AlphaBeta: computational inference of epimutation rates and spectra from high-throughput DNA methylation data in plants

Stochastic changes in DNA methylation (i.e., spontaneous epimutations) contribute to methylome diversity in plants. Here, we describe AlphaBeta, a computational method for estimating the precise rate of such stochastic events using pedigree-based DNA methylation data as input. We demonstrate how AlphaBeta can be employed to study transgenerationally heritable epimutations in clonal or sexually derived mutation accumulation lines, as well as somatic epimutations in long-lived perennials. Application of our method to published and new data reveals that spontaneous epimutations accumulate neutrally at the genome-wide scale, originate mainly during somatic development and that they can be used as a molecular clock for age-dating trees.

Hematopoietic lineages diverge within the stem cell compartment

Hematopoietic stem cells (HSCs) produce a highly diverse array of cell lineages. To assay hematopoietic differentiation with minimal experimental perturbation, non-invasive methods for heritable labeling1–3 or barcoding4–7 of HSCs in vivo have recently been developed and used to study lineage fate of HSCs in physiological conditions. However, the differentiation pathways leading from HSCs to mature cells remain controversial8, with suggested models ranging from gradual lineage restriction in a branching cascade of progenitors to HSCs already making ultimate lineage decisions. Here we show, by iterating HSC fate-mapping, mitotic history tracking, single-cell RNA-sequencing and computational inference, that the major differentiation routes to megakaryocytes, erythro-myeloid cells and lymphocytes split within HSCs. We identify the hitherto elusive self-renewing source of physiological hematopoiesis as an HSC subpopulation co-expressing high levels of Sca-1 and CD201. Downstream, HSCs reduce Sca-1 expression and enter into either thrombopoiesis or erythro-myelopoiesis, or retain high Sca-1 levels and the ability to generate lymphocytes. Moreover, we show that a distinct population of CD48−/lo megakaryocyte progenitors links HSCs to megakaryocytes. This direct thrombopoiesis pathway is independent of the classical pathway of megakaryocyte differentiation via multipotent progenitors and becomes the dominant platelet production line upon enhanced thrombopoietin signaling. Our results define a hierarchy of self-renewal and lineage decisions within HSCs in native hematopoiesis. Methodologically, we provide a blueprint for mapping physiological differentiation pathways of stem cells and probing their regulation.