 John is Professor of Cognitive Neuroscience at the Sainsbury Wellcome Centre within UCL and he's a Principal Research Fellow of the Wellcome Trust. Anybody who doesn't know him will realise when he starts speaking that he's not a native Londoner. He came to the UK from America via Canada, but he arrived in 1967 at UCL. He stayed there ever since. I find that quite remarkable for all sorts of reasons, but basically because all the career advice I was ever given was that you should move around to get on in the world in terms of a scientific career. You couldn't stay in one place all the time. Now, as I'm sure you've realised, John has focused on hippocampal formation in spatial and episodic memory and navigation. In 2014, as the culmination of this research so far, he was awarded both the Cavley Prize for Neuroscience and the Nobel Prize for Physiology or Medicine for the discovery of so-called place cells. He stayed in the same place, but he's done very well. These cells, which create an energy PS, enable us to orient ourselves, and John is now applying this research to the study of Alzheimer's dementia, which again, I guess, as we get older, is going to more and more of us with urgency. So I'm not going to say any more before he gives a lecture or read out what's written in the programme, but it's time to come to welcome John. I'm delighted to introduce him for his lecture. What rodents have taught us about spatial cognition and memory? Thank you very much, Jeremy. It's a great pleasure and an honour to be here to give this even credit lecture. If you look back through the list of previous lectures, it's actually quite daunting. And if you go no farther back than just the previous two lectures, it's very daunting. Four years ago, Clyde Page gave the lecture on the role of animal research in the development of medicines for asthma, and particularly the work of David Jack for developing cell-bure-rol and cell-bure-rol. And I have to say I particularly have been a beneficiary of that research because I spent about nine days in hospital several years ago suffering from an acute asthma attack. And when I found out that it was David Jacksworth that had been instrumental in providing the medicines, particularly that, that was quite daunting. And of course, following Colin Lever, sorry, Colin Blake, even at a distance of two years, is also quite daunting. In his lecture, Colin gave four stories about brain research and how they relied on animal research. And then, in addition, eloquently addressed some of the ethical and practical problems and issues involved in use of animals in brain research. And I hope to address some of them and at least speak to the issue of how we see those issues and how they're addressed in our own research. The title of my talk is What Roads of Tories about Spatial Cognition and Memory. Hopefully this will work. Sorry if things frozen here, so I must be doing something more. My interest in the brain structure called a campus really started with the work of Brenda Miller. Brenda Miller was already a luminary when I was at McGill in the 1960s. And she was known best for her work on patient HM and for hallucinating the psychological and particularly the memory problems that HM exhibited. HM was a severe epileptic and he was operated on by the surgeon Scovel and in fact to try to alleviate some of the disabilities from his epileptic seizures. To some extent the surgery which involved removal on both sides of the inferior temporal lobes was assessed in that his seizures were reduced and the amount of medication that he needed to take was reduced. On the other hand, he presented immediately after the operation with a severe and dense amnesiaid. He for all intents and purposes had lost his memories for the events that have happened to him everyday life. Brenda Miller, who's still going strong, she's having her 100th birthday this year, she's having several parties, one in Montreal and one in New York, spent many years trying to understand what aspects of memory were lost in HM and her best student Suzanne Corkin, who unfortunately had died recently, was also involved in that surgery. Suzanne summarised all of the work on HM to date in this very important, very readable book and what she concluded was that the problem with HM was that he cannot recall anything that relied on personal experience such as a specific Christmas gift his father had given to him. He retained only the gist of personal experience events, playing facts, but no recollection of specific episodes. So he not only could not learn new facts and store information about his experiences after the operation, but he had lost quite a few of the things that he knew before the operation. He seemed to have lost what we would now call episodium, the ability to recall what you had experienced yesterday, the day before last year. So when I started thinking about the research enterprise that I would engage in, I thought this is really a golden opportunity. We know that this part of the brain must have something to do with memory, but of course exactly what memories would look like at the cellular level was something that nobody knew about. So I decided to actually see if I could see what memories looked like by recording the neural activity of rodents, specifically rats, and see what memories looked like. Now the justification for that is that if we look at the structure that was removed in HM, the main structure was removed with the hippocaptus, nestled here inside of the temporal lobes, and here is the same structure in rodents. If we take cross sections, anatomical cross sections, through the hippocampus, either of humans or rodents or monkeys, what we see is anatomically they look very, very similar, and they seem to have many of the cells, cell types and many of the same connections. So you see here in the rodents, you see that the major cell type, the pyramidal cells, all line up in this wonderfully simple layer here and there's another layer here, and they're essentially formed two interlocking seats. And if one looks at the monkey one sees a similar kind of architecture and similar view. So at least in the first pass, one might think that these two structures in humans and rodents might have some similarity in their function. But I have to say this research was, and in some cases, in some sense, is still curiosity-difficultation. I really was interested in trying to understand what the neural basis of memory was. And as we are already, that the rat is probably not going to be a perfect model. It's not a perfect model for many functions in humans. It's certainly for the cognitive and memory capacities of humans. So the question is, what role does the hippocampus play in the life of the rat? I wasn't trying to, as it were, see the rat hippocampus as a model for human memory. I was trying to see what it does for the rat. And I believe then is that understanding this might give insights into human memory. It wasn't going to be a simple one-to-one correspondence or translation between the two. But if you knew what this system was doing for the rat and its importance in the rat's life, then we might be able to see how, with simple modifications, it might function in human memory. But probably, in this case of time, this was only going to be possible as an indirect extension of our understanding. This, of course, involves an ethical conundrum. What we would like to do if we want to understand certain aspects of the human brain and certain aspects of the human mind is we would like to find an animal which actually captures many of the aspects that we're interested in. But insofar as it captures all of them and insofar as we're looking at animals which are approximately approaching the human capacity either in cognition or emotion, then we're faced with this ethical dilemma where, of course, we then have to have a very clear understanding and attitude towards the ethical use of those animals. So it's a kind of what we would like to do, is find animals which are close enough to the human being and the human function, but not so close that we are faced with severe ethical concerns. Well, it turned out that, I think my lecture will demonstrate that that's in fact the case with the rat memory system and particularly the spatial memory system, as I'll tell you. So what we did is we placed here's again the rat woman to the campus and if you look at a cross-section to it, this is what it looks like, the separation, and we placed tiny electrical recording electrons into the hippocampus under deep surgical anesthesia and taking care of the animal's post-operative recovery. And then after the animal had returned to normal function, we looked at the activity of one or a small number of neurons in this part of the brain as the animal was going about different spatial and non-spatial behaviors. And of course we were particularly interested in memory so we gave animals a number of memory tasks. And what we found to make a long story short is shown in this slide here in the first publication of our work on the spatial functioning of the hippocampus and spatial attitudes and the correlates of hippocampal cells. What we found is that the cells weren't interested in actually what the animal was doing so much. They weren't interested in why the animal was doing it, but they were actually interested in where the animal was doing the behavior. So here's an animal wandering around on a tabletop essentially. Here's a curtain to map, to sort of to give the animal only a limited view of the rest of the laboratory and to include all the other aspects of the laboratory. And what we saw is that as the animal went around this environment visiting different places, that the cells and particularly this fellow here were essentially silent over most of the environment and only became active when you came over to this part of the environment here. You can see that here in this histogram that shows you the number of action potentials per year at a time. So if he's in A or B you get quite a bit of activity but when he's in the other areas here there's no activity. Subsequently we've refined and other people have refined the methods and experiments. We've introduced overhead cameras so that we can automatically track the animal's position by looking at and monitoring the lights on the animal's head. We can tell where the animal is as it explores an environment in this case a square environment looking for little bits of food and we can tell where the cell is fired as the animal goes around that environment. And what we see as shown here is that for example as the animal moves around this square environment as shown in this black line here the cell is quiet most of the time but only becomes active when he comes over to this southwest core. And we can represent this by a false colour map here a heat map where the red colours show how active the cell is and where it's most active and warmer colours so you can see that there's a very nice representation of this corner of the box by this type of activity. This is a cartoon that we made which shows the activity it's real activity. Here's the animal running around and when he comes here we put little red dots and you can hear the cell firing as he comes to that particular location. You can see he runs all over the box he's looking for food which is thrown in randomly and it's only when he comes here that he actually causes the cell to fire. One of the things you might notice is that the cell fires irrespective of the movement of the animal through that location. It doesn't matter whether he will be north or south or east or west telling us that this is not a simple this is not a simple cell responding to a simple visual stimulus but a cell which is constructing some notion about space and we believe that the idea of places is something that the brain constructs and it's not there are places in the outside world it's a construct of the brain which enables the animals to locate itself and to locate other objects in there. Now if we look at not one cell but many cells at the same time we see something interesting and here are the activity of these different cells what we see is that each of them has its own preferred area in the environment and if we map where in the environment each of these cells fire then we see that every place the animal goes whether it's running along the north edge of the box or the south edge of the box or the west of the east edge of the box there's one representation in one of these cells the activity moves around on the surface of the hippocampus to track where the animal is and provide a representation for every location in the environment now this is a small number of cells it's only about 32 cells there are hundreds of thousands of cells in the hippocampus of the rash and so the representation of an environment is carried out over and over again in these cells well thinking about this a chap called Le Del and myself you can see there's a picture taken from the 1970s decided that perhaps the hippocampus was actually some sort of a map that it was providing the animal with information about where it was in the familiar environment where other things were objects of desire such as food or places to avoid and that these place cells as we call them were actually acting as the anchors for locations and locations signals in that system so we said in that environment there's a set of place representations if you have a lot of place representations ABC that will provide you with the bare bones of a mapping system and they could be activated just by the information that's coming in to the animal in a particular location in the environment however, in addition to that having just a bunch of place representations does not make a map because what you need to know in addition is the spatial relationships between those place representations and if you're not enough to know that you're here or there you need to know the spatial relationships between those different