Thursday, November 16, 2006

Session 1: Stem Cell Research Update

Hans Robert Schöler, Ph.D.
Director, Cell and Developmental Biology, Max Planck Institute for Molecular Biomedicine

 

DR. PELLEGRINO: And so to that end, we have dedicated the first session to that subject.  You have the agenda before you, and as in the past, we have not engaged in extended introductions.  So I hope you'll forgive us for that, but material is available, and obviously many people around the table know our first speaker, a distinguished investigator in the field of cellular biology and related issues on stem cells.

So I would like to ask you to come to the podium and to begin the session.  When Professor Schöler is finished, we have had agreement by a member of our Council, Dr. Floyd Bloom, who will open the discussion.  I want to thank you in advance for your willingness to do so.

Dr. Schöler.

DR. SCHÖLER:  First of all, I would like to thank you very much for this invitation.  It's a big honor for me to be here, and I'm happy to see some friends here in the audience.  I hope I can provide you with an idea of what I think has been interesting with respect to stem cell research over the last, let's say, one or two years since you had the Alternative Sources of Pluripotent Stem Cells published as a white paper.

The way I would like to start this is by raising an important point that you will see again and again.  That is, our body — soma — is something which is not lasting forever.  As you can see in this scheme, our bodies are aging, and if you think about what is maintained from us, that is our germline, that information which is passed from one generation to the next.

And with respect to regenerative medicine, the germ line has turned out to be extremely important, and a couple of publications on that issue did come out in the last few years, and I'm going to emphasize their importance.

To understand the mammalian germline, scientists are mostly using the mouse, and as you can see here, the highlighted germline of mammals, in order to see the parts of the germline, and it's obvious to all of you that the germ cell lineage giving rise to sperm and eggs is part of the germline.  Some people think that is the germline in mammals, but that's not true because you have cells which give rise to the germ cell lineage and will also give rise to the bodies.

You see here three mice.  We're talking about cloning today.  These have been cloned by the computer, by copy and paste not by nuclear transfer.

Now, these cells that give rise to the three germ layers, ectoderm, mesoderm, endoderm, and the germ cell lineage, these are the pluripotential cells that are in the focus of science and also public discussions, and both together, the pluripotential cells and the germ cells, comprise the mammalian germline.

That's different for other model species, like drosophila or C. elegans, where the germ cell image is set aside very, very early.

Here the germ cell image isn't used until a rather late stage.  In the case of a mouse, it's like one-third of development before birth.  So let's say seven days up to 20 days that it takes a mouse to be born.  That's when the germ cell image is induced.  Before that, there are no germ cells or progenitors of germ cells.

And we can look at this not only in a linear way, but in a cyclical way.  You have these cycles giving rise to new individuals after fusion of sperm of oocyte.  You basically can say the germ line lineage is the only lineage of a cyclical nature in development.  All others terminate at some stage.

So you have these two phases here.  The first phase, the phase where you have pluripotential cells; beginning even totipotential cells come to that, and here at the time that the embryo starts to gastrulate, when the three germ layers are formed, that's when the primordial germ cells are distinguishable.

And then they migrate as the embryo and the fetus develop from a posterior position in the embryo to the gonads and then eventually will have sperm and oocytes to start the cycle again.

So you have these two phases, the germ cell phase and this first phase, the phase of pluripotential cells, and it has, you know, been extremely fortunate for scientists that from this early phase, cells can be derived from different stages, pre-implantation stages.  Cells can be derived that can be cultured in the dish.

And the amazing thing is that these cells in development only show up for a very short period of time.  Once the embryo gastrulates, there are no pluripotential cells, these cells that can give rise to the three germ layers and to germ cells.

But you can take these into culture, and you can basically maintain these cells for an extremely long period of time, and ifyou think about the first embryonic stem cell lines that have been derived by Jamie Thompson from human blastocysts, these, the three lines that are mostly used called H1, H7, H9, three embryos that would fit on the tip of a needle have generated embryonic stem cells distributed all over the world, which I think if it would take them all together, you would have in the grams or even higher numbers.  Maybe you can even have in the range of kilograms by now embryonic stem cells that are derived from these three embryos.  So they have an enormous proliferation potential.

Just to at least mention that here — I will not get into that today — one of the focuses of my research is to try to get the germline, the mammalian germline cycle, into the dish, and that's for scientific reasons, but also for practical reasons that we can derive from zygotes eight cell embryos that we can use to derive embryonic stem cell lines from these, derive oocytes from the oocytes, derive metaphase II oocytes that we can use for nuclear transfer.

So if that cycle is completed and the only missing link for us is that from these oocytes we have not derived from embryonic stem cells, we have not succeeded in getting oocytes that are good enough so that we can do nuclear transfer with these oocytes in mouse, and so that is the major focus of the research of my lab in Minster currently, to fill that gap.

All of these others, nuclear transfer with mouse is something which we do routinely.  These steps and these steps here have all been done at the lab, and we'll start next year to try to do this cycle from embryonic stem cells to oocytes here with human embryonic stem cells.  So far we have been only working with mouse embryonic stem cells.

Now, if you look at embryonic stem cells, you have a very simple definition.   You have cells that make themselves again at more different stage of cells, but there are different levels of stem cells, and you can take the first cells, the mother of all stem cells, the oocyte, that after being fertilized forms a zygote, which is totipotent.  You have pluripotent cells, multipotent, and then eventually you have unipotent cells. 

So there's a restriction in potency during development, and that makes sense.  You'd rather not have a totipotent or pluripotent cell in muscles because you might risk to form a tumor.  Potency goes along with potential to form all of these different lineages. 

So at the end you rather have something which is more restricted and specialized, starting from here, this all-rounder as I call it, and in specific, you want to have a specialist at the end which is doing its job and it's not doing everything.  You want to have somebody who can do the job.  So you have these specialists at the very end.

