Somatic epigenetic drift during shoot branching: a cell lineage-based model
Plant architecture is shaped by the continuous production of new organs, most of which emerge post-embryonically. This process includes the formation of new lateral branches along existing shoots. Shoot branching is fundamental to plant development, plant-environment interactions, and vegetative propagation. Current empirical evidence supports a "detached meristem" model as the cellular basis of lateral shoot initiation. In this model, a small number of undifferentiated cells are "sampled" from the periphery of the shoot apical meristem (SAM) to act as precursors for axillary buds, which eventually develop into new shoots. Repeated branching thus creates a series of cellular bottlenecks (i.e. somatic drift) that affect how de novo genetic and epigenetic mutations propagate through the plant body during development. Somatic drift could be particularly relevant for epigenetic changes in the form of stochastic DNA methylation gains and losses (i.e. spontaneous epimutations), as they have been shown to arise rapidly with each cell division. Here, we formalize a special case of the "detached meristem" model, where pre-cursor cells are randomly sampled from the SAM periphery in a way that maximizes cell lineage independence. By following a population of SAM cells through repeated branching processes, we show that somatic drift gives rise to a complex mixture of cellular phylogenies, which shape the evolution of cell-to-cell DNA methylation heterogeneity within the SAM over time. This process is dependent on the number of branch points, the strength of somatic drift as well as the epimutation rate. For many realistic cell biological settings, our model predicts that cell-to-cell DNA methylation heterogeneity in the SAM converges to non-zero states during development, suggesting that epigenetic variation is an inherent property of the SAM cell population. Our insights have direct implications for empirical studies of somatic (epi)genomic diversity in long-lived perennial and clonal species using bulk or single-cell sequencing approaches.