risk of rascate thrombosis. We haven’t done that just because, I think it’s much riskier. And so my answer is we don’t know in terms of what you’re asking. It’s an important question, you know once we, once people get more and more familiar with these poly genetic disease they will eventually move into these polygenic disease to see if these IPS cells recapitulated what’s being seen in the clinic.

– And to move into that direction, I’m almost thinking, I’m wondering if you can comment on whether or not each individual patient you’re going to have to sort of carry out a large scale microarray identifier of kind of what’s going on.

– Yeah.

– So that you can make sense ultimately of what the results are. Is that true?

1:02:50 I think at this point that’s what we’re thinking about, again just getting back to the element of time right, so, it is foreseeable a hundred years from now for example. I take you IPS cells, I take your skin and make IPS cells, and lets say a sample, god forbids, you have Alzheimers and we don’t know what Alzheimers medication we should give you because there’s 10 drugs or 15 drugs out there. It is foreseeable that we can differentiate these IPS cells into neural cells and screen on your newer cells the 10 different Alzheimer drugs. And have some kind of readout to say that, ah ha! Drug number five is much better than drug one, two, three, four, six, seven, eight, nine, 10. And give you a catered therapy. In that case there’s really what we call a personalized medicine. Like wise the same thing could happen with cardiac drugs using your personalized IPS cells derived epithelial cells, you pernalized IPS cardiomyocytes. Is it gonna happen in the next 10 years? Probably not. Is it gonna happen in the future, yeah there’s a higher probability that this is where the futures gonna move.

1:04:11 Fantastic. Does anyone have any last before we end the call? Well I’d like to give you the last chance to perhaps give use your perspectives on what the next big question is that your lab is trying to address, I think you probably touch on it already but if there’s anything else like, I haven’t had a chance to, to shout at us yet. We’d love to hear about it as we close things down here.

1:04:37 Yeah. I think there’s several questions that many physicians and scientist will want to know for the adult stem cells that are already in clinical trials. These include for example, What is the best cell type or cell types, to improve cardio function, how best to deliver these cells into patients and how do you monitor these cells without falling to temptation into patients. And how do you some how coelate cell ingraphment in patients with functional efficacy. For IPS cells which is what we were just talking about. You can use to model cardiovascular disease and for drug screening. I think it would be important for us in the next decade or so to further optimize cardiac differentiation process so that we can obtain more mature cardiac cells, that better resemble adult cardiac well physiology and phenotype. This is a question that several of the students have raised as well. For ES cells which are already in clinical trial for treatment of immaculate degeneration and perhaps many other diseases in the future. The questions that need to be addressed are optimizing the differentiating process, and including, avoiding the tumorigenicity and maybe solving the immunological barriers of allogeneic transplantation. In fact I would think that these three hurdles differentiation, tumorigenicity, and immunogenicity would also apply toward allogeneic IPS cell based therapy. Which I think as I eluded to earlier is much more feasible than autologous IPS cell therapy simply from the commercialization stand point. So these are my closing comments and I’m very proud and happy that you guys have come up with these excellent questions. Many of which the experts in our field are discussing constantly and that we have no answers for it. And I think it’s testament to how smart you guys are and also how well your professor Mike has prepared you guys for. So I would say keep working hard and I hope many of you guys will eventually decide to go into academic research and help us answer some of these questions.

– Thank you so much Dr. Rue, we really appreciate your time.

– Thank you! Best of luck and bye bye.

– We’ll be in touch thank you.

– Okay bye bye.

– Cheers.