Advances Towards Multi-Tissue Human Cell Atlas

The latest studies published by The Atlas of Human Cells are still making progress toward their goal of mapping every type of human cell, and these four papers focus on multi-tissue cell analysis.

The Human Cell Atlas (HCA) was founded in 2016 and is an international consortium of 2,300 members from 83 countries working to map every type of cell in the healthy human body. The HCA created detailed maps of over one million cells collected from 33 organs and systems and focused primarily on individual organs and tissues, or smaller subsets of tissues. Now they have developed methods to collect the data needed for multi-tissue cell atlases. The resulting cell atlases are freely available, meaning researchers can compare specific cell types and their functions throughout the body.

Study immune cells through tissues

So far, HCA has focused on immune cells that are carried in the blood; however, tissue immune cells also play an important role in the immune system. HCA researchers have created an immune cell catalog after sequencing the RNA of 330,000 unique immune cells to understand their function in different tissues. [1] From this catalog, they developed a machine learning tool called CellTypist to automate cell identification. Using this tool, they identified around 100 different immune cell types and their distribution in tissues, for example T cells, B cells and macrophages.

“By comparing particular immune cells in multiple tissues from the same donors, we have identified different flavors of memory T cells in different areas of the body, which could have great implications for the management of infections,” says Sarah Teichmann, head of cell genetics Welcome Sanger Institute (Cambridge, UK) and co-author of the article. “Our freely available data will contribute to the HCA and could serve as a framework for the design of vaccines, or to improve the design of immune therapies to attack cancers.”

The second published study looks at the tissues involved in the formation of blood and immune cells and reveals the types of cells lost from childhood to adulthood. This could inform in vitro cell engineering and regenerative medicine research. [2]

Freezing tissue for analysis

A single-cell atlas would be beneficial in identifying and mapping the specific cell types in which disease genes act. To create this, all cell types must be profiled, including those that are difficult to collect, for example fat cells or cells from skeletal muscle or neurons. Additionally, it is essential to profile cells from many different individuals, so tissue freezing prior to analysis is necessary.


Reference map of retinal pigment epithelium cells created with AI

The researchers identify five distinct cell subpopulations in the retinal pigment epithelium, which could explain different severities of retinal degenerative diseases.


HCA researchers have developed a single-nucleus RNA sequencing method using frozen cells. [3] They then used this method to create an inter-tissue atlas and analyze 200,000 cells from a frozen tissue bank containing genes for rare and common diseases. A new machine learning algorithm was used to match cells in the atlas to 6,000 single gene diseases and 2,000 complex genetic diseases and traits to identify cell types and genetic programs in disease. This could lead to new starting points for health and disease studies in the future.

Aviv Regev (Genentech early research and development; CA, USA), lead author of the paper, explains, “Our single-core HCA study demonstrates a powerful, large-scale way to analyze cells from frozen tissue samples across the body with advanced learning computations in depth and paves the way for tissue studies from entire patient cohorts at the single-cell level.We have been able to create a new roadmap for multiple diseases by directly linking cells to human disease biology and disease risk genes in the tissues.

The Tabula Sapiens dataset

The fourth and last article published in Science of this collection produced an inter-tissue atlas from living cells. [4] The resulting dataset is called “Tabula Sapiens”. This was done using single-cell RNA sequencing of live cells to analyze multiple organs from the same donors. The Tabula Sapiens has been used to characterize over 400 specific cell types, distribution and variations in gene expression. This will provide researchers with a large resource of annotated cell types and the Tabula Sapiens has enabled the first large-scale analysis of alternative gene splicing in a single-cell atlas.

“The Tabular Sapiens is a reference atlas that provides a molecular definition of hundreds of cell types in 24 organs of the human body,” said Stephen Quake, lead author of this paper and professor at Stanford University (California, USA). “It represents the efforts of over 150 authors across multiple institutions; the scientific community will discover new insights into human biology from this resource for many years to come.

Together, these four studies contribute to the creation of the consortium’s unique Human Cell Atlas and could have therapeutic implications such as understanding common and rare diseases, vaccine development and anti-tumor immunology.

Leave a Comment