 and has held the position of associate of the Neurosciences Research Program of Rockefeller University in New York since 1990. Professor Bjorkland's VEDA lists several hundred publications giving testament to his research breadth and of productivity. Professor Bjorkland has been presented with many awards and honors over his career. He has been a member of the Royal Swedish Academy of Science since 1989. It was a little more recent than that. He shared the 1990 Ipsen Prize in your own plasticity for his work with Drs. Fred Gage and Albert Aguayo. In 1993, Professor Bjorkland was presented with the Charles A. Dana Award for pioneering achievements in health and education. His pioneering work on the repair and replacement of damaged neural tissue holds tremendous promise as a way of treating such neurodegenerative diseases as Parkinson's and Huntington's. The effects of these diseases were long thought to be irreversible, but the work of Professor Bjorkland has opened up new exciting frontiers. Professor Bjorkland's recent work has focused on techniques for the transplantation of neural tissue and on the use of genetic engineering of healthy cells to create or clone new populations of cells for transplant. Please join me in welcoming Professor Bjorkland. Thank you very much. I want to thank the organizers for inviting me to come to this college. As a Swede, it's a very special feeling for me to come here. I grew up in a region of Sweden called Småland. My father was a Lutheran clergyman in a small town called Växjö. And from this region, between roughly 1850 and 1920, about a third of the population emigrated to North America. And I know many arrived here in this part of the Midwest and looking for a new life. And it remains a strong romantic feeling in my area for all those people who dared to leave our then quite poor country to find a new life here. And I think it's of course something special to come back and see all the Swedish flags mixed with the American ones, see all Swedish names. I haven't seen Bjorkland yet, but it must be around the corner somewhere. Otherwise, it feels like a transplant, I think, of a piece. And transplants are not all bad. I'm convinced about that. Now, my theme is brain repair. It remains a challenge in neuroscience research to utilize the knowledge that is generated by modern neuroscience about the build-up and function of the brain and the nervous system to understand mechanisms that may be utilized to repair, promote functional recovery, restoration of function in conditions of disease. It can be neurodegenerative diseases like Alzheimer's or Parkinson's disease or brain trauma, stroke, and other devastating illnesses. It's a very early science brain repair, I must confess. And it means that I can't in this lecture tell you how to go about it. We don't know how to do this yet. Some things work quite well in rats. So if you have a sick rat you want to care for, we may be able to do something about that. But the human brain is, if nothing else, very much larger. It's about 1,000 times larger. And it contains a vast number of connections that together, of course, function in a very complex manner. So it means that the explorative techniques that we are investigating in small mammals, they can't automatically be transferred to practice in humans. There is, if nothing else, a scaling up problem. There's also a problem of complexity that one has to face. Now, this idea of repairing the brain is not entirely new. This is a quotation from a famous series of books, Ten Books of Surgery by Ambrose Parry, who was one of the pioneers of surgery, perhaps the father of surgery. And he tells this story about the gentleman who was otherwise well. But he has the idea that his brain was rotten. So he went to the king and begged him to command the physician, the surgeon in ordinary, and Parry himself to open his head, remove his deceased brain and replace it with another. We did many things to him, but it was impossible for us to restore his brain. You can see that just the wish to repair the brain is not enough. Parry was insightful enough to realize that the brain is a precious organ. One is better to be careful about it. It was 300 years later before the first identifiable paper related to transplantation to the brain appeared. And it had this very, I would say, optimistic title, successful brain grafting. It was published by an American physiologist, Jelman Thompson, who was working in New York City. And it's notable because of its formulation of the problem. He saw no reason why the brain tissue and brain cells would have to be different from other cells in the body. He felt that it hadn't been explored sufficiently whether or not brain tissue could take transplantation. In the 19th century, one realized and discovered that tissues were transplantable. One identified cells as the elements of living tissue. And one realized by the end of that century that these cells could be alive also outside the body. Now Jelman Thompson was in fact totally unsuccessful, but it just shows the art of publishing. He would probably not have got the paper accepted otherwise. Nowadays we know the restraints and conditions under which brain tissue and brain cells can survive transplantation. One of the most important restrictions is that the cells, the neurons have to be immature. They have to be plastic cells, young neurons or progenitor cells, which are in an active state of differentiation and growth. Such cells, they survive removal from the parent tissue well and they can be put into tissue culture media, but they can also be put in another animal. The three principle kinds of procedures that are being explored are illustrated here. I don't know if you can pick it up in the back there. I can't mention it briefly. These have to be small pieces. One cannot take a lot of chunks of tissue. They will not be sufficiently nourished from the environment. So one has to take small fragments of pieces and they can be put either into the spaces that are available in the brain already. These are the ventricular spaces that are flooded by the cerebrospinal fluid, which can serve as a kind of nutritional medium for the cells to survive during the first day or so before any vascularization takes place. An alternative procedure is to make cavities, make little chambers in the recipient brain where the tissue can be cultured. This requires that there is support of vessels that can supply a new circulation to the tissue. It requires also that there is a direct communication with the cerebrospinal fluid spaces that can flood the cavity. This has been utilized experimentally to explore survivability of tissue, the electrophysiological properties and the way they interact with the host. The third alternative is to work on the tissue as a cell suspension dissociated into its cellular constituents and then use syringes and thin needles and inject the cells as deposits in the depth of the brain. This has several practical advantages. It means that one can reach any site in the recipient brain where one can do multiple implantations and it's also easier to control when it comes to cell numbers and quantity of tissue implanted. Now just to put immature neural cells into the brain doesn't mean that they necessarily have to do something useful. There are several possible outcomes of such a procedure. One is of course that it could be totally negative, that the presence of foreign elements could somehow interfere with the function of the recipient target and thus cause negative, let's call it side effects. Another possibility is that they would do nothing. They may be walled off, recognized as foreign elements and be isolated from the rest of the brain which could very well be a likely outcome because the brain would then have a chance to protect itself from these elements. The third possibility which we're more interested in is the possibility of these immature neural elements, immature neurons and immature glia to participate in repair processes, to