Breaking the brain barrier: cell competition in neural development and disease

General information on cell competition: Social behaviors are the basis of biological life. Like species and populations, cell communities experience Darwinian ecological interactions, and in case space and nutrient availability are not uniform throughout the tissue, they begin to compete for ground occupancy. In the 1970s, studies carried out by Drosophila geneticists pioneered the concept of cell competition, observing that genetic heterogeneity in a developing tissue led to the elimination of suboptimal cells (called losers) and the concurrent expansion of the fittest (accordingly named winners). This mechanism fits the logic underlying organ development, where tissue heterogeneity due to the intermingling of cells from different compartments may disrupt the entire process. Decades after, cell competition was associated with different MYC levels in confronting cells: high-MYC-expressing cells repeatedly eliminated low-MYC-expressing neighbors and grew up to colonize the whole territory (Gallant, 2005). Subsequent studies led to the identification of MYC-mediated cell competition (MMCC) as a central process in embryonic development, from flies to mammals (Penzo-Mendez and Stanger, 2014). Over time, different molecules and pathways have been associated with cell competition, and now this mechanism is known to be active in different tissues and organs, engaging different cell histotypes, from flies to humans. Losers forfeit by death, cannibalism, displacement, differentiation or quiescence, and winners achieve by proliferation, survival, hypertrophy or stemness maintenance (Baker, 2020). Independent of where, when, and how it occurs, cell competition pursues the stereotyped functional principle of selecting and expanding the most appropriate cells in each specific condition. In this perspective, we discuss the most relevant findings on the role of cell competition in neural cells.

Cell competition in brain development and decline: The “Flower Code” consists of three different isoforms (Ubi, Lose-A and Lose-B) encoded by the Drosophila flower locus. Starting from the findings obtained in larval epithelia, Merino et al. (2013) investigated if cell competition may be involved in the removal of supernumerary neurons during development. Using the Drosophila retina as a model, they showed that, in this case, the Flower^Lose-B isoform marks prospective loser neurons (Merino et al., 2013). Given the presence of at least four different Flower isoforms in mammals, these molecules may play a conserved role in post-mitotic tissues, promoting a process in which only the winners contribute to the adult organ. Several years later, the same research group used a fly model of Alzheimer's disease to investigate possible functional roles of cell competition in cognitive and motor decline. They expressed the human amyloid β₄₂ (Aβ₄₂) peptide in retinal neurons and observed they upregulated Flower^Lose-B and were eliminated by adjacent wild-type (wt) siblings. The same happened upon the expression of human HuntingtinQ128, which encodes a toxic peptide associated with Huntington's disease (HD), suggesting that the tissue is implementing a common safeguard mechanism rather than a specific response to a given insult. Aβ₄₂ expression in the adult brain promoted the accumulation of degenerative vacuoles, associated with apoptotic death and cognitive and motor decline. Notably, an extra copy of azot, a gene known to accelerate the elimination of unfit cells, restored brain architecture, long-term memory formation and locomotor activity in Aβ₄₂-expressing animals; conversely, azot silencing provoked an increase in dysfunctional neurons and faster brain deterioration (Coelho et al., 2018). These observations suggest that an efficient removal of damaged neurons can ameliorate cognitive performance by protecting neural circuits from aberrant synaptic transmission. Shifting from Drosophila to mammals, Jam et al. (2020) established an in vitro system in which mouse wt neural progenitor cells (NPCs) were co-cultured with siblings expressing oncogenic RAS, which is known to induce senescence. While viable in monoculture, RAS^act NPCs succumbed by apoptosis when mixed with wt NPCs. The researchers had previously identified a pool of factors highly expressed in juvenile brain cells, and when monitoring the levels of one of these, namely Srsf7, in single and co-cultured NPCs, they found that it was downregulated only in co-cultured RAS^act cells. Overexpression of Srsf7 prior to co-culture was sufficient to restore cells' capability to bypass competition (Jam et al., 2020). In this case, cell competition preserves progenitors' performance by culling those that, for different reasons, lose the juvenile phenotype and no longer meet the metabolic requirements of a developing tissue. Sun et al. (2023) approached the study of cell competition in vivo by manipulating Axin2 levels in NPCs. Axin2, known to be expressed in early brain development, is a Wnt-responder molecule, and the Wnt pathway has been associated with cell competition in many different organs. Experiments of genetic mosaicism were conducted by generating twin wt- and Axin2-deficient NPCs, followed by long-term tracing of the resulting progenies. These experiments revealed that Axin2-deficient NPCs succumb to the wt neighboring NPCs by apoptotic death. They also found that Axin2 regulates murine brain size through physiological cell pruning in early development, extending the functions of cell competition to mammalian neurogenesis (Sun et al., 2023). A recent study by Vieira et al. (2023) focused on glial cells' dynamics; several neurological conditions, such as Huntington's and Parkinson's diseases, are indeed severed by glial dysfunction. The researchers engrafted human wt glial progenitor cells (hGPCs) in mice whose brains were chimerized at birth by the injection of hGPCs expressing the mutant form of Huntingtin. Tracing the two lineages, the researchers observed that, by the time of engraftment (36 weeks), when the striatum was almost completely humanized by HD glia, wt hGPCs gradually eliminated the HD counterparts and repopulated the organ. MYC and its targets were upregulated in wt hGPCs with respect to the HD siblings, suggesting these cells were undergoing MMCC. As the engrafted wt hGPCs were much younger than the resident HD cells, the researchers hypothesized those aged cells were however disadvantaged, independent of the disease status. Should this be the case, it may imply that, in damaged or aged brains, reactivation of physiological cell competition could alleviate cognitive and motor decline. Of note, wt hGPCs were able to eliminate and replace wt aged siblings (Vieira et al., 2023).

