Monday, 24 December 2012

The Neuroscience of Navigation - Christmas Symposium Review

by Jonathan Smith

On the 19th December in the Royal Society, the British Neuroscience Association (BNA) held a special Christmas symposium on the subject of the neuroscience of navigation, featuring topics ranging from ants and bird flocks to computer simulations for rodents! After these exciting talks, a concluding session of wine and mince pies went down a treat!

Not being overly familiar with the area around Pall Mall, I was forced to put my own neglected navigation skills to the test in order to arrive at the prestigious venue, the Royal Society, in time for the introduction by Professor David Nutt who is the current president of the BNA. In the introduction, he explained some of the background of research into navigation and outlined some of the latest developments that were being made by researchers. These covered a wide range of life, embarking from more basic organisms like the ant, passing through flocks of birds, crossing the development of navigation in mammals, traversing the fields of mammalian cognitive maps and arriving at the age-related changes in human navigation. Here, I try to summarise some of the fascinating presentations which deserve much more than a single article to review!

The first speaker, Dr Paul Graham from the University of Sussex, talked about the humble ant. Ants need to find food. They also must know where their own nest is in order to transport the food back home. But how do they remember where it is? In a series of experiments on the Australian desert ant, Dr Graham’s team worked out that these ants use visual panoramic cues to encode the locations of the nest and of food sources. Not only that, but they quickly set up a route between the two locations that becomes hard-wired and idiosyncratic, just like a human travelling the same route to work and back. That way, it seems that ants do not have a cognitive map of the area around the nest, but instead store information of food sources in relation to familiar cues (e.g. the location of the nest). It is thought that this system could even be the origin of our spatial cognition!

Dr Laura Biro from the University of Oxford presented her research into flock navigation. The research began with studies of individual pigeons establishing routes and expanded into simulating the flight paths of over 10 pigeons! How do these flocks decide which route to take? Firstly, individual pigeons develop idiosyncratic routes, similar to those in ants, that are based on visual landmarks. If you train two individuals with different routes and release them as a travelling pair, the results vary from either bird compromising its own route to them falling out and going their separate ways! Clearly there are complex leadership issues at work here. In flocks of more than two pigeons, there is a definite leader whose route is followed by the rest. This leader is not always at the top of the pecking order in social issues, but may possibly be the most efficient navigator of the flock.

But how do the flock decide who is the leader? This is a complex decision-making process that Biro et al have made strides in simulating. It may be that there is a hierarchy of each pigeon asserting its dominance over another in a similar fashion to winning Wimbledon - the champion proves that he plays better than the runner up and all of the runner up’s previous opponents. There is still much work to be done. Biro et al are currently working on good simulations for flocks of thousands such as those of starlings that form incredible shapes in the sky!

We then moved on to mammals. Next to present was Dr Emma Wood from the University of Edinburgh who dealt with the subject of encoding an intended destination into a memory. Firstly, mammals have neurons that fire only when the organism is in a specific location in an environment. These are called “place cells” and it is thought that these help to encode our location in space. There are also many types of these place cells e.g. some that fire at a boundary and others that fire when the mammals are travelling to an intended destination, called goal-dependent place cells.

Wood et al found through many behavioural experiments that these goal-dependent place cells were more active when the animal (in this case, a rat) was strategically planning to run to an area containing a reward. Additionally, through further experiments, they found that instead of mainly encoding the location of the destination, the goal-dependent place cells principally encoded the route to the destination. From this, a pattern is emerging that remembering routes is easier for an organism than just remembering locations and recalculating the route every time!

Dr Francesca Cacucci from University College London then talked about her research into the development of spatial cognition in rats. After birth, a rat takes roughly three weeks to develop skills needed for exploration of its surroundings. Interestingly, at approximately 19 days after birth, rats shift from being couch potatoes to intrepid explorers practically overnight! Cacucci et al think that somewhere in this transition the capacity for encoding spatial maps is developed. Rats are able to perform spatial memory tasks after around 20 days of age. This is largely dependent on a brain region called the hippocampus, so the implication is that the hippocampus is sufficiently developed to encode spatial maps. As navigation and memory uses many other navigational functions such as orientation, distance and boundaries, place cells (encoding the rat’s current location) in the hippocampus must be connected with orientational cells called head direction cells, map-encoding cells called grid cells and boundary-encoding cells called boundary vector cells. Work by Cacucci et al shows that these are not all connected before 20 days of age and could be the reason why spatial memory and exploration only kicks in at this age.

Dr Carlo De Lillo from the University of Leicester made a presentation based on searching systems in space. This was based on the concept that when making an efficient search of objects in an area, you can either form a structured way of searching them e.g. from left to right, depend on simply remembering which objects you have already visited or a bit of both. Experiments investigating how efficiently different species searched a set of objects in a room found that in comparison with rats, four-year-old children and capuchin monkeys made the most structured searches. Other experiments by De Lillo et al showed that humans in fact use structured searching as a complement to memory retention much more than other species. Put another way, it is like someone is searching through boxes and doesn’t have to remember exactly which boxes he has searched because he is working through them from left to right. This research could lead to many new tests of memory and executive function that could help in the diagnosis of conditions such as dementia and schizophrenia.

Next to present was Dr Jan Wiener from the University of Bournemouth. His research consisted of giving human participants the task of navigating a virtual maze with a set route and then retracing and rejoining the route from unfamiliar directions. Egocentric navigation is the strategy of recalling a route and exploring until your surroundings resemble this route. On the other hand, allocentric navigation involves being able to consider the spatial map independently from your own location and is needed for retracing a route and rejoining it. Experiments by Wiener et al on younger and older participants indicate that as we age, we become less able to use allocentric navigation than younger people and use egocentric navigation more and more, even when it fails to help us navigate well. Don’t panic just yet though - this ageing effect can be reduced by regular training!

The final speaker that day was Professor John O’Keefe from University College London who discovered hippocampal place cells in 1971 and was a member of the BNA in its infancy as a pub meet up! He outlined his most recent research into virtual reality for rodents. In order to generate a virtual reality for mice, his group set up a floating ball on which the animal is placed. Two screens project a scene which the mouse can move in while walking on this ball, similar to a hamster ball except with a virtual backdrop! O’Keefe et al found that a virtual scene resulted in similar amounts of place cell activity to an actual environment. Additionally, the team found that passively moving the mouse through the scene resulted in much less place cell activity, suggesting that place cell activity is largely based on active movement through a location. O’Keefe hopes now to expand the experiments to record hundreds of place cells in future as a better measure of encoding location.

Overall, the day was fascinating to attend and the discussions afterwards were very engaging. I benefited from the presentations as a way of glimpsing the world of neuroscience outside of my own studies to get the bigger picture of our progress. Oh, and the mince pies and wine afterwards didn't hurt either!