SCIENCE AT THE SHINE DOME canberra 4 - 6 may 2005

Symposium: Recent advances in stem cell science and therapies

Friday, 6 May 2005

Professor Gesine Kögler
Professor for Immunology in Medicine, University of Düsseldorf, Germany

Gesine Kögler is Professor for Immunology in Medicine at the University of Düsseldorf, where she also received her PhD. After completing a postdoctoral fellowship in Immunology and Stem Cell Transplantation Immunology, she was the Director of the Histocompatibility Testing Laboratory. She is currently the Director of the Stem Cell Processing Laboratory at the University of Düsseldorf. Her current main research topics are the characterisation of non-hematopoietic unrestricted somatic stem cells from cord blood and the ex-vivo expansion of hematopoietic progenitor cells from cord blood. In addition, she is responsible for the stem cell processing of autologous bone marrow mononuclear cells for the repair of acute myocardial infarction and chronic ischemia. She is the Founding member of the European organisations for public cord blood banking, EUROCORD and NETCORD, and a member of the FACT/NETCORD Committee for cord blood standards and numerous other international societies.

ES-equivalent adult stem cells from cord blood

First of all I would like to thank the Australian Academy of Science for inviting me, and giving me the chance to present some of our data.

(Click on image for a larger version)

Eliot Marshall, in his paper in Science 282: 1014, 1998, said: ‘Imagine being able to reach into the freezer, take out a cell culture, treat it with growth factors, and produce almost any tissue in the human body.’ Several scientists thought seven years ago that it would be very easy, with embryonic stem cells, to regenerate liver, kidney, heart, skin, neurons and other tissues of the body.

(Click on image for a larger version)

And we all thought that maybe embryonic stem cells are great, since they are able to produce endodermal tissue, mesodermal tissue, they are able to produce neural cells.

(Click on image for a larger version)

However, even recent work showed that highly differentiated neural nerve cells from mouse embryonic stem cells, transplanted into a syngeneic organism – into another mouse, into the brain of the mouse – showed teratoma formation. And, when we looked at this tissue (this was the work of one of my colleagues, as of 2003, in collaboration with the Max Planck Institute in Cologne) we did not observe neurons in the brain but we observed cartilage in the brain. This means for us as scientists that we have to solve the problem of tumour formation first, because otherwise we don’t see any way that these cells can be used in a clinical application.

And there was the question: are there alternatives?

(Click on image for a larger version)

It has been known for more than 20 years now that cord blood is a very valuable source of hemopoietic stem cells, and that the hemopoietic stem cell compartment contains many more long-term culture initiating cells, or mouse repopulating cells, compared with adult bone marrow or adult peripheral blood.

(Click on image for a larger version)

It was the pioneering work of Professor Gluckman, in Paris, to transplant the first patient with Franconi’s anaemia in 1988.

(Click on image for a larger version)

This was the reason we were motivated, in Düsseldorf, to establish both a related and an unrelated cord blood bank. We have stored, up to today, more than 8000 unrelated and related cord blood units, and have transplanted 262 patients.

(Click on image for a larger version)

I would like to point out that cord blood is a very safe source of hemopoietic stem cells. It can be obtained after the delivery of the baby and it can be processed up, under sterile conditions – under industrial conditions. In Germany and in many other countries, cord blood is a pharmaceutical drug, and there are several international applications making the use of cord blood very safe.

(Click on image for a larger version)

This has motivated the whole world, not only to establish cord blood banks, including in Australia, but also to go for cord blood transplantations for unrelated and related ones. And up to today more than 5000 cord blood transplantations have been performed in patients with leukaemia, haematological and metabolic disorders and genetic diseases.

However, the hallmarks of cord blood are that the stem cell compartment is very immature, compared with adult cells; the cells have a higher proliferative potential, associated with an extended life span – and there are many publications out; they are always available, there is no risk for the donor; there are no ethical issues; infectious agents such as cytomegalovirus or Epstein-Barr virus are rare exceptions; and immunological immaturity is of great importance. That means we can transplant over histocompatibility or HLA barriers.

