High-Dimensional Analysis of Type 2 Lymphocyte Dynamics During Mepolizumab or Dupilumab Treatment in Severe Asthma

Abstract

Background: Although the type 2 biologics mepolizumab and dupilumab show clinical efficacy in severe asthma, their influence on circulating lymphocytes is largely unknown. Here, we studied their impact on type 2 lymphocytes in severe asthma.

Methods: We performed high-parameter flow cytometry analysis of peripheral blood mononuclear cells from 40 patients with severe asthma before, and after 4 and 12 months of mepolizumab (n = 33) or dupilumab (n = 7) treatment, focusing on type 2 lymphocytes. Additionally, we performed single-cell RNA sequencing (scRNA-seq) (n = 3) and stimulation experiments of type 2 lymphocytes (n = 3) to explore transcriptional and functional changes associated with mepolizumab treatment.

Results: Mepolizumab treatment increased circulating type 2 innate lymphoid cell (ILC2), type 2 T helper (Th2) and type 2 cytotoxic (Tc2) cell frequencies, skewing ILC2 towards a CD117^low signature with high CD62L expression, and Th2/Tc2 cells towards a CD45RA^-CD62L⁺ central memory phenotype. Dupilumab-treated patients also showed increased frequencies of total ILC2 and CD117^low ILC2. Mepolizumab treatment reduced the expression of tissue homing receptors CXCR4 in ILC2, and GPR183 in ILC2, Th2, and Tc2 cells while enhancing their type 2 cytokine producing capability in response to alarmins.

Conclusion: Mepolizumab increases the frequencies of circulating ILC2, Th2, and Tc2 cells, with reduced tissue homing receptor expression but increased type 2 cytokine production potential. This reveals a potentially new mechanism for how mepolizumab reduces airway inflammation by re-directing trafficking of inflammatory type 2 lymphocytes away from airway-homing, with implications for the possibility of achieving biologics-free remission in asthma.

Keywords: ILC2; Tc2; Th2; biologics; severe asthma.

FIGURE 1

(A) Study schematic. (B) Patient treatment response to mepolizumab therapy (n = 33). Statistical analysis of the change in exacerbation frequency at baseline and after 1 year of biologics therapy was performed using the Wilcoxon signed‐rank test. (C) Total lymphocyte counts from whole blood before and after starting mepolizumab therapy (n = 33). (D) Total ILC, and bulk CD4⁺ and CD8⁺ T cell frequencies in patients before and after starting mepolizumab therapy (n = 33). (E) Final flow cytometry gates used for the identification of ILC2, Th2, and Tc2 cells. (F) Scatter plots and correlation analysis (two‐tailed Spearman) between blood eosinophil numbers and CRTH2 expression on CD4⁺ T or Tc2 cells or (G) Th2 frequency or (H) between ILC2 frequency and total serum IgE at study baseline (n = 33). (I) Eosinophil numbers and type 2 lymphocyte frequencies in SA patients with and without OCS medication at study baseline (n = 40). Both groups were compared using Mann–Whitney U test. (J, K) ILC2, Th2 and Tc2 cell frequencies in patients before and after starting mepolizumab therapy (n = 33). If not stated otherwise, all statistics were performed using Friedman test with Dunn's multiple comparisons. Violin plots include the median and interquartile range. The percentage change in cell frequency between 1 year and baseline mean is included for selected figures.

