CoChat

Single-cell transcriptome and cell type-specific molecular pathways of human non-alcoholic steatohepatitis

Open access
Fred, Rikard G., Steen Pedersen, Julie, Thompson, Jonatan J., Lee, Julie, Timshel, Pascal N., Stender, Stefan, Opseth Rygg, Marte, Gluud, Lise Lotte, Bjerregaard Kristiansen, Viggo, Bendtsen, Flemming, Hansen, Torben, Pers, Tune H. · 2022 · Scientific Reports

The DOI matches its Crossref record; 40 of the 41 references it cites were found in a registry or an index.

Abstract

The aim of this study is to characterize cell type-specific transcriptional signatures in non-alcoholic steatohepatitis (NASH) to improve our understanding of the disease. We performed single-cell RNA sequencing on liver biopsies from 10 patients with NASH. We applied weighted gene co-expression network analysis and validated our findings using a publicly available RNA sequencing data set derived from 160 patients with non-alcoholic fatty liver disease (NAFLD) and 24 controls with normal liver histology. Our study provides a comprehensive single-cell analysis of NASH pathology in humans, describing 19,627 single-cell transcriptomes from biopsy-proven NASH patients. Our data suggest that the previous notion of ”NASH-associated macrophages” can be explained by an up-regulation of normally existing subpopulations of liver macrophages. Similarly, we describe two distinct populations of activated hepatic stellate cells, associated with the level of fibrosis. Finally, we find that the expression of several circulating markers of NAFLD are co-regulated in hepatocytes together with predicted effector genes from NAFLD genome-wide association studies (GWAS), coupled to abnormalities in the complement system. In sum, our single-cell transcriptomic data set provides insights into novel cell type-specific and general biological processes associated with inflammation and fibrosis, emphasizing the importance of studying cell type-specific biological processes in human NASH.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41598-022-16754-7

