Aim 1: Cerebellar contribution to cognition and autism
We trained mice to perform a complex working memory decision-making task. Using mice allowed us to manipulate activity in specific brain regions as well as perform detailed high-quality recordings of brain activity which are not feasible in humans, thus getting a better insight into the exact neuronal mechanisms during specific behaviours. Suprisingly, when we specifically perturbed neuronal activity in the cerebellum during this cognitive task, either by using a cerebellum-specific mouse model of autism spectrum disorder or by using optogenetic activation of the cerebellum, we found enhanced function. While most animal models of disease lead to impairments, we discovered improved learning of the evidence accumulation decision task and altered activity in forebrain areas. We propose that the cerebellum regulates sensory reactivity at a brainwide level to regulate task persistence and learning. The results from this aim have now been published as a pre-print (Oostland et al., BioRxiv, 2021), and presented in a short video format which was prepared for the Society for Neuroscience 2021 meeting which is now available online (
https://scholar.princeton.edu/sites/default/files/oostland_sfn_2021_withcc.mp4(öffnet in neuem Fenster)). In addition, we published a book chapter about the contributions of the cerebellum in decision-making (Deverett & Oostland, 2023).
Aim 2: Automated cell-type classification in the cerebellum
This part of the project turned into an international collaboration, as multiple labs across the world had individual efforts with unique approaches to reach the same goal: developing an automated cell type identification from Neuropixels recordings from the cerebellar cortex. Thus, we teamed up into the C4 collective: the Cerebellum Cell type Classification Collaboration (C4). The specific contribution of this fellowship to C4 was the development of a ground-truth dataset with characteristics from five different cell types acquired using Neuropixels recordings in mice. To acquire this ground-truth dataset, we further developed the optotagging method and performed numerous recordings. Together, C4 presented four linked posters at the Society for Neuroscience (SfN) annual meeting in 2022. We are now developing this project into one manuscript from all labs combined. This combined approach will yield the best result for an open-source classifier to be used by the whole cerebellum community, not limited to specifics of one individual lab, but useable across a wide range of methods, approaches, and species.
Aim 3: Kin-avoidance in cannibalistic homicides
The results from this aim have now been published in a peer-reviewed paper (Oostland & Brecht, Frontiers in Psychology, 2020) and more information for the general public about this topic as well as the full data set is now available on www.cannibalismresearch.org for anyone to browse. To investigate whether kin recognition is still in place in offenders of cannibalistic homicides, we generated a unique data set of information about 121 cannibals with approximately 631 victims, operating worldwide since 1900. We found that cannibalistic homicides are a distinct category of homicides with a unique pattern of murder methods, offenders, and victims, and only rarely ate kin. The preserved kin-recognition and kin-protection in offenders of cannibalistic homicides is not unlike the behaviour of cannibalistic fish or tadpoles, and points to anti-kin-ingestion mechanisms evolved from kin-selection many hundred millions of years ago.
