A study from the Ramanathan Lab, helmed by graduate student Yusuf Ilker Yaman, has shed light on how developing human embryos take shape. The paper appears in the latest issue of the journal Cell.
As the embryo elongates along the head-to-tail or anterior-posterior axis, cells in the tailbud known as neuromesodermal progenitors generate both the cells of the posterior neural tube, which develops into the spinal cord, and the paraxial mesoderm, which eventually gives rise to tissues of the locomotor system including the vertebral column and skeletal muscles of the back. The paraxial mesoderm is then segmented into structures called somites. In embryos, a wave of gene expression manifests at the tailbud, which then moves anteriorly. When it reaches the anterior-most part of the unsegmented paraxial mesoderm, a new pair of somites are generated.
Studying developing human embryos at these stages is not possible, prompting the use of human embryonic stem cells to model these developmental processes in the lab. However, stem cell models are extremely variable, rendering genetic studies difficult. Yaman wondered if it was possible to get stem cells to generate a neural tube and paraxial mesoderm that undergoes somitogenesis and generates segmented somites in the same way as in the embryo. Further, he asked if one could do so hundreds of times in an experiment with no errors.
Following a novel approach in a companion paper by Anand and colleagues, Yaman found that he could spatially arrange human stem cell aggregates on a coverslip and use the crosstalk between the aggregates through secreted signals to accurately generate axially elongating structures. By doing so, Yaman was able to generate hundreds of such structures, each with a neural tube and paraxial mesoderm that shows anteriorly moving waves of gene expression activity leading to periodic and sequential somite segmentation. Through subsequent single-cell transcriptome sequencing, he validated that the expected cell types were present and organized correctly along the axis.
The ability to generate hundreds of such structures in one experiment allowed Yaman to perform genetic manipulations and signal perturbations on a scale that was previously impossible in vitro, and well beyond what is possible in mouse embryos. Using this system, Yaman’s project investigated the molecular signals that govern the formation of somites, in particular questioning the “clock and wavefront” model that explains the spatiotemporal regulation of somite formation. The model suggests that the periodic somite generation along the axis is temporally coordinated by the oscillatory expression of “clock” genes and spatially controlled by a “wavefront”, defined by gradients of signaling molecules. The new stem cell model allowed Yaman to investigate the interactions between the “clock” and the “wavefront”.
Yaman perturbed the structures at precisely timed intervals as they were being patterned and was able to show that the key signaling pathways WNT and FGF played separate roles. Specifically, FGF not only defined the wavefront, but also controlled the frequency of the oscillating gene expression patterns constituting the clock.
“The clock consists of oscillating gene expression patterns that travel along the axis of the structures,” Yaman explains. “We found that what was thought of as the wavefront was also controlling the dynamics of the clock oscillations.”
While Anand’s team studied the signals that lead to axial elongation and shaping of the neural tube, Yaman’s work studied the patterning and segmentation of the neighboring mesoderm. Anand’s and Yaman’s papers debut alongside each other in the same issue of Cell.
“These two tissues are developed in the first month of pregnancy, and it is impossible to study these processes in vivo,” Yaman says. “Modeling them in a dish is an opportunity to study the development of these tissues, somites and neural tubes.” Defects in these developmental events lead to defects from spina bifida to congenital scoliosis.
“The ability to generate complex tissues that are accurately patterned both spatially and temporally give us an unprecedented opportunity to study human development,” says MCB faculty and study co-author Sharad Ramanathan. “The ability to do this at scale opens avenues to use the system to study the molecular mechanisms that govern development in normal and disease states in a way that was not possible before. The field of stem cell biology gives us exciting opportunities to test the understanding we have gained through more than fifty years of developmental biology by trying to generate complex tissues akin to those seen in the embryo.”
Yaman adds, “Both papers show us a way to study development in a different way, using stem cells and high throughput methods to gain mechanistic insights.”