And there was a question: are there other stem cells in cord blood?

(Click on image for a larger version)

There were some observations by Ende et al. that so-called ‘Berashi cells’ are responsible for clinical effects observed in SOD mice and mice with a neuronal teratoma [inaudible] defect, as so-called amyotrophic lateral sclerosis. But he was never able to detect these Berashi cells.

There was another observation, by Chen et al. in 2001, that intravenously administered cord blood has been shown to reduce behavioural defects after a stroke in rats. But it was never clear what kind of cell it was from cord blood.

(Click on image for a larger version)

Last year we described the new human somatic stem cell from placental cord blood referred to as unrestricted somatic stem cell, or USSC.

(Click on image for a larger version)

This cell can be generated from 30 per cent of all cord blood samples. This cell has a very low frequency: we can obtain in medium only four with a frequency of one to eleven per 100 to 200 in our cord blood. However, they are easily expandable in culture. It is very easy to generate one to 10⁸ up to 10¹⁵ cells[JEM 2004]. That means it is a very rare cell but it can be amplified.

(Click on image for a larger version)

The telomere lengths of the cells correlates with the age of the cells. A long telomere means a young cell. From this slide we can see that the telomere length of USSC is much longer, compared to adult mesenchyma cells harvested from bone marrow donors.

(Click on image for a larger version)

As of today we are able to generate these USSCs from 30 per cent of fresh cord blood specimens. We do not see a correlation to gestational age; also I would like to point out here that due to ethical considerations we are only allowed to take cord blood starting in week 36. I do not see a correlation to the volume, to the number of nucleated cells or to the hours after cord blood collection. Nevertheless, I have tried to work with very young and fresh cord blood.

(Click on image for a larger version)

The immunophenotype of the USSCs is quite similar in all the cell lines. We never see the expression of CD34; the cell does not express CD45, which is the marker for hemopoietic antigens.

(Click on image for a larger version)

However, the cells seem to express some markers of endothelial formation as FLK-1 and von Willebrand factor.

(Click on image for a larger version)

And even in immunochemistry you can never observe the marker CD45, which is a hemopoietic antigen. That means it is really a non-hemopoietic cell in cord blood.

(Click on image for a larger version)

It is very important to point out here that USSCs expanded in culture up to 19 or 20 passages; they have a very normal karotype 46XX or 46XY. That means they are very stable in culture. However, one of the main features of the cell is the differentiation potential.

(Click on image for a larger version)

All the USSCs from cord blood tested so far were able to be triggered into the mesenchymal cell differentiation. Under certain differentiation conditions they were able to perform osteoblasts and to develop into bone, and under different conditions they were able to develop into chondrocytes or into adipocytes. In a collaboration with Arnold Caplan, in Cleveland, we used an atumoric rat model in order to show whether the USSCs are able to repair a critical size femur defect. In the first of these X-rays on the slide you can see it is a porous ceramic cylinder without any cells, and after several weeks no healing is observed. If the USSCs are allowed into the cylinder you can see, in the lower X-ray, a healing between four and eight weeks after implantation.

If we introduced the USSCs into gelatine sponges and introduced them under the skin of NOD/SCID mice, a chondrogenesis in the mouse was observed and here you see the staining toluidine blue for an extracellular matrix which the cell produced.

(Click on image for a larger version)

In addition, in order to look for the safety of the cell in NOD/SCID mice we used several approaches, and we also injected cells intrafemurally into these NOD/SCID mice.

(Click on image for a larger version)

And it was interesting for us to see that we injured the femur at this time, so when USSCs were applied intrafemurally the USSCs were sitting there and were repairing the bone. They did not go out to the spleen or to any other organs.

In contrast, when hemopoietic cells were infused into the mouse, we observed colony formation in the injected femur but also in the spleen. That means we can really see that in this model the cells are sitting there and repairing the injury.

