Learning and Memory: How memories are encoded in the brain of the fruit fly

When thinking of tiny fruit flies, one doesn’t usually have their brainpower in mind. But even these small insects, like all animals, can learn behaviours in response to different stimuli.
 
Prof André Fiala studies the learning behaviour of fruit flies, aiming to dig deeper into the computational principles underlying the encoding of learned information. 

Find more from Prof Fiala at his University of Goettingen website, and follow him on Twitter. 

Read the original paper this episode is based on : https://doi.org/10.1016/j.neuron.2020.03.010

Transcript

Hello and welcome to Research Pod. Thank you for listening and joining us today. In this episode we will be looking at how nervous systems can learn and form memories, and how this is studied using the brain of fruit flies, a topic investigated by André Fiala at the University of Göttingen in Germany. 

When thinking of tiny fruit flies, one doesn’t usually have their ability to learn in mind. But even these small insects, like all animals, can adjust their behavior according to experience, which is called learning. And they can also memorize their experience. For example, fruit flies can learn that an odour they have been exposed to is followed by a rewarding sugar stimulus. In the future, they will approach that smell. Conversely, they can learn that an odour precedes a negative experience and will avoid it. We all know this type of learning from our own experience. Strong smells have always had the power to evoke vivid memory recall. For some, the smell of vodka can bring a shudder-inducing memory of a night spent sleeping on the bathroom floor after one too many. For others, cinnamon can bring back childhood memories of Christmas baking. These strong associations are a form of learning and memory that is essential for our survival. They teach us which sensory stimuli have been good to us in the past, and which have not. In a rudimentary sense, this would have aided our survival as we would have learnt to avoid eating the sweet-smelling poisonous berries again.  

Of course, the brains of flies and humans are quite different in size and complexity. Can one actually compare the machinery that underlies learning in humans with that of flies? Let’s look at how smells are actually encoded by the human brain. The processing of odours in our brain occurs in much the same way as the processing of other sensory stimuli. The brain is mainly composed of neurons that communicate with each other via electrical pulses called action potentials. Action potentials cause the release of transmitter substances at the synapses of neurons. These synaptic connections determine which neurons act together as neuronal circuits. Information about sensory stimuli - like smell - is transmitted in a hierarchical manner between circuits within different brain regions. The further up you go, the better the neurons and circuits are able to encode and extract important information. For instance, in the human brain, odour information is detected by sensory neurons in the nose and transmitted to a first processing center, the olfactory bulb of the brain. From there, the information is directed via mitral cells to higher brain regions such as parts of the cerebral cortex.  

Interestingly, there are very similar neuronal connections and circuits in the much smaller brains of insects. In Drosophila melanogaster, or the fruit fly, odour information is first passed from sensory neurons on their antennae and mouthparts to olfactory projection neurons in the antennal lobe, a first processing center similar to our olfactory bulb. From there, information is transduced to so-called Kenyon cells in the mushroom bodies. The mushroom bodies are higher-order structures of arthropod brains that integrate incoming sensory information with positive or negative experiences, such as rewards or punishments. Interestingly, the transmitter substance that informs the neuronal circuits about rewarding or punishing experiences is the same in the insect brain and the mammalian brain - namely dopamine. Decades of research have revealed that in the insect brain, odours evoke activity in small groups of Kenyon cells of the mushroom body. One could say that the exact group or “pattern” of Kenyon cells that is activated tells the brain which odour is smelled. This principle is very similar to how odours are encoded in parts of the cortex of the mammalian brain, but at a much smaller size. If an odour is smelled and at the same time dopamine is released to indicate a reward or a punishment, a memory for that odour is formed.  

As a consequence, the pattern of activated cells will be modified if the odour has been learned as positive or negative, and this modification represents a so-called memory trace. It is believed that reactivation of the pattern that can occur in the future, for example by smelling an odour that has previously been learned as positive or negative, informs the learner such that they react in an appropriate way. It is the plasticity of synapses that underlies the formation of such memories: during learning, physical changes at the implicated neuronal synapses occur, such that the activity patterns of neurons are changed to represent a memory trace. This induces the appropriate behavior if the same stimulus is perceived again at a later date.  

However, learning and memory are not simple, and studying what exactly changes in the neurons to form a memory trace is tricky. It turns out that the myriads of neurons and synapses involved are hard to follow and visualise experimentally. Therefore, the fruit fly represents an excellent tool for investigating the principles of learning and memory. As an organism, their brains are complex enough to extract information meaningful to neuroscience, while at the same time representing a system that is simple enough to be experimentally measured. The fruit fly also has a manipulatable genome allowing complex experimental techniques to be developed. In this way, the fruit fly can be used to address this key issue in learning and memory research. 

Professor André Fiala and his colleagues at the University of Göttingen, Germany, have utilised the insect brain to enhance our knowledge of learning and memory. The group have taken advantage of the fruit fly to generate a genetically altered model system to study how synapses change in the course of learning. To put it simply, they have generated flies that express a green fluorescence calcium indicator in a few Kenyon cells of the mushroom body. This calcium indicator is valuable because calcium levels increase when neurons are active. Monitoring the indicator’s fluorescence using high-resolution microscopy, also called calcium imaging, allows researchers to have a readout of neuronal activity that can be used to infer functional properties. Prof Fiala and his colleagues were able to open the heads of the tiny fruit flies and monitor the activity of single Kenyon cell synaptic boutons in the brains of the living fruit flies in real time.  

Synaptic boutons are small, specialised structures of some neurons that contain synapses. This brain structure is well-characterised as the site of associative learning. Importantly, the synaptic boutons in the brains of living flies were fluorescently imaged while the flies were subjected to training called olfactory conditioning. The flies were presented with a particular odour, a conditioned stimulus, and an electric shock as a negative signal, an unconditioned stimulus. Typically, the flies learn to avoid this odour in the future. The group wanted to understand how these cells and synaptic boutons are responsive to odours, and how exactly their responses are modified to form memory traces.  

The research team was able to demonstrate that when odours are presented to the flies, different synaptic boutons on the same Kenyon cells would respond. This was shown by light emission of the green, fluorescent calcium indicator from the synaptic boutons. But pairing of the odour with a negative experience resulted in a complex alteration of calcium signalling at the boutons. They changed their concerted, coherent pattern of activity and became more desynchronised in their activity. As a result, they rendered the “code” for the perceived odour unique, and more distinct from the code for other odours that have not been learned. This may be how different odour association memories are encoded as relevant. These results are significant as they indicate that the concerted and orchestrated actions of numerous, widely distributed synapses of many neurons that encode a signal such as a particular smell contribute to the formation of memories. In the future, these experiments can be progressed further to investigate the precise molecular processes underlying memory trace formation.  

That’s all for this episode – thanks for listening, and stay subscribed to Research Pod for more of the latest science. See you again soon.