And it makes sense if you just think about how an organism develops.  So you're starting off with the totipotent zygote, which can form an organism, but then you come to a stage where you have cells that potentially can form all different cell types, but they don't have to do that in a concerted way.  You can show today that a pluripotent cell forms ectoderm tomorrow, and mesoderm and endoderm and germ cells.

And in vivo this would be shown by moving the cells around in the embryo, transplant the cells from one position in the embryo to another one.  That's how you can show that they are still pluripotent.  They can still do all of these different things.

And, again, from these stages here, that's where you can derive embryonic stem cell lines.  You can't get them from a later stage.

And as a summary to my introduction, pluripotential cells, if somebody tells me I've found a new pluripotential cell, then I ask him can it form derivatives of the three germ layers and can it form germ cells.

Germ cells are mostly forgotten in that in proving that these cells are pluripotential and the best way to prove that they are pluripotential is to show this both in vivo and in vitro.  That's something that has been done for embryonic stem cells at least for mouse and partially for human embryonic stem cells.

If you just concentrate for a second on adult stem cells, these are extremely useful cells because these are specialists that can be used to restore some tissues, but not all, and also, they might be able to augment survival after damage, like after heart attack.  If you provide them at the right time, they might help so that the heart cells, the cardiomyocytes will survive.

Even if they are not forming cardiomyocytes, they might help other cells to migrate to that area and help them survive.  So you have to take hematopoietic stem cells.  You all know that is the best system, the best stem cell with respect to therapies.  People have since many years been using after chemotherapy or radiation.  Before the chemotherapy, they took the hematopoietic stem cells and brought them back.

But this has not been shown that you can use hematopoietic stem cells or other cells, for example, to form neurons in a way that these cells then can be used for treating, for example, Parkinson's.  So I think that is something which still has to be explored.

But the potential, if you just think about what I said at the very beginning, the potential is very limited.

On the other hand, if you would try to use embryonic stem cells for therapies, uses all around us, you had better know what the specialists can do and what the specialists are so that you can convert these embryonic stem cells to neural stem cells or hematopoietic stem cells and then bring them back.

If you would try to do this right away, then you would risk that these cells form tumors, and that is an outcome of quite a number of experiments that people do not really know what kind of intermediate, what kind of specialists.  They haven't even tried to inject the derivatives of embryonic stem cells and are surprised that tumors are formed.

In that respect, I was very pleased to see this paper published by Austin Smith last year, in September 2005.  The reason why I thought this is a key paper for me, that he succeeded with mouse embryonic stem cells to derive neural stem cells.  So basically he converted an all-rounder to a specialist.

And this is important.  We can't see it down here.  It's a stable intermediate.  He can culture these cells almost like a cell line and can take these cells and inject them into the brains of mice and then can get functional derivatives and does not risk — as far as I know, there was no tumor formed after these injection experiments, transplantation experiments.

So that is something which I think is very crucial if you would like to benefit from embryonic stem cells.  You need a thorough understanding of adult stem cells.

So I think what I would like to stress here is that both adult and embryonic stem cell research has to go side by side.  If you just concentrate on one or the other, you will not be able to unravel the full potential of either.  I think that's a statement, one of the very strong statements I want to make that you have.  If you even want to think about developing therapies, or develop their full potential, you have to study both side by side, and this is something that we can discuss later.

Now, from now on I will concentrate on pluripotential cells.  And the question is:  how can they be obtained?

And for that reason, it was important for me to show you the distinction between soma and germline because these are the two different sources for obtaining pluripotential cells or how people think pluripotential cells can be derived.

One way is deriving pluripotential cells from germline cells.  The other one is reprogramming of somatic cells.  That means non-germline cells.

This is a picture, which might remind one or the other here about Waddington schemes.  Here you have at the very beginning of this mountain, you have the zygote, which then will form an embryo which contains this inner cell mass, and then this totipotent cell is kind of rolling downhill to eventually form a germ cell.  That will be down here.

And on its way, it's forming all of these different lineages, which leave the mammalian germline.  So you see here the trophectoderm.  You see your hypoblast, and then here are the three somatic lineages.  And the primordial germ cells from here on would then normally not form any of these lineages.  That's at day seven, as I said, in mouse.  That's the time point when the germ cell lineage has been allocated.

What happens here, as you concentrate on the germline, you have an inner cell mass of pluripotential cells.  That is cells of the inner cell mass at a different shading to primordial germ cells which are unipotent.  That means primordial germ cells will give rise to germ cells, but not to somatic lineages.

The pioneer of transplantation of germline cells is Ralph Brinster.  This pioneering work started more than ten years ago where he showed that spermatogenesis following male germ cell transplantation can be done with mouse, in mouse, but also with rat spermatogonial stem cells in mouse test.  So we have complete rat spermatogenesis in mouse.

And this work has been proliferating enormously over the years, and one of his post-docs after he started back in Japan, Takashi Shinohara, he actually showed that you can use not only spermatogonial stem cells from testis, from the testis, from the adult testis and from the neonatal, but you can use primordial germ cells, those very early cells that, as I said, around day seven or later, they can be transplanted into postnatal mouse testis and could even go a little bit further back.

So it's not only here spermatogonial stem cells, primordial germ cells, but also epiblast cells, which I would position right here, he could use for transplantation in testes.

But in general you would say that's fine.  That's going the right direction from, you know, soma cells to primordial germ cells.  Still that was a big surprise.

What I want to say here is that along this germline axis, there's some freedom, experimental freedom to move these cells around from what position here straight to such a position, and you can get sperm, and the sperm can give rise to viable offspring without any apparent problems.

Now, that was germline cells and transplantation in this direction.  Are we going uphill?  And that's where it comes to germline cells and pluripotency, the focus of today's talk.