locations and we thought that the way to do this was by creating a set vector representations in mathematical locations which connect the places in the map of environment together in terms of the distance and directions between them connected together by vector AB which represented the distance and direction in some framework between those two locations animals using such a map will be able to locate themselves in an environment and locate objects such as the rewards and punishments in that environment and move from one place to another by any available okay, that's a theory Theories are only so good as they're testable and in order to test it we made two very very specific statistics we predicted that animals which had damage to the hippocampus would have deficit in place learning knowing that they were in a particular location, navigation, how to go from that location to some other desired location and exploration which we thought might be the basis on which maps were constructed we thought that maps would probably be constructed not because the animal was hungry or thirsty but because it actually had an internal system for driving curiosity and which said to it said to the rest of the brain I don't really understand this environment I don't have a map of this environment why don't you explore and actually provide the information for creating that map we also, and I'll come back to this in the second part of the talk predicted that if this really were a map then there would have to be not only these place cells in the hippocampus formation either in the hippocampus itself or related areas which were providing information about these vectors about distances and directions in the environment and as I'll tell you we now have information about both of these predictions which strongly confirm those predictions sometime very soon after you were propagating you were propagating this theory Richard Morris who was then at Edinburgh came into the lab and said I think I know how to test at least one part of your prediction I think I can construct a behavioral test which will show whether the animals reply with the hippocampus no longer can navigate the way that you suggest and Richard's idea was actually to come up with what could colloquially be called a rat swimming pool very simple tub of water filled with milky water where the animal can't see below the surface and in which is placed a small platform that when the animal gets to the platform if it knows where it is and if you find it it can climb up onto the platform and then escape from the pool and it turns out that animals learned this test very quickly so one of the nice things about these spatial navigation tasks is that animals seem to know exactly what they're supposed to do they don't take a lot of training and they know very well that if they're putting the pool and there's a platform there they can go and within about eight trials in the best animals the animals can learn to go directly from where they're placed any place in the in the in the water maze to this location here and of course they must do that by using some kind of navigation by looking around the pool looking at the objects in the room and saying that's the location over there that I need to go to and we were delighted to find that animals which had damage to this part of the brain of the campus lost that ability to do that and didn't lose their ability to do most of the things in our hands I thought I would show you just a picture, a little video that Richard has put onto the web this is some flavor for this task, here is the room and there's the swimming pool and here's the milky water and here's the hidden platform and the idea is that the animal is placed down and he swims, this is an animal which is trained to find the platform and he goes and after he's on the platform he's taken off you can do all kinds of manipulations to see what's going on in his mind by for example removing the platform and starting from another location and he swims directly to the place where the platform should be but isn't and of course he's a clever rat so he swims around and around looking for the platform and of course it's a clever rat and you can show where he's going by looking at this overhead tracking system after a while he says actually this is probably a psychological experiment so they probably need the platform and he starts swimming off the proof of the platform in that area but it gives you a very strong question that he knows where that platform is and he knows the direction of that platform okay well Richard's work sits in a long series of experimental work on rats in their way around mazes and this work started with Willard Small in the Clark University lab at the turn of the century 1901 and Clark and his colleagues were interested in trying to understand the workings of the rat line did they have a mind and what would they use it for and he was advised by his thesis advisor to find a maze and see how rats loaded the maze and then he took, they have to quote maze as a model and then of course here's the maze and you can see people wandering around it I don't know how many people have actually experienced it it's quite a difficult task because there are all kinds of blind alleys index and what he did was he essentially used the layout but regularized it he linearized it and studied how rats found their way well that gave rise to a large amount of rat running literature much of it in the united states much of it on either the east coast or the west coast in which people tried to modify mazes to see if they could actually understand what it was that the animal was doing and how it was solving the maze so this is one of the mazes this is the stone's multiple teammates and what stone realizes that the best way of assessing what the animal is doing is by giving it a series of left and right choices so you start the animal at the entrance here and he runs up to the first choice point and then he has a choice between going left or right if he makes the correct move he goes here and then he's given another essentially teammate's choice and so on and so forth but what this means is that the animal can make mistakes but he doesn't have to pay the price later on for making mistakes earlier on in the maze if he goes wrong here then he still can go back and get this from writing so on and so forth out of a lot of experimentation like this although many psychologists originally thought that the rat was a very simple learning machine a stimulus response learning machine so what he was doing was going up to the choice point and saying go left, going to this will say go right and he was essentially learning a series of left, right turns they began to see evidence that the animal was doing something much more sophisticated this by the way is just a what maze looks like from the animal's point of view and what they noticed is that if you look at the mistakes of the animal it became pretty clear that they were not just learning