Skip to main content
Premium Trial:

Request an Annual Quote

Liver Cell Atlas Assembled From Single-Cell Transcriptome Data

NEW YORK – A team from Germany and France has used single-cell RNA sequencing to assemble an atlas representing new and known liver hepatocyte and endothelial cell types from healthy human liver tissue, producing a baseline that is expected to help in understanding liver disease in the future.

"Our human liver cell atlas provides a powerful resource to enable the discovery of previously unknown cell types in normal and diseased liver," corresponding authors Dominic Grün, at the Max Planck Institute of Immunobiology and Epigenetics and the University of Freiburg, and Thomas Baumert, at the University of Strasbourg, and their colleagues wrote.

As they reported online today in Nature, Grün, Baumert, and colleagues did scRNA-seq on almost 10,400 individual cells from nine healthy donors, defining more than half a dozen cell type clusters. These included endothelial, Kupffer, and hepatocyte cell subtypes not described in the past, they reported.

For a proof-of-principle analysis using this liver atlas, the team incorporated new sequence data for liver tumor cells from three individuals with hepatocellular carcinoma (HCC) to identify suspicious shifts in cell type composition, gene expression, and biological activity in the liver cancer, validating these results with insights from mouse tumor xenograft model experiments.

"As demonstrated by our HCC analysis, the atlas provides a key reference for the investigation of liver diseases," the authors reported, "and will contribute to the development of urgently needed human liver models, including organoids and humanized liver chimeric mouse models."

Despite the diverse roles the liver plays in producing bile and protein, metabolizing lipids, and regulating glucose, the full repertoire of cell types performing these and other functions remains relatively uncharacterized. The authors reasoned that such profiles may be particularly important for understanding not only normal liver biology, but also liver cancer and other forms of liver disease.

Using Illumina short-read sequencing instruments and a previously described mCEL-Seq2 protocol, the researchers successfully generated transcriptome sequences for 10,372 individual cells from fresh or cryopreserved, disease-free liver samples. The study participants had liver samples removed during treatment for metastatic colorectal cancer or the bile duct cancer cholangiocarcinoma, they noted, but had not been diagnosed with chronic liver disease.

Along with single-cell transcriptomes from randomly sampled cells, which typically fell into hepatocyte or immune cell buckets, the team sequenced cells from populations enriched for endothelial cells and cells expressing EpCAM, a protein implicated in everything from cell adhesion and signaling to differentiation, proliferation, and migration.

Using gene expression markers, the researchers looked at how these liver cells clustered, flagging populations of hepatocytes; EPCAM+ bile duct cells called cholangiocytes; endothelial cell subtypes known as liver sinusoidal or macrovascular; immune cells (including B cells, T cells, natural killer cells, natural killer T cells); and other cell types.

The team went on to characterize the liver cell types and subtypes in more detail, searching for potential liver progenitor cell subpopulations and bringing in data from prior mouse hepatocyte analyses to retrace the spatial relationships between cells in the liver. The results revealed a subset of bile duct cells suspected as serving as liver progenitors, for example, and highlighted heterogeneity and "zonation" across cell populations that appears to contribute to the liver's tasks in both mice and humans.

"Our atlas reveals transcriptome-wide zonation of hepatocytes and endothelial cells, and suggests that different liver cell types may cooperate to carry out essential functions," the authors wrote, though they cautioned that additional, large-scale gene expression analyses are needed to further validate these and other findings.