FIGURE 2

(A) UMAP of ILC2 from SA patients treated with mepolizumab (all time‐points combined) with Phenograph clusters overlaid. (B) Distribution of surface marker expression across the ILC2 UMAP. (C) 2D kernel density estimation across the ILC2 UMAP (all time‐points combined). (D) Nearest‐neighbor analysis, determining the association for each cell within the overall ILC2 UMAP with areas enriched for cells from either baseline (blue) or 1 year of mepolizumab treatment (red). (E) Flow cytometry gating on CD117^high and CD117^low ILC2 with CD117 FMO. (F) CD117^high and CD117^low ILC2 frequency in biologics‐naïve SA patients (n = 40). Statistics was performed using the Wilcoxon signed‐rank test. (G) CD117 expression level and CD117^low and CD117^high ILC2 frequency (H) before and after starting mepolizumab treatment (n = 33). (I) CD117^low/CD117^high ILC2 ratio dynamics in patients during mepolizumab treatment (n = 33). (J) Frequency of CD62L⁺ and CCR6⁺ cells within CD117^high or CD117^low ILC2 in biologics‐naïve SA patients (n = 40). Groups were compared using the Wilcoxon signed‐rank test. If not stated otherwise, all statistics were performed using Friedman test with Dunn's multiple comparisons. Violin plots include the median and interquartile range. The percentage change in cell frequency between 1 year and baseline mean is included for selected figures.

FIGURE 3

(A) UMAP of Th2 cells from SA patients treated with mepolizumab (all time‐points combined) with Phenograph clusters overlaid. (B) Distribution of surface marker expression across the Th2 cell UMAP. (C) 2D kernel density estimation across the Th2 cell UMAP (all time‐points combined). (D) Nearest‐neighbor analysis, determining the association for each cell within the overall Th2 UMAP with areas enriched for cells from either baseline (blue) or 1 year of mepolizumab treatment (red), depicting changes in Th2 composition in the context of mepolizumab treatment. (E) CD62L expression levels in Th2 cells and (F) CD62L⁺ Th2 cell frequency during mepolizumab treatment (n = 33). (G, H) Frequency of Th2 memory cell subsets before and after the start of mepolizumab treatment (n = 33). All statistics were performed using Friedman test with Dunn's multiple comparisons. Violin plots include the median and interquartile range. The percentage change in cell frequency between 1 year and baseline mean is included for selected figures.

FIGURE 4

(A) UMAP of Tc2 cells from SA patients treated with mepolizumab (all time‐points combined) with Phenograph clusters overlaid. (B) Distribution of surface marker expression across the Tc2 cell UMAP. (C) 2D kernel density estimation across the Tc2 cell UMAP (all time‐points combined). (D) Nearest‐neighbor analysis, determining the association for each cell within the overall Tc2 UMAP with areas enriched for cells from either baseline (blue) or 1 year of mepolizumab treatment (red), depicting changes in Tc2 composition in the context of mepolizumab treatment. (E) CD62L expression levels in Tc2 cells and (F) CD62L⁺ Tc2 cell frequency during mepolizumab treatment (n = 33). (G, H) Frequency of Tc2 memory cell subsets before and after the start of mepolizumab treatment (n = 33). All statistics were performed using Friedman test with Dunn's multiple comparisons. Violin plots include the median and interquartile range. The percentage change in cell frequency between 1 year and baseline mean is included for selected figures.

FIGURE 5

(A) UMAP visualization of scRNA‐seq data of sorted ILC2, Th2, and Tc2 cells from 3 SA patients at baseline and after 1 year of mepolizumab treatment. (B) Expression of hallmark genes for the identification and distinction between ILC2, Th2, and Tc2 cells in the scRNA‐seq data. (C) Volcano plots showing the differential gene expression in ILC2, Th2, and Tc2 cells between baseline and 1 year of mepolizumab treatment. (D) Venn diagrams depicting the number of selectively and shared differentially expressed genes after 1 year of mepolizumab treatment as compared to baseline. (E) Expression of chemotactic receptors GPR183 and CXCR4 before and after 1 year of mepolizumab therapy. (F) Expression of HPGDS in ILC2 before and after treatment. (G) Urinary levels of type 2 lipid mediator biomarkers at baseline and after 1 year of mepolizumab treatment. (H) Expression of AP‐1 family transcripts in type 2 lymphocytes before and after 1 year of mepolizumab treatment. (I) Expression and (J) quantification of IL‐5 and IL‐13 in type 2 lymphocytes after 6 h of ex vivo stimulation with alarmins and PMA/ionomycin of whole PBMC samples from patients at baseline and after 1 year of mepolizumab treatment (n = 3).