Aberrant immune programming in neutrophