Perry Bartlett - Symposium 2003 - BMA Home -
Regulating Neuronal Production In The Adult Brain: Opportunities for Restoring Brain Function.
Perry Bartlett
The Queensland Brain Institute
University of QueenslandThank you very much- and congratulations Max on organizing this wonderful event. I think we are going to look back on this as a pivotal meeting, given that so many of us have shared your vision of somehow opening up a research dialogue between the basic neurosciences, cognitive neuroscience, behaviour and mental illness. However, until now, this has been almost impossible to achieve, so it was a delight this morning to hear Christos Pantellis talking about neuronal turnover (in schizophrenia), a subject which I am also going to address today.
Essentially, I want to focus on the concept that the adult brain is capable of change. What Chris pointed out, of course, is that in some cases, e.g. schizophrenia, this change may in fact be detrimental, in that it involves a loss of neurones. Nevertheless, ten years ago we never envisaged that the level of cell loss or the ability of the brain to make new connections could be so dramatic or so robust. The realization that this is the case has revolutionized the way we view the adult brain. By contrast, the developing brain, a system in which many of us cut our teeth, was always acknowledged as a highly plastic organ; however, we thought plasticity would all be over by the time we got to adulthood and, when we eventually reached the age that Max or I now are, we were sure it would be downhill all the way. So the ability of the brain to adapt or respond to external stimuli, predominantly by making new connections, opens up the exciting possibility of harnessing this potential to rewire or fine tune the circuitry. There are incredibly large numbers of interconnections, perhaps 104 per cell, which appear amenable to being manipulated and reinforced in a way that Mike Calford is probably going to talk about next. Mike really was a pioneer in this area of plasticity, some ten to fifteen years ago, and I'm sure we're going to look back and realize what a major contribution he made.
I am going to talk more today about a second aspect of plasticity, i.e. cellular plasticity, which is how we might actually generate new nerve cells in the adult. In other words, how the adult brain might be able to adapt to a changed environment and, instead of accepting an irrevocable loss of neurones, as Chris talked about this morning, we might be able to use the innate capacity for cellular plasticity to drive repair or regeneration processes. The potential inherent in plasticity is enormous, and there is no doubt that if we could understand some of these attributes we could enhance learning and creativity. I've described this in my slide as re-tuning functional networks. I have to confess I have no idea what that really entails, but it's clearly the sort of thing that we want to achieve in mental illness and, of course, in relation to damaged or aging neurones, particularly in conditions such as Alzheimer's disease. I think all these mechanisms are going to bear fruit in the next ten years, at least in part. It reminds me of the most florid illustration of plasticity at a synaptic level, in which Mriganka Sur rewired the optic nerve into the auditory cortex instead of into the visual cortex. This was an experiment that took many years to perfect, but it revealed not only that auditory cortex now acted as visual cortex, but also that an architectonic map and orientation columns were established similar to those in the visual cortex. Not a simple experiment, but I think Sur convinced everyone that the cortex is in many ways a tablarosa on which various functions can be imprinted. Now, admittedly, this paradigm involved a ferret at a very immature stage of development, and not an adult animal. Nevertheless, I think understanding the rules by which we can reinforce synapses, and perhaps integrate new neurones, is something that is going to revolutionize neuroscience over the next ten years.
