Work performed
We did a detailed study of vertebrate brain evolution using neuroimaging, histological data, and computational neuroanatomy methods, and creating collaborative platforms for neuroimaging. This included building a large open database of vertebrate brain MRI, with a focus on primates, and generating precise 3D brain reconstructions to extract measurements like brain volume, surface area, and folding. Using phylogenetic comparative methods, we tested evolutionary scenarios, estimated ancestral phenotypes, evolutionary changes, and multivariate relationships across brain structures. The project was a collaboration between the Group of Theoretical and Applied Neuroanatomy at Institut Pasteur, the National Natural History Museum in Paris, and the Institute of Biology at École Normale Supérieure.
Results
In this project, we have collected a large comparative dataset and studied neuroanatomical diversity across more than 150 species.
In the first part of the project, we studied the diversity and evolution of cerebellar folding across mammals which was published in eLife. We analysed histological data from 56 species, and created tools to study the geometry of cerebellar folia and to estimate the thickness of the molecular layer. We identified two distinct sets of phenotypes. The first set comprised "diverse" traits, such as body weight, brain weight, and the area and length of cerebellar and cerebral sections. These traits showed a huge variation across species, spanning multiple orders of magnitude in relation to body size. In contrast, the second set consisted of "stable" traits, including folial width and the molecular layer thickness, which showed minimal variation and changed only marginally with differences in brain size.
Phylogenetic comparative methods showed that the evolution of cerebellar and cerebral neuroanatomy across mammals follows a stabilising selection process, where phenotypes vary randomly and converge around one value. Ancestral phenotype estimations indicated that size and folding of the cerebrum and cerebellum increase and decrease concertedly through evolution. They confirmed that the common ancestor of mammals had a folded cerebrum and a correspondingly folded cerebellum, comparable to the brain of a Rock hyrax. Our results confirmed a strong correlation between the size of the cerebellum and the cerebrum, and revealed a disproportionately higher degree of folding of large cerebella compared to smaller ones. Furthermore, it showed a relationship between cerebellar fold width and molecular layer thickness. Across this large range of species, these patterns were remarkably consistent, hinting at a shared mechanism that governs the folding of the cerebrum and cerebellum in mammals that can be explained by buckling and where wavelength of folding depends on cortical thickness.
In the second phase of the project, we examined neuroanatomical variability across a wide range of primates. By collecting and analysing MRI data for 70 species – the largest sample to date – and incorporating endocast data, we studied neuroanatomical diversity in 105 primates. We created tools to segment the data and reconstruct precise 3D brain surfaces for all species, many of which have not been available before. We computed traits like cerebral volume, surface area, and folding metrics, and applied phylogenetic comparative methods to study evolutionary patterns. Results showed brain evolution in primates follows a Brownian motion model, where phenotypes vary randomly, with some branches (e.g. humans) showing brain volume increases, others (e.g. galagos) showing decreases, and mixed patterns elsewhere. The common primate ancestor, 74 million years ago, likely had a folded brain similar to an Aye-aye. Our findings highlight strong links between folding and brain volume and reveal conserved fold width across species, offering valuable insights into primate brain evolution, including humans.
Exploitation and dissemination
Two peer-reviewed articles have been published, with one featured in a press release, one submitted, and others are in progress. Additionally, four articles resulting from collaborations related to this project have been published. The results were shared through talks by the MSCA fellow at 8 conferences (4 international, 4 national) and several talks by the supervisor, as well as in a poster, a science festival, and on the project website.