As I told you before, you can derive embryonic stem cells from the inner cell mass of blastocysts, and we know now that can be done even as early as the eight cell stage embryo, that you can derive embryonic stem cells.  I'm going to come to that later again.

Now, at the time, it was a big surprise that you can derive embryonic germ cells from primordial germ cells.  That was a big  surprise because these cells are unipotent, and by culturing these cells, Peter Donovan and co-workers, Brigid Hogan and co-workers have been able to push these cells basically uphill to convert a unipotent germ cell to a pluripotent cell which has many features in common with embryonic stem cells.

And more recently, two years ago, Takashi Shinohara, the one I have just already mentioned, has been working together with Ralph Brinster.  He succeeded in getting neonatal spermatogonial stem cells to be converted to what he calls germline stem cells.

He had to do a trick once he had these spermatogonial stem cells, but before he got them from testes, he could just culture the testis under certain conditions, and then has seen colonies of pluripotential cells in these testes which we think are derived from these neonatal spermatogonial stem cells, but that is something that still has to be explored.

And even more recently, that's the work of Takashi Shinohara, published in Cell, December 2004.

More recently, a German group, Engel in this collaboration with Hasenfuss, who is the cardiologist; he is the germ cell, the reproduction biologist.  He has obtained pluripotent cells from spermatogonial stem cells from adult mouse testes.  That was a big surprise at the time, and this has to be further explored, but here you would also see this is something where these are pluripotent cells derived from germline cells.

There are a couple of points that have to be discussed with both.  I'm going to come to that later.  There's still uncertainty with respect to stability, imprinting, and cancer.  The question, if they are really pluripotent, and I will come to the litmus test later, what a cell also has to do to be considered a pluripotent cell.

So basically, to complete that section you can derive cells which are pluripotent as far as one can tell at this stage from any given stage here up to the adult testis.  I don't think it is possible from any stage.  I would doubt at this stage that spermatocytes can give rise to pluripotential cells, but this is something that will have to be shown.

Definitely you can get pluripotential cells from all the different time points, stages that I just mentioned.

The second part is reprogramming of somatic cells.  These are now non-germline cells, and one reason to do this, besides the scientific interest, the interest that scientists have in this topic, is how to deal with the problems of rejection of transplanted cells.

And one major issue is that scientists try to derive cell lines, stem cell lines that would allow them to study disease in the dish or at least certain aspects of disease in the dish.  Patients with the known genetic disease would provide genetic information for reprogramming of somatic cells, regardless if it's done by nuclear transfer or reprogramming by fusion as I will tell you in a minute.

That's something which I think will lead to a broadening of an understanding of disease, which then eventually can lead, of course to therapy.  But this first is like the basic understanding of disease in the tissue culture dish.

And then there's of course a huge interest in generating allogenic stem cell banks, as I'll mention later, and the major question here is not only with germline cells, but also with somatic cells, can you convert these specialists, these tissue specific specialists or their derivatives to all-rounders.  Can you go uphill with respect to the potency of a cell?  Can you unravel that?

And my personal view with respect to here when it comes to somatic cells, just somatic cells, it's my personal view of what is in the pipeline, what scientists are doing and trying, is highlighted in this picture, and we start off with oocytes and tissue culture oocytes and then come to the other topics.

And as you've seen probably many times, nuclear transfer is so far the only other way to derive embryonic stem cells, to derive embryonic stem cells with the genetic information of a certain mouse in this case, not possible in humans so far.  People are trying hard to do this, but to replace the genetic information of an oocyte by that of another organism, another mouse is working out very, very well in the lab.

And if it comes to human, this search for alternative oocyte sources, people right now, there's a lot of discussion based on what groups in Newcastle have been asking for and applying for, using oocytes from other species. Then there are ways that oocytes may be derived from the ovaries of corpses and biopsies and so on and also egg donations have been discussed.

But one thing that we are concentrating on is in vitro, deriving oocytes from embryonic stem cells in the dish.  This is something that might work out one day, but we can't say that this will work out in the near future.  It's something we are trying hard, but we don't know and others are trying as well.

And if you look at this scheme where I've shown you that from pluripotential cells down to Petri season, down to germ cells, of course, that works very well in vivo and has been shown that you can push cells uphill.

So for us it was not a big surprise that we can use embryonic stem cells to let the cells basically roll downhill to obtain follicle-like structures, and out of these follicle-like structures, structures which resemble preimplantation embryos.

And of course, there's a huge interest in deriving such structures from human embryonic stem cells, and the only thing basically that they ought to do is to be able to reprogram an incoming nucleus. 

I think at the end this will be easier than fertilizing an artificial oocyte, but that's something we really have to see.  The outcome is at this stage completely unknown. 

And as I've mentioned, using embryonic stem cells to develop therapies which understand disease and identifying drugs is something that a lot of scientists are dreaming of.  There are a lot of attempts, as you know, I guess much better even than I, what is happening currently in the States and other countries, Singapore, England, that you derive, for example, neurons from patients with a certain specific disease, and then use them, for example, for small chemical compound screens to see if that disease can be changed to the better.

I have been using now the white paper and also what I have been provided with as kind of a frame to mention a couple of recent publications which fit into that frame, and here in that scheme that has been provided to all of you, there are cells that are obtained from the adult body and which have markers of pluripotential cells, like Oct4.

And in that respect, I would like to mention some interesting papers that these cells or possibly these cells or related cells have been shown to give rise even to male gametes, and here you see this is actually the same group that had published pluripotency  of spermatogonial stem cells from adult testes.  The same person, Karim Nayernia, had three major papers.  Here he was co-first.  He was first here and here.  In three major publications he could show the derivation of male germ cells from bone marrow stem cells.