replace neurons that have been lost or alternatively exert functional effects through release of active compounds, functional molecules that can improve either the function of the damaged brain or promote recovery regenerative processes that may occur in the damaged host. This is the possibility we've been exploring over the last 20 years and it's only the last decade I should say where this technique has been reasonably well controlled to be used for more precise experimentation. I will be talking about two aspects of transplantation. One is to supply a missing neurochemical dopamine in individuals with Parkinson-like condition, either experimental Parkinson in animals or Parkinson's disease in humans. And secondly I'll talk about experiments where one has utilized cells to supply a growth factor, the so-called neurotrophic factor in order to promote functional recovery in neuronal elements that are still left in the brain but are dysfunctional, are functioning less than normal. So first I would like to touch on the Parkinson's disease condition. This is a classic drawing by Gowers illustrating the problem that the Parkinsonian patient has. There are three cardinal symptoms, the bradykinesia which is the difficulty of movement. These patients have a difficulty to initiate movements and hence difficulty to walk, use their hands or use any motor problems. Secondly they have muscle rigidity, stiffness in the muscles that also complicate movement. And thirdly, often they have a tremor that can be severe in some cases. It's known that the area of the brain that's affected is a deep motor control structure called the stratum. And this is now what is always referred to as Hubble's cartoons, which means that they are not too more than to a limited extent. And this is a very crude simplification of the system. But it illustrates the main portions of the brain that are participating in motor acts. There is a cortical area which you heard Dr. Geutjopoulos talk about, the motor cortex and the associated areas that are involved in motor planning. There is the deep structure, the stratum, which is a critical structure in the Parkinsonian brain because there the storage of a quiet motor program is importantly stored there. Then downstream structures in the brainstem and the spinal cord are the executing parts of the CNS, of the central nervous system. And there we know we have some motor programs we have in-night in our development. There can be locomotion or breathing, similar kinds of automatisms. What we do is to command the system from the cortex and it's executed through the stratum. Now the part of the brain that is affected in Parkinson's disease is the green structure here. It's a small subset of neurons located in the brain brain in a nucleus called the substantia nigra which controls the function of the stratum. It sets the level of threshold for activation of the motor systems, the motor programs in the stratum and downstream from the cortex. And to make this simplistic, I've compared this here by a gas pedal where we drive our motor system from cortex and the dopamine system then works like a clutch for an engine. That is a gadget that allows us to put the motor system into gear or take it out of gear. Now this system in Parkinsonian patients is lost or severely de-enerated. And because of this the clutch doesn't work and the patient has difficulty to utilize its motor system. Now the interesting implication of this is that the engine, the motor driving system is in fact intact. So all the subsystems that are necessary to execute motor acts are present but they cannot be properly used because of the lack of these dopamine neurons. The standard treatment for Parkinson's disease is a drug that stores dopamine in the stratum. It's an amino acid called L-dopa which is the main ingredient in the most commonly used medicines. And the L-dopa amino acid is converted into dopamine in the stratum and in a pharmacological way reinstates function and allows the use of the motor system. The idea behind the transplantation procedure in Parkinson's disease is to restore dopamine transmission in the stratum by putting new dopamine neurons into the stratum. And that's a substitute for the loss of the dopamine input, the clutch system. And this substitute is then supposed or hypothesized to be able to do a job similar to the normal pathway. In Parkinson's disease and the Parkinson model it doesn't seem all that forfeit that this will work because the motor system is intact as I said. All the complex motor command and motor driving systems are probably not damaged, at least not severely damaged in the Parkinson's patient. Whereas the control system is. And this is a very good test condition then to see whether neurotransplants can exert effects. There is an animal model that mimics this. Animals don't have Parkinson's disease which is a problem for experimentation in this disease. But one can generate a similar condition by damaging the dopamine system with a neurochemical, a neurotoxin. And then you get a rat that has analogous problems to the Parkinson's patient. The rat that has difficulties to initiate movement is virtually achonetic. And this is illustrated in this picture where you have a rat that is immobile, has a hunched posture. When you put it into this activity cage it moves into a corner like this and then you can come back 15 minutes later and it hasn't moved. You can activate the rat but it requires very strong activating stimuli. The procedure we use then is to take dopamine neurons from mid-stage rat fetuses. There are dopamine neuroblasts that have just undergone differentiation from the precursor cells. They have not yet extended any long processes. And they have a period in front of them of high rate growth differentiation and formation of connections. And these cells they are taken, dissociated into a milky cell suspension and then injected into the target site. For those who are observant here they can notice that we put them in the wrong place. And that's something we realized early on that we couldn't put the cells in the place where they should be because the neurons do not show any ability to extend their processes up to the critical target sites in Australia. And thus the effective or the non-activated receptors would not be reached by the implanted cells. So instead we put the cells near the target in what's called an ectopic site which is an obvious limitation of the procedure so far. And it may possibly be a serious limitation. We don't know that but this is a matter of fact. These cells they survive well provided that the right fetal age window is used. This is E14 that is 14 to 15 days of gestation in the rat. Then the cells can be identified with histochemical technique which is used here. This visualizes the dopamine in the cell bodies. And one can see that the neurons extend axons, processes, out into the dinovated target area. They make proper synaptic contacts. And with biochemical and micro dialysis techniques it's been shown that they actively release and synthesize the transmitter at a rate that's similar to the normal straighter. So these cells they are spontaneously active. They connect up with the target cells in the normal way. And they have many other features that resemble the mature dopamine neurons. So what about the functions? Well I'm not going to go much into that but in a brief one can restore motor functions quite well in rats. They become mobile again when tested for spontaneous motor behavior and motor initiation in different tests. They come either close to normal range or they are partly recovered in their motor performance. So on basis of these experimental data one has over the years started to test this technique in patients. And this is a short summary which I'm sure the people in the back won't be able to read so I'll take you through this briefly. The first demonstration of functional effects of these fetal dopamine neuron grafts was obtained in 1979. That's about 15 years ago. The first clinical trials that was used done with grafting technique was not done with fetal dopamine neurons but with an alternative tissue namely the adrenaline medulla which contains not neurons but an endocrine type of cell that produces dopamine and related compounds. These cells can be isolated from the patient himself or herself because we have most of us at least have two adrenals and we can spare one. So these can be autographs which was one of the major arguments when trying this kind of tissue. It means that one has limited ethical problems, the patient is his or her own donor. However the adrenal cells have proven not to be particularly suitable for this purpose. They survive poorly in the stratum. The functional effects that are seen are not long lost. After about three years most of the initial effects have disappeared and probably because of the wealth of other compounds released by these cells there are quite a number of side effects that have made the use of adrenal grafting of limited interest and it's hardly pursued any longer. It was however popularized in 1986 by Mexican neurosurgeon Ignacio Madraso and many of you will have heard about the trials using this technique over a number of years after his initial reports. However the summary reports that have been obtained from the registry in Chicago I think it's about two years ago now concluded that without further improvement this technique is not really useful. I should emphasize though that there are several laboratories that are trying to improve the functional properties of adrenal medulla cells these chromatin cells to make them more suitable for this purpose. The first clinical trials using graphopethanigrates was performed in 1987 in Lund in our university and in Mexico by Dr. Madraso. This is as you realize an ethically problematic issue and not least here in the United States it has generated lots of discussions. This is by no means settled but as you are aware of now there is a possibility also for the National Institute of Health to sponsor research and experimentation in this field and this is important because it is a technique that requires considerable improvement yet. I'm going to say a few words about the results we've obtained in the Lund program just to give you an idea on how far the technique has reached. In our patients and we've done two series of patients one a first series of six and now another series that's underway and we use multiple donors. First we have in our preclinical experiments found that single donor tissue material doesn't give us sufficient cell numbers to be able to expect any effects from. So we use in each surgical session tissue fragments obtained from three to four donor fetuses. They have to be young. This is another insight that has been generated from preclinical experiments in rodents that between approximately six and nine weeks of the station that's the period when the neuroblasts are in the plastic stage where they can take the grafting procedure. If they are older the survivability of the cells drops dramatically. We use a stereotaxic grafting procedure similar to the one I described in the rat experiment. We use multiple site implantation. In the first patients which I'll be talking about mentioning we made unilateral implantations. This is an MRI picture taken a month after implantation in one patient where you can see the caudate nucleus and containment here. You can see the implantation site as white spots here. This is not implanted tissue but it's the edema around the needle track which allows one to control or check the site of implantation. So these are unilateral implantations. In some patients the tissue has been focused into one containment on one site. This is then a deliberate restriction which makes it such that these are partial engulfments and I'll come back to this. We believe that one has to work up the procedure to make complete engulfments probably on both sites. This is one patient that's been followed on this slide here three years and is now four years into evaluation. And there is a marked effect of the transplant that according to Ulle Lindvall who is the neurologist in charge of this program is therapeutically meaningful. This is his self-scoring of the time spent enough. These patients who've had their disease for seven to ten years they have run into a complication phase where the drug treatment doesn't give a satisfactory effect despite attempts to optimize those treatment schedules. And consequently they spend a large part of the day in what's called off. This patient between around 60% of the day. After about four months, three to four months there is a first improvement and then beyond one year there is a second phase of slow improvement and by the end of the third year the medication could be removed and he's now without medication functioning well. I should already at this point say that this doesn't represent the cure by no means. He has a well-diagnosed Parkinson's disease. The disease process continues. I can pick that up in several ways. But the impression is that it brings back function to a level that he had seven years earlier approximately. Rigidity, which is one cardinal parameter here. This is rigidity in the arm. It's totally eliminated on the side opposite to the implant and reduced also on the ipsilateral side. And this is an interesting observation because despite that the graft is unilateral there is a distinctive bilateral effect although the effect is clearly more pronounced on the side contralateral to graft, which is the expected side. The way we've been able to study graft survival which is a very important parameter to follow is to use positron emission tomography. This PET scanning technique which you heard about yesterday. In this procedure one uses a ligand, a tracer, fluorodopa that can be seen by the PET camera. And this fluorodopa amino acid will be converted in the brain to fluorodopamine. And it's the accumulated dopamine that is visualized over time in the structure. This patient has hardly any or very low signal over his tube botanist. Some residual signal in the cortex which is up here. Over time you can see that there is now growing up a hot spot here that by three years is dominating the picture. This is over the implantation site in the botanist where he received his implants. If we look at this with combined MRI PET image and I should now say that all these positron emission tomography studies have been done at Hammersmith's Hospital in London by Richard Prokowiak, Dysol and David Brooks who's running the PET center there. And they have, Dysol has in this picture combined the MRI picture and the PET picture in the computer. And here you can see these are two levels from the same patient. You can see on the non-graphed side the essentially empty corded nucleus and pertainment which have low signals. And then on the other side over the grafted pertainment the distinctive as we then believe graph derived signal. This is the sagittal picture that's generated by the computer. The computer can reconstruct pixel by pixel a plane that's running through the injection tracks. And there you can see the three needle tracks going down to the pertainment where the fluorodopa uptake area is. And that fits then with the site of implantation of the cells. So this we believe provide proof for surviving dopamine neurons that the dopamine neurons are actively synthesizing dopamine and storing it, illustrated in the signal here. And we believe also that the level of the signal is an indicator for the number of cells surviving. Despite that we have put as much tissue as we can here the signal is still below that seen in non-symptomatic individuals or normal individuals which means that we have not been able to get sufficiently large grafts to bring dopamine synthesis back to normal. Nevertheless we see a clear effect of the transplants. The other pair of patients I want to mention about has been done in collaboration with William Langston in California