Final Report Summary - NEUROBAT (Neural codes for space in complex multi-scale environments: Insights from the bat) The main objective of this ERC grant was to develop innovative neurotechnology and record ‘place-cells’ and ‘grid cells’ from the brain of bats crawling in 2D and flying in 3D, on different spatial scales – in order to shed novel light on the neural basis of spatial memory and navigation in the mammalian brain. Our work focused on the hippocampus and entorhinal cortex – brain areas that are crucial for spatial memory in animals and humans; notably, these are the brain areas that are the first to be damaged during Alzheimer’s disease. Our first step was to show the existence of small-scale, 2D ‘place cells’ and 2D 'grid cells' in the entorhinal cortex of Egyptian fruit bats. In 2011, we demonstrated for the first time grid cells in a non-rodent species – but, surprisingly, we found that these grid cells existed without the presence of continuous theta oscillations. These results provided the first causal dissociation between the two major theories of grid formation – one theory that relies on network interactions and does not require theta oscillations, whereas the other theory relies on integration of the theta oscillation – and our data provided very strong evidence against the latter theory. This study was published in NATURE (Yartsev, Witter, Ulanovsky, Nature 2011). Next, we increased the spatial scale of our experiments, and trained bats to fly back and forth along a ‘linear flight track’ using vision without sonar – in a lit environment, or using sonar without vision – in the dark. We then conducted in-flight tetrode recordings from hippocampal place cells. Many cells exhibited ‘remapping’ when switching between vision and echolocation (which reverted when switching back to vision) – suggesting different cognitive maps for different sensory modalities. This was the first demonstration of hippocampal global remapping in a non-rodent species. This study was published in NATURE NEUROSCIENCE (Geva-Sagiv et al., Nature Neurosci. 2016). Finally, we have embarked on a set of projects to examine how 3D space is represented in the mammalian brain. We developed novel wireless neural-recording technologies, which allowed us to record the activity of individual neurons from freely-flying bats – in fact, the first recordings of single neurons in any flying animal. We found that neurons in the bat hippocampus had their activity limited to a restricted 3D region of space – a 3D ‘place field’ – and in a series of experiments we have shown that the representation of 3D space by hippocampal neurons is uniform and nearly isotropic. This study was published in SCIENCE (Yartsev and Ulanovsky, Science 2013). We also found 3D head-direction cells, that act as ‘3D neural compasses’; this study was published in NATURE (Finkelstein et al., Nature 2015). We also discovered in the hippocampus of 3D-flying bats a new class of neurons which represent the direction and distance to navigational goals – a vectorial representation of spatial goals. This novel result is published in SCIENCE (Sarel et al., Science 2017). Our innovative project has thus provided a uniquely novel understanding of the brain mechanisms of 3D spatial codes in the mammalian brain.
