Temporal constraints for developmental cell competition in mosaic gastruloids

In early mammalian embryogenesis, the selective elimination of suboptimal cells is critical for developmental integrity. Cell competition is a non-autonomous quality control in which “winner” cells eliminate viable but suboptimal “loser” cells based on their relative difference in fitness. Due to its central role in fitness perception, loss of p53 results in the emergence of “supercompetitor” cells, which stand at the apex of cell competition and induce apoptosis in neighboring wild-type (WT) cells. Here, we investigate cell competition dynamics using mosaic 3D mouse gastruloids, an embryonic stem cell (ESC)-based in vitro model of gastrulation, composed of defined numbers of WT and p53-KO cells. In mosaic gastruloids, even low numbers of p53-KO cells robustly outcompete WT cells, and introduction of as few as two p53-KO cells is sufficient to measurably impair neighboring WT cell growth. Cell competition in gastruloids is independent of proliferation rates, nutrient availability, or reactive oxygen species, and not influenced by Nodal and ERK signaling. However, we observe that Wnt and BMP signaling protect from cell competition, which is exclusively driven by intrinsic apoptosis, as indicated by Bcl2-mediated complete rescue of WT cells. During gastruloid development, cell competition is temporally restricted to a window of 48–96 hours after aggregation, mirroring embryonic days E5.5–E7.5 in the mouse. Heterochronic mosaic gastruloid experiments demonstrate that relative differences in pluripotency levels are neither necessary nor sufficient to cause supercompetition, but that cell competition is contingent on both competitors residing within the developmental window permissive to competition. Neither pluripotent mosaic 3D aggregates, nor 3D EpiGastruloids, which model more advanced developmental processes, display any competition, supporting the hypothesis that developmental cell competition is specific to the onset of gastrulation. Our findings offer insights into the mechanisms of cell fitness evaluation in mammalian embryogenesis and establish gastruloids as a powerful 3D model for investigating developmental stage-specific cell competition. Overall design: Mouse E14tg2A embryonic stem cells were cultures in serum+LIF conditions. Three clones were generated, one clone referred to as "F" carried an H2B-mCherry transgene, a clone "W" carrying an H2B-emiRFP670 transgene, and a daughter clone "K" which had the same H2B-emiRFP670 transgene and additionally a p53-KO by +1 frameshift insertion in exon 4 of the Trp53 gene. Each one of these clones was sequenced in pluripotent conditions before gastruloid formation. (o h). Then these clones were used for the generation of mosaic 3D gastruloids consisting of either 150 "F" cells with 150 "W" cells, or 150 "F" cells with 25 "K" cells. Mosaic gastruloids were dissociated and sorted into their respective clone of origin groups by FACS using their mCherry or emiRFP670 fluorophore markers. Gastruloids were harvested at 48h, 72h, 96h, 120h, 144h after gastruloid aggregation to generate a timecourse. Sorted cells were lysed and RNA was extracted and used for library preparation for RNA-TruSeq mRNA- sequencing on a NovaSeq6000-S4 with 2x150- 50 M read/sample conditions.