CoChat
Sign Up
Cited by 684Sarcoma Diagnosis and TreatmentSingle-cell and spatial transcriptomicsBone Tumor Diagnosis and Treatments

The DOI matches its Crossref record; 96 of the 96 references it cites were found in a registry or an index.

Single-cell RNA landscape of intratumoral heterogeneity and immunosuppressive microenvironment in advanced osteosarcoma

Open access
Zhou, Yan, Yang, Dong, Yang, Qingcheng, Lv, Xiaobin, Huang, Wentao, Zhou, Zhenhua, Wang, Yaling, Zhang, Zhichang, Yuan, Ting, Ding, Xiaomin, Tang, Lina, Zhang, Jianjun, Yin, Junyi, Huang, Yujing, Yu, Wenxi, Wang, Yonggang, Zhou, Chenliang, Su, Yang, He, Aina, Sun, Yuanjue, Shen, Zan, Qian, Binzhi, Meng, Wei, Fei, Jia, Yao, Yang, Pan, Xinghua, Chen, Peizhan, Hu, Haiyan·2020·Nature Communications
DOI 10.1038/s41467-020-20059-6PMC PMC7730477
Open this paper in CoChat — free accountOpen PDF

Abstract

Osteosarcoma is the most frequent primary bone tumor with poor prognosis. Through RNA-sequencing of 100,987 individual cells from 7 primary, 2 recurrent, and 2 lung metastatic osteosarcoma lesions, 11 major cell clusters are identified based on unbiased clustering of gene expression profiles and canonical markers. The transcriptomic properties, regulators and dynamics of osteosarcoma malignant cells together with their tumor microenvironment particularly stromal and immune cells are characterized. The transdifferentiation of malignant osteoblastic cells from malignant chondroblastic cells is revealed by analyses of inferred copy-number variation and trajectory. A proinflammatory FABP4 + macrophages infiltration is noticed in lung metastatic osteosarcoma lesions. Lower osteoclasts infiltration is observed in chondroblastic, recurrent and lung metastatic osteosarcoma lesions compared to primary osteoblastic osteosarcoma lesions. Importantly, TIGIT blockade enhances the cytotoxicity effects of the primary CD3 + T cells with high proportion of TIGIT + cells against osteosarcoma. These results present a single-cell atlas, explore intratumor heterogeneity, and provide potential therapeutic targets for osteosarcoma.

Overview

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

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

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

Integrity signals

Registry

Crossref · 10.1038/s41467-020-20059-6

Cites

96 of 96 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Single-cell transcriptomic analysis of OS lesions. a Graphical view of the study roadmap. Single-cell suspensions were collected from OS lesions of 11 patients followed by scRNA-seq on 10× Genomics platform, respectively. A total of 100,987 qualified single cells were recovered. The peripheral blood CD3+ T cells were isolated for cytotoxicity analysis for TIGIT blocking experiments. b The t-distributed stochastic neighbor embedding (t-SNE) plot of the 11 identified main cell types in OS lesions. c Violin plots showing the normalized expression levels of eight representative canonical marker genes across the 11 clusters. d Dot plots showing the 21 signature gene expressions across the 11 cellular clusters. The size of dots represents the proportion of cells expressing the particular marker, and the spectrum of color indicates the mean expression levels of the markers (log1p transformed). e Relative proportion of each cell cluster across 11 OS lesions as indicated. The values of the detailed relative proportion of each cell cluster are provided in the Source Data file.
Fig. 1 Single-cell transcriptomic analysis of OS lesions. a Graphical view of the study roadmap. Single-cell suspensions were collected from OS lesions of 11 patients followed by scRNA-seq on 10× Genomics platform, respectively. A total of 100,987 qualified single cells were recovered. The peripheral blood CD3+ T cells were isolated for cytotoxicity analysis for TIGIT blocking experiments. b The t-distributed stochastic neighbor embedding (t-SNE) plot of the 11 identified main cell types in OS lesions. c Violin plots showing the normalized expression levels of eight representative canonical marker genes across the 11 clusters. d Dot plots showing the 21 signature gene expressions across the 11 cellular clusters. The size of dots represents the proportion of cells expressing the particular marker, and the spectrum of color indicates the mean expression levels of the markers (log1p transformed). e Relative proportion of each cell cluster across 11 OS lesions as indicated. The values of the detailed relative proportion of each cell cluster are provided in the Source Data file.CC BY · PubMed Central
Fig. 2 Distinct clusters of malignant cells in OS lesions. a Seven main malignant OS cell subclusters were identified by t-SNE analysis. b Feature plots for marker genes of osteoblastic (COL1A1, RUNX2, and COL3A1) and chondroblastic (ACAN, COL2A1 and SOX9) tumor cells. The color legend shows the log1p normalized expression levels of the genes. c The heatmap of the average expression of top 15 DEGs among six subclusters of osteoblastic tumor cells. The color legend indicates normalized gene expression levels among the subclusters. d The heatmap of GSEA of the 50 hallmark gene sets in MSigDB database among the six osteoblastic cell subclusters. e The scatter plot of the DEGs between osteoblastic tumor cells from lung metastasis (upper panel) or recurrent lesion (lower panel) versus primary lesion. The top 10 DEGs in each comparison were labeled in red. f The t-SNE plot of the four subclusters of chondroblastic cells. g The heatmap of GSEA of the 50 hallmark gene sets in MSigDB among the four chondroblastic clusters. h The scatter plot of the DEGs between chondroblastic and osteoblastic malignant cells. The top 10 genes in each subcluster were marked in red. Relative GSEA scores for each gene set across the cell clusters (d, g) and detailing average normalized gene expression (log1p transformed) values in different tumor sites (e) or types (h) are provided in the Source Data file.
Fig. 2 Distinct clusters of malignant cells in OS lesions. a Seven main malignant OS cell subclusters were identified by t-SNE analysis. b Feature plots for marker genes of osteoblastic (COL1A1, RUNX2, and COL3A1) and chondroblastic (ACAN, COL2A1 and SOX9) tumor cells. The color legend shows the log1p normalized expression levels of the genes. c The heatmap of the average expression of top 15 DEGs among six subclusters of osteoblastic tumor cells. The color legend indicates normalized gene expression levels among the subclusters. d The heatmap of GSEA of the 50 hallmark gene sets in MSigDB database among the six osteoblastic cell subclusters. e The scatter plot of the DEGs between osteoblastic tumor cells from lung metastasis (upper panel) or recurrent lesion (lower panel) versus primary lesion. The top 10 DEGs in each comparison were labeled in red. f The t-SNE plot of the four subclusters of chondroblastic cells. g The heatmap of GSEA of the 50 hallmark gene sets in MSigDB among the four chondroblastic clusters. h The scatter plot of the DEGs between chondroblastic and osteoblastic malignant cells. The top 10 genes in each subcluster were marked in red. Relative GSEA scores for each gene set across the cell clusters (d, g) and detailing average normalized gene expression (log1p transformed) values in different tumor sites (e) or types (h) are provided in the Source Data file.CC BY · PubMed Central