stimulus response associations they were doing some more cognitive processing that couldn't be reduced to just simple stimulus response associations and one of the best people who had studied these kinds of changes was a man called the shield and so for example if they looked at the errors that an animal made or the animals tended to make on this maze what they found is that they tended to make this choice very, they found this choice very very difficult or they found this choice very very difficult or they found choices that are very far away from the goal very difficult so thinking about that and analysing what the mistakes were they realised has shown that there's a considerable evidence that there were a principle of learning in the maze of situations that blinds are more difficult to find as they are away from the goal not all choices were equivalent but importantly other factors that appear to be important in determining the order of difficulty of blind alleys are the absolute direction of the animal from and the choice from the goal the animals seemed to have some knowledge which couldn't just be reduced to simple behaviourist associations it seemed to know where the goal was in particular the direction of the goal thinking about those that whole long history thinking about the success of the water maze and I have to say the water maze is probably the most successful piece of behavioural apparatus that's ever been invented those two original publications from Richard Mars have 15,000 citations if you go into any drug company or you go into any university of Barotry you'll find the water maze it's a very simple piece of apparatus but it is very elegant in its conceptual concept that it's firing and also in terms of what it tells us about the animal thinking about that we thought it would be nice if we could meld the positive aspects of the water maze one of the standard mazes that have been used historically the stones maze for example and what we came up with was a version of the water maze which is a land based water maze in which in this case consists of 37 individually movable platforms which can all be up at one time if we so wish or where only one or several can be up at the same time and what this enables us to do is to give the animal a water maze-like task ask it to do what it has to do with the water maze which is to find some particular location in the environment but where we can control the choices that are given to it at any given point so for example if the animal wants to go over here we can actually if it is sitting on this platform we can give it these two choices or as shown here where one of the choices if the animal is on the blue platform one of the choices actually represents the direction to the goal and another is a very poor choice from that point of view but we can also ask it to go to the goal and take one of two choices nine of which are very good so can the animal actually compute which is the best path to the goal even though neither of them are perfect let me show you what it looks like here is an animal sitting on the platform he's been trained to go to a goal down here and you'll see what he does and I think what's nice about this maze is you can actually almost read what's going on in his mind you can see what he would like to do you can see the knowledge that he has about where the goal is and even though we are actually forcing him to go somewhere else he's telling us by his behaviour that he understands the direction to the goal so the goal is down here and what you'll see is he's presented with a series of choices a sequence of choices which eventually get into the goal but not always by the move that he would actually prefer and here he is and he knows the goal is here you can see he's looking in that direction and then eventually he said and he knows that he's going to be given several choices so he eventually turns around and looks at this is 135 degrees from the direction of the goal and this is 180 degrees so this is in fact the right way to go actually the right thing to do would be if you could just stay where he is he knows that but he reluctantly knows that he can't stay there forever we can get him to go wherever he wants if I just give him one platform and here he goes now and he chooses the correct one here because he wants to go down here there's another good choice having to recognize it there's another one here and he almost doesn't even have to wait for the platform to look before he chooses it and every once in a while he actually stops and thinks quite a bit and I think he'll almost see them and then finally he chooses the correct one and then this is the goal and at that point when he goes to the goal platform we then put food on the goal platform so when he's running the task there is no food on the goal he has to know the correct platform without actually knowing that we're okay well this is a very good task like the waterways for assessing damage in the hippocannabis normal rats learn quickly and make more mistakes when you're farther from the goal and when asked to head away from the goal so it's exactly like the historical results told us we should be finding that there is something about knowing the direction to the goal and also something about actually the farther the animal gets away from the goal the harder it is to do and hippocampal animals animals with damage from the hippocampus having paired performance and are selected by the effectors they're even worse when they have to make a choice here away from the goal hippocampal animals are removed okay well back to the theory and the theory also predicted that there should be the existence of hippocampals in the hippocampus or around it again of those signals which were telling the animal and the hippocampus where other locations were so when he's in one location what's the direction of the distance between those locations where's the information about these vectors and I'll summarize about 40 years worth of research into his resetances because in our own laboratory and many other laboratories around the world there has been intensive search for those signals here in the hippocampal place cells as we call them which are located in the hippocampus here these two interlocking C shaped cell by these structures and those cells by and large in open field environments don't care about the direction in which the animal is facing as I showed you they're happy to fire if the animal is heading north or south or east or west they're interested in the place and not directions on the other hand there is another group of cells not in the hippocampus proper but in this part of the whole hippocampal formation which are called head direction cells and were discovered by Jim Long in New York in the 19th and they're not interested in where the animal is they're interested in exactly the opposite the converse type of information they fire all over the place as shown by this but they only do some of the animals pointing