Cites

40 of 40 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Single-cell analysis of human liver from obese patients with NASH. (A) Schematic diagram of the single-cell analysis workflow and analyses. Wedge biopsies from obese patients with NASH were dissociated enzymatically followed by single-cell barcoding, sequencing and data analysis. (B) UMAP projection of 19,627 cells showing the assigned identity for the main cell types in the liver, annotated by transfer of labels from previously published data sets. (C) Bar graph showing the proportion of cell types in each donor sample. (D) Left: Violin plots showing expression of selected marker genes for 12 distinct coarsely clustering cell types. The expression values are normalised, corrected for sequencing confounders and batch corrected for differences in average and variance between donor samples. Right; median number of RNA molecules, median number of genes and total number of cells in each cluster.
Figure 1 Single-cell analysis of human liver from obese patients with NASH. (A) Schematic diagram of the single-cell analysis workflow and analyses. Wedge biopsies from obese patients with NASH were dissociated enzymatically followed by single-cell barcoding, sequencing and data analysis. (B) UMAP projection of 19,627 cells showing the assigned identity for the main cell types in the liver, annotated by transfer of labels from previously published data sets. (C) Bar graph showing the proportion of cell types in each donor sample. (D) Left: Violin plots showing expression of selected marker genes for 12 distinct coarsely clustering cell types. The expression values are normalised, corrected for sequencing confounders and batch corrected for differences in average and variance between donor samples. Right; median number of RNA molecules, median number of genes and total number of cells in each cluster.CC BY · PubMed Central
Figure 2 Expression of co-expressed gene modules is associated with both inflammation and fibrosis in NASH. (A) Weighted gene co-expression network analysis workflow used to identify disease-associated gene modules. For each of the 12 major cell types, co-expression gene modules were identified. After quality control, 22 high-confidence modules were tested for association between module activity and levels of fibrosis, lobular inflammation and steatosis using a published RNA-seq data set with 24 healthy controls and 160 fatty liver disease samples. (B) Bar plot showing strength of module co-expression preservation in the MacParland et al. scRNA-seq liver data set. P-values were corrected for multiple testing (Benjamini-Hochberg) and the dotted line marks the significance threshold (P < 0.05). Modules are coloured by the cell cluster in which they were identified. Modules from Dendritic cells and Hepatic stem cells, which were absent in the MacParland et al. data set, were omitted in this step. (C) Cell type-specific gene modules whose activity pattern is associated, either positively or negatively, with the level of inflammation and fibrosis in fatty liver disease, in multivariate linear regression models with module activity as the outcome variable. The vertical and horizontal axes correspond to standard deviation differences in module activity for each unit change in lobular inflammation (encoded as 0, 1 or 2) or fibrosis (0 or 1). The plot includes only modules with statistically significant association to at least one condition using bootstrap percentile P-values corrected for multiple testing within each phenotype (Benjamini-Hochberg). Modules are coloured by cell cluster.
Figure 2 Expression of co-expressed gene modules is associated with both inflammation and fibrosis in NASH. (A) Weighted gene co-expression network analysis workflow used to identify disease-associated gene modules. For each of the 12 major cell types, co-expression gene modules were identified. After quality control, 22 high-confidence modules were tested for association between module activity and levels of fibrosis, lobular inflammation and steatosis using a published RNA-seq data set with 24 healthy controls and 160 fatty liver disease samples. (B) Bar plot showing strength of module co-expression preservation in the MacParland et al. scRNA-seq liver data set. P-values were corrected for multiple testing (Benjamini-Hochberg) and the dotted line marks the significance threshold (P < 0.05). Modules are coloured by the cell cluster in which they were identified. Modules from Dendritic cells and Hepatic stem cells, which were absent in the MacParland et al. data set, were omitted in this step. (C) Cell type-specific gene modules whose activity pattern is associated, either positively or negatively, with the level of inflammation and fibrosis in fatty liver disease, in multivariate linear regression models with module activity as the outcome variable. The vertical and horizontal axes correspond to standard deviation differences in module activity for each unit change in lobular inflammation (encoded as 0, 1 or 2) or fibrosis (0 or 1). The plot includes only modules with statistically significant association to at least one condition using bootstrap percentile P-values corrected for multiple testing within each phenotype (Benjamini-Hochberg). Modules are coloured by cell cluster.CC BY · PubMed Central
Figure 3 Pervasive down-regulation of genes related to ribosomal translation in inflammation. (A) Heatmap showing the proportion of genes in a module (columns) shared with other modules (rows). The overlap of a cluster with itself was set to 0 and columns and rows were clustered hierarchically. (B) Top; bar plot shows standard deviation differences in module activity for each unit change in lobular inflammation (encoded as 0, 1 or 2) in a multivariate linear regression model with module activity as the outcome variable. Error bars show 0.95 confidence intervals, obtained using bootstrap percentiles with bootstrap replicates. Bottom; module enrichment for three top-shared KEGG terms identified using GERR. (C) Protein–protein interaction network of the 32 overlapping genes correlated with oxidative phosphorylation. (D) Bar plot showing the top ten associated gene sets from Gene Ontology, based on the 32 overlapping genes correlated with oxidative phosphorylation. Significance notation refers to P-values obtained using a hypergeometric test and correcting for multiple testing (Benjamini-Hochberg, n = 30). ns: P > 0.05; *: P ≤ 0.05; **: P ≤ 0.01; ***: P ≤ 0.001; ****: P ≤ 0.0001.