(Click on image for a larger version)

However, for me the quality control is ultimately not only that we were able to differentiate the cell in the mesenchymal cell direction but that we were able to differentiate them under certain conditions into neurons and astroglial cells. In this slide we see an expression of the so-called neurofilament, and some of the neurofilament positive cells do express sodium channels. In addition we can observe the expression of synaptophysin and the inhibitory transmitter GABA. It is more interesting that some of the cells, upon differentiation, are positive for tyrosine hydroxylase, which is a precursor enzyme for the dopimergic pathway. And I can say today that even in the hBLc we observed the production of dopamine.

In addition, some cells are able to be expressed for choline acetyl transferase, and upon transplantation in a rat brain, several months after transplantation we observed human cells in the brain.

(Click on image for a larger version)

In order to look for repair in other organs we used a model system which is known from Professor Zanjani. Here we are able to introduce the cells at 60 days. At 60 days the sheep cannot reject the human cells; that means it is a perfect model for human cells. It is not an injury model, it is just a model to follow the fate of the cells in a normal organism. Several months after transplantation, both the heart and the liver of these animals were harvested.

(Click on image for a larger version)

It is very clear here that human cells integrated in the right atria and they are integrated into the left ventricle. Here we have double staining both for a human antibody and for cardiac myocyte antibodies, the human dystrophin, and you can see here that the cells are overlapping. That means the cells really integrated in the heart and produced cardiomyocytes. In addition to cardiomyocytes we saw a staining for the so-called Purkinje fibres. These are cells which are important for the electric coupling in the heart.

(Click on image for a larger version)

Furthermore, we showed that cells were also carrying the so-called ryanodine receptor, which is a functional receptor in the heart. That means that in this model system, in the preimmune sheep, the cells were able to regenerate cardiomyocytes and Purkinje fibres.

(Click on image for a larger version)

On analysing the liver of the animals, it was obvious that 80 per cent of the liver cells stained positive for a human hepatocyte antibody. When we look at the section of the whole liver shown at the left of this slide we see that of the whole liver, 20 per cent were of human origin. And they did not only stain positive for human hepatocyte antibody but they also produced albumin.

In addition we formed Weston blot experiments to look at the serum of these animals, 17 months after transplantation. And the cells here also produced human albumin. And then we asked whether the human hepatocyte cells were fused with sheep cells or stayed alone and just produced human factors.

(Click on image for a larger version)

Therefore we used single cell PCR on microdissected cells. And here it was clear that human cells were really stained positive only for human markers and not for sheep markers. That means that in this non-injury model we never observed fusion to the sheep cells.

(Click on image for a larger version)

Since the in vivo results were very convincing, we tried to follow the fate of the cells towards endodermal differentiation in vitro. And therefore we looked at the embryonic developments described for mouse and human, and we defined primers important for the liver, as for instance α-1-antitrypsin, α-fetoprotein, albumin, hepatocyte growth factor, and markers for the human pancreas, as ISL-1, PDX, NKX6.

(Click on image for a larger version)

However, the results we have up to today show that the USSCs differentiated in vitro under certain conditions were negative for several markers, including insulin, Pax4, PDX-1. But they were positive for HNF4a, α-GATA4, albumin, α-1-antitrypsin, ISL-1 and several other markers known to be expressed in the liver and the pancreas but, of course, also in other tissues. That means we are able to generate at this stage only precursor cells.

(Click on image for a larger version)

Here you see an example of how it looks. We have a nice expression of Cyp3A and of ISL-1, which are precursor cells for both endodermal and mesodermal development.

(Click on image for a larger version)

Therefore we are trying now to establish in vitro models to co-cultivate the USSCs together with a biological dish of liver tissue, injured liver tissue, rat hepatocytes and sheep hepatocytes. But these experiments are ongoing.

For me there was still a question. If the cell is so beautiful as to go towards several differentiation lineages, for me as a cord blood transplant physician it was also important to find out whether the cells can really support hemopoiesis in a transplant situation.