in a particular direction so they're conveying the information about the direction within the wound framework that the animal is pointing and just like to play cells different cells have different preferred directions and if you manage to rotate the heading direction prefer heading direction of one of them they all rotate so they look like they're forming something like a compass but it's not a geomagnetic compass it's a compass which is which is related to the frame of reference of the laboratory that the animal finds itself in we also know there are cells which monitor the distance from the walls how far away am I from the walls so we think this is part of the sensory information which is telling one of these play cells where you're at so if the animal knows he's a certain distance from that wall and a certain distance from this landmark in this direction if you put those together you can create this notion of place just on the basis of sensory information and finally and notably the information about how far the animal is going in a particular direction information that you need to construct these vectors which relate to places together was discovered by the Moses in Trondheim in 2005 so it took that long almost 30 years to discover the last important important part of the jigsaw puzzle these are cells which have the most important part of the jigsaw puzzle these are cells which have place fields but not just one many of the most these cells have only one place field in the small standard environment that we use but a multiplicity of these fields and you can see there are several of these fields and if you look closely you can see that the architecture of the firing of these cells is such that they lay out a grid-like pattern which is why they call grid cells and this is a pattern which is highly regular and highly symmetrical and essentially the firing of these cells occurs at the peaks of Ithosulis triangles and we think those are the cells at least one of the jobs of those cells is actually to tell the animal and the rest of the system how far the animal is going as it travels in a particular direction you can imagine the animal is moving in this direction and it says I've done three drops and it's some sort of metric that we're not quite sure yet whether that's the whole story and we know that at least part of this symmetry is controlled by the actual shape of the environment that the animal finds itself in we usually use highly symmetrical environments by circles or squares and when you start changing the shape of the environment then this symmetrical pattern is distorted slightly so we still have a lot to learn about it one of the jobs we think it's doing is to provide the metric system so I've been talking for a little while now about this system in Romans and the question has always been there I won't even say the back of our mind the front of our mind what does this have to do with HM what does this have to do with human memory and we found it very difficult to think of how we would actually test human memory's spatial memory of the sort that we were testing in animals because what we're looking at is the animal's ability to navigate around large scale environments and it was only until sometime in the mid 1990s that these first person shoot them up games for finding your way around and finding monsters and finding various politically incorrect aspects of the environment that became available and they came with editors so that you could take the game apart and you could get rid of many of the aspects that were not relevant to your interest and this was done mainly by Neil Burgess who was a mathematical physicist who undertook to take apart one of these fairly large scaled games and to leave us with a very complex environment this is the environment of 70 by 70 meters here which consisted of many rooms and main thoroughfares and you can see one of the main thoroughfares here but it included things like a movie so that you could go and watch it it included places like bigger rooms and bars so it's a very complex environment which is presented on a video screen and which it takes about an hour for most people to get the feeling that they know their way around it of course what you have then is you have an environment where you can test large scale navigation for example when the person has the head of the scan and ask the question is the hippocampus or other parts of the brain involved in this type of special navigation so Neil and myself got together with Eleanor McGuire who was very clever very clever psychologist who we also are now in London and we did just that we scanned people's brains originally in PET scans more recently in FMRI scans and asked people to go from one place in the environment to another place we showed them a picture of the other place and they had to go by the best the best route that they could advise and what we found is that when you do the appropriate controls for movement and things like that there are really only two parts of the brain which are involved in this kind of behaviour there's the hippocampus here right here in the mesial part of the temporal lo here's a broader picture of it and two lateral views and another part up here in the parietal cortex which is also known to be involved in localising things in the world but really localising them not in terms of this spatial cognitive mapping framework but in terms of where they are when they're taken to the body and it's part of the way that you have to find your way around the environment by moving relative to obviously the environment and it turned out that the better these people were at navigating the more activity there was in the hippocampus so if they took a we plotted the routes that they took so if they took a very very direct route from say A to B in this environment then you got quite a bit of activity as measured by blood flow in the hippocampus and if they took more securities routes like this fellow here chaplain around the corner here and came back to the correct goal the correct location but did it by a securities route or this one went to completely a lie and never got there they had less activation in the hippocampus so there was not only evidence that the hippocampus was involved in this way finding in this navigation but the better that people were at the navigation the more active there were in the hippocampus now Eleanor is famous because she then went on to ask a question well the hippocampus really is crucial to navigation is there a group of people who are better navigators than all the rest of us and the answer of course is yes the London taxi cab drivers have spent at least two or three years learning learning their way around the the streets of London and they do so to the level that they can actually find their way from any one of the 23 25,000 streets in London to any other one and they can even do it take a little account traffic conditions and the weather so they have this extraordinary knowledge of the