Figure 3 Pervasive down-regulation of genes related to ribosomal translation in inflammation. (A) Heatmap showing the proportion of genes in a module (columns) shared with other modules (rows). The overlap of a cluster with itself was set to 0 and columns and rows were clustered hierarchically. (B) Top; bar plot shows standard deviation differences in module activity for each unit change in lobular inflammation (encoded as 0, 1 or 2) in a multivariate linear regression model with module activity as the outcome variable. Error bars show 0.95 confidence intervals, obtained using bootstrap percentiles with bootstrap replicates. Bottom; module enrichment for three top-shared KEGG terms identified using GERR. (C) Protein–protein interaction network of the 32 overlapping genes correlated with oxidative phosphorylation. (D) Bar plot showing the top ten associated gene sets from Gene Ontology, based on the 32 overlapping genes correlated with oxidative phosphorylation. Significance notation refers to P-values obtained using a hypergeometric test and correcting for multiple testing (Benjamini-Hochberg, n = 30). ns: P > 0.05; *: P ≤ 0.05; **: P ≤ 0.01; ***: P ≤ 0.001; ****: P ≤ 0.0001.CC BY · PubMed Central
Figure 4 A subset of pro-fibrotic macrophages associates with the M-Mac-3 module. (A) UMAP projection of the four macrophage clusters. (B) Violin plots of normalised and batch-corrected expression of macrophage specific marker genes. (C) Violin plots of the activity of macrophage modules, where some modules appear to be cluster-specific. (D) M-Mac-3, of which the 30 most central genes are seen in the network diagram. (E–F) Normalised and batch corrected gene expression of M-Mac-3 in the single-cell RNA-seq samples from individuals with NASH (E) and normalised (but not batch corrected) gene expression in normal liver samples from the MacParland et al. data set (F). (G) M-Mac-1, of which the 30 most central genes are seen in the network diagram (right panel), enriched for processes related to antigen presentation (left panel). In the network diagrams, node sizes represent gene centrality in the module (‘kIM’), and edge widths represent the strength of gene–gene co-expression. For the bar plot, gene sets from the Gene Ontology (GO), Kyoto Encyclopaedia of Genes and Genomes (KEGG) and Reactome databases were queried. Only significant results were plotted (P < 0.05).
Figure 4 A subset of pro-fibrotic macrophages associates with the M-Mac-3 module. (A) UMAP projection of the four macrophage clusters. (B) Violin plots of normalised and batch-corrected expression of macrophage specific marker genes. (C) Violin plots of the activity of macrophage modules, where some modules appear to be cluster-specific. (D) M-Mac-3, of which the 30 most central genes are seen in the network diagram. (E–F) Normalised and batch corrected gene expression of M-Mac-3 in the single-cell RNA-seq samples from individuals with NASH (E) and normalised (but not batch corrected) gene expression in normal liver samples from the MacParland et al. data set (F). (G) M-Mac-1, of which the 30 most central genes are seen in the network diagram (right panel), enriched for processes related to antigen presentation (left panel). In the network diagrams, node sizes represent gene centrality in the module (‘kIM’), and edge widths represent the strength of gene–gene co-expression. For the bar plot, gene sets from the Gene Ontology (GO), Kyoto Encyclopaedia of Genes and Genomes (KEGG) and Reactome databases were queried. Only significant results were plotted (P < 0.05).CC BY · PubMed Central
Figure 5 Activated human hepatic stellate cells are divided in two functionally specific subpopulations. (A) UMAP projection of the three hepatic stellate cell clusters. (B) Violin plots of normalised and batch-corrected expression of hepatic stellate cell marker genes. (C) Violin plots of the activity of hepatic stellate cell modules. (D) M-HSC-2, of which the 30 most central genes are seen in the network diagram. In the network diagram, node sizes represent gene centrality in the module (‘kIM’), and edge widths represent the strength of gene–gene co-expression. For the bar plot, gene sets from the Gene Ontology (GO), Kyoto Encyclopaedia of Genes and Genomes (KEGG) and Reactome databases were queried. Only significant results were plotted (P < 0.05).
Figure 5 Activated human hepatic stellate cells are divided in two functionally specific subpopulations. (A) UMAP projection of the three hepatic stellate cell clusters. (B) Violin plots of normalised and batch-corrected expression of hepatic stellate cell marker genes. (C) Violin plots of the activity of hepatic stellate cell modules. (D) M-HSC-2, of which the 30 most central genes are seen in the network diagram. In the network diagram, node sizes represent gene centrality in the module (‘kIM’), and edge widths represent the strength of gene–gene co-expression. For the bar plot, gene sets from the Gene Ontology (GO), Kyoto Encyclopaedia of Genes and Genomes (KEGG) and Reactome databases were queried. Only significant results were plotted (P < 0.05).CC BY · PubMed Central
Figure 6 WGCNA reveals that multiple circulating markers of NAFLD are co-regulated. (A) Network plot showing the 30 most central genes of M-Hep-8, as well as ALDOB, APOM, VTN, and AFM, four out of six recently identified NAFLD plasma proteome markers. Edge width is proportional to gene–gene co-expression and node size is proportional to gene ‘kIM’ centrality (Methods). (B) Violin plot showing the activity of M-Hep-1 in 12 cell clusters. (C) The biological function of module M-Hep-1, identified through gene set analysis using the Gene Ontology, Kyoto Encyclopedia of Genes and Genomes and Reactome databases. Up to five of the top hits from each database shown. Only significant results were plotted (P < 0.05).
Figure 6 WGCNA reveals that multiple circulating markers of NAFLD are co-regulated. (A) Network plot showing the 30 most central genes of M-Hep-8, as well as ALDOB, APOM, VTN, and AFM, four out of six recently identified NAFLD plasma proteome markers. Edge width is proportional to gene–gene co-expression and node size is proportional to gene ‘kIM’ centrality (Methods). (B) Violin plot showing the activity of M-Hep-1 in 12 cell clusters. (C) The biological function of module M-Hep-1, identified through gene set analysis using the Gene Ontology, Kyoto Encyclopedia of Genes and Genomes and Reactome databases. Up to five of the top hits from each database shown. Only significant results were plotted (P < 0.05).CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 40 references