So I would assume that these cells have been positive for markers of pluripotential cells or due to the culturing of these have developed features of pluripotential cells.  And these are the first to succeed in using embryonic stem cells to give rise to male gametes to fertilize an oocyte, to then generate offspring mice, which were not viable for a long time, but this is as a proof of principle, that you can obtain sperm from embryonic stem cells in the dish.

These publications are complemented by others, one by that of Paul Dyce's lab where he has shown that there is in vitro germline potential of stem cells derived from fetal porcine skin. 

Here he has obtained structures which are very, very similar to oocytes from skin and you certainly have heard about Jonathan Tilly's work where he claims that oocyte generation in adult mammalian ovaries, might occur by putative germ cells in bone marrow and peripheral blood.

It looks like from what I've heard from him at a recent meeting that these cells at this stage are not capable of forming a functional follicle, functional oocyte, but they can kind of develop in a way that the program of oogenesis is developed.

And so the question really is if you have these cells which are Oct4 positive, stage specific antigens, positive, have these been originating from the skin or are these, for example, cells like PGCs that came to certain niches in the adult, then eventually showed up there in the adult body, and then originally were germ cells, were derivatives of the germline, and that is something that has to be studied.

It's not sure at this stage if these are really adult stem cells that we're talking about, but it could be, again, germline stem cells.

So the second part, embryonic stem cell-soma fusion and then segregation, is something that is, again, taken from your overview here, is something that has been studied for quite a number of years in the mouse and then eventually Kevin Eggan's lab has reproduced what has been shown in the mouse also for human embryonic stem cells, and that is that embryonic stem cells can reprogram adult cells and don't have to be adult stem cells.  They can reprogram them after fusion because the embryonic stem cells are dominant.  They take over the program and by all means can convert the adult program to a pluripotential program.

The problem here is that we will still have the chromosome of embryonic stem cells.  So that is something that people are trying to get rid of, and I'll show you one way how people are succeeding and doing that at least to some extent.

So what we have been doing, for example, is to study that process by using cells, different cells from the mouse, fusing them with embryonic stem cells, and we are just looking at the green color being turned on, and by doing this we could actually show that this activity is found in the nuclei of embryonic stem cells.

And a method that has been published by the group of Paul Verma in cooperation with Alan Trounson is that they have been using embryonic stem cells to reprogram adult cells by not allowing the nuclei to fuse.  So you have one that is the adult cell, the other one, the embryonic stem cell.  That's 4N.  So it's twice the number of normal chromosomes, and before the nuclei fuse, they centrifuge these cells.  So the 4N nucleus would be lost during the centrifugation process.

And apparently that appears to be enough to reprogram these adult chromosomes in a way that they acquire features of pluripotential cells.  It's an extremely, from what I can tell from the publication, an extremely inefficient way and has to be optimized to see if, indeed, these are pluripotential cells that are of therapeutic value.

But that would be a way how the nucleus here of the embryonic stem cell can kind of force the adult cell to be reprogrammed.  And here the recent publication which just came out just a week ago or two.  That is that people are trying to get rid of the chromosomes of the embryonic stem cells, and Azim Surani and Takashi Tada have developed a chromosome elimination cassette that would eliminate certain chromosomes of the embryonic stem cells.

So you could use this, for example, to eliminate those chromosomes which would result in host rejection.  Still you would have all of these other chromosomes.  So at that stage, that's an interesting proof of principle study, but it has to be shown if this can actually lead to pluripotential cells that are of therapeutic value.

But I just want to mention these publications, that there are major attempts to have the embryonic stem cell reprogram adult cells and then try to get rid of the chromosomes afterwards.  Of course, that would be something wonderful if this approach would work.

Now, the last group of procedures are here, the cellular vesicles or artificial vesicles or, at the end — I'm not going to talk about this — in situ reprogramming where people will aim to try to bring certain factors to certain organs to reprogram cells to become stem cells in a certain organ.  But this still too (speculative) at this stage.

So pluripotential cells via somatic cell differentiation, you have mentioned this paper, which in my eyes is a key paper, but before I come to this, basically the ideas — a lot of you know this picture better from Kronau— is that you use not human as here, but cells, by buffering cells in a certain cocktail of factors to turn back the program so it would become an umbrella program just by the factors.

And the first paper on that topic has been published by Phillippe Collas, and what he was doing is to use extracts of carcinoma and embryonic stem cells and use this to put adult cells in this cocktail made pores in the adult cells so that the factors of these cells could enter the cells, and he succeeded induction of de-differentiation genome-wide transcription of programming and epigenetic reprogramming by these extracts.

These cells look very promising, but there are still so many tests to be done to see if these indeed will fulfill these hopes that one would have if you look at this publication, which I think is worth reading.

The only problem with this publication, I think, is that they didn't have the rigid biological tests.  Otherwise it would have been published  in Cell and not in Molecular Biology of the Cell.

This is the paper that you have here  and the paper that you have distributed, and I, indeed, consider this one one of the key papers of the last years: "Induction of Pluripotent Stem Cells for Mouse Embryonic and Adult Fibroblast Cautions by Defined Factors."

So in contrast to what I've mentioned before, nuclear transfer or fusion, in this case he has been using defined factors which have been provided to these cells by viruses and has succeeded by using four different factors, c-Myc, Klf4, Oct4 and hidden here is Sox2, to convert a differentiated cell to an undifferentiated.

So of course, he would see that there are many problems with providing viruses and so on, but just the idea by having to have a defined set of factors and converting one stage to another stage is, I think, a major step in understanding, and now people will say, "Okay.  I don't want to have c-Myc.  I think this factor is better than c-Myc or Klf4, which have oncogenic potential.  So you might not want to have this if you want to think about therapies.

And you don't want to have viruses in them, and you don't want them to have them consistently be expressed.  You basically want to bring them as proteins to a cell and then convert it, just like Collas did it with the extract, with defined factors converting one cell type to another.

