Research

Developmental roadmap

We are systematically mapping the branching lineage decisions through which human pluripotent stem cells differentiate into two dozen different cell-types spanning the endoderm, mesoderm, and ectoderm germ layers. At each differentiation step, we are defining the combinations and timings of extracellular signals that have to be turned on or off to induce a given cell-type, while repressing formation of alternative cell-types. This is creating an informative reference map for human development. Instead of focusing on the development of a specific cell-type, we are broadly exploring many different lineages simultaneously, hoping to learn general principles underlying developmental biology.

Image: Human vascular roadmap draft 

Developmental reconstitution

Cells differentiate within the inaccessible 3D environment of the embryo. The complete set of signals that drive differentiation at any given step is usually unknown in vivo. Historically, we studied development by deleting or silencing single genes to identify those necessary for a specific cell-type’s development ("developmental genetics”). This is akin to disabling a car by removing a single part (e.g., the engine). Yet this single part does not explain how the car works: a car has many parts that must work together. Analogously, multiple signals converge to induce a human cell-type, but we do not know all the signals that induce which cell-types and when. We and others are employing a reconstitution approach, where we build human cell-types in vitro to discover which signals are sufficient to generate a given cell-type from human pluripotent stem cells. Put another way, if we truly know how a specific human cell-type is born, we might build it from scratch in the Petri dish, outside the embryo. This feat would be analogous to assembling a car from its multiple constituent components. Reconstitution might seem quixotic, but is familiar to biochemists and engineers, who define the minimum components sufficient to reconstitute complex biological processes (e.g., DNA replication) in vitro.

In vivo veritas

We are reconstituting the development of specific cell-types in vitro from human pluripotent stem cells in order to learn more about how they form. For certain cell-types, we discover unexpected developmental progenitors that form them in vitro, or hitherto-unknown extracellular signals that specify them in vitro. We test these in vitro hypotheses through complementary analyses of mouse embryos, using classical techniques in developmental biology, including in vivo lineage tracing, genetic perturbations, and marker staining. 

Image: Whole-mount stain of an 8.5-day-old mouse embryo

Generating nearly-pure cell populations from stem cells

Generating pure batches of human cells is crucial for regenerative medicine. However, converting human pluripotent stem cells into desired cell-types with high efficiency and precision has remained challenging. At each branching lineage decision during stem-cell differentiation, we provide signals to specify the desired cell-type, while simultaneously inhibiting signals that otherwise specify unwanted fates. This "forced" stem cells to differentiate into a pure population of a desired cell-type by blocking off alternate paths. We have differentiated hPSCs into enriched batches of human blood progenitors (Fowler & Zheng et al., 2024; Developmental Cell),  human artery and vein endothelial cells that could build blood vessels (Ang et al., 2022; Cell), human liver progenitors that could engraft the mouse liver (Ang et al., 2018; Cell Reports; Loh & Ang et al., 2014; Cell Stem Cell), and human bone progenitors that could form an ectopic human bone in mice (Loh et al., 2016; Cell).  

Image: Human artery endothelial cells generated from pluripotent stem cells