A time and single-cell resolved model of hematopoiesis

Abstract

The paradigmatic tree model of hematopoiesis is increasingly recognized to be limited as it is based on heterogeneous populations and largely inferred from non-homeostatic cell fate assays. Here, we combine persistent labeling with time-series single-cell RNA-Seq to build the first real-time, quantitative model of in vivo tissue dynamics for any mammalian organ. We couple cascading single-cell expression patterns with dynamic changes in differentiation and growth speeds. The resulting explicit linkage between single cell molecular states and cellular behavior reveals widely varying self-renewal and differentiation properties across distinct lineages. Transplanted stem cells show strong acceleration of neutrophil differentiation, illustrating how the new model can quantify the impact of perturbations. Our reconstruction of dynamic behavior from snapshot measurements is akin to how a Kinetoscope allows sequential images to merge into a movie. We posit that this approach is broadly applicable to empower single cell genomics to reveal important tissue scale dynamics information.

Unveiling Clonal Cell Fate and Differentiation Dynamics: A Hybrid NeuralODE-Gillespie Approach

“…Many works in the field of mathematical modeling for hematopoiesis have assumed that adult physiological hematopoiesis is in perfectly homeostatic conditions (referred to as "steady state"), where transition rates are constant and so are the populations' sizes. On the other hand, more recent models 33,63 have started incorporating the idea that, given that the number of HCSs increases over time and the relative abundance of the different populations varies upon aging, there is no steady state. CLADES can partially address this issue thanks to the constant vs dynamic mode option.…”

“…This method brought new insights to multi-clonal time series data analysis with respect to the identification of early cell fate bias at the cellular level. Third, discrete state space over cell types is also commonly used to ensure better interpretability, higher robustness, and computational efficiency 33 . Finally, other approaches have also significantly contributed to this field by modeling individual genes or phenotypes as a different type of cell state space, for example, upon learning smoothed transcription and regulatory dynamics 34 , or explicitly analyzing the potency bias during HSCs' reactivation following platelet depletion 35 .…”

HOXA9 and β-catenin safeguard HSC integrity

“…Multilineage hematopoiesis is sustained by self-renewing HSCs and a complex dynamic hierarchy of progenitor cells that give rise to mature blood lineages. [2][3][4] To preserve their integrity under steadystate conditions, HSCs remain quiescent with low metabolic activity. 3,4 However, HSCs are required to temporarily exit quiescence to replenish progenitor pools and thereby maintain hematopoietic homeostasis.…”