streets of London and their interrelationships so Eleanor said well what does the hippocampus look like and what she did was she just did structural scans of taxi cab drivers brains particularly if you are looking at the hippocampus and compare them to all the non-taxi cab drivers eventually she prepared them to bus drivers and of course that's a good comparison because taxi cab drivers have to find their way by using these circular routes and by taking them to Kendall whereas bus drivers travel around London a lot but they always go along more or less the same way and what she found was that this part of the hippocampus here showing the yellow is actually larger in taxi cab drivers of course there's no free lunch so the other part of the hippocampus is slightly smaller but it certainly is the case that if you use your hippocampus a lot to find your way around then you can expect it to grow you're not going to be a taxi cab driver it's not that you've got a big hippocampus and that these you will predispose to take a job which utilizes that ability because the longer you've been a taxi cab driver the larger the structure grows and if you stop being a taxi cab driver it relaxes back to the economical side well another strong piece of evidence that we're on the right track by using the Rathic the campus to tell us something about the Rathic the campus comes from studies in which patients who are epileptic and candidates for epilepsy surgery have implanted electrodes in the hippocampus and we're for a short period of time in addition to looking for epileptic activity one can carry out an experiment to see what the cells in the human hippocampus are doing this work is only done in a very small number of centers in this case it's a center in Pennsylvania there's another center in California and people see lots of different things I don't want to lead you astray to tell you that all they see in the hippocampus of humans is something to do with space one notable finding is there are cells in the human hippocampus which is sponsored collectively to Jennifer Aniston and you can present the name Jennifer Aniston you can present pictures of her you can present pictures of people she knows but there is certainly a number of research now where if you use a virtual reality environment and hear a couple of screen shots and hear what it looks like and ask people to find their way around in a virtual environment and monitor their activity in the hippocampus here you find that there are cells for example this one which fires here or this one which fires here selectively very much as we would have predicted so at least in so far as the cellular activity and activations in fmri and imaging studies is concerned one of the functions of the human hippocampus of the memory functions is to form some sort of a spatial happy assessment okay and in the last few minutes I want to just tell you a little bit about where we are going with this and as I said when we first started this work it was purely curiosity driven we were interested in what this part of our team did and how it operated in memory studies and what cells were doing when an animal flew in things but sometime in the middle of the 1980s it began to be obvious that the hippocampus might be a very good model of Alzheimer's disease Alzheimer's disease is actually growing in prevalence and it's people over 60 years these are already a little out of date in terms of the figures but people over 60 years old have an incidence of about 5.5% in Western Europe and 6.5% in North America in terms of coming down with Alzheimer's and we asked that question to show you what was the interest in this in the next slide can we use our understanding of the chemical function to study Alzheimer's disease in animal and human models so this is a good example of where we studied the structure to try to understand a competence and a type of memory for its own sake but when we now begin to see that this might help us in beginning to address a major disease and the reason that's the case is that if you look at the toxic proteins which are involved in Alzheimer's and particularly the tau these are tau tangles and these are animal and plant and you look at the different stages of Alzheimer's what you see is according to BRAC at the earliest stages these tangles and the tau are localized in exactly the hippocampus formage they are located in exactly the places in the brain where the the grid cells are found and that what happens then it seems as though these malignant proteins actually spread out over time and move into the hippocampus proper and then spread out from there into surrounding cortex until they invade the whole of the cortical mantle so that let us think well maybe we can use our knowledge of the spatial functions and the navigation functions of the hippocampus to do two things in the first place perhaps we can actually construct memory tasks and particularly spatial memory tasks which enable us to attack into the earliest stages of the disease and I think it's pretty well believed in the community that Alzheimer's is there for many many years prior to its clinical manifestation and the idea is could we find a task which is actually centered enough to show the earliest stages of the disease and of course if we ever do find some therapy and right now there isn't really a successful therapy then it would be nice to be able to select people who are at the earliest stages in providing with that therapy so this is one of the tasks that people in particularly Neil Burgess and other people including Dennis Chan have come up with to test spatial memory functions in normal and in patients Dennis is a dementia he's a dementia neurologist at Cambridge and Neil Burgess is the one who first came up with the virtual reality environment and it's a very simple task actually what you do is you present a picture of four mountains shown from a particular perspective and then after a delay you take that picture where you present four other pictures and you say which of these are the same mountains not necessarily seen from the same perspective and it's not an easy task because the mountains are roughly the same color there they have different shapes and you have to actually say well what's the spatial relationship between these four mountains and how would it look from different perspectives and it's this one which is actually the correct one and they validate this by actually giving this task to a series of patients and controls patients with hippocampal frank hippocampal damage so these are people where scans show that they have damage in the hippocampal and what you see is these none of the controls or the patients have a problem with simple visual perception being able to identify our pictures simple visual memory remembering that they had seen this picture before but when you start asking them to actually do this task where they have to actually identify the pictures from a different location then they have great trouble if there's a two second delay and even some of them have trouble if they're all presented at the same time and it turns out that this is a very sensitive task to what looks like the onset of AD so if you test control patients people with frank Alzheimer's the major and then two other groups people with MCI who have positive signs of biomarkers and there's a three with spinal cord and who don't and you see that there's a big difference between the controls and the AD patients and you see that that the test actually distinguishes not perfectly but quite nicely between MCI patients who have positive and negative biomarkers and this they've done lots of work but there's a good correlation between their memory of this task and say the count of all of the volume and the latest work is looking as though this task is better a better predictor of Alzheimer's than the actual hippocampal volume which is one of the standard clinical science that's used and the other thing we're doing is trying to see if we can test and specifically use spatial tests to see if we can see what's going on in rodent models mass models of Alzheimer's disease so there are many models out there transgenic models and knock-in models where you can create some but not all of the aspects of the disease this is an easier animals which had overexpressed the and what we asked is what's happening to the place cells as these animals get older what we found is that when the animals are young these are the transgenics the cells look for all intents and purposes like the cells that you see in the wild type and then as the animals get older there's not much change in the spatial place cells in the wild type but some of the transgenic animals begin to show severe signs of disruption of the place function of the spatial functions of these cells some of them are all right and some of them are showing a really severe decrease in the amount of information they are providing and there's a big correlation between the plaque birth in the hippocampus and the spatial information that you see in these animals so it looks as though there's some correlation some relationship between how much plaques you get and how much plaque birth in the animals actually activity in hippocampus cells and we also looked at spatial memories in town that was correlated to my story and I want to go back to those first points that I made right at the beginning as I said this is curious originally curiosity driven research we were trying to find out what this part of the brain did in animals and to see how it functioned in memory for space and it turned out it was a specific kind of memory a memory for spatial and that the animal could use to navigate and to locate objects in a familiar environment we now see that having learned quite a bit about what this part of the brain is doing we can use that information and begin to now engage in translational research as well as continuing to engage in curiosity driven research so the brain is not a perfect model and we see that the hippocampus provides a cognitive map which animals such as the rat can use to identify locations and to navigate and we know quite a bit about how it does it in terms of the neural architecture and what the neurons are doing we don't have perfect computational models yet which put together all of these pieces to tell us exactly how it works but it's also clear that in addition to that function in humans the hippocampus in humans has additional functions so it's clear for example that in addition to being involved in spatial memory the hippocampus in humans is also involved in episodic memory so something is at and we think that we have the bare bones of what the machine can do and that something has been passed in human beings which takes the hippocampus into other areas of function in addition to what it does in animals and you can imagine that some of these areas are things like it's not clear that rats and rodents in general have a sense of linear time it's a very highly contested area of research some people see cells in and around hippocampus which look like they're at least coding for short periods of time but if you think about what episodic memory is it's the ability to actually say not only that you've been in a particular place and had a particular experience but to do so had a particular time so you have to be able to say I did this in the past it's happened to me now maybe I'll do it in the future the position of a linear time sense might be one of the things that changes the function of the hippocampus from a purely spatial system to something more like the episodic memory system another thing of course is that animals don't have the sense of themselves they probably don't have a representation of themselves as an entity and that's of course one of the important additions so not only do you know that you have a representation environment but you know where you are in the environment and your place in the environment and lastly of course there is a function of language and you think that the left hippocampus has something to do with language you don't know what but clearly language has a whole of a layer of function which takes it beyond the realm of animal brain functions finally we thought that maybe somewhere down the line we might be able to address diseases of memory if we knew enough about how the hippocampus works and I think it's clear that this is giving us some sort of purchase on the disease by our understanding of how the hippocampus operates in normal moments will give us the ability to design powerful tests of memory in particular the spatial memory is a dysfunction of the hippocampus in these animal models and it's also going to be used and we use it as a guide to develop special tasks for the early diagnosis of the disease in humans and finally these are the funders of our research the Gatsby Foundation and the Wellcome Trust and as you heard John say I've been a UCL my whole career and we now have built a new institute for the study of the neural circuits indicator since we welcome the trust as I said we welcome the centre and I thank you for your attention Thank you John for a fantastic lecture I'm not going to say much about it because I want to say some time for people to ask questions that are more informed than me about the right questions to ask all I want to say is I found it absolutely fascinating it gave me some insight into how really careful thinking allows you to design experiments that give you insights into answers that lead you to more thinking and details that maybe elude me but I felt that's what it gave me so thank you very much and this link between finding that perhaps a rat knows where it is and how to get to it but doesn't necessarily experience you call it episodic memory I think I was thinking given the technical structure and how this kind of thing gets teased out of these experiments absolutely fascinating but I want people to ask some questions for John I want to be somebody you must remember something what he said yes I'm interested to know how you know that a rat doesn't have episodic memory that's a very good question and I some of my colleagues would disagree with me but if you think that episodic memory involves actually knowing not just experiences you've had but where you've had them and also when you've had them it's been very difficult to demonstrate that rats can do that there's some work that's been done in Cambridge in terms of terms of pros who seem to know how it's been that they buried something in a particular location but we haven't been able or nobody's been able to live in in Lawrence so you would need to have evidence so it's the absence of evidence it's not it's not proof that they don't have it it's just until someone comes up with demonstration that they do have it one can leave it as an open question given some popular stereotypes around gender differences in spatial navigation I wonder if you've any evidence of your rat work on any sex differences in these behaviours so it's been a subject of study right from the first May study by Small and people looked to see whether there were differences in males and females in learning mazes and so on and every once in a while there is some evidence I don't think there's much evidence in the Romney literature for differences in different space there's beginning to be many people are asking the same question about human spatial abilities and it's still a very much an open question and I guess one thing you have to keep in mind is that and I didn't go into this whereas the rat uses both of its hippocampuses to find its way around humans have hit upon the trick of actually using only half the hippocampus to find the way around and so there is this lateralisation of function not only in hippocampus but also in the neocortex overall and a rough a very very rough dichotomy between the two functions of the two hemispheres is that the right hand side is doing very much something like the rat is finding its way around remembering where objects are located relative to each other whereas the left hand side is more devoted to language so you could almost think about possibly that you can choose how much of your hippocampus you are going to devote to one or the other and there is some suggestion that females devote more of not only of hippocampus but the neocortex to something like cremish potions yes Is there any evidence that suggests that London-Titian cap drivers have lower levels of Alzheimer's to be found? It's a good question and I don't know the answer I'm not sure anybody has done that yet it's certainly part of the problem is we're not exactly sure what that increase in size is it almost definitely isn't an increase in the number of cells the hippocampus is one of the few places in the brain which actually generates cells throughout life and in fact involves another one but in this case it's not it's very unlikely that the increase in size is due to more cells it might be due to more contacts between cells or synergies it might be more more glian cells other kinds of non-neuronal components but it's not going to be there but I don't expect that to be a protective it's not done but it's a very good question and we don't know the answer There was a mention about another difference between rodents and climates that you made about being able to understand then cells in the space is it almost like a self a condition of cells self-awareness How would you test it if you were looking at the behaviors of rodents and how would you would you see signals in relation to this test or whether other animals are in the space with them do they be able to place other animals within their framework So it's a good question we have the beginnings of partial answer to the second part which is that a group in Israel working on bats of course bats are really like flying rodents and have many of the same cells with the same properties as rats in mice it reported it hasn't been confirmed in other laboratories they reported that there were cells in the campus which actually monitored the location of other animals so and your question is a very good one it would be very very difficult to see how you would experimentally demonstrate this self-awareness and the awareness that you are in a particular location has any of this work been done in the primates and if so there are more similarities with the activity of the primate in the campus so I think this work is a bit of an advertisement for looking at animals that are free to move around in space are not self-awareness much of the work in primates is done in animals which are very much more constrained and people are now beginning to use virtual reality to try to look at cells on campus but it's really early days and they haven't really done very much of it You may point that the computational models aren't quite up to putting all this together to make us understand how things work in your view how much of that is due to the programme that we're not having enough of certain amount of information to plug in That's a very good question and I think one of the things that's lacking at least in the moment is how the goal is represented in this special representation and how if it exists it is a theoretical construct and how the representation of the direction to the goal is represented So again going back to the work of the bats the group in Israel suggests that they have found cells which represent the heading direction to the goal So these are cells which fire whenever the animal is actually pointing to the goal We haven't seen those cells nor has anyone else and so right now part of the problem is that we have found most of the components of the system but it's still lacking at least an understanding of how you would represent how you would represent the location of the goal and the direction to the goal That's what the other organisation is I think it's just a question of a bit of rock for somebody to figure out exactly how and the minute they find that it may be that we're not looking at it in the right way and that means not seen in individual cells but in the distribution of activity across the whole environment and then it's at a network population level and not at the individual cell level So in past few weeks we've been asking a lot of discussion about the general week of shuffling in neuronal cells that they managed to change their DNA code to expand their repertoire Is that something that is not seen in individual cells or is it happening in all types of neuronal cells It's a good question and I don't know the answer We like to believe that if it's happening it's also happening in the epicampus but I don't know the answer to that I'm certain I think we should draw this to a close Can I just ask you to thank John again for that It really illustrates the fact that you've suddenly arrived at the stage talking about humans and I think if we leave we might have a reception and something to drink presently out here Thank you