Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Jun 5;12(1):58.
doi: 10.1186/s40364-024-00605-w.

Single-cell and spatial transcriptomics reveal a high glycolysis B cell and tumor-associated macrophages cluster correlated with poor prognosis and exhausted immune microenvironment in diffuse large B-cell lymphoma

Liyuan Dai #  1 , Guangyu Fan #  2 , Tongji Xie  2 , Lin Li  3 , Le Tang  2 , Haizhu Chen  4 , Yuankai Shi  5 , Xiaohong Han  6
Affiliations

Single-cell and spatial transcriptomics reveal a high glycolysis B cell and tumor-associated macrophages cluster correlated with poor prognosis and exhausted immune microenvironment in diffuse large B-cell lymphoma

Liyuan Dai et al. Biomark Res. .

Abstract

Background: Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous malignancy characterized by varied responses to treatment and prognoses. Understanding the metabolic characteristics driving DLBCL progression is crucial for developing personalized therapies.

Methods: This study utilized multiple omics technologies including single-cell transcriptomics (n = 5), bulk transcriptomics (n = 966), spatial transcriptomics (n = 10), immunohistochemistry (n = 34), multiple immunofluorescence (n = 20) and to elucidate the metabolic features of highly malignant DLBCL cells and tumor-associated macrophages (TAMs), along with their associated tumor microenvironment. Metabolic pathway analysis facilitated by scMetabolism, and integrated analysis via hdWGCNA, identified glycolysis genes correlating with malignancy, and the prognostic value of glycolysis genes (STMN1, ENO1, PKM, and CDK1) and TAMs were verified.

Results: High-glycolysis malignant DLBCL tissues exhibited an immunosuppressive microenvironment characterized by abundant IFN_TAMs (CD68+CXCL10+PD-L1+) and diminished CD8+ T cell infiltration. Glycolysis genes were positively correlated with malignancy degree. IFN_TAMs exhibited high glycolysis activity and closely communicating with high-malignancy DLBCL cells identified within datasets. The glycolysis score, evaluated by seven genes, emerged as an independent prognostic factor (HR = 1.796, 95% CI: 1.077-2.995, p = 0.025 and HR = 2.631, 95% CI: 1.207-5.735, p = 0.015) along with IFN_TAMs were positively correlated with poor survival (p < 0.05) in DLBCL. Immunohistochemical validation of glycolysis markers (STMN1, ENO1, PKM, and CDK1) and multiple immunofluorescence validation of IFN_TAMs underscored their prognostic value (p < 0.05) in DLBCL.

Conclusions: This study underscores the significance of glycolysis in tumor progression and modulation of the immune microenvironment. The identified glycolysis genes and IFN_TAMs represent potential prognostic markers and therapeutic targets in DLBCL.

Keywords: Diffuse large B-cell lymphoma; Glycolysis; Metabolism; Single-cell transcriptomics; Spatial transcriptomics.

PubMed Disclaimer

Conflict of interest statement

The authors report there are no competing interests to declare. All authors agreed to submit for consideration for publication in this journal.

