SCIENTIFIC EDUCATIONAL CENTER science idea

An international group of scientists, which included a Moscow State University student, proposed a new tool for mapping cell types in body tissues based on cellular and spatial transcriptomics data. The use of information about the amount of RNA synthesized by each cell type, combined with modern computational methods, allows us to study the fine structure of tissues with a sensitivity that was not previously available. The researchers demonstrated the effectiveness of their development in experiments with mouse brain tissue by localizing a previously unknown subtype of astrocytes.

All cells in the body have the same genome, but differ significantly in both structure and function. The development of single-cell RNA sequencing (scRNA-seq) technologies has shown that human cells are much more diverse than previously thought - hundreds of different types are now known. The complexity of the organism, however, is provided not only by the presence of a large number of cells performing different roles, but also by their interaction at the level of tissues and even entire organs.

The task of mapping cell types in tissues is important both from the standpoint of fundamental science and medicine. The latest technologies of spatial transcriptomics help to solve it: different types of cells synthesize a different set of RNA (transcriptome) and proteins (proteome), which makes it possible to distinguish them from each other. If you divide a tissue sample into sections, then, based on the analysis of the composition of RNA from each, it is possible to assume what type of cells are in one place or another. This approach, however, creates a number of additional difficulties. First, a tissue can contain many cell types that differ slightly in RNA composition, for example, of stromal origin, which significantly complicates their identification. Secondly, the size of the studied tissue sections is usually much larger than the average cell size, as a result of which a mixture of RNA from different types gets into each, which does not make processing the results easier. Taking into account these limitations, manual labor of researchers cannot be dispensed with; scalable computational methods are needed.

The international team, which included Artem Shmatko, a student from the Faculty of Fundamental Physical and Chemical Engineering of Moscow State University, developed the cell2location tool, which reveals the spatial distribution of cell types based on single cell sequencing and spatial transcriptomics data. The system compares the amount of RNA in the spatial data with the reference RNA expression profiles for the types present in the tissue, determining the exact number of different cells in each of the experimentally studied sites. Cell2location effectively corrects for various sources of experimental error, allowing integration of cellular and spatial transcriptomics with higher sensitivity than existing tools.

The use of the proposed tool has already led to new scientific discoveries. Thus, the researchers were able to localize in the mouse brain a new type of astrocytes, the second most important (after neurons) type of brain cells that perform structural, homeostatic, and reparative functions.

“Our experiments have shown that cell2location is versatile and allows you to find rare cell types that cannot be detected by traditional histological methods. Now researchers have a powerful new tool in their hands, and I believe that it will help us move further in understanding the fine structure of tissues, especially the brain,” commented Artem Shmatko.

Article published in the journal Nature Biotechnology

PHOTO © Andrej Vodolazhskyi/shutterstock
Source: msu.ru, sci-dig.ru