and his collaborators and concern a very special group of patients that developed Parkinson's due to intake of a synthetic heroin. These individuals are fairly well known. They developed this and they got this in 1982 and were identified or the causative drug was identified by Bill Langston in 1982 and they have been followed over the years. The advantage of these patients is that the causative agent one nose is no longer there. They took the drug a few times before they developed the symptoms then they developed severe disease within a few weeks and have since then had an essentially stable non-progressive condition. And this is in contrast to the idiopathic and normal Parkinsonian patients that have a progressive disease. So we believe that these patients are particularly interesting to test the functional capacity of the neuro-growth system. One can also believe that these patients have as pure a dopamine-dependent symptom picture as is possible to obtain without contributions of other pathologies. So two patients were operated this way. They had received the or taken the drug in 1982. Without medication they were severely incapacitated. They were bedridden and unable to move. Under medication they regain mobility but had on the other hand severe on-off fluctuations and severe dyskinesias and in one of them also visual hallucinations. So it means that the therapeutic window for the elder treatment was very narrow. So that made treatment less meaningful or less useful for them. These patients both of them showed marked improvement in their mobility and illustrated here with an ACAMNESHA measure that is the arm velocity and the computer test. You can see that prior to transplantation they had very slow movement. This is under what's called defined off that is they had been off medication since the night previously. This is between 5 and 10 seconds to 5 and 10 centimeters per second if you think about it is a very very slow moment. Over the months this is 6, 12 and 22 months. The mobility is recovered and now they are in the range almost at the lower end of normals. These are the two sides left and right side. Perhaps I forgot to mention that these two patients they receive transplants on both sides which is important to point out here because the improvement is bilateral here. Also rigidity which is another cardinal feature gradually tapered off. You can see it took quite some time. In one patient here it appeared within the first three months but in the other patient was evident only after a year. Now in the steady state condition rigidity is minor. Another important aspect of the effect of the transplants have been on the discount issue. In both patients who had marked this condition when they took the L-Dopa medication this condition has been greatly reduced and in one of the patients almost totally eliminated which as I understand from Bill Langston is an important thing for them. Also in this case we have been able to see distinct recovery of L-Dopa uptake. These are pictures taken in Vancouver by Barry Snow who's done the test scanings on these two patients. These are two levels and you can see here with this high resolution of the camera there appears almost as a string of hotspots which might fit to each implantation site. The similar thing on the other side although it looks a bit less. If we look at the quantitative assessment I hope you can see also the orange lines here. This is a combined score or combined chart where the improvement in the neurological motor scores this is a rating scale that has been followed over time in the patients and the PET scanning at L-Dopa level with the orange curve here. There is overall correlation between the two curves that is to say that although it seems as if the neurological score shows a trend to improvement when PET scanning or PET signal is not clearly increased it's on the other hand so that when they at the time when the scores are markedly changed there is also a clear increase in PET signal. However there is clearly a continued improvement this has continued also after two years while the PET signal remains seemingly on the same level. But we feel that there is a direct correlation between the dopamine synthesis capacity of the transplants and the changes in the neurological status of the patients. So then I come to my first conclusion that I phrase as challenges for future studies. Again you can't read at the foreender so I'll have to tell you. The first challenge is to increase the yield of surviving dopamine neurons above the low range that we have with the human cells. We think it's in the range of 5% possible. Temp fold increase up to about 50% seems quite realistic and that would help us in quantitative terms to get good growth. We have to reach larger areas of the strata complex. This is the scaling up problem I mentioned initially that the large human brain is quite a different structure to influence by transplants and this is a problem hard to reach as many of the subregions as possible that we know are involved in motor control. And then we have to reach a more complete restoration of dopamine function in all parts of strata up into the range seen in asymptomatic patients. As I mentioned we know from both the MPTP patients and the idiopathic focusing patients that the PET signal, the fluorodopa uptake level does not reach the normal level and this has to be worked up. And then finally and importantly I think that one in the long run have to find ways to produce more efficient dopamine producing cells for grafting either by using cell proliferation techniques perhaps in combination with the enrichment of the cells or by directly engineering cells for transplantation purposes. And in the long run I think that's the only satisfactory way to do this to have cells that are well characterized that are known to be viable and functional and known to be harmless that is not carrying any infectious agents or any other thing that could be negative for the patients. This is however still in its infancy. We have however been interested in pursuing this and in this chart I've indicated possibilities how to engineer cells. I realize that you have difficulties to read this. These are alternative strategies that are pursued for engineering of cells for transplantation in the brain. On the left side here it's illustrated that one could engineer the neurons themselves. There is a possibility to improve their growth capacity for example or control their proliferation in vitro conditions by genetic manipulation and in this way be able to generate either pure neuronal populations or more of the cells and more effective cells. There are also ideas to use precursor cells as indicated in the middle here to use progenitor cells that are present in the early neuroepithelium of the individual and then produce mature differentiated cells that can be used for transplantation and they could then be produced in large numbers. There's also a line here indicated where one use encapsulation that is to produce tumor cell lines make them a secret dopamine for example and then encapsulate them to prevent them from proliferating in the brain and use this as a kind of delivery device. This would be a semi-artificial gadget which could be implanted as a mini-pump a biological mini-pump and supplying the missing compound that way and then on the right here I have the conditionally immortalized cells. These are cells that with a genetic trick is made to proliferate under controlled conditions. These are generated supposedly from neuroepithelial stem cells and the ones we've been using and which I will talk about now they have been produced in Ron Mackay's lab in Washington with modified oncogenic protein that keeps the cells dividing under low temperatures 33 degrees lower than body temperature but is switched off at body temperature 37 to 39 degrees. In this way one can have a controlled proliferation of the cells one can put the cells in culture grow them up under 33 degrees transduce the cells