Figures

Fig. 1
Fig. 1
Flow chart of this study and identification of malignant B cell subgroups and CNV score comparison of scRNA-seq in GSE182434. A UMAP plot of cell types and samples distribution. B. Hallmark and pathways of different cell types determined by GSEA. C. UMAP plot of PCA clustering result of B malignant cells and other cell types grouping. D. Dot plot for expression levels of cell markers across B malignant subclusters (B0-B4). E. Chromosomal landscape of inferred CNVs among B malignant subclusters. F. Comparison of inferred CNV scores across B malignant subclusters. G. UMAP plot of cell types including high and low malignant B cells. H. Comparison of inferred CNV scores between high and low malignant B cell types. I. Hallmark and pathways of high and low malignant B cell types determined by GSEA. (Abbreviation: CNVs: copy number variations; scRNA-seq: single-cell RNA-sequencing; UMAP: uniform manifold approximation and projection; GSEA: gene set enrichment analysis; PCA: principal component analysis; DLBCL: diffuse large B-cell lymphoma; MB: malignant B cells)
Fig. 2
Fig. 2
Metabolism altas of samples in single-cell RNA-sequencing. A Metabolism enrichment of different cell types by heatmap. B. Correlation between inferred CNV scores and metabolism pathway scores. C. 18 metabolism pathway scores correlation with inferred CNV scores (r > 0.3 and p < 0.05). D. Clusters of metabolism pathway (n = 79) from benign B cells, low malignant B cells, to high malignant B cells. E. Comparison of 18 metabolism pathway scores among benign B cells, low malignant B cells, and high malignant B cells. F. UMAP plot of glycolysis / gluconeogenesis pathway score. G. Barplot of glycolysis / gluconeogenesis pathway scores in benign B cells and B0-B4 subgroups (Abbreviation: CNV: copy number variation; MB: malignant B cells; UMAP: uniform manifold approximation and projection. Mann-Whitney test was performed between groups.)
Fig. 3
Fig. 3
Identifcation of glycolysis / gluconeogenesis maker genes in high malignant B cells. A Volcano plot of differential genes among the benign B cells, low malignant B cells, and high malignant B cells. B. Functional analysis of highly expressed genes in high malignant B cells by Metascape. C. Identification of eight commonly genes overlapped across three groups and glycolysis hallmark genes. D. Module trait correlation showed the relationships between modules, CNV score, and Glycolysis score. E. Network visualization of 10 modules of high maligant B cells.(The modules highlighted in red and underlined are modules associated with CNV score and Glycolysis score.) F. The first 25 eigengenes of each module. G. Trajectory of different malignant B subclusters predicted by monocle. H. Genes expression level in single spot ordered along the pseudotime for MKI67 and seven glycolysis / gluconeogenesis gene markers (STMN1, ENO1, LDHA, TPI1, CDK1, PKM, and PPIA). (Abbreviation: HMB: high malignant B cells; CNV: copy number variation; UMAP: uniform manifold approximation and projection. *** p < 0.001.)
Fig. 4
Fig. 4
Macrophage subgroups identification and Cell-cell Communications in single-cell RNA-sequencing. (A) UMAP plot of PCA clustering result of macrophage and samples clustering. (B) Dot plot for cell marker expression levels. (C) Heatmap representation of top 20 highly variable transcription factor activities. (D) Top 5 higher metabolic pathways in IFN_TAMs compared with LA_TAMs. (E) Comparison of CXCL10, CCL2, CCL8, PD-L1, PD-L2, IL4I1, PFKFB3, TGFB1 and CD44 gene expression in LA_TAMs and IFN_TAMs. (F) Top 5 higher metabolic pathways in LA_TAMs compared with IFN_TAMs. (G) Comparison of CCL18, PTGDS, CHI3L1, APOE, APOC1, and ACP5 gene expression in LA_TAMs and IFN_TAMs. H-I. Heatmap of cell-cell communication network for incoming and outgoing signaling analysis. J. Heatmap of the relative importance of cell groups in the MIF signaling network based on four network centrality degrees. K. Circular plot of the quantity or intensity of interactions among various cell groups in PD-L1-PDCD1 and PD-L2-PDCD1 networks (Abbreviation: UMAP: uniform manifold approximation and projection; PCA: principal component analysis; IFN_TAMs: interferon-primed tumor-associated macrophages; LA_TAMs: lipid-associated tumor-associated macrophages; MIF: macrophage migration inhibitory factor. Mann-Whitney test was performed between groups. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns, not significant.)
Fig. 5
Fig. 5
Performance of seven glycolysis / gluconeogenesis markers in predicting OS in GSE181063 (n = 802) and GSE10846 (n = 164), and relationship between risk score and immune landscape in bulk-RNA seq. A Comparisons of seven glycolysis / gluconeogenesis genes’ mRNA expression in DLBCL (n = 47) and HCs (n = 491). B and E. Univariate and Multivariate Cox analysis for OS in GSE181063. C-D. Scatter and heatmaps for the seven markers-based risk score and Kaplan-Meier curves of OS in GSE181063. F and I. Time-dependent ROC curves for OS of the seven markers-based risk score in GSE181063 and GSE10846. G. Kaplan-Meier curves of OS based on seven markers-based risk score in GSE10846. H. Multivariate Cox analysis for OS in GSE10846. J. Comparisons of Estimate, Stromal and Immune scores among high riskscore and low riskscore patients in GSE181063. K. Distribution of 28 immune cell types in high and low risk groups in GSE181063. L. Correlation between riskscore and activated CD8+ T cell in GSE181063. (Abbreviation: DLBCL: diffuse large B-cell lymphoma; HC: healthy controls; OS: overall survival; ROC: receiver operating characteristic; ECOG: Eastern Cooperative Oncology Group; IPI: International Prognostic Index. Mann-Whitney test was performed between groups. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns, not significant.)