The origins of cellular plasticity, rather than synaptic plasticity, go back almost a hundred years. We've known that cell division occurs in the adult brain since the days of Allan around 1912. Then, in 1962, Altman showed fairly convincingly that new neurones were being produced. Unfortunately, due to a lack of unequivocal markers for neurogenesis, no-one believed him at the time, and it was generally thought that the cells he had identified were astrocytes rather than neurones. The breakthrough really came in 1984 with the publication of Fernando Nottebohm's beautiful work using songbirds, including some Australian species. He showed that the male songbird's centres could be replaced during song learning periods in spring by new neurones that had undergone radial migration. It took a while for a few of us to realize that this might also occur in mammalian species, but in 1992 my group, including graduate students Trevor Kilpatrick and Linda Richards, together with Brent Reynolds and Sam Weiss in Canada, were able to show that in fact one could grow cells from the adult brain which were capable of giving rise to new nerve cells. Although we recognized the significance of this finding, we still had no real understanding of the role these cells played in brain function, suspecting that perhaps they were simply developmental remnants. We have since come to realize that these cells (which have popularly been called stem cells, although I prefer to refer to them as precursor cells) represent a resident population within the adult brain with the everyday potential to proliferate, migrate and differentiate into new neurones and glia in various parts of the brain. The assay we now use involves taking a single cell from a mouse or from a human and culturing it over the course of a week to obtain a neurosphere (a tightly packed ball containing 5,000 or 6,000 cells). You can then differentiate these neurospheres into neurones, astrocytes and oligodendrocytes or, alternatively, disaggregate them in order to grow secondary neurospheres. In fact, from a single mouse precursor cell you can grow up to 1010 cells, although these numbers are yet to be achieved using adult human cells. Consequently, the potential to generate large number of neurones from a single cell is staggering, making it unlikely that there are deficiencies in stem cells or precursor cells in the adult. We also know that these precursor populations, and hence neurogenic potential, are found throughout the brain. However, the highest numbers appear to be in the forebrain, in the subventricular zone or lateral ventricle, which is the region I am mainly going to talk about today. In the ten years since our 1992 discovery, we have come to realize that these cells are functioning in everyday life. The first functional studies were carried out in 1993 by Marla Luskin and Alvarez-Buyla, both of whom did beautiful experiments revealing the existence of a population of precursors residing in the lateral ventricle which generated neurons that migrated along this rostral migratory stream into the olfactory fold. In fact, these cells lie cheek to jowl as they move en masse toward the olfactory bulb. We now know, given the use of BrDU in terminal cancer patients, that this process also occurs in humans, and that it probably involves as many as 10,000 neurones a day. There is also unpublished evidence to say that those olfactory neurones which receive olfactory stimuli capable of activating their receptors preferentially survive, while those which don't receive such input die, a bit like clonal selection in immunology. In other words, use it or lose it. Whether this selection also occurs in the olfactory bulb interneurones produced by the ventricular stem cells is unknown. More relevant to the idea of environmental selection were the findings by Gould's group and Gage's group who demonstrated cellular plasticity in the adult hippocampus, especially in the dentate gyrus. In mice, although a magnitude lower than in the subventricular zone, a couple of hundred cells turn over every week or so. In fact the whole dentate gyrus would probably turn over in a mouse every five years- if it lived that long. This finding has been followed up over the last three or four years with quite staggering results. Experiments using a retrovirus expressing GFP to label dividing cells and their progeny have revealed that these cells integrate into the dentate gyrus; in slice preparations one can record from individual cells and show that GFP-positive cells have in fact integrated to some degree. Gage's group has just shown that it may take up to fourteen weeks for these cells to fully integrate and put out complete dendritic arbours- so the functional importance of making new neurones clearly has to be some time distant from time of birth.This is particularly important because the number of neurones in the hippocampus is able to be modulated by various external forces. Stress, exercise, depression- all these factors have been shown to have an effect on neurogenesis. A number of particularly interesting findings emerged from a series of experiments published a couple of years ago, in which mice were placed in either a normal cage, an enriched environment (i.e. a few black tubes were added to the cage, which doesn't appear to be all that enriched to me, but apparently a mouse finds it that way; I've tried to redecorate my house this way but my wife wasn't so keen), or put in a cage with a running wheel so that they were free to run (and it turns out mice like to run a lot- miles in fact). The first surprising finding was that the most neurones, based on BrDU labelling, were produced under running conditions; that is, exercise seemed to provide the greatest stimulus for making new neurones in the hippocampus. Unfortunately, at least half of the newly generated neurons died within the first few weeks, although the final total number was still more than in the controls. The more interesting finding, however, was that, although no more neurones were generated in the enriched environment vs controls, a greater percentage of these neurones survived than in the other groups. So again the simplistic hypothesis would be that if you are making new neurones most of them will die except for the ones that happen to be stimulated by appropriate environmental inputs i.e. electrophysiological input. So environmental input is clearly a major way in which we can control the production of neurones in the hippocampus. I guess one way of self medicating would be to do a long-distance run and then solve the cryptic crossword- preferably before you cool down- and hope that this preserves those cells being generated. The significance of these newly formed cells to hippocampal function is still unclear; it certainly doesn't mean that they are recording memory since, as I said, they take many months to fully integrate. The possibility also exists that their integration clears the network of short-term memory and allows new memories to be formed in the hippocampus -it's going to be very exciting over the next two or three years to work out what's actually going on.