And then it has to be stable and pluripotent.  There are a little bit of techs who can, because I think this key paper has a couple of features which have to be really understood because based on this paper, there might be too many hopes at this stage, and there are so many things that you still have to understand before something like this can lead to something that can be used with respect to therapies or at least to obtain pluripotential cells.

So I think that the results suggest that Takahashi and Yamanaka, the two authors of that paper have successfully reprogrammed terminally differentiated cells to a state that has features in common with those of pluripotent cells.  I would not call them pluripotent.  I would say that they have features in common with those of pluripotent cells.

However, several observations indicate that as they call them, induced pluripotent stem cells are similar but not identical to embryonic stem cells, and there are three major differences that I want to go through.

One is the absence of any contribution of these cells, these induced pluripotent cells to postnatal animals following blastocyst injection suggests that the cells have a limited capacity to stably integrate into normal tissue in vivo.  That is something that has to be studied more thoroughly and at this stage is a problem.

Although rare induced pluripotential cell clones showed expression patterns of known embryonic specific genes that were very similar to the controls, embryonic stem cells as controls, a substantial degree of clone-to-clone variation was observed, and some clones failed to reactivate a number of the genes assayed and notably none were found to express embryonic stem cell-associated Transcript 1, Ecat1, which apparently is an important player.

Transcription profiling experiments revealed that although these cells cluster more closely to embryonic stem cells than they did to their parental fibroblasts, they still present a distinct gene expression signature.

And the third point is that DNA methylation of the Oct4 promoter as one marker and the post-translational modification of histones positioned there suggested that these cells are caught in an epigenetic state that is intermediate between their somatic origins and fully reprogrammed embryonic stem cells.

So there are things missing, and I think the next months and years will have to be used to find what is missing, but I think these guys are on the right way.  They are on the right way to become, as I trust, to become pluripotential cells.

And so in summary — that's the last slide with a lot of text — in summary, the nuclear reprogramming observed by introduction of these four transcription factors into somatic cells is substantial, but it differs from the more complete reprogramming that is observed after transfer of nuclei from somatic cells into oocytes or after fusion of somatic cells with embryonic stem cells.

By all means, this here, these two ways are resulting in a complete reprogramming.

Several important questions remain. Are these cells trapped in an intermediate state between somatic cells and embryonic stem cells or are they actually some other pluripotent cell type, for example, those that correspond to cells of the epiblast?

And one possibility is that they are, instead of being embryonic stem cells, they could have more features that come with embryonic carcinoma cells.  These are still questions that have to be solved before we can even think about using such cells in organisms.

Now, basically what this type of research is trying to do is to convert the unipotent somatic cell to a pluripotent IPS, induced pluripotent cell, and this at the end might not lead to therapies, but I think it is right now one of the most exciting fields in biology, to try to use this system as a way to understand how a cell is converted from one stage to the other, and you have to do this by using defined factor to understand the molecular biology behind that.

And I guess many groups are going to concentrate on this work based on what Shinohara has published in that key paper.

Now, we have here this scheme, but there's a big "but" here, and the reason for this big "but" is that for some reason a lot of people think that their face is getting older and older, but the DNA is staying young.  This is a major problem.  We are aging, and with us our DNA is aging, and if you think about cloning of an aged person, just by knowing a little bit of biology, it is ridiculous.

But even therapies might be very problematic if you would like to use the genetic material of an aged person, and here is one scheme that I took from a review article, and that is the increase as you see here of mutations in the human population based on what people have outlined in that paper.

So you see that from the very beginning of our life we are accumulating as a human population, accumulating mutations, and statistically seen that is resulting in an increase of tumors in the population, and statistically seen a young person has less risk of getting a tumor than an aged person.  We all know that.

But what we sometimes forget is that there is a time point where there is almost like an exponential increase, and this is called in the literature — it's not my terminology — the end of warranty.

(Laughter.)

DR. SCHÖLER:  Well, I'm beyond this end of warranty because that's 45 years.

So if you think about this, what that means, if you would like to use that genetic material for reprogramming studies, I would say you either have an extremely good screening procedure or you're risking that you're causing problems by therapies and that there are genetic problems.

I just show you one example that we can actually show by cloning.  These are two clones from the same mother, two mouse clones.  It's pretty obvious, and that's why mouse geneticists love this kind of phenotype.  It has a short tail.

And this is interesting because you don't have to open the mouse to see that there's a genetic problem.  There is a genetic problem.  They both originate in the same genetic material, and the offspring, as you can see here, some of them actually have a short tail.  They have a normal length.  They have a short tail.  So this is genetic because it's passed from one generation to the next.

At birth the mice are naked, don't havefur, and they develop this like after two weeks or so, that they get fur.  So that's why they look like small pigs instead of mice in that picture.

So that is genetic, and if you just look at the chromosomes, either this way or by chromosome painting, you won't see that they have a genetic aberration.  It's not obvious from this.  So if you would like to use genetic material of an aged person, you would, I think, run into many more problems than this one I've been showing you, and you still wouldn't be able to pinpoint before you do this that there is a problem or there's not a problem.

And the same note of caution I would raise if it comes to such procedures, which is using — and this is also nicely described in the white paper — I think there might be a reason why these embryos arrest; that if you're not sure that the arrested embryos that are obtained here, like Miodrag Stojkovic has succeeded in deriving embryonic stem cells; if the embryonic stem cells that you have are as perfect as the ones that you derive from a nonarrested embryo, you might be risking that at the end this attempt is a failure.

It's important to follow this, I think, but there are a couple of question marks that you have to be aware of.