with the gene that is of interest and then check the transduced cells with appropriate techniques for their function grow them up in sufficient numbers and then grow them and this can once the cells are characterized they can be grown up in unlimited amounts. This is an attractive source of cells for several reasons first it's CNS dried secondly they can be controlled with respect to their proliferation and thirdly they can be generated in unlimited numbers. The final advantage is that these cells in fact can differentiate in both neurons and glia depending on the conditions. So what we've done is to start with one of these cell lines it's called high B5 it's generated from embryonic rat hippocampus in Romach-Eisland and these cells proliferate nicely at 33 degrees as seen here and when they are moved to 37, 38 they stop dividing and they start to change shape they become more elongated send out longer processes under normal in vitro conditions they show only partial differentiation when they are put in the brain depending on the condition one can either see them turn into glial cells, most are astrocytes or into a mix of neurons and astrocytes. We have transplanted these cells into the rat brain and followed them over time and this is just to show perhaps it's almost uninformative to most of you these are thymine labels so each dot here represents a cell what I want to show is that if you implant these progenitor cells as a bolus injection in one spot they will not remain there over a few days they will migrate out settle in an area up to about two millimeters away from the implantation site and there differentiate most of them into astrocytes when it comes to this cell line here so we think we have here a cell that can be turned into glial cells they can migrate the mix with the host tissue and they can as it seems totally and completely integrate into the texture of the host environment now these glial cells may not be possibly particularly interesting for supplying dopamine but what we have been more interested in is to use them to supply growth factors these molecules, protein molecules that are called neurotrophic factors or neurotrophins that can support survival and growth of certain populations of neurons in the brain and this we've done by transducing the cells with the gene for nerve growth factor, nerve growth factor is the best studied of these neurotrophin molecules this is the construct here it's a simple retroviral vector that is infected into the dividing high-bified cells and after repeated infections with the retroviral vector there is an increasing production of NGF in vitro this is illustrated here so after four to five infections there is a high secretion rate of NGF then this heterogeneous infected cell line have been subcloned and different subclones are illustrated here by subcloning these cells and I can't focus you now this one here is the highest secreting subcloning it's called E8 here it has tenfold higher secretion rate than the mix so by subcloning we can increase the secretion rate by one further order of magnitude and this is independent of whether cells are differentiated or undifferentiated these are the differentiated cells and these are the undifferentiated cells these cells are then implanted into the rat brain and I'll limit myself now to telling about two experiments which we have done to evaluate the usefulness one is simply to put them down in the vicinity of the cell group that is sensitive to NGF this is a colonergic neuron called nucleus basalvis that projects to the entire cortex and that's known to be involved in memory learning and it's known to be denerated in Alzheimer's disease it's one of the interesting target neuron systems for neurotrophic actions in the brain we put in 100,000 it should be there zero missing cells per deposit on each side and in order to compare we put control cells on one side this is the control side where you see a normal appearance of these colonergic cells on the implanted side the NGF receptor protein which is illustrated here it's much increased in its expression and the cell increasing size indicative of a hypertrophic response this is the typical response for seeing of the infusion of NGF into the nucleus basalvis this is the cell increase seen here after one week, four week we know it's last at least for ten weeks about a 40% increase in cell size which is the maximum hypertrophy response that's seen after direct infusion of NGF into this nucleus this now we've taken to test in old animals or aged animals this is a model we've been interested in for several years you've heard this commented upon before that with aging part of the aged rat population declines in learning and memory performance and this is correlated with distinctive changes in the colonergic nucleus basalvis system which undergoes what appears to be an atrophy possibly also a partial cell loss the way we've tested their learning ability is in a water tank this is a test called the Morris Water Mace it tests for the rat's ability to learn a place navigation task that is to swim to find a platform that's just under the water surface from any position in the tank the only way the rat can learn to do this is to learn the position in the tank relative to the external cues in the environment and this is then something that normal young rats learn quite quickly this is three groups of animals here the triangles at the bottom illustrate the learning curve of young animals they start off searching the pool they don't know where the platform is and then they are tested each day for five days and this follows then Eric Kandel's rule that you'd better let the rats sit down and think about what they've learned for a while before you test them again so if you test them over a working week they are pretty good at the end of that week now aged animals they fall into two main categories one category is rats that are fairly similar in learning to the young animals these we call non impaired another group which is about 30 to 40% of our two year old rat population they learn little or nothing over the week this is very little learning they can show some more learning as you will see in the later curve this is what it appears like if you plot this which is done by a computer you can see the first time you put the rat into the tank it does what probably you and I would do also swim around the edge to see if there is a ladder or something then when they failed after the test period which is here was 120 seconds they are put on the platform to be aware of the fact that there is something to escape to by the end of the week they have acquired the task and they now navigate virtually directly onto the platform and most critically when you remove the platform in one trial here you can see that the rat knows fairly well where it should be it's swimming across this site thus marking that it has acquired a genuine knowledge of where the platform site was now an impaired aged animal and this is now upside down here so you have to read the other way I hope you don't mind then you see it starts off as the young rat but by the fifth day which is over here it doesn't really have a clue where the platform is it doesn't navigate towards it it sweeps around aimlessly as it seems and when you remove the platform it doesn't influence their swimming pattern because they don't know where it is anyway and this is then reflected in impaired learning and it doesn't matter how long you train those impaired animals they will never learn so what we have done in the first step here is that we've shown that these neurotrophic factors these neuron directed growth factor proteins they can improve learning in this test and this is the design of it they are first selected in a pretest we identify the impaired animals then we take the impaired animals and put in a pump which is underneath the skin and there is a cannula going into the ventricular system and then they are tested twice by the end of the second week and by the