Thus it seems that there are cells in your and my brain that are capable of producing new neurones in quite significant numbers. The question is can we somehow use this capacity to address clinical problems- including issues such as Chris talked about this morning in relation to the loss of cells in the grey matter of schizophrenic patients. We think the problem in most cases is that the disease process, possibly inflammation, prevents this. The other thing that happens in disease is that the precursor cell is redirected to make astrocytes, rather than neurones, so the 64 billion dollar question (and its pharmaceutical response) is whether we can stimulate these cells to give rise to new neurones. The proof-of-principle for this has come from Jeff Macklis's group at Harvard, in a very elegant study involving retrograde labelling of neurones with laser-sensitive nanobeads that release toxic substances when activated by the laser, killing the neurone. Essentially this process leads to a sterile death with little inflammation. Under these conditions, cells that normally migrate down the rostral migratory stream into the olfactory bulb start to migrate into those areas where cells have been lost. Here they integrate and put out processes that reconnect with the thalamus; in other words, they form functional processes. In some recent unpublished work, Jeff has shown that this also occurs after ablation of layer V corticospinal neurones- the neurones that send axons all the way from the cortex down the spine to hook-up with motor neurones. He has found that in the adult brain the same thing happens following such an ablation: new neurones integrate, putting out processes through the midbrain and hindbrain into the spinal cord. So the extent to which one can take advantage of this system in the adult seems to be unlimited, although the number of neurones replaced and the functional outcomes of this are still unknown. So that's basically where we stand at the moment. It is certainly true that you and I have the ability to make new cells- we're doing it all the time- but can we take advantage of this by driving precursors in the right direction? We've spent five or six years trying to answer this question. What we wanted to know first is what the precursor cell is actually like- what's on its surface- so that we can start thinking about how we might drive the production of neurones. To do this we use a fluorescence-activated cell sorter, such as the one we have just set up in Queensland. In essence, a stream of cells is passed through an electromagnet and deviates to left or right, depending on whether the cells fluoresce in a laser beam. This allows us to collect high purity populations, cells that are labelled with marker x or y, and then look at their ability to form neurones. A graduate student, Rod Rietze, found two markers- markers that aren't related to function- in a very empirical experiment which involved testing several thousand markers before he found the right combination and was able to sort the cells to high purity. In fact, as you can see from this slide, approximately 1 in every 1.28 cells gave rise to one of those huge neurospheres I mentioned earlier, indicating stem cell activity and that this population constituted the majority of the stem cells one could get out of a mouse. This wasn't a minor population; it constituted the major population from the subventricular zone of the mouse which allowed us to start looking at what the cell was like and how we might be able to manipulate it. What we found is that it didn't have any of the markers associated with mature neuronal types and was not like any stem cell found in other tissues. So does this cell that we have isolated give rise to new neurones in vivo, or is it just an artefact of the in vitro system? Well, we are lucky to have a mouse that has a mutation in which the olfactory bulb doesn't grow after birth compared with the wild type animal. When we looked at these mice in terms of their stem cell populations we found a five fold deficiency in the identified population compared with the wild type, so it appears that an insufficiency of stem cell activity results in fewer migratory cells that go into the olfactory bulb. Thus, based on our data, it does seem that the stem cell we have isolated is the functional cell in vivo. Now, can we find out exactly what's going on in terms of regulating this population and making new nerve cells? Well, as Assen talked about, there are certainly exciting ways of addressing this. We have chosen the gene microarray technique, and have isolated approximately 20,000 cells from about 1000 mice- there are not too many stem cells per mouse. One of the real problems of microarray is that you generate so much data and, if one doesn't have purified populations to compare, the contamination leads to such a distortion of the results that it really wastes everyone's time. However we have been able to generate a profile of gene expression of the stem cell. For example, here is a profile readout of transcription factors and here is the stem cell population compared with other cell populations in the brain: red indicates very high levels of a transcription factor that is expressed exclusively by the stem cell and not by other cell types. There are approximately 122 of these genes that are preferentially expressed by stem cells. However, only 12 of them are surface receptors and these are the receptors we are currently looking at to see if we can use their cognate ligand to regulate neuronal production. So let me talk just briefly about two of the receptors we think are important in terms of stem cell regulation. Kaylene Young, a graduate student in the lab, has found that the receptor for neurotrophins like NGF, BDNF and NT3, the p75 neurotrophin receptor, is expressed on a small population of cells in the lateral ventricle, and that, using the same sorting technology, approximately 0.4% of the cells (vs. 0.27% for the other markers I showed you) are p75 positive. So here's a positive marker we were able to use to sort these cells, showing that virtually all the stem cell activity occurred in the p75-positive population and that there was virtually no stem cell activity in the p75-negative population. Obviously, we are very excited about this, not only from the point of view of it being a functionally important molecule in terms of regulating stem cells but also because it seems to be the first positive marker identified.