And that brings me to my vision, how I think what should be used as genetic material, and that is umbilical cord blood, and not because these cells are pluripotential.  Umbilical cord blood cells are limited in their potential.  They are not like embryonic stem cells.  They might have a bigger potential than originally thought based on the publications that have been out there since during the last few years, two or three years, but I would find them extremely interesting because the DNA is very young, and you would not risk to the same extent that you introduce problems by genetic mutations if you take one of these procedures to reprogram these cells so that they will be pluripotent.

And so umbilical cord blood or another way to use nuclei of HLA-compatible donors, to use any of these procedures to convert these cells into banks of pluripotent and/or multipotent stem cells.  I think that is something that at least in Germany I'm trying to get that established in a network with other researchers working on umbilical cord blood.  Peter Wernet in Düsseldorf is the one person who has these banks and should be with whom we collaborate, and I'd like to see if we can get from these, let's say, at best multipotential or alipotential cells to pluripotential cells, but we'll see if that works out.

Now, the point here that I would like to make is if you ask if you can go back from these unipotential cells to pluripotential cells, I stressed enough that this, I think, is one of the most exciting topics in biology, and the therapeutic potential needs to be explored.

However, we currently only have as a source for useful pluripotential cells and embryonic stem cells those cells which are derived from embryos, and these cells are the gold standard.  And any other cell that you obtain by reprogramming, you have to be able to compare it with these embryonic stem cells.  If they are as good — I doubt they would be better, but they have to be as good, and we don't know if such cells once available can actually replace embryonic stem cells.

There might be genetic/epigenetic problems, cause tumors, and you can see this down here.  So we'll skip right to the next slide.

The crucial litmus test at least in mouse is that these cells have to be able to give rise to a mouse in this tetraploid aggregation experiment.  I took this scheme from Janet Rossant's technical report here.  So basically what has to be done to show that these cells are pluripotent is that you use a clump of embryonic stem cells that you have obtained or embryonic stem cells or pluripotential cells obtained after reprogramming and combine them with tetraploid host embryos, and the host embryo would then form trophoblast and establish the yolk sac, and the rest here, this diploid part, would then give rise to the embryo proper, the mesoderm of the yoke sac, but the embryo proper then has to be born.

If that's not working, these cells are not as good as embryonic stem cells.  That's a standard procedure with embryonic stem cells, and even the report which I think is extremely well done, the one from Takashi Shinohara where he showed generation of pluripotential cells from neonatal mouse testes, he hasn't done this experiment.  Many people who are doing the studies, they either do not report them or don't even do them, these complementation studies.

What he has done here, after deriving these pluripotential or these induced pluripotential cells, a total of 92 tetraploid embryos were created by electrofusion.  So they went ahead with that procedure, and aggregated with these AS cell-like cells and transferred to pseudopregnant ICR females.

When some of the recipient animals were sacrificed at day ten and a half, we found one normal looking fetus and several resorptions with normal placentas.  The normal placenta, of course, is coming from the tetraploid part.  It has nothing to do necessarily with this part.

The fetus showed some growth retardation, but clearly expressed this gene, and none of these were born.  So if you even have this problem with cells of the germline, I am not surprised that people who have been trying to do these experiments with reprogrammed cells are not reporting their failures.

This is something which has to be really worked out, and this, as I mentioned here, has to be the test.  If you have pluripotential cells and claim you have them, you don't only show that the three germ layers and germ cells are formed, but you have to go through this test in mouse and then you know that the procedure is I would say very good or even perfect.

And that brings me to the only way I think one can go ahead at this stage, and that is by pluripotent stem cells derived from biological artifacts, and I would like to provide you with some data from our lab, which I think is making a good case that the proposal, the ANT proposal, is a procedure that at this stage in my eyes is the best way of going ahead if it comes to trying to provide an embryo stage.

And I'm going to show you this data, and it's something we can discuss.  It's what Guangming Wu in the lab has done with the help of a couple of other people in the lab, is to use Cdx2.  That's the gene that has been widely discussed in this group, to knock Cdx2 down, not out.  This is the knock-down approach, and he has done it by siRNA, not like Rudolf Jaenisch has published it, by a viral infection of the nuclei that are transplanted into the oocyte, but in this case, we have been using fertilized oocytes.  See here?  That would be the female pronucleus and that would be the male pronucleus, and injected siRNA against Cdx2.  That's like small, 23 base pair RNAs are scrambled.  The same nucleotides were used, but scrambled.

And then you look at what happens when the zygote is formed and the embryo is developed, and this is a very efficient way of knocking down a gene.  You can see here in this scheme this is quantitative realtime PCR.  That means you can really look at levels.

Cdx2 in normal development with scrambled RNA would increase more and more.  As you see here, this is the eight cell embryo.  The early morula, the morula, the early blastocyst and blastocyst.

You see here that the Cdx2 knock-down experiment reduced levels more than 95 percent.  There's just a little bit left here after that knock-down experiment just by injecting this RNA once at that early stage.

And what you can see here if you look at the development of stages, you see here pictures of early blastocysts and late blastocysts, and these are the ones that have been control treated with the scrambled RNA.

Now, you look here at the Cdx2 treated and you see these stages look very similar.  The eight cell, the early blastocyst and this, the late blastocyst, as we can see here — I hope you can see it from the back — all of these embryos here failed.  They are all intact in the zona pellucida.  They have not hatched in comparison to the late blastocyst that you have here in the controlled treated one, and that's something that none of these in any of the embryos that we have obtained did hatch.

And I've said embryos, but I rather would not even call these embryos.  These stages which correspond to late blastocyst I should say.

Now, if you look here for the protein, this is now by immunocytochemistry.  So you can actually look at the Cdx2 protein here.  You see there is, of course, protein in the control treated one, and you see that there's no protein here in the knock-down experiment.