Fig. 3 Copy-number variation and clonal evolution analysis of OS cells. a The summary CNV profiles of the OS cells for the 11 OS samples inferred from inferCNV analysis. The CNV levels were categorized by the chromosome arm and simplified as gain or loss in single cells. Color in the heatmap indicated the percent of the CNV events in the single cells from each individual sample. b Clonality trees of the single cells from each patient. The branches are delineated according to the percentage of cells in the subclone containing the corresponding CNVs. The canonical CNV events in each lesion were labeled in the clonality tree. c The hierarchical heatmap showing large-scale CNVs in chondroblastic OS lesions form one primary (BC22) and one recurrent (BC20) OS sample (see Supplementary Fig. 8 for the other nine OS samples). d The percentage of chondroblastic and osteoblastic OS cells in each branch of clonality tree as indicated in (b) for the two lesions (BC20 and BC22). The percent value of the chromosomal CNV events (a) in the single cells from each individual sample is provided in the Source Data file.
Fig. 3 Copy-number variation and clonal evolution analysis of OS cells. a The summary CNV profiles of the OS cells for the 11 OS samples inferred from inferCNV analysis. The CNV levels were categorized by the chromosome arm and simplified as gain or loss in single cells. Color in the heatmap indicated the percent of the CNV events in the single cells from each individual sample. b Clonality trees of the single cells from each patient. The branches are delineated according to the percentage of cells in the subclone containing the corresponding CNVs. The canonical CNV events in each lesion were labeled in the clonality tree. c The hierarchical heatmap showing large-scale CNVs in chondroblastic OS lesions form one primary (BC22) and one recurrent (BC20) OS sample (see Supplementary Fig. 8 for the other nine OS samples). d The percentage of chondroblastic and osteoblastic OS cells in each branch of clonality tree as indicated in (b) for the two lesions (BC20 and BC22). The percent value of the chromosomal CNV events (a) in the single cells from each individual sample is provided in the Source Data file.CC BY · PubMed Central
Fig. 4 Trajectory analysis of osteoclast cells (OC) in OS lesions. a t-SNE plot showing the three main subclusters of osteoclasts. b Feature plots showing the normalized expression levels of myeloid and osteoclast markers CD74, CD14, ACP5, CTSK, and TOP2A in these subclusters. c, d The Monocle 2 trajectory plot showing the dynamics of osteoclast subclusters (c) and their pseudotime curve (d). e The DEGs (in rows, q-value < 10−10) along the pseudotime were hierarchically clustered into four subclusters. The top annotated GO terms in each cluster were provided. f Heatmap hierarchical clustering showing differentially expressed transcription factor genes along with the pseudotime curve in (e).
Fig. 4 Trajectory analysis of osteoclast cells (OC) in OS lesions. a t-SNE plot showing the three main subclusters of osteoclasts. b Feature plots showing the normalized expression levels of myeloid and osteoclast markers CD74, CD14, ACP5, CTSK, and TOP2A in these subclusters. c, d The Monocle 2 trajectory plot showing the dynamics of osteoclast subclusters (c) and their pseudotime curve (d). e The DEGs (in rows, q-value < 10−10) along the pseudotime were hierarchically clustered into four subclusters. The top annotated GO terms in each cluster were provided. f Heatmap hierarchical clustering showing differentially expressed transcription factor genes along with the pseudotime curve in (e).CC BY · PubMed Central
Fig. 5 Clustering and identification of mesenchymal stem cells (MSCs) and cancer-associated fibroblasts (CAFs) subclusters in OS lesions. a t-SNE plot of MSCs identified in the 11 OS lesions, colored by the three subclusters of cells as indicated. b The mean percent of the 3 MSC subclusters in primary, lung metastasis and recurrent samples. c Violin plots showing the normalized expression levels of marker genes across the clusters. d t-SNE plot of three subclusters of CAFs identified in the 11 OS lesions. e The mean percent of CAF subclusters in the three types of lesions. f Violin plots showing the normalized expression levels of marker genes across the three subclusters of CAFs. The values of mean proportions of MSCs (b) and CAFs (e) subclusters are provided in the Source Data file.
Fig. 5 Clustering and identification of mesenchymal stem cells (MSCs) and cancer-associated fibroblasts (CAFs) subclusters in OS lesions. a t-SNE plot of MSCs identified in the 11 OS lesions, colored by the three subclusters of cells as indicated. b The mean percent of the 3 MSC subclusters in primary, lung metastasis and recurrent samples. c Violin plots showing the normalized expression levels of marker genes across the clusters. d t-SNE plot of three subclusters of CAFs identified in the 11 OS lesions. e The mean percent of CAF subclusters in the three types of lesions. f Violin plots showing the normalized expression levels of marker genes across the three subclusters of CAFs. The values of mean proportions of MSCs (b) and CAFs (e) subclusters are provided in the Source Data file.CC BY · PubMed Central