end of the fourth week during continuous infusion of nerve growth factor or baby for a thing here this is what it looks like if you look at the pretest where the impaired animals here learn a little but it's still very bad the non-impaired aged animals they learn slower but they get down to the same escape latency as the young ones which is the bottom curve there and then they are retested on week one and week two and you can see that infusing vehicle that is just the carrying of fluid does not make much difference now if nerve growth factor NGF is included in the fluid the rats will learn considerably and significantly better here you have the NGF group as a control or matching protein one can say we've used brain derived neurotoxic factor which is another member of the same family which had no effect and there was no additive effect of BDNF over NGF the combined infusion had the same effect as NGF alone we saw a similar effect of two other neurotofins the ones that are called neurotofin 3 and neurotofin 4 which means that the NGF effect is not unique for that subtype of neurotofin it's probably shared by other ones so there is then a clear effect of exogenous neurotofin in this model and how about then the possibility to achieve this correction of behavior by using the transduced cells this is an experiment where we have implanted the NGF secreting high B5 cells the ones that I showed this so called E8 clone that we know produced the hypertrophic effect of the nucleus basaldis and again we select the animals in a pre-test similar to the infusion experiment we have two impaired groups one receive control cells that do not secrete NGF and the other two groups I should say receive NGF cells either in nucleus basaldis and the septal region or only in nucleus basaldis these are two sub portions of the colonergic neuron system and you can see that there is here an effect on acquisition of the task that is very similar to that obtained by exogenous NGF it developed over the first test week and it's maintained over the second test week so it's at least maintained for four weeks with these cells so these are then very initial experimentation in rats it's done with a model system that we think is interesting to explore and it shows at least the usefulness of these kinds of cells for delivery of neurotrophic factors, growth factors to the brain we think that these kinds of progenitor cells are very useful for gene transduction because they are easy to handle and easy to manipulate genetically they show good survival and integration of the growth thing in the adult brain and we know that they survive for very long times and are stable over time in both cell number and in their cell type there is no negative effects at least on the level of analysis we have done so far on the host tissue, there are no tumors there are no immune responses despite that they are derived from other subtype of sprague d'orio rats and they maintain efficient transgene expression and transgene delivery with the same efficiency as high doses infused by minipods and we know that these cells can deliver their NGF at least over a 10 week period and we are now exploring longer term there are however a number of aspects of this use of engineered cells that have to be explored first one has to devise gene in long term experiments to see how long can they be functioning how can they maintain function over many months, preferably years and we have to look at the cells to them over considerably longer period than we have now, that's years and also establish and clarify the transgene expression over long time we have to look more at what they do as cells if they integrate the way we think they do and to what extent they can participate in reparative processes in the brain our hope is that these kinds of cells would be able to join in to the regenerative mechanisms that are induced either by lesions or by degenerative processes and that they can repopulate the tissue in a virtually normal manner and become normal constituents of the tissue but supply factors that may improve functional recovery the aged model indicates a particular usefulness of this this is a condition where one can expect that any supply of a growth factor or a neurotrophic factor that would be useful in such a condition has to be sustained over long periods infusional proteins for therapeutic purposes will always be problematic as long as it's not short term we think that the cell implantation, the cell transplantation procedure can circumvent that limitation and provide for a new and possibly also very versatile repair technique that may be possible to use also in clinical conditions welcome we're going to arrange a couple of comments from our panel come in response to a job following that we will address questions for Dr. Bjorkman's talk Dr. Churchill okay, just start by asking Dr. George Offless a couple of questions about his talk which I found absolutely fascinating I was interested in focusing first on the case where the neurons fire during the delay now part of what puzzles me is this that the firing looked much the same during the delay as when there was no delay or when they were remembering where to go as opposed to having the signal in front of them so what's different when the action is actually performed why isn't, or let me put it this way if the neurons are firing during the delay why doesn't the movement occur, what else do you need and I guess the other sort of related question then is where does the learning take place is it in the motor cortex itself or do you think it's more likely to be in the basal gagglia and if so what are the connections what is the role of the motor cortex here altogether if it's not doing the learning if it's not doing the deciding if it's not doing the initiating what exactly does it do and where are these other things happening there's a second much more general question I think motor cortex is doing practically everything instead of nothing but let me answer the first question first that is more factual one what in the motor cortex during the delay tasks about 50% of the cells will be activated and another 50% will be inactive unless there is a movement that is going to happen now in what I showed for the delay task the instructor delay task I had simply the population vectors during the delay and I didn't follow that the length of the population vector is about one third during the delay relative to what we get after the go signal which was actually seen in the memorized delay example where you could see that the vector is quite smaller the reason that it is smaller is that these extra 50% of cells are not active during the delay period so these add up that are recruited strictly after the go signal now there is also strictly speaking mean that there is activity in the motor cortex it's not obligatorily mean that the movement is going to happen and there are a couple of explanations for that one is that the activity is gated downstream and we know for a couple of places especially in the spinal cord and brainstem where there is convergence from other structures all to the same neurons that these signals are directed and the gating if you want there has to be a somehow approval from other structures for the movement to happen on the other hand it is also clear that if you look at what is happening in the spinal cord during the delay with sensitive techniques like the Hoffman reflex when you look at the excitability of the motor neurons there seem to be changes during the delay so there is something a downstream signal that is leaking if you want but it's not enough to engage the alpha motor neurons and start making a massive contraction now concerning the second question I think it's now a process somehow quite unrealistic to think of anyone's structure doing everything in a way I think especially the motor system where with all due respect to Dr. Hubel if you count the neurons in the brain probably three fourths or more are devoted to the motor system