Spatial and single cell tools open the black box of early human development
What happens when two powerful technologies join forces to resolve the “what” and the “where” of complex biology? We share those stories in our “When single cell meets spatial” series and highlight studies that capture the best of both technologies—and how they can be even better together.
Mapping the human embryo without a reference atlas
Four researchers from the Sloan Kettering Institute and Tongji University credit maintaining their sanity during the COVID-19 pandemic to something that tested mine: Zoom meetings (1).
During biweekly Zoom meetings in 2020, researchers spanning the United States and China developed a framework to annotate 157 papers that characterized cells and the genes they expressed in mice and humans in the earliest stages of organogenesis.
This wasn’t just a quarantine pastime; these researchers were on a mission. They were preparing for an experimental expedition into the scientific unknown—the molecular landscape of 4- to 6-week-old human embryos.
Embryos undergo a huge transition in their fourth week of development. They exit gastrulation—the earliest stage of embryonic development where cells are partitioned into specialized tissues known as germ layers—and enter organogenesis. Cellular diversity skyrockets during this phase as cells start committing to cardiac, neuronal, dermal, and other tissue types as they differentiate to progenitors.
The biological stakes are high during organogenesis. Molecular missteps that cause developmental defects that contribute to miscarriages and birth defects often occur at this stage.
Our understanding of organogenesis is shaped by corresponding developmental phases in non-human models, such as mice and zebrafish embryos, due to various technical and ethical issues related to studying human embryos at this stage. But it’s unclear how well these findings translate to human embryos.
When this team of researchers from Sloan Kettering Institute and their collaborators at several institutions in China set out to answer questions about the cellular makeup of seven 4- to 6-week-old human embryos using Chromium Single Cell Gene Expression and Visium Spatial Gene Expression, they realized their insights could be limited by using reference data from non-human models at this stage.
“The biggest challenge in this study was certainly the annotation of cell types, as there is not a comprehensive and curated list of cell types, marker genes, and lineage relationships at this developmental stage,” said researchers Yichi Xu, PhD, and Weiyang Shi, PhD, from Sloan Kettering Institute and Ocean University of China, respectively (1).
Using the 157 papers they identified, they developed a curated list of cell types and gene expression expected at 4–6 weeks in human embryos based on data from mouse embryos at a corresponding developmental stage and older human embryos. This gave them the framework to answer questions about conservation of developmental trajectories between other mammals and the unique features of human development.
They recently published a spatially resolved single cell atlas of 313 cell types built on the analysis of 180,000 single cell transcriptomes in Nature Cell Biology (2). Their work ultimately provides not only a benchmark to evaluate how well other models mimic human development, but a working model for the molecular mechanics powering the earliest stage human organogenesis.
“Early embryogenesis, especially in humans, is very much underexplored and has only recently started to reveal its secrets, following the advent of new technologies,” said Stylianos Lefkopoulos, PhD, an Associate Editor at Nature Cell Biology (1). “This work by Xu et al. stood out to me because it characterizes a developmental window that spans early organogenesis in human embryos. [The atlas] could contribute to the understanding of both normal and disease-linked developmental events that have been so far a mystery.”
Building an atlas of the human embryo
The researchers first analyzed the human embryos using Chromium Single Cell Gene Expression. They dissected each of the seven embryos into four major parts: head, trunk, limb, and viscera. They combined common dissection segments and created a single cell library for each. Separate analysis of major tissue sections in the embryos ensured that the expected cell populations were present in the expected locations.
Ultimately, they captured over 185,000 cells from 22 dissection parts—an average of 7,732 transcripts and 2,338 genes per cell. They identified 313 cell clusters across 18 developmental systems (cell lineages, organs, and tissues) based on 234 biomarkers identified in the 157 studies analyzed.
Most of the cell types aligned with the reference atlas they developed; 213 of the identified clusters matched known cell types and 98% of cells came from the expected dissection part. The authors commented that the unique cell types uncovered in the head mesoderm would be technically challenging to isolate and detect in any embryo.
The researchers then complemented their single cell analysis with Visium Spatial Gene Expression analysis. While single cell analysis required the researchers to dissect the embryos to map gene expression in the embryo, Visium provided more detailed, spatial information through whole transcriptome analysis of intact, whole embryo, tissue sections alongside hematoxylin and eosin (H&E) staining.
The spatial resolution allowed them to take a closer look at the previously unidentified cell types in the head mesoderm identified with Chromium. Mesoderm cells are talkative, constantly sending signals to interacting cells, often to promote differentiation. They identified 134 signaling interactions based on a signaling analysis—a statistical method developed for this study—of both the Chromium and Visium data. They uncovered known developmental signaling pathways such as Bone morphogenic protein (BMP) signaling from the myocardium.
But they also discovered unique cell-to-cell interactions between the newly identified cell types and with otic vesicles—the early structure of the inner ear—via extracellular matrix interactions, primarily between collagen and integrin. Extracellular matrix signaling often promotes differentiation and contributes to morphogenesis, indicating these new cell populations are likely involved in early stages of head formation.
The trajectory of human embryo mapping
The researchers leveraged their analyses to both identify new cell interactions critical to human organogenesis and build on previous findings with improved resolution of Visium compared to in situ expression such as whole-mount in situ hybridization. They developed detailed maps of limb buds, identified developmental landmarks based on gene expression analysis across other vertebrate datasets, and integrated their analyses with data previously collected from older embryos.
However, only single cell analyses of 10- to 26-week-old embryos are available. The researchers suggest this gap should be filled, but acknowledge the ethical and technical challenges to accomplishing this goal. Analyses before four weeks exist, but are sparse.
“Our best hope in the foreseeable future to study the earliest stages may be in vitro models. In this regard, our dataset provides the in vivo benchmark to evaluate how well different models mimic development: the end product of such a model should match the in vivo cell types,” wrote the authors.
This study provides not only an indispensable resource for developmental biologists, but a strategy for other researchers to combine single cell and spatial analyses for other embryonic samples, no matter what model type or developmental stage.
Learn more about our Chromium and Visium technology, and explore our Xenium In Situ datasets to explore what you can do with improved spatial resolution.
References:
- Xu Y and Weiyang S. Single-cell and spatial transcriptomics during human organogenesis. Nat Cell Biol 25, 522–523 (2023). doi: 10.1038/s41556-023-01113-z
- Xu Y, et al. A single-cell transcriptome atlas profiles early organogenesis in human embryos. Nat Cell Biol 25, 604–615 (2023). doi: 10.1038/s41556-023-01108-w
Social share image credit lunar caustic (CC BY 2.0).