Cell competition in brain cancer: Given the relevance of MYC to human malignancies, several studies have investigated MMCC in many aspects of cancer initiation and progression, from Drosophila to humans (Paglia et al., 2020). Glioblastomas are aggressive brain cancers that exhibit different cell origins, remarkable heterogeneity, high recurrence rate and resistance to therapy. Two recent studies investigated the dynamics by which glioblastoma cells overwhelm normal neighbors, and the mechanisms underlying cancer evolution. Glioblastoma can arise from adult neural stem cells (NSCs) with driver mutations that, over time, escape niche restrictions. Lawlor et al. (2020) investigated the dynamics occurring between wt and transformed NSCs. Following co-culture, the researchers noticed that wt cells' proliferation rate was significantly reduced compared to that observed in monoculture, suggesting some paracrine signals were counteracting their expansion. A comparison of the transcriptional profile of mono- and co-cultured wt NSCs highlighted in the latter the upregulation of genes associated with cell quiescence, and further experiments showed that those cells are driven back to an inactive state by neighboring transformed NSCs, which so favor their own self-renewal and niche colonization (Lawlor et al., 2020). Since tumor initiation, when mutated cells are very few in the field compared to the wt counterparts, is a limiting step in neoplastic transformation, these findings reveal a novel mechanism through which sporadic transformed cells can outcompete and overtake wt neighbors. Very recently, Ceresa et al. (2023) analyzed in vivo the mechanisms underlying clonal selection since the early stages of gliomagenesis. They performed intracranial injections of retroviral libraries into late mice embryos, with each vector carrying, in addition to the common elements, a unique sequence-based barcode. By doing so, the researchers retrospectively assigned each cell to a given clone. By dissecting brains at different points after injection, they observed that clone number dramatically decreased, with most of them disappearing within 1 month after injection. Clone number inversely correlated with clonal mass and cellularity, suggesting a process of selection that in time extinguished prospective loser clones while promoting the expansion of the dominant one. An RNAseq analysis of tumors harvested at different times after injection found a predominant deregulation of genes associated with MYC signaling. To directly link MMCC to clonal extinction, the researchers performed intracranial injections of MYCKD glioma cells, either alone or mixed with MYCwt siblings. While MYCKD cells were able to generate secondary gliomas, they were not recovered from tumors induced by mixed MYCKD and MYCwt cells, demonstrating that they were outcompeted by the MYCwt cells during cancer evolution. This provides functional evidence that clone selection in brain cancer is driven by MMCC (Ceresa et al., 2023).

