Single-cell immunophenotyping revealed the association of CD4+ central and CD4+ effector memory T cells linking exacerbating chronic obstructive pulmonary disease and NSCLC

Abstract

Introduction: Tobacco smoking generates airway inflammation in chronic obstructive pulmonary disease (COPD), and its involvement in the development of lung cancer is still among the leading causes of early death. Therefore, we aimed to have a better understanding of the disbalance in immunoregulation in chronic inflammatory conditions in smoker subjects with stable COPD (stCOPD), exacerbating COPD (exCOPD), or non-small cell lung cancer (NSCLC).

Methods: Smoker controls without chronic illness were recruited as controls. Through extensive mapping of single cells, surface receptor quantification was achieved by single-cell mass cytometry (CyTOF) with 29 antibodies. The CyTOF characterized 14 main immune subsets such as CD4+, CD8+, CD4+/CD8+, CD4-/CD8-, and γ/δ T cells and other subsets such as CD4+ or CD8+ NKT cells, NK cells, B cells, plasmablasts, monocytes, CD11c^dim, mDCs, and pDCs. The CD4+ central memory (CM) T cells (CD4+/CD45RA-/CD45RO+/CD197+) and CD4+ effector memory (EM) T cells (CD4+/CD45RA-/CD45RO+/CD197-) were FACS-sorted for RNA-Seq analysis. Plasma samples were assayed by Luminex MAGPIX^® for the quantitative measurement of 17 soluble immuno-oncology mediators (BTLA, CD28, CD80, CD27, CD40, CD86, CTLA-4, GITR, GITRL, HVEM, ICOS, LAG-3, PD-1, PD-L1, PD-L2, TIM-3, TLR-2) in the four studied groups.

Results: Our focus was on T-cell-dependent differences in COPD and NSCLC, where peripheral CD4+ central memory and CD4+ effector memory cells showed a significant reduction in exCOPD and CD4+ CM showed elevation in NSCLC. The transcriptome analysis delineated a perfect correlation of differentially expressed genes between exacerbating COPD and NSCLC-derived peripheral CD4+ CM or CD4+ EM cells. The measurement of 17 immuno-oncology soluble mediators revealed a disease-associated phenotype in the peripheral blood of stCOPD, exCOPD, and NSCLC patients.

Discussion: The applied single-cell mass cytometry, the whole transcriptome profiling of peripheral CD4+ memory cells, and the quantification of 17 plasma mediators provided complex data that may contribute to the understanding of the disbalance in immune homeostasis generated or sustained by tobacco smoking in COPD and NSCLC.

Keywords: CD4 central memory T cells; CD4 effector memory T cells; exacerbating COPD; non-small cell lung cancer; single-cell mass cytometry; stable COPD; tobacco smoking.

Figure 1

Demonstration of the studied main immune subsets. (A) Unsupervised clustering and the heatmap of expression intensities of the analysed immune cells. (B) The FlowSOM (Self-Organizing Maps for flow cytometry) and UMAP (Uniform Manifold Approximation and Projection for Dimension Reduction) analysis identified 14 main subsets as follows: CD4+ T cells, CD8+ T cells, CD4+CD8+ T cells, DN (double-negative) T cells (CD4−CD8−), γ/δ T cells, NK cells, CD4+ NKT cells, CD8+ NKT cells, plasmablasts, B cells, monocytes, CD11c^dim cells, plasmacytoid dendritic cells (pDCs), and monocytoid dendritic cells (mDCs). The analysis was carried out on the entire dataset in the R software including 33 FCS files.

Figure 2

The population frequency (A–C) and marker expression profile of CD4+ T cells (D, E). The differences in the percentage of CD4+ CM (F) and CD4+ EM cells (G). *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3

The population frequency (A–G) and marker expression profile of CD8+ T cells (H–L). *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4

Marker expression (A) CD25, (B) CD27, (C) CD8 profile of CD4+CD8+ T-cells. p *<0.05, **<0.01.

Figure 5

The population frequency of the metaclusters in CD4−CD8− (DN) T cells (A–C) and the marker expression profile of (D) CD28 and (E) CD183 on the surface of DN T cells. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

The population frequency of (A) CD27+ or (B) CD183+ MCs in γ/δ T-cells. p *<0.05, **<0.01.

Figure 7

Results of gene set enrichment analyses of CD4+ central memory and effector memory T cells. The figure shows those Hallmark and KEGG gene sets that had similar trends in NSCLC vs. smoker healthy controls (SmHCs) and exacerbated COPD vs. SmHC comparisons. Positive normalized effect score (NES) values indicate enrichment in control. Point sizes indicate false discovery rate (FDR) value ranges.

Figure 8

The analysis of the concentration of immuno-oncology mediators of the plasma of the patients. (A) Analyte concentrations in different samples are shown color-coded. Concentration levels have been normalized using the Z-score transformation. Three main sample clusters were identified using hierarchical clustering. (B) The distribution of patient categories in the three clusters. (C) The distribution of clusters among different patient categories. Clusters 1 and 2 can be characterized by the highest and lowest analyte levels, respectively. Cluster 3 shows intermediate analyte levels and contains the largest fraction of NSCLC samples.

Figure 9

The schematic cartoon of our result focusing on three levels: genes, proteins, and cells. The transcriptome of CD4+ CM and CD4+ EM cells showed a correlation in exCOPD and NSCLC. The three most differentially expressed genes are highlighted. Next, soluble mediators were measured, where CD86 and GITRL were higher in the plasma of exCOPD patients, while TIM-3 and PD-L2 were lower in exCOPD patients. Lastly, single-cell immunophenotyping revealed alterations in the composition of peripheral compartments, where the frequency of the immune subsets was quantified and demonstrated characteristics for exCOPD and NSCLC.