(Click on image for a larger version)

Therefore we measured the cytokine production of these cells. It was obvious that the cells produced a wide range of cytokines, as for instance SCF, which is a well-known stem cell factor, to produce LIF; they produced such other factors as VEGF.

(Click on image for a larger version)

And when we irradiated the cells and put CD34⁺ cells from cord blood in it, we saw a high proliferation of other hemopoietic cells on the feeder layers. That means the USSCs can also serve as a biological niche for hemopoietic cells.

(Click on image for a larger version)

And we saw a very nice amplification of total cells, of pre-hemopoietic precursor cells and of myeloid precursor cells.

(Click on image for a larger version)

Also – and this is important for people working in haematology – we saw amplification of some long-term culture-initiating cells in vitro.

(Click on image for a larger version)

For us, the major points in the future of the ongoing projects are whether we are able to differentiate the USSCs also in other tissue, for instance muscle; we will use, for example, an mdx-deficient mouse to analyse whether the cells are able to regenerate skeletal muscle, and of course we are looking for animal models suitable to accept human cells. In addition, one of the hallmarks will be that we analyse the immune reactivity of these cells. We are developing, together with Anne Dickinson in Newcastle, UK, a human in vivo skin explant model for evaluation of USSC alloimmunocompatibility.

(Click on image for a larger version)

And it looks to us that, compared with the control, the USSC might immunosuppress an immune reaction. This would mean that we can go via more HLA or histocompatibility barriers than expected – or that we must match the generated USSCs for HLA to the patient.

However, I would like to point out at this stage, already today, we are not only performing cord blood transplantation and we are not only using embryonic stem cells for research, but we can use so-called adult bone marrow – autologous bone marrow nuclear cells – at certain ages to repair an infarcted myocardium and in chronic ischaemia. This is very helpful for patients between 40 and 60 years.

(Click on image for a larger version)

I think later on it will be quite difficult to harvest enough stem cells from the bone marrow. But at this stage I would like to point out that already, in Düsseldorf, we have transplanted 120 patients with these kinds of diseases by using autologous adult stem cells into the infarcted area, and the results are very convincing.

(Click on image for a larger version)

We have here the initial scans of the heart wall, at the time of the myocardial infarction, showing the so-called dead tissue. And after three months we see a regeneration of this tissue. That means today we already have a tool to help patients with some diseases.

(Click on image for a larger version)

As the end of my presentation I would like to thank my international collaborators, Esmail Zanjani and Judith Airey from Reno, Arnold Caplan from Cleveland. I have many cooperations with John Dick and with Anne Dickinson in Newcastle.

And I would like to point out that for us it is very important to compare the data we have today with people generating tissue out of embryonic stem cells, because I personally feel we need many embryonic stem cells to learn the differentiation pathways from them. Otherwise, also in adult transplantation medicine, if I may put it this way, it is very hard to analyse the function of the generated cells.

(Click on image for a larger version)

This is my team. I would like to point out that the people in white are the people who are doing routine work in the clinic but the people in black are the researchers that we have, because we feel it very important that we have many tasks to fulfil during the next years.

Questions/discussion

Chair – Well, Gesine, that is even more fantastic than I thought it was going to be. Don’t you get this amazing feeling that really what Gesine has shown us is that the fetus is swimming with stem cells in its blood? Maybe we should have known this long ago, because in 1912 two scientists in Vienna showed that in cattle the freemartin condition – where, if you have twins, the female is born sterile – occurs because there is a vascular anastomosis between the two fetal circulations. And we have known for 50 years, in cattle, that if you have twins there is a fetal vascular anastomosis in 90 per cent of cases and the twins are born chimaeric, with one another’s leucocytes and erythrocytes.

So the evidence has been staring us in the face, but it actually took this fantastic piece of work from Düsseldorf to bring it home to us that here are ethical stem cells with an enormous potential, because they are human stem cells and they are ones that we make ourselves. Is there a single person in the audience here who would think that it is unethical to study stem cells from human cord blood? Could you raise your hand? Can we please report to parliament that there was not one single person at this meeting who could find any objection, ethically, to the work that has just been presented!

Question – Do these stem cells behave like side population cells, and express ABC cassette transporters? And do they give rise to symmetric or asymmetric cell division, or both?

Gesine Kögler – This is the work we are just doing. We are looking now for the ABC transporters and how the ABC transporters are down-regulated upon differentiation, for instance, in the osteogenic differentiation.

On the other question, I am not sure if they divide asymmetrically. This is work that we are also doing now, because we have now the experience with 81,133 cells from cord blood that do divide asymmetrically. So it was first to establish the method and now we will adapt this to USSCs.

Question – Can I ask why you are cautious about going into clinical practice? Cord blood is the one tool that is known to be safe. As you said, there have been 8000 transplants, there have been no teratomas, it works. You are looking at an audience, many of whom are at high risk of heart attack. Why are we not actually using cord blood today – if it is HLA matched – for therapy?

Gesine Kögler – I think there are two issues. One issue is that the USSCs we have now can only be generated from fresh cord blood. That means that at the time of collection we have to cultivate the cells immediately and cryopreserve them down. This costs a lot of money, because all the serum we cultivate them in is very expensive. So we would need sponsoring. I have a cord blood bank of 8000, and I would need at least several hundred of these USSC-generated cell lines, because at this stage I do not get them from cryopreserved cord blood. I can easily cryopreserve the USSCs, but if I try to get them from frozen cord blood it is very difficult, or I do not get them.

But I can tell you that we are planning with our cardiology department that this year, in patients with a severe cardiomyopathy who are unable to get a heart, we will try to use cord blood in an allogenic setting where we isolate the AC133 compartment in order to get endothelial cells, and then implant this into these patients without immunosuppression. And for me it is just a question of immune reaction.

I personally feel we cannot harm the patient, because in a case of a heart transplant he would also get, for instance, immunosuppression, but we will try for the first time to do it without immunosuppression. But I have to bring it through the ethics committee. I think from regulatory points of view, since it is a pharmaceutical drug it will not be such a problem. I think people will try it, and I expect they will try soon.

Question – There are a couple of issues that are worth discussing a bit further. One is that in the liver you have a lot of polyploid cells, so it is normally a tissue which contains polyploid cells as a normal function. Markus Grompe has shown very nicely in the mouse that a hemopoietic stem cell/macrophage cell will actually fuse with the liver cells, be polyploid and correct liver disease. So I think there is already some evidence that a hemopoietic stem cell will do that, and in animals that is well accepted.

I am a little hesitant in thinking about these cells as being pluripotential, despite the Chair’s enthusiasm. I think the need to demonstrate that it is not a trophic effect or a fusion effect is what is challenging a lot of the publications in this area at the current time. I suppose what I would ask as a question is this: do you really have evidence that these cells are residing in the tissue without fusion? Clearly, I accept that mesenchymal stem cells will do that.

Gesine Kögler – Yes. I have shown one slide with one example, but it took us a while – it took us two referees, so to speak – to analyse each cell. We later captured one single cell and we presented this also to the referees of the paper. We later captured a single cell; this can be controlled, and then you can do a single cell picture. And we used both T-cell receptors, different in sheep and human beings, and in addition immunoglobulin strains. We never had both a human cell detected by an ovine genome or vice versa. So it was very clear. In this case it was a non-injury model where we never observed fusion. The majority of models where fusion was observed were injury models, where it is normal, it is physiological that the cells fuse. However, in clinical situations, I wouldn’t care if it was fusion or not.

Question (continued) – I agree with you. Please don’t take it as a criticism. I am just exploring some of the basic biology with you. Do you have evidence that these cells actually ever form cardiac muscle cells, cardiomyocytes? That has really been very difficult to show in many other situations.

