Decoy fitness peaks, tumor suppression, and aging

Abstract

Recent reports by Martincorena et al and Yokoyama et al reveal unanticipated dynamics of somatic evolution in the esophageal epithelium, with clonal expansions apparently driven by mutations in Notch1 dominating the epithelium even in middle-aged individuals, far outpacing the prevalence of these mutations in esophageal cancers. We propose a model whereby the promotion of clonal expansions by mutations such as in Notch1 can limit more malignant somatic evolutionary trajectories until old ages.

Figure 1

Proposed model for fitness landscape changes with age. (a) The young fitness landscape favors cell occupancy of the evolved adaptive peak (with “wild‐type,” WT, phenotype), but disfavors more malignant phenotypes. Some selection for potentially oncogenic mutations occurs, but with a bias toward “decoy peaks.” While progression up the decoy peak may be limited by the required passage through lower fitness intermediates, the small size of epithelial progenitor pools could facilitate such transitions through neutral drift. Alternatively, a single mutation, such as in Notch1, could mediate the “jump” to the other peak. (b) By middle age, fitness landscapes engender greater selection for the phenotypes that occupy decoy peaks (often with Notch1 mutations); while partially transformed, phenotypes of cells on these decoy peaks are more benign with reduced malignant potential. (c) At older ages, further tissue degradation and damage accumulation result in a landscape that increases the odds of mutational adaptation toward both benign (decoy) and more malignant phenotypes. Clones with more malignant phenotypes are more likely to progress toward cancer. Arrow thickness reflects hypothetical probabilistic phenotypic and fitness effects of mutation. The x‐y plane represents potential genetically and epigenetically encoded somatic cell phenotypes. Note that for simplicity, this model does not incorporate roles for age‐dependent mutation accumulation, which should clearly contribute to cancer risk with age. We also note that while experimental and observational data support changes in somatic selection in aging tissues, the shapes of these landscapes are hypothetical