In preprints: the deep evolutionary roots of Cajal-Retzius cells

Vertebrates show a striking diversity of brain size and morphology. Mammals, in particular, have an especially large cerebral cortex, presenting with a characteristic six-layered structure that is believed to have enabled the evolutionary emergence of complex behaviors and cognitive abilities. How this peculiar organization arose remains a puzzling question.One hypothesis that was put forward is that the expression of the glycoprotein reelin by Cajal-Retzius (CR) cells in the most superficial layer of the developing cortical primordium (the dorsal pallium) contributed to mammalian-specific layerization (Causeret et al., 2021; Goffinet, 2017). From that perspective, the evolutionary emergence and amplification of CR cells can be considered as key events that contributed to the expansion and complexification of the mammalian brain. However, the extent of conservation of CR cells beyond mammals has been a matter of debate for decades, and the mechanisms by which they appeared remained elusive.Two recent preprints investigating shark (Quintana-Urzainqui et al., 2025 preprint) and salamander (Gumnit et al., 2025 preprint) provide compelling evidence that CR cells are much more ancient than anticipated. Using single-cell transcriptomics and cross-species comparisons, the preprint authors identify when CR cells emerged (at least in the last common ancestor of jawed vertebrates) and how (by divergence from an ancestral olfactory cell type). Together, these studies allow us to better understand how the evolution of developmental processes shaped brain morphogenesis in vertebrates.A defining feature of CR neurons is their transient lifespan, underpinning their exclusively developmental function. Their low abundance necessitates large single-cell datasets at developmental stages for reliable detection, and the incomplete characterization of their birthdate and death dynamics further complicates the issue.In a recent preprint, Quintana-Urzainqui et al. (2025 preprint) used single nucleus and spatial transcriptomics to build an atlas of the developing shark telencephalon during key stages of neurogenesis and neural differentiation. They identified a cluster of pallial neurons, characterized by the expression of transcription factors Tp73, Ebf3 and Lhx1, reminiscent of the molecular signature of mouse CR cells (Moreau et al., 2021). Cross-species integration with murine embryonic single cell RNA-sequencing data confirmed that CR cells are indeed an evolutionarily conserved cell type that already existed in the last common ancestor of jawed vertebrates, 450 million years ago. Consistently, one should expect to find CR cells in most (if not all) clades that subsequently emerged. Indeed, the analysis of previously published amphibian data (Woych et al., 2022) led to the identification of an unannotated cluster, showing robust transcriptomic similarity with both shark and mouse CR cells. Spatial RNA profiling indicated shark CR cells, like their mammalian counterpart, occupy a superficial position throughout the pallium, and are found enriched in medial pallial regions, corresponding to one of their demonstrated origins in mice.In mammals, one of the key roles of CR cells during development is to secrete the glycoprotein reelin that controls the radial migration of cortical glutamatergic neurons and enables the inside-out layering of the neocortex. Quintana-Urzainqui et al. therefore investigated the presence of reelin and found it to correlate with CR cells, suggesting that the blueprint for cortical layering already existed in the last common ancestor of jawed vertebrates. Along these lines, the authors performed electroporation experiments to label the progeny of telencephalic progenitors and observed their radial distribution at distinct stages. This led them to conclude that a rudimentary layered organization exists in the dorsal pallium of the shark telencephalon, supporting the hypothesis that the emergence of CR cells may have contributed to this process.A concomitant preprint by Gumnit et al. (2025 preprint) reports comparative studies across jawed vertebrate species, with a strong emphasis on salamander. Using single-cell transcriptomics, they confirm that a cell type homologous to mouse CR cells is present in amphibians and additionally identify species-specific signatures. The study also provides histological evidence that salamander CR cells share the same progenitor domains (namely the hem, septum and prethalamic eminence) and specification program, involving multiciliogenesis genes, as previously described in mice (Moreau et al., 2023).The emergence of new cell types during evolution is thought to occur through the gradual separation or divergence of functions originally carried out by a shared ancestral cell type, ultimately giving rise to two more specialized sister cell types (Arendt et al., 2016). In this context, Gumnit et al. investigated the heterogeneity of Tp73+ neurons in salamander, through comparison with developmental and adult mouse datasets. Beyond CR cells themselves, they identified a population with high transcriptomic similarity that corresponded to murine external tufted cells (ETC), an olfactory bulb neuronal cell type. ETC were found to be conserved across jawed vertebrates, and despite showing a specific transcriptomic signature, they also expressed a number of transcription factors typically considered as CR cells marker (including Trp73, Ebf3, Nhlh2, Lhx1, Lhx5 and Nr2f2). These results suggest that CR cells and ETC are sister cell types that evolved in stem vertebrates from an ancestral population involved in olfactory processing; CR cells progressively acquiring new specialized roles in developmental signaling.Taken together, these two preprints propose a new scenario in which CR cells arose by individuation from an ancestral olfactory-related cell type in the vertebrate lineage at least 450 million years ago. Further studies in more distant vertebrates such as lampreys are now needed to refine the evolutionary timeline of the appearance of CR cells. The subsequent acquisition of species-specific features and the amplification in the mammalian dorsal pallium then likely contributed to the emergence of the complex architecture of the neocortex as we know it today. This point remains speculative and would require a better understanding of the precise levels of reelin expression by CR cells, for which both studies reach different conclusions in two chondrichthyans (sharks and little skate). Although the developmental stage could account for those differences, reelin expression appears robust across amniotes. This supports the notion that the conserved role of CR cells in all vertebrates is reelin-independent and that increased reelin expression by CR cells was an evolutionary innovation that enabled the transition from outside-in to inside-out corticogenesis. Studies of reelin-dependent and -independent functions for CR cells (Elorriaga et al., 2025; Vílchez-Acosta et al., 2022) as well as evolution of cortical layering (Deryckere et al., 2025 preprint) are providing valuable insights that will need to be extended.Another point that deserves further investigation is the degree of CR cell diversity across species. In mouse, several subtypes have been characterized transcriptomically, correlating at least in part with their site of birth (Moreau et al., 2021). Although the progenitor domains from which CR cells are generated appear to be relatively well conserved across vertebrates, it remains to be established whether another round of diversification among CR cells occurred in mammals. Unfortunately, such studies are hindered by the scarcity of CR cells in the developing telencephalon and their migratory behavior that additionally blurs the picture.We are grateful to all members of the Pierani lab for stimulating and helpful discussions, as well as support services from Imagine Institute and IPNP.