Gesine Kögler – I am not an expert on cardiomyocyte in vitro differentiation. The only data we have at present is from the sheep model. However, I plan to have a pig model this year, and so we will see. But in vitro differentiation of cardiomyocytes is a very difficult topic anyway, I think.

Question – To what degree do these cells enter the maternal circulation, and if they are there, are they likely to have some positive effect, a negative effect, or neutral?

Gesine Kögler – Ah, it’s a very good question. But at the beginning, when I started the cord blood bank in 1992–93, we looked at whether we had contaminating cells of the mother in the cord and of the cord in the mother. It seems to be that you always have a traffic between maternal cells and fetal cells, but this is a very low frequency, below one in 100,000. And the cells we generated are really from the child and not from the mother, because this you can detect by HLA. But there is always a traffic between both cells, and of course we all know that some maternal cells can make some damage in the child and vice versa.

Question – When you injected the human cord blood into the sheep fetus, you harvested the heart and the liver and found the human cells there. Did you look in any other organs?

Gesine Kögler – We tried, but at this time when we harvested the organs we had the problem that there were not good markers – or no markers – available distinguishing between ovine and human, for instance in the pancreas. It was very frustrating. And we harvested some of the animals two and a half or three years later, because we thought maybe there was something left, but we only found some cells in the kidney.

Question (continued) – You found cells in the kidney?

Gesine Kögler – Yes, and this was surprising. But not in the other organs, maybe because of the high turnover.

Question (continued) – Can you tell us what other organs you did look for these cells in?

Gesine Kögler – After two and a half years we only looked in the liver again, the heart, the kidney. Yes, that’s it.

Question (continued) – Can you tell us what cells in the kidney, what part of the kidney?

Gesine Kögler – I don’t have any idea. I have already discussed this with some people. Single cells, I don’t know what.

Question – The question of blood cells making neurones, as Martin Pera raised, has been very controversial. What you showed us was cells that had neurotransmitters. A neurone is a neurone when it functions as a neurone. Do you have any evidence that it has electrophysiological properties at all in these cells?

Gesine Kögler – We tried quite hard to look at the electrophysiology also. I showed that they have sodium channels. But we only got not very good signals, and what we are planning now is to do cold cultures with the cells, with neuronal tissue, because Professor Kottmeyer, in our physiology department, who I think is an expert in this field, has told us that this might improve the function.

Chair – Maybe I could end with a challenge for all these young research workers-to-be who are in the back row. Any one of you could start some stem cell research tomorrow morning which would be invaluable. Go to your nearest slaughterhouse, wait until a pregnant cow comes through, and if it has twin fetuses just check that there is a vascular anastomosis, which as I said in 90 per cent of cases there will be. Then look at those two fetuses and, using some simple serology, see how chimaeric they are with respect to organs other than blood and white cells. Sitting in the old freemartin maybe is the answer to all these questions. Does each twin share part of its co-twin’s brain? Does it share its co-twin’s germ cells? There is some evidence for that. Does it share its co-twin’s skin?

And if really, through Gesine’s research, we have got this new concept that the fetus is swimming in stem cells, and that if fetal blood gets into another fetus you will have a massive transfusion of stem cells, it gives a whole new concept to biology.

Gesine, why is it, do you think, that the fetus is swimming with spare parts?

Gesine Kögler – I think we know this from the hemopoietic system already. You have many circulating hemopoietic stem cells going on from fetal liver to bone marrow, and this is just a resource because the fetus has to undergo remodelling all the time. I personally think this is just a stem cell resource a fetus has, because we know from the cord blood that the hemopoietic cells are only in cord blood. if you take the baby blood one day after the delivery, they are gone. It is not the same quality any more. So there must be a reason, at the end of the pregnancy or during the whole pregnancy, that you have so many stem cells in it.

Chair – As a final thought, could we get our terminology straight? How would you like to call these cells in cord blood? Do we call them ‘newborn stem cells’?

Gesine Kögler – Neonatal.

Chair – Shall we agree on that? This would actually be a very important definition to establish. Let’s call them NSC, neonatal stem cells.