Spatial and temporal signatures of cell competition revealed by K-function analysis.

Abstract

Cell competition is often categorised into mechanical competition, during which loser cell elimination is induced by long-range mechanical effects, and biochemical competition, during which loser cell elimination results from direct cell-cell contacts. Before confluence, proliferation of winner cells has often been hypothesised to gov- ern competition. Conversely, elimination of loser cells is thought to induce cell proliferation in its vicinity. However, causality is challenging to establish. To address this, we compute spatiotemporal signatures of competitive interac- tions using K-function clustering analysis. For this, we acquire long-term time lapses of two examples of mechanical (ScrKD) and biochemical (RasV12) compe- tition. We then segment cells, track them, and detect mitoses as well as elimina- tions. Finally, we perform K-function clustering to highlight spatiotemporal regions in which wild-type cell proliferation is enhanced or repressed around an elimination event. Our analysis reveals striking differences between the two types of competition. In the ScrKD competition, elimination seems driven by diffuse proliferation that does not cluster near the immediate elimination site. In contrast, RasV12 cell elimination is preceded by clustered proliferation of wild-type cells in the vicinity of the eventual RasV12 extrusion. Following loser elimination, an increase in local wild-type cell proliferation is observed in both competitions, although the timing and duration of these responses vary. This study not only sheds light on the diverse mechanisms of cell competition but also underscores the complexity of cellular interactions in tissue dynamics, providing new perspectives on cellular quality control and a new quantitative approach to characterise these interactions.