The other thing I'll just briefly tell you about is another receptor molecule that we think is very important in regulating these cells, and this is the LIF family of receptors which signal through GP130. This receptor is found on the stem cell population and we have been looking at what might be the cognate ligand that regulates it. Moreover, we have been investigating how the signalling might actually regulate the production of neurones, given that paradoxically you can get different outcomes of stimulation, depending on cell type. The molecule is regulated through Janus kinases and STAT, which in turn upregulate a molecule called suppressor of cytokine signalling (SOCS), a type of negative feedback molecule that turns off the phosphorylation of STAT and, in this way, is able to turn off signalling through cytokines. What we started to look at is whether the reason one could get different outcomes in terms of stem cell stimulation was dependent on the way that this downstream signalling was handled. It turns out that neurones have one form of suppressor of cytokine signalling, SOCS-2, which is upregulated by LIF, whereas astrocytes have the other two canonical forms of SOCS. It therefore appears that there is lineage-specific expression of the downstream regulator which may affect outcomes. We tested this hypothesis in neurospheres and showed that, if one knocks-out SOCS-2 from these populations, you get about a 60% reduction in the number of neurones. Conversely, if you overexpress SOCS-2 you get a two to three fold increase in the number of neurones, so SOCS-2 does appear to be vital in terms of the lineage direction of stem cell differentiation down the neuronal pathway. However, the story is much more complicated than that. SOCS-2 doesn't actually work by regulating LIF signalling; it works by regulating another STAT-signalling molecule and, surprisingly, the molecule that it regulates is in fact growth hormone, something we never expected. We thought growth hormone would have very little to do with neurogenesis or neuronal regulation, but it turns out that growth hormone is in fact a very strong inhibitor of neurogenesis- so anyone taking growth hormone for personal fitness reasons, you may be feeling better but new brain cells aren't being made. So growth hormone is a very strong regulator, and it turns out that when you add growth hormone to precursors it acts by suppressing the expression of neurogenin, a transcription factor which is vital for making neurones. If you don't have SOCS-2, you get virtually no neurogenin. You can overcome the inhibitory effect of growth hormone by adding SOCS-2 back in, so it's clear that SOCS-2 is regulating growth hormone's ability to suppress the production of neurones. More interestingly, another paper came out about six months ago showing that another STAT-signalling molecule which also has something to do with neurogenesis is prolactin. This paper showed that pregnant females were making more neurones. It also showed that sex increases the number of neurones being made, which is interesting. We're not there yet, but I think we are getting closer to being able to understand what might drive cellular plasticity, and I really believe that we will have new pharmaceutical-based therapeutics which will address many brain diseases by promoting plasticity and repair. Now, whether we can use some of these things to promote plasticity to restore brain function in mental illness is something that we hope to be able to explore in collaboration with many of you here today, and I thank you for providing this forum in which to discuss such issues.
I think it is time to invite the next speaker, Mike Calford from the School of Biomedical Sciences, University of Newcastle.
© 2004, Brain and Mind Australia Inc. - Copyright Notice -