Research works printed but not linked: 1

  1. Vujkovic, M. et al. A genome-wide association study for nonalcoholic fatty liver disease 1 identifies novel genetic loci and trait-relevant candidate genes in the 2 Million Veteran Program. 3 Themistocles L 12 Assimes. 36, 49 (2021).

Cited by

See all 51

Metadata

Authors
Fred, R. G., Steen Pedersen, J., Thompson, J. J., Lee, J., Timshel, P. N., Stender, S., Opseth Rygg, M., Gluud, L. L., Bjerregaard Kristiansen, V., Bendtsen, F., Hansen, T., & Pers, T. H.
Affiliations
  • University of Copenhagen (DK)
  • Frederiksberg Hospital (DK)
  • Hvidovre Hospital (DK)
Corresponding author
  • Bendtsen, Flemming
  • Hansen, Torben
  • Pers, Tune H.
Venue
Scientific Reports
Published
2022-08-05
Volume
12
Issue
1
Pages
13484-13484
ISSN
2045-2322 (ISSN-L)
OpenAlex
W4289885671
Cited by
44
References
41
Citations by year
2026: 11 · 2025: 14 · 2024: 8 · 2023: 11
FWCI
6.32
Type
journal article
Language
en
Open access
Licence
Locations
  • Scientific Reports (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Research at the University of Copenhagen (University of Copenhagen) (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • Novo Nordisk Fonden (NNF18CC0034900)
  • Lundbeckfonden (R19020143904)
Topics
  • Health Sciences › Medicine › Hepatology › Liver Disease Diagnosis and Treatment
  • Health Sciences › Medicine › Hepatology › Liver physiology and pathology
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
Keywords
nonalcoholic steatohepatitis, non-alcoholic fatty liver disease, hepatic macrophages, hepatic stellate cells, liver fibrosis, complement system, inflammation, cell type-specific gene expression
MeSH terms
Fibrosis, Humans, Liver, Biomarkers, Genome-Wide Association Study, Transcriptome, Non-alcoholic Fatty Liver Disease

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Single-cell transcriptome and cell type-specific molecular pathways of human non-alcoholic steatohepatitis. https://app.cochat.ai/papers/10.1038/s41598-022-16754-7
The paper
Fred, R. G., Steen Pedersen, J., Thompson, J. J., Lee, J., Timshel, P. N., Stender, S., Opseth Rygg, M., Gluud, L. L., Bjerregaard Kristiansen, V., Bendtsen, F., Hansen, T., & Pers, T. H. (2022). Single-cell transcriptome and cell type-specific molecular pathways of human non-alcoholic steatohepatitis. Scientific Reports. https://doi.org/10.1038/s41598-022-16754-7

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.