Fig. 6 Comprehensive dissection of myeloid cells in OS lesions. a t-SNE plot separated 10 subclusters of the myeloid cells in OS lesions. b The violin plots showing the normalized expression levels of signature genes across the myeloid subclusters. c Dot plots showing cluster signature genes in myeloid cells. The size of dots represents the proportion of cells expressing the particular marker, and the spectrum of color indicates the mean expression levels of the markers (log1p transformed). d The proportion of the myeloid cells in the OS lesions with different types of lesions. e The heatmap of GSEA of 50 hallmark gene sets in MSigDB database between the 10 subclusters of myeloid cells. f Heatmap of the gene sets or signaling pathways specific for M1-activation, M2-activation, and IFN-γ activation for the macrophages in OS lesions, based on GSVA enrichment scores. g Heatmap of the gene sets or signaling pathways specific for each of the four subclusters of DC-activation based on GSEA enrichment scores with MSigDB database. The detailed cell proportion values of (d) and the relative GSEA scores for each gene set across the cell clusters of (e, g) are provided in the Source Data file.
Fig. 6 Comprehensive dissection of myeloid cells in OS lesions. a t-SNE plot separated 10 subclusters of the myeloid cells in OS lesions. b The violin plots showing the normalized expression levels of signature genes across the myeloid subclusters. c Dot plots showing cluster signature genes in myeloid cells. The size of dots represents the proportion of cells expressing the particular marker, and the spectrum of color indicates the mean expression levels of the markers (log1p transformed). d The proportion of the myeloid cells in the OS lesions with different types of lesions. e The heatmap of GSEA of 50 hallmark gene sets in MSigDB database between the 10 subclusters of myeloid cells. f Heatmap of the gene sets or signaling pathways specific for M1-activation, M2-activation, and IFN-γ activation for the macrophages in OS lesions, based on GSVA enrichment scores. g Heatmap of the gene sets or signaling pathways specific for each of the four subclusters of DC-activation based on GSEA enrichment scores with MSigDB database. The detailed cell proportion values of (d) and the relative GSEA scores for each gene set across the cell clusters of (e, g) are provided in the Source Data file.CC BY · PubMed Central
Fig. 7 Cell clustering and functional annotation of tumor-infiltrating lymphocytes (TILs) in OS lesions. a t-SNE plot for TILs in OS lesions, and the cells were classified into seven subclusters. b The violin plots showing the normalized expression levels of 8 signature genes across the TIL subclusters. c Dot plots showing 14 signature genes among the TIL subclusters. The size of dots represents the proportion of cells expressing the particular marker, and the spectrum of color indicates the mean expression level of the markers (log1p transformed). d The t-SNE plot showing the expression profiles of the four selected well-known marker genes for exhausted T cells. e Heatmap of the gene sets of T-cell cytotoxicity, exhaustion, regulatory cytokines and receptors, naive T cells, and T-cell costimulation, based on GSVA enrichment analysis. f Blockade of TIGIT increases the CD3+ T cells mediated cellular cytotoxicity activities on U2OS and 143B cells derived from patients BC3 and BC16, but not for BC5 and BC6 (n = 3). Error bar: mean value ± SD. P values were determined by paired two-sided Student’s t-test. The source data for the relative T cellular cytotoxicity activities on OS cell lines (f) are provided in the Source Data file.
Fig. 7 Cell clustering and functional annotation of tumor-infiltrating lymphocytes (TILs) in OS lesions. a t-SNE plot for TILs in OS lesions, and the cells were classified into seven subclusters. b The violin plots showing the normalized expression levels of 8 signature genes across the TIL subclusters. c Dot plots showing 14 signature genes among the TIL subclusters. The size of dots represents the proportion of cells expressing the particular marker, and the spectrum of color indicates the mean expression level of the markers (log1p transformed). d The t-SNE plot showing the expression profiles of the four selected well-known marker genes for exhausted T cells. e Heatmap of the gene sets of T-cell cytotoxicity, exhaustion, regulatory cytokines and receptors, naive T cells, and T-cell costimulation, based on GSVA enrichment analysis. f Blockade of TIGIT increases the CD3+ T cells mediated cellular cytotoxicity activities on U2OS and 143B cells derived from patients BC3 and BC16, but not for BC5 and BC6 (n = 3). Error bar: mean value ± SD. P values were determined by paired two-sided Student’s t-test. The source data for the relative T cellular cytotoxicity activities on OS cell lines (f) are provided in the Source Data file.CC BY · PubMed Central

