Swammerdam Institute for Life Sciences

Focus on research: theoretical physicist and neuroscientist Francesco Battaglia

The four walls are draped in black ceiling-to-floor curtains, and Francesco Battaglia makes sure the door is covered too once we have sealed ourselves in the room. The laboratory is now pitch black. During experiments, he explains, even the lights are switched off. Battaglia studies the behaviour of rats and mice: nocturnal animals.

Italian-born Battaglia was originally trained as a theoretical physicist. During his Master's studies and doctoral research he already focused on applied biology, specialising in the modelling of neural systems. He first conducted behavioural experiments on real animals during a post doc in the US. Today, as a university lecturer at the Swammerdam Institute for Life Sciences, Battaglia still combines this research with his theoretical background.

Despite the fact that he is not a biologist, Battaglia explains, he is fascinated by the brains of living creatures. They contain an infinite amount of fascinating characteristics, to which he eagerly applies his mathematical skills. He is especially intrigued by the workings of the memory. ‘How does the brain process events? How does it record them?'

Memories in the brain

A memory, he explains, is like a trail in our brain. ‘It's a dynamic trail that develops over time. It's not like a computer that stores information in the form of static ones and zeros.' These memory trails consist of connections between cerebral neurons, which conduct electrical impulses. The strength of these connections may differ. However, a memory involves far more than two neurons: it involves an entire mass of intricate neural connections that can send feedback from one neuron to another. Battaglia: ‘You need a trigger in order to retrieve a memory: some minor part of the event, like a smell or some visual element. Memories are also interconnected. A trigger can activate all sorts of memory trails in your brain, which then trigger other memories. For example, when I look at you, I think about our appointment, which makes me think about my agenda, which reminds me of someone else I'll be meeting later today.'

Battaglia studies communication within and between two areas of the brain: the hippocampus and the cerebral cortex. As we now know, the former is involved in the process of creating new memories - surgeons in the 1950s would remove this and other parts of the brain from patients suffering from epilepsy. After surgery, the patients still had their old memories, but were unable to remember anything new. Our long-term memories are stored in the cerebral cortex. ‘Once a memory has been transferred to the cerebral cortex, we no longer need the hippocampus to recall it.'

The darkened laboratory is just one of the resources at Battaglia's disposal as he works to discover how these two parts of the brain communicate. He has placed a Plexiglas rat maze on a table in the centre of the room. Cameras and wires hang down from the ceiling above. The wires, Battaglia explains, are connected to electrodes. These electrodes can be connected to the inside and outside of a laboratory animal's brain in order to make EEGs of its hippocampus and cortex. Other electrodes can be connected directly to neurons inside the brain. Single electrodes are connected to a group of five or six neurons until approximately two hundred neurons have been linked up to the computer. The electrodes immediately measure electrical communications between the individual neurons. When combined with images from the camera, the results offer detailed insight into the genesis of memories.

Rats in a maze

Click for full size picture.<br> Test arrangement learning experiment: 'The animals are placed in a corridor that leads to an intersection where they can choose to go left or right....the test is always conducted on the basis of a system: the food is always on the left when the lights are off. Then, he changes the rule.'

As part of the test set-up, Battaglia has the rats carry out a basic learning test. The animals are placed in a corridor that leads to an intersection where they can choose to go left or right. Battaglia and his colleagues place morsels of food in either the left or right corridor. In some tests, they will switch on the light for a few seconds, in others the rats must negotiate the maze in darkness. However, the test is always conducted on the basis of a system: the food is always on the left when the lights are off. Then, he changes the rule: ‘At first, the rats become frustrated, and climb the walls, but they must learn to be flexible. After a few more tests, they will have grasped the new situation and know how to find the right corridor effortlessly.'

Battaglia shows us EEG scans of the tests: waveform graphs representing electrical activity in the hippocampus and cortex. The rhythm of the peaks and troughs between the two brain regions is usually inconsistent, he explains. ‘As soon as the rat approaches the intersection where it must make a decision, however, they become coherent. Like the different instruments in a band: the tempo is suddenly synchronised.'

However: this only occurs after the rats have learned the rule for finding food. Battaglia: ‘This means the correlation between the different patterns must be related to learning. If the animal is expecting a reward, we see a strong correlation between the patterns in the cortex and hippocampus.'

Click for full size figure.<br> The convergence of the patterns is related to learning.

Battaglia also analyses the neuron data on this crucial decision-making moment. He shows us an overview of active and inactive neurons that looks somewhat like a barcode. ‘I've placed them in different groups, and we can see that all the neurons in a specific group - divided over the cortex and hippocampus - emit a pulse at a specific moment. This correlates with the moment at which the rats arrive at the intersection, and it only occurs in rats that have learned the rule.'

In order to explain his findings, Battaglia refers to a study conducted by French researchers. In an experiment with rats, they injected dopamine - a substance that alters the way in which neurons are activated - into highly specific parts of the brain. The French observed the same patterns Battaglia had identified at the intersection. ‘Like my experiment, we can see the dopamine signalling a reward. The animal expects a reward, and that information is important enough to remember. We believe this process is the basis for our ability to form memories.'

Recording memories

Battaglia suspects such memories mature further during sleep. He studied the neuronal activity of sleeping rats and found the test animals showed major peaks in electric activity after having learned the reward rule. ‘The activity is spontaneously reactivated and ‘replayed' in the brain. This process probably helps to record the memory trails in the cortex.'

However, these assumptions are yet to be proven: our knowledge of memory is still rudimentary, Battaglia explains. We will need advanced techniques to truly understand what happens in the brain as we form new memories. Battaglia aims to apply a larger number of electrodes in future in order to gather more data. He also hopes to apply optogenetics, whereby laboratory animals are genetically modified so that activated channels of the brain emit a light pulse. The use of optic fibres would then allow us to gather large quantities of data.

His background as a theoretical physicist will certainly come in handy, he explains. ‘The analysis phase is crucial: the study will require a great deal of calculation. A single trial session yields a gigabyte of data. I spend two months doing lab work, and the rest of the year sitting behind a computer. I have data from ten years ago that's still waiting to be analysed.'

However, the many calculations haven't put him off. ‘It's a dynamic system, and there's still so much we don't understand when it comes to memory. In twenty or thirty years, we may be starting to grasp how information is encoded and recorded in the brain.'

Published by  Faculty of Science

21 August 2012