Conclusions: The studies described above have investigated different types of neural cells from different species, indicating that cell competition is a universal driving force underlying the selection of the fittest, either in physiological or pathological contexts. The ongoing characterization of different ways that prospective winners achieve, other than over-proliferation, opens up investigations of cell competition in the post-mitotic nervous tissue. From the studies described above, it appears that the brain is a much more dynamic structure than previously recognized. Indeed, studies in Drosophila have shown how aberrant neurons confronted with wild-type siblings are removed from the tissue, and in pathological contexts, this leads to overall rescue due to the restoration of functional cell-cell contacts. Prospective studies are warranted to extend these findings to mammalian brain degeneration, where cell competition is expected to play similar roles (Figure 1A). The glial component deserves separate consideration: glial cells indeed conserve the ability to proliferate across adult life, especially in response to acute and chronic injuries, and their condition is known to profoundly influence neuronal physiology. Moreover, different populations of glial cells inhabit several anatomic districts, and glial functions are currently being investigated in disparate pathologies of unrelated organs. Given the functions of these cells in controlling inflammation, innate immunity and tissue structure, the evidence that cell competition governs glial homeostasis may provide an unprecedented opportunity to protect tissue health by modulating a common process. However, the characterization of cell competition in neural cells is still in its infancy; indeed, we need to determine how cell competition is triggered and proceeds in the central nervous system and peripheral organs and tissues, whether it employs specific mechanisms in different cells that make up the nervous tissues across the body, and which genetic or metabolic contexts are resistant to this process. The ability to define cellular- and time-specificity in detail will in the future provide us with the opportunity to find predictive and prognostic markers and more specific and personalized treatments for neurological diseases and cognitive decline. Given the role of cell competition in brain cancer, glial cells are once again key players: their ability to proliferate, migrate and transdifferentiate even under physiological conditions makes them particularly prone to transformation. As with other solid tumors, clonal selection during brain cancer evolution is coordinated by MMCC, a finding that gives us the opportunity to decipher this process in neural tissues starting from a reliable knowledge (Figure 1B).

Figure 1.

Expected roles for cell competition in brain disease.

(A) Potential role of cell competition in mammalian neurodegeneration. The scheme illustrates an atrophic brain (right half) compared to a normal brain (left half). The pathological phenotype can be due to different diseases, depending on the affected cells. The magnifications on the right describe how cell competition may modulate tissue texture. HD: Huntington's disease; PD: Parkinson's disease; AD: Alzheimer's disease. (B) Cell competition in brain cancer. The scheme illustrates a brain undergoing cancer development. The magnifications on the right describe how aberrant cell competition can favor cancer grow, while its inactivation may rescue organ structure. The different neural cell types are displayed at the bottom of the panel. Created with BioRender.com.

For example, we have extensively characterized the link between cell polarity, MMCC and cancer in epithelial tissues (Grifoni et al., 2013). Since some polarity determinants are essential for the asymmetric division of NPCs, it is expected that the expansion of NPCs following polarity disruption can promote brain cancer initiation. Finally, it is worth highlighting the enormous and, in our opinion, undeniable contribution of Drosophila to the characterization of this phenomenon. Also in this case, studies in the fruit fly have represented the starting point for human-oriented investigations. In the future, basic and translational research will hopefully find the missing pieces in the mosaic of neuronal competition and provide new tools to restore organ architecture and function in a variety of disease conditions.

This work was supported by a collaborative project between “INFN - Laboratori Nazionali del Gran Sasso” and University of L'Aquila, Dept. “Life, Health and Environmental Sciences” (to DG).