Verification

Claims the paper makes

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

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

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 96 references

Cited by

Sorted by citations

Related

See all

Metadata

Authors
Zhou, Y., Yang, D., Yang, Q., Lv, X., Huang, W., Zhou, Z., Wang, Y., Zhang, Z., Yuan, T., Ding, X., Tang, L., Zhang, J., Yin, J., Huang, Y., Yu, W., Wang, Y., Zhou, C., Su, Y., He, A., Sun, Y., Shen, Z., Qian, B., Meng, W., Fei, J., Yao, Y., Pan, X., Chen, P., & Hu, H.
Affiliations
  • Shanghai Jiao Tong University (CN)
  • Third Affiliated Hospital of Nanchang University (CN)
  • Second Military Medical University (CN)
  • Southern Medical University (CN)
  • Jinan University (CN)
  • Key Laboratory of Guangdong Province (CN)
Corresponding author
  • Yao, Yang
  • Pan, Xinghua
  • Chen, Peizhan
  • Hu, Haiyan
ORCID
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2020-12-10
Volume
11
Issue
1
Pages
6322-6322
ISSN
2041-1723 (ISSN-L)
DOI
10.1038/s41467-020-20059-6
PMC
PMC7730477
OpenAlex
W3110998212
Cited by
672
References
96
Citations by year
2026: 106 · 2025: 136 · 2024: 152 · 2023: 139 · 2022: 96 · 2021: 41 · 2020: 2
FWCI
40.35
Type
journal article
Language
en
Update notices
correction (2021-04-30)
Open access
Licence
Locations
  • Nature Communications (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Natural Science Foundation of China (81372873, 8187215, 81503396, 81770173)
  • Please indicate the funder(s) of the research described in this manuscript, and the associated grant reference numbers
Topics
  • Health Sciences › Medicine › Oncology › Sarcoma Diagnosis and Treatment
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
  • Health Sciences › Medicine › Orthopedics and Sports Medicine › Bone Tumor Diagnosis and Treatments
Keywords
osteosarcoma, intratumoral heterogeneity, tumor microenvironment, immunosuppression, single-cell transcriptomics, cell of origin, osteoblastic osteosarcoma, chondroblastic osteosarcoma, transdifferentiation, copy number variation, macrophage infiltration, FABP4, osteoclasts, lung metastasis, recurrent osteosarcoma, TIGIT, CD3+ T cells
MeSH terms
Cancer-Associated Fibroblasts, Adolescent, Adult, Cell Aggregation, Child, Clone Cells, Dendritic Cells, Female, Humans, Immunosuppression Therapy, Macrophages, Male, Neoplasm Staging, Osteoclasts, RNA, Neoplasm, Osteosarcoma, Lymphocytes, Tumor-Infiltrating, Stromal Cells, Genetic Heterogeneity, Myeloid Cells, Young Adult, DNA Copy Number Variations, Single-Cell Analysis, Tumor Microenvironment, Transcriptome, Mesenchymal Stem Cells

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