Now, we look taking a marker of pluripotency.  That is Oct4, and you will see here in the control group Oct4 is where it's lying.  It's in the inner  cell mass, in that area which will give rise to the embryo proper.

In this case, it's all over the place, and if you have an overlay, you'll see here Oct4s all over the place, and there is an Oct4 restriction here in the control treated embryos.

It's important to stress at this point that these look very similar to blastocysts.  If you would look at these here, you would say these are blastocysts, but they aren't.  They look like blastocysts because the oocytes already have RNA and protein which would pump in fluid into these structures.

But this is just a pumping activity which are depending on proteins and RNA laid down in the oocyte.  That you would get regardless if this is an embryo or not.

And as I mentioned none of these embryos — you've been using large numbers — none of these embryos actually hatched out of the zona pellucida.

Now, when we tried to understand, if these embryos at an earlier stage are any different from the control embryos, we looked at the whole genome by RNA profiling.  So we used eight cell mouse embryos that were obtained from eight-cell stage embryos by the control and here compared them with what happens if Cdx2 is knocked down.

And this was done with Kuniya at the RIKEN Institute in Japan.

And if you look here just at this scheme, this is just a comparison.  You will see that even at the eight cell stage, there are differences between the two types of eight cell stages.  You see here even at that early stage, you have like 300 which are higher and 300 which are lower than normal, supporting the idea that the development programs of the two are different, and this is based on the fact — and this has been published by others in the meantime, Dr. Roberts — that there is an early expression of Cdx2.

Here we show this again by quantitative view on PCR, and I should mention, stress that this is a logarithmic scale.  So these are always jumps of ten.  So that actually means that there are very low levels in the metaphase II oocyte.  That's what is used for nuclear transfer, then the two-cell, even lower in the four-cell.  It is really so low that it's a base level and you need a couple of embryos to really be sure about the numbers here.

But that's the nice thing about the siRNA, that you can use large numbers.  You know that you have a group of embryos which behave the same way.

So this goes down and then you have an increase, and you see that there is expression of Cdx2 RNA, and we have been looking also for the protein because that's what's actually important if you want to express genes to see your Cdx2 protein at the zygote stage, and it's very, very difficult to really prove that this is not unspecific.  It's much easier than if you can do the knock-down, if you can look at the result of the knock-down experiment.

Here we see the eight cell stage, and now we have to help you.  I can convince you that if you treat these zygotes with siRNA and compare this to the control which shows weak expression of this protein in the nucleus, you see that there is no expression in the nucleus in the case of the Cdx2 knock-down.

So RNA protein and the profiling data are all in agreement with the fact that at this stage the embryos, the control embryos are different from these knock-down stages.  And since this is a transcription factor, you don't need a lot of transcription factor to turn on these 300 genes and turn off other genes, other 300 genes.

And we wanted to know what's happening here at later stages to see quite nicely when it's strongly expressed, and that's what people have been mainly looking at, Janet Rossant and others.

You see quite nicely that the expression in the nucleus is much stronger, and you see here that there is no expression or there's basically no signal detectable in the knock-down.  That's now the morula stage.

And this is the first time that we see something like asymmetry. This is for Bill Hurlbut.  We had a discussion on that yesterday.  That's the first time that we see something, and it's actually not always like they're fore it in one correct.

At the four cell stage, we don't have any evidence that one nucleus has more protein than the others, what we see at a later stage.

And to get an idea of why the embryos fail, why do the embryos degenerate at a later stage?We again did a profiling experiment with Kuniya, and now at the early blastocyst stage, and there you can see that there are tremendous differences.  You see that about here more than 2,000 probes, more than 2,000 proves are below this level to indicate that there is differential gene expression.  This because there is no trophoblast being formed.  These embryos don't have a trophoblast.  These mainly are trophoblast genes or genes which are expressed  in the trophoblast.

And since there are a couple more pluripotential cells or cells which have features in common with pluripotential cells.  You have a couple more genes about this level here, but this is indicating that there is lineages missing, that these cells, that there's structures here that you can see here by using two different pluri Tarticipation.  This group of cells is now over all the place with Oct4 expression that is present for all cells, and you have the same for a non-knock.

And the reason why they are failing, we think, one reason for that is that the cells don't have tight junctions as they should have. The cells are not linked together as they should, and that's indicated by ZO-1 you see is missing to quite some extent in comparison to the control, and another one, E-Cadherin, which he had nicely distributed in the embryo — see the green color, quite nicely distributed here.  You see that it is a problem with respect to E-Cadherin, and this with Cdx2 knock-down, the phenotype is even stronger than with the knockout that was been published by Janet Rossant.

And now this is really, I think — when we started doing electron microscopy, this was for me an eye opener of what's happening.  We wanted to look at the tight junctions, and you can nicely see here that the way the cells in  trophoblasts are linked together, see here?  These are really tightly knit together here.  Here you can see them and here.

Now, look at the knock-down.  You see that basically they are kind of sticking together, but they're not really tightly linked as you have them here and here.  Here you actually see that this is opening, and that's why it's no surprise that such embryos would pump, but they would collapse because they don't have these tight junctions.

But look at something which is even more exciting, which I did not expect.  Look at the mitochondria.  These are the energy departments in the cells.  You see here these are mitochondria, as they should look like in trophoblasts.  They are long, longitudinal, and have a lot of what is called crista, these structures inside which are providing energy, which are generating ATP.

And look at those here in the knockdown.  These are round mitochondria, which have an embryo appearance here, and you can see them here.  These are not energy producers.  They have more of a resemblance to those of pluripotential cells.

This is quite nice.  I just found this publication, "Energy Metabolism of the Inner Cell Mass and Trophectoderm of the Mouse Blastocyst."  The trophectoderm consumes significantly more oxygen producing more ATP and contained a greater number of mitochondria than the inner cell mass.  These data suggest that trophectoderm produces about 80 percent of the ATP generated, and responsible for 90 percent of the amino acid, not as a turnover compared with inner cell mass. In conclusion, the pluripotent cells of the inner cell mass displays a relatively quiescent metabolism in comparison to the trophectoderm.

