Madin-Darby canine kidney (MDCK) epithelial cells are a well-characterized model for investigating epithelial repair and leader cell migration. The genetic tractability of these cells and their ease of use for imaging assays allow for in-depth molecular dissection of epithelial cell biology. Prior work has shown that leader cells have a characteristic flattened morphology and distinct cytoskeletal properties and that they activate specific migratory pathways. In this study, we sought to identify how the leader program is activated in epithelial cells. We took advantage of the fact that in untreated MDCK cultures, a few cells spontaneously display a morphology reminiscent of leader cells and, on contact with neighbor cells with epithelial morphology, lead directed migration. These “spontaneous leaders” behave similarly to leaders observed at the edge of injured MDCK epithelial sheets. We investigated the mechanisms that drive spontaneous leader cell behavior, as an entry point to identify what invokes leader cell specification upon epithelial injury.
We found that spontaneous leaders exhibited elevated cellular tumor antigen p53 levels. Indeed, inducing p53 activation, either by inducing binucleation (which is a common feature among leader cells) or with the DNA-damaging agent mitomycin C (MMC) or the Mdm2 inhibitor nutlin-3, was sufficient to instruct leader behavior. Ablation of p53 by CRISPR mutagenesis strongly inhibited the emergence of leader cells upon MMC treatment, indicating that p53 is necessary for cells to be leaders. Working downstream of p53, we found that its target gene p21WAF1/CIP1 (p21) was also elevated in spontaneous leaders. Indeed, p21 elevation and its functional output, cyclin-dependent kinase (CDK) activity inhibition, were sufficient and necessary to induce leader cell behavior. Up-regulating p21 was sufficient to elevate integrin β1 and phosphoinositide 3-kinase (PI3K), which are known markers of spontaneous leader cells and are required for their migration.
Next, we investigated whether the p53-p21-CDK inhibition pathway that we identified in spontaneous leaders was also responsible for leader behavior upon epithelial injury. We found that scratch-induced leader cells experienced cell cycle delay, consistent with CDK inhibition, and showed high levels of both p53 and p21. Using a live reporter of p53 activity, we showed that injury itself induced p53 elevation at the edge of the damaged epithelium. This induction likely resulted from the mechanical insult, as it was dependent on the stress kinase p38, which activates p53 in response to mechanical stress. We then found that p53 and p21 both promoted cell migration in monolayers undergoing repair. Indeed, activating p53 at the migration front by laser-induced DNA damage accelerated cell migration, an effect that could be rescued by p53 inhibition. Conversely, inhibiting p53 or p21 slowed down migration.
We then followed the fate of leader cells as the injury resolved itself. Prior work has shown that cells with moderate p53 activation are eliminated by mechanical cell competition. Accordingly, we found that, once the epithelium was repaired, leader cells with high p53 activity were cleared by cell competition, undergoing extrusion or apoptosis. Failure to remove leader cells compromised the regular cobblestone-like morphology of the epithelium.