Fig. 6
Fig. 6
Performance of IFN_TAMs and LA_TAMs in predicting OS and association with PD-L1 in GSE181063 (n = 802) and GSE10846 (n = 164). (A) Kaplan-Meier analysis for OS based on IFN_TAMs and LA_TAMs (calculated by ssGSEA), and correlation between IFN_TAMs and PD-L1 in GSE181063. (B) Kaplan-Meier analysis for OS based on IFN_TAMs (calculated by ssGSEA) and LA_TAMs (calculated by ssGSEA), and correlation between IFN_TAMs and PD-L1 in GSE10846. (C) OS stratified by the glycolysis markers-based risk score combined with the IFN_TAM. (Abbreviation: IFN_TAMs: interferon-primed tumor-associated macrophages; LA_TAMs: lipid-associated tumor-associated macrophages; OS: overall survival; HH: IFN_TAMs high and glycolysis markers-based risk score high; M: IFN_TAMs high and glycolysis markers-based risk score low or IFN_TAMs low and glycolysis markers-based risk score high; LL: IFN_TAMs low and glycolysis markers-based risk score low.)
Fig. 7
Fig. 7
Metabolism altas and GLC score prognosis value validation of DLBCL in spatial transcriptomics. A-B. Metabolism enrichment of different cell types, and B cells and normal cells by heatmap. C. Highly expressed hallmark pathway scores of B cells using UMAP plot. D. Violin plot of glycolysis / gluconeogenesis pathway score across cell types. E. Comparison of GLC score in B cells and non-B cells. F. Spatial plot of GLC score in 10 DLBCL samples. G-H. Violin plot of GLC score in all cells and B cells between NR (n = 4), R (n = 2) and other samples (n = 4) (Abbreviation: GLC: glycolysis; DLBCL: diffuse large B-cell lymphoma; UMAP: uniform manifold approximation and projection; NB: non-B cell; R: relapsed patients, patients without EFS24; NR: non-relapsed patients, patients with EFS24. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns, not significant.)
Fig. 8
Fig. 8
DLBCL samples with high glycolysis / gluconeogenesis activity were characterized by immunosuppressive microenvironment in spatial transcriptomics. A-B. Vlnplot and representative spatial plots of IFN_TAMs between NR (n = 4), R (n = 2) and other samples (n = 4). C-D. PD-L1 expression and GLC score of TAMs in samples by violin plot. E. Vlnplot of PD-L1 expression in all cells between NR (n = 4), R (n = 2) and other samples (n = 4). F. Representative spatial plots of PD-L1 expression in NR (S3) and R (S10) group. G-H. UMAP and violin plot of activated CD8+ T score in NR (n = 4), R (n = 2) and other samples (n = 4), along with the correlation between GLC score and activated CD8+ T score. I. Representative spatial plots of activated CD8+ T score in NR (S3) and R (S10) group (Abbreviation: DLBCL: diffuse large B-cell lymphoma; UMAP: uniform manifold approximation and projection; TAMs: tumor-associated macrophages; GLC: glycolysis; R: relapsed patients, patients without EFS24; NR: non-relapsed patients, patients with EFS24. Mann-Whitney test was performed between groups. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, ns, not significant.)
Fig. 9
Fig. 9
Prognostic value of four glycolysis / gluconeogenesis (STMN1, ENO1, CDK1, PKM, and PPIA) proteins in IHC cohort (n = 34, 100X) and IFN_TAMs (CD68+CXCL10+PD-L1+) in mIF cohort (n = 20, 10X). A-B. Kaplan–Meier curves of OS and PFS according to STMN1, CDK1, ENO1 and PKM proteins expression. C. Representative IHC staining of STMN1, CDK1, ENO1 and PKM in patient 1 (PFS = 7 months, OS = 9 months) and patient 2 (PFS = 135 months, OS = 135 months). D. Kaplan-Meier analysis for PFS based on IFN_TAMs intensity. E. Comparison of IFN_TAMs intensity in relapse and non_relapse groups. F. Correlation of IFN_TAMs’ intensity, CD8+ T cells’ intensity, and TGFβ1 intensity. G. Representative mIF staining of IFN_TAMs and CD8+ T cells in patient #1 (PFS = 2.7 months) and patient #2 (PFS = 90 months). (Abbreviation: IHC: immunohistochemistry; IFN_TAMs: interferon-primed tumor-associated macrophages; mIF: multiple immunofluorescence; OS: overall survival; PFS: progression -free survival, Relapse: relapsed patients, patients without EFS24; Non_Relapse: non-relapsed patients, patients with EFS24. Mann-Whitney test was performed between groups.)
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Sehn LH, Salles G. N Engl J Med. 2021;384:842–58. doi: 10.1056/NEJMra2027612. - DOI - PMC - PubMed
    1. Ruppert AS, Dixon JG, Salles G, et al. International prognostic indices in diffuse large B-cell lymphoma: a comparison of IPI, R-IPI, and NCCN-IPI. Blood. 2020;135:2041–8. doi: 10.1182/blood.2019002729. - DOI - PubMed
    1. Calvo-Vidal MN. Cerchietti the metabolism of lymphomas. Curr Opin Hematol. 2013;20:345–54. doi: 10.1097/MOH.0b013e3283623d16. - DOI - PubMed
    1. Caro P, Kishan AU, Norberg E, et al. Metabolic signatures uncover distinct targets in molecular subsets of diffuse large B cell lymphoma. Cancer Cell. 2012;22:547–60. doi: 10.1016/j.ccr.2012.08.014. - DOI - PMC - PubMed
    1. Hanahan D. Weinberg Hallmarks of cancer: the next generation. Cell. 2011;144:646–74. doi: 10.1016/j.cell.2011.02.013. - DOI - PubMed