So sine you don't have any power houses in these embryos, as you can see here, the control, this is an assay.  It's called JC1 assay, which is kind of showing where the active mitochondria are.  So these red dots there indicate there are active mitochondria.

In this case, the knock-down, even if you have a longer exposure, you at best see a very, very weak signal.  So what I think happens here is that these are pluripotential cells or cells which have a lot of features in common with pluripotential cells, but they  need energy to further develop, and the trophoblast is providing this since this is basically one lineage instead of two.  This is not, to my understanding, an embryo, but is something which is just a number of pluripotential cells.

And now the way that we're trying to show that these are one lineage, just one lineage of pluripotential cells that comes out of this Cdx2 approach is here by visualizing pluripotency, and that is by using the green color, the green fluorescence protein, which has been integrated into the gene of Oct4.

And if you now look at these three stages here, it's an early blastocyst and late blastocyst, and you want to derive embryonic stem cells, you see that these early blastocysts from the control treated ones, in mouse you get about 90 percent embryonic stem cell lines.

In this case, since you know that these are degenerating structures, you can get one out of — you get one line out of 50.  That means two percent which is a tremendous drop, which means that at this stage they degenerate.

Now, if you then ask what you get out of the eight cell stage, here you see that the green color is distributed like a lost egg.  There's some green cells here, but there are a lot of other green cells.  Of course they are because they are two different lineages, one which will give rise to the trophoblast and one which will give rise to the inner cell mass.

And if you take these embryos in culture, that's that you get, derivatives of the outer cells and the inner cells.

Now, look here if you take these ones, which is where I've been claiming that this is just one lineage.  Here you see that this is one glowing green ball of cells.

And if you look at numbers now, you have 22 percent embryonic stem cell lines, which is the same range as you have here with the eight cell embryo, and here you're going up to 34 percent.  So it's not only much better than the two percent, but it's even better than the control treated.  That means if you use Cdx2 in that type of experiment, you get more cells, I think, that have features of pluripotent cells, and my interpretation is that's why you have a higher efficiency of deriving embryonic stem cell lines, and that these are by all means as good as normal cell lines.

Now, the last two slides.  Here, first of all, is section through here.  You see that.  Just look at it.  These are different cells.  If you look here, all of these nuclei, they look over similar.  So this is a more uniform type of cell that you have here, which I think if you do it this way, derive cells at the eight cell stage embryo, you have basically a group of pluripotential cells.

And here, this is the embryonic stem cell line that has been derived from one of these that can give rise to germ cells, that you can form chimeras, and they have even long lasting effects on these chimeras.  And as you can see here, these are stem cell niches where if that would be in a transient, that would not exist.

So in the end, I would just like to highlight again that this here coming from here to here by the Cdx2 knock-down is a very, very efficient process.  So we have now going forward been using them to derive oocytes.  We're trying to get them useful for nuclear transfer so that we can do all of this in the tissue culture dish, but at this stage, I think if you would like to derive embryonic stem cell lines without generating embryos, I think we have to go through a procedure where a gene like Cdx2 is affected.

I think I'll leave this for the discussion.  This is the procedure that Rob Lanza has published.  I had a lot of problems with that procedure because he has been, as a proof of principle, has been destroying so many embryos to show that the procedure is working and selling this as something of high ethical standards that I had a major problem with that.  But that's something we also can discuss.

At the end by reprogramming and by looking at embryonic stem cells, we're always thinking about therapies, but this work and the work that was from Stewart Arc and Rudolf Jaenisch, Doug Melton and quite a number of people will at the end show us what pluripotency is, and that is very important, that we don't forget the basic science behind all of these approaches, that we understand actually what a pluripotential cell is.

And I think that many excellent groups are now working on that topic, and I think once we understood that, we also have a better way of developing therapies.  My credo is that good basic science is an important step towards applied science, and along these lines something has been published by Peter Donovan that neutrophins mediate human embryonic stem cell survival.  By understanding this here, he, for example, was able to show why or giving one reason why  trisomies happen when embryonic stem cells are cultured, because something like this is missing, and we hope that this something that we have just published with collaboration with Sheng Ding and Peter Schultz, that we can obtain substances that can maintain cells in the pluripotent state by repressing differentiation.

All of these approaches I think are required if we at the end would have a pluripotential cell in hand, and maybe a substance like this which is freezing in the pluripotent state might also help us to derive embryonic stem cell lines from other species.

So this is my international group of people.  You can see here all of the different countries, a lot of European countries, but also we have no problem of Iranians in my lab working next to Americans and Chinese and South Koreans, Indians, Greek and so on.

Sine this is 13 and 13 is not a lucky number, we have the Kingdom of Bavaria as number 14, and finally we moved into a new institute.  Whoever come close to Minster, please come visit me.  It would be a pleasure for me to host any of you at the new institute.  We just moved in there three weeks ago. 

That is the Max Planck Institute for Molecular Biomedicine, which brings me to this slide, Rembrandt, where I think some people have the feeling they know everything.  That's like this person, but I'm one of these guys.  I'm still looking, and I'm completely confused with what's going on.  I try to get a better understanding.

Thanks for your attention.

(Applause.)

DR. PELLEGRINO:  Thank you very much for a very complete overview.

I think we'll have a break of about 15 minutes before we ask Dr. Bloom to open the discussion, if that's okay with you, Dr. Bloom.  So let's take a break and be back in 15 minutes and a little shorter if you can make it that way, please.

(Whereupon, the foregoing matter went off the record at 10:27 a.m. and went back on the record at 10:43 a.m.)