if you count the granules in the cerebellum it's just very I think naive to think that all of that producing a movement is a function of just a single structure whether it's motor cortex, cerebellum, basal ganglia or any other now we know that all of these structures are interconnected they are dynamically interacting at many different levels studies in behaving primates do show that changes in activity for example in the motor cortex in cerebellum are almost simultaneous very early and very simultaneous in the basal ganglia they definitely come later but I think what we need to develop is a more coherent or if you want a theory that takes into account the interaction between areas and structures rather than in any single one of them now what is the learning taking place I really can't answer that certainly memory signals that is when you retain something in motor memory that seems to be distributed in these areas which has been seen in the motor cortex has been seen in the basal ganglia definitely that's where it started actually and I'm sure if you look at it in the cerebellum you'll find that there are two so motor memories or things relating to keeping in memory what you are going to do is locally seen in the areas that are involved the same is true for signals relating to eye movements that are withheld and memorized and you see signals in the parietal cortex and frontal cortex and so on and so forth you don't seem to have one structure other than the local somehow motor areas that is delegated the keeping of short term or ongoing sort of memorizing if you want I was struck by the comment that you made and perhaps I could engage you and David to discuss this issue you made a very nice point when you were looking at cells responding to a specific axis of movement that you expected them to have properties similar to the axis of orientation neurons which David and Thorsten described in area 17 and as David pointed out those are very narrowly tuned so if you present a ball from 12 o'clock to 6 o'clock the animal will give you a high frequency response if you change that 20 degrees it'll respond much less effectively maybe even not at all by contrast you find that your cells are extremely broadly tuned and as a result you suggest that a single cell participates not only in one kind of movement but really in a family of movements and as a result any given movement is determined by a population of cells some of which will have their maximum response in the preferred movement but many will have responses that are often preferred movement that would suggest that there are in fact two kinds of codes in the brain those that govern the rules of sensory systems and those that govern the rules in the motor system and that may very well be so but then one has to reconcile that with the Vilniusum experiment in which Vilniusum is asking an animal to judge changes in the direction of movements and spots that move across the oscilloscope screen and you can stimulate a small population of neurons and influence the animal's judgment its evaluation, its decision as to whether or not it sees movement in one direction or another so I wonder whether the two of you might comment on these alternative views of the brain and how the Nusum experiment fits into that view the idea that I suppose would be that you have cells, two cells broadly tuned but slightly differently tuned and then by subtracting the activity of the two you can achieve narrow tuning that is an idea first proposed as far as I know by Talbot and Marshall way back in the 1950s we had always thought that when we saw broad tuning we hadn't found the appropriate stimulus for the cell and an example of how you can have such a thing is a narrowly tuned cell a cell that responds very sharply to an appropriate orientation if you look at its responses to movement and very rate of movement you find that it's very broadly tuned to rate of movement similarly with other variables like spatial frequency of gradients would be a good example our conclusion always was that in those cases we hadn't found the stimulus that the cell is really tuned to and when we did find the stimulus it would be very sharply tuned I can't see any evolutionary advantage in pooling things so as to make every cell a crummy analyzer of some variable and then have to do a subtraction to get the precision back there's one case where this is known to happen and known absolutely securely as well as anything that we know and that is in color vision you have three receptors, three cones each very broadly tuned to color and by subtracting their activities you get sharp analysis of color but there's a very good reason for doing that you can't every spot of the retina have a whole gamut of sharp tuning for all sorts of different colors so you have to subtract the activity at higher levels but my prejudice and it's certainly a prejudice because one has to look at different systems and try every variable one can possibly find think of my prejudice is that the broad tuning is an unusual thing if it exists at all in a central nervous system of course it's a prejudice that's hardly falsifiable because it's a negative impression that would be my answer such as it is or let's say proposal an alternative proposal there's so many things one can do to at least within the motor system you study movements or you can study isometric contractions or you can study some natural movements or train movements and that's what essentially what we have done and we have seen a regular behavior it's a matter of calling it not terribly good coding on the other hand it could be a very nice coding that is a certain function that the shape of the function doesn't necessarily define the goodness of it this kind of behavior sometimes single cells have been seen not only in the motor cortex practically in every motor system or structure that has been looked at that is the pre-motor cortex the parietal cortex cerebellar nuclei and the cerebellar cortex the other system of course the sensory system but you have very similar sort of aspect is the vestibular system you don't have any sharp coding there and I think there is from the motor system point of view I think there is an evolutionary advantage to that and that is that it's very resistant to cell loss by design you can lose a lot of cells and still you can produce a rather accurate motor response without having to rely on any specific element which if you would lose you would have a scotoma somehow for that particular movement so it's a judgment call if you want on the other hand this does not mean that there are no very specific cells for very specific patterns and we have identified such cells ourselves recently when we described cells that are very specific for particular learned memorized movement trajectories where we found that there is a number of cells that behave like usually that is that relate to the movement that are being memorized as previously but there are others that would be very specific to a particular trajectory of movement that is evolved from memory so these cells are indeed very specific but the majority of them would be quite active during regular movements and I would be hesitant to somehow reject all of them as being irrelevant after all they proceed the movement they produce the movement to some extent they interact with other cells that have the same properties and within a small ensemble they provide very good information without being at the single cell level but at the level of a small ensemble interestingly it isn't work in the frog by Emilio Mitvici and his group they have identified very similar behavior of spinal interneurons that seem to control the limb in a broadly turned fashion and which actually would be very appropriate recipients of motor cortical signals that we have seen I think we'll commission from everyone make a transition to the questions about Dr. Bjorklin's talk by bringing up a couple of questions from the audience and then we'll ask for comments from the participants and the panel how to choose to here one person from the audience has asked has higher mammal tissue other than human fetuses been considered for transplantation other so higher mammals higher mammals the use of possible use of animal cells in humans is something that's much debated it seems to replace one ethical argument and discussion with another one the first problem I will have is that grafting across species barriers will inevitably create immune rejection problem so the first thing one would have to do is to find a way to get around that I think that although we have been using immunosuppression as a precaution in the studies we've done so far I don't think that one should in the long run accept any procedure that would require long term immunosuppressancy treatment that would be a negative aspect there are attempts to modify cells again with genetic techniques to make them innocent in immunological sense so that they would be accepted across species that would be for example to engineering human-like surface antigens so that they would not be recognized as foreign by the human immune system if such an approach succeeds then one would eliminate one clear obstacle for using animal cells now even given even if one has assumed that this was possible I'm not sure that animal cells would be a better choice than human cells I think that human cells could be generated under conditions that would be ethically acceptable at least in most communities and provided that one does not depend on access to aborted people material I think that this would be okay I'm just wondering whether one couldn't take from a given patient that you want to do this procedure on skin fibroblasts grow them in culture and then transfect them and use them in the brain would that be possible? Well that is an active line of research that's pursued in a couple of laboratories here in the United States this is quite feasible fibroblasts do survive after transplantation to the brain the weakness is that these cells are not normal constituents of the brain so they will remain relatively isolated and they will remain as a kind of cell core at the site where you put them in this will create some geometric problems when it comes to diffusion of factors that you want to reach deep into the brain with and it may also represent a problem over a long time because nobody is known that the fibroblasts are turning over and it's unclear if these cells will in fact survive several years and they may also gradually become more and more walled off and thus isolate from the brain I think there is much to say that the ideal cell for to use for growth in the brain should be of another system Just a brief question I wondered when you showed the slides of the implants the needle tracks were really quite evident and so my question is is there any damage that's caused by the tissue that is destroyed in the implantation procedure itself? Do you test for that? Does anything show up? There is a damage associated with the needle penetration the needle has two dimensions the upper part is about two and a half millimeter wide and the lower part is about one millimeter These dimensions are neurosurgeons refer to other experience they have when they say that use of instruments of that size does not produce a damage that can be picked up Now obviously from where to expand the implantation side it's difficult to avoid the impression that this will also add to the total trauma and I think that if this is a way to distribute tissue over larger areas which I think is necessary here then one may also have to develop finer thinner instruments and this is actually done in some of the programs ongoing now in the United States they do have that Another question from the audience here from what area of the human fetal brain are cells taken? Are there differentiation problems or do all of these neurons produce dopamine? What we do is to dissect us closely as we come the region that contains the developing dopamine robot but the estimation we do on our preparations suggests that it's about 5-10% of the total number of cells that are neurons the other 90-95% is a mixture of other neuron types and progenitors of the ideal cells that we develop later this is clearly let's say an imperfection of the technique we would like to have more control and pure cell preparation than we have now it might be that the supporting cells, the ideal cells are valuable constituents of the cell mix because they what may speculate or hypothesize that these cells helped the neuroblast to survive and differentiate whereas we think that the other normal that is just accidentally included in our dissection may there will be a negative and this is another factor that tells us that the current technique is not particularly sophisticated one has to accept this one more I'm going to start with the end of this one and then go to the beginning the person starts by thanking you for the understanding of individual cells as a way of avoiding fetal tissue transplantation starts the question with a quotation from Robert Oppenheimer who said at one point in the development of this bomb according to the source when the technology is sweet you go ahead and do it and the person wants to know if you could comment on the NIH's recent decision on fetal tissue research and the ethical problems that you see in this area I have a very definitive view that NIH made a mistake in banning experimentation using fetal tissue currently early fetal tissue is the richest source of progenitor cells and cellular elements with high growth capacity in the nervous system but in other tissues in the body it is the same for the hematopoietic system for some of the endoprime systems for example pancreatic arguments by banning the use NIH made a disservice to this research enterprise and pushed it out into the wilderness and generated the kind of among surgeons with less than serious interest in the scientific development of scientific aspects of the technique and this has to some extent discredit a field that is very important to develop in a good way there is so much to gain from serious careful development of the cell transplantation technique much of the diseases that are in focus for this are diseases which where we need radically new or entirely new approaches to treatment and of course cell transplantation is only one of these possible developments but the development that holds much of flexibility and potential usefulness if one cannot develop this technique within well controlled serious programs that are controlled by the most influential bodies here in the United States like NIH then one can be sure that the future of this field is not going to be as good as it could be now by lifting the ban I don't think one has solved of course all the problems except of other ethical problems and that is to say which kind of experiment should be allowed under which conditions should be allowed and who should be allowed to do it my experience from working with this approach in Sweden is that this can and should be solved in open society in full dialogue between scientific community and the public opinion and representatives of the world it took a lot of so much I was just going to add briefly because I agree wholeheartedly with what you say but the additional thing I guess I would want to say is that for those of us who see abortion before viability as a morally appropriate decision and who view the conceptus not as persons but as tissue that it's still before the formation of a nervous system then I think for us it is not a moral problem and it seems to me that so long as abortion is permitted and so long as it is a morally legitimate thing to do then using the tissue rather than simply disposing of it and destroying it but using it for some useful purpose is also legitimate I guess at this point checking down the line here it's appropriate to stop and to thank the panel for their patience and for their energy and for the excitement of their work