NEW YORK – Researchers have generated a molecular and cellular atlas of the human heart based on single-cell and single-nuclei transcriptome data.
To gain better insight into the makeup of the healthy heart, an international team of researchers generated transcriptome data from hundreds of thousands of cells from half a dozen heart regions. The heart has four chambers separated by septa that need to coordinate with one another to function properly. As they reported in Nature on Thursday, the researchers found that the types of cardiac cells present varied between heart regions and comprised numerous cell states.
"We have created the most detailed atlas of the human adult heart to date combining single-cell technologies with artificial intelligence methods to characterize almost half a million single cells," co-first author Carlos Talavera-López from the Wellcome Sanger Institute said in a statement. "For the first time, we could see exactly what each cell is doing in the human heart. This atlas shows that the cells in each of the four chambers of the heart behave differently to each other, mirroring the different functions of each area and helping us understand the healthy human heart."
Talavera-López and his colleagues isolated single cells, nuclei, and CD45+ enriched cells from the left ventricle, right ventricle, left atrium, right atrium, and interventricular spectrum from 14 adults for transcriptome and deep learning analysis. By including single nuclei, the researchers sought to ensure high enough representation of large cardiomyocytes, and they relied on their single-cell sampling to enrich for endothelial and immune cells. In all, they analyzed 78,023 CD45+ enriched cells, 45,870 other cells, and 363,213 nuclei, representing 11 major cell types, including cardiomyocytes, fibroblasts, and endothelial cells.
The prevalence of the major cardiac cell types varied by heart regions. For instance, the researchers noted that atrial tissues were 30 percent cardiomyocytes and 24 percent fibroblasts, while ventricular tissues were 29 percent cardiomyocytes and 21 percent mural cells, but only about 16 percent fibroblasts.
In particular, the researchers noted a high level of heterogeneity in cardiomyocytes. Among ventricular cardiomyocytes — which are enriched for gene transcripts encoding sarcomere proteins and transcription factors — the researchers identified five distinct populations. One of them, vCM1, makes up nearly 64 percent of the left ventricle, they noted, but only about 37 percent of right ventricle cardiomyocytes. Meanwhile, vCM2 is enriched in the right ventricle compared to the left ventricle.
The researchers similarly uncovered five atrial cardiomyocyte populations that varied in their gene expression. They noted that the gene expression differences between atrial cardiomyocytes and ventricular cardiomyocytes pointed to different developmental origins. Additionally, they uncovered 17 different populations of endothelial cells, smooth muscle cells, pericytes, and mesothelial cells and, in all, found more than 60 different cell states across the 11 cell types investigated.
To illustrate how their cardiac cell atlas could be applied, the researchers examined where in the heart the SARS-CoV-2 receptor ACE2 is expressed. They found it to be most highly expressed in pericytes, followed by fibroblasts, and least expressed in cardiomyocytes, though ACE2 transcription was higher in ventricular cardiomyocytes than in atrial cardiomyocytes.
"We mapped the cardiac cells that can be potentially infected by SARS-CoV-2 and found that specialized cells of the small blood vessels are also virus targets," co-senior author Michela Noseda from Imperial College London said in a statement. "Our datasets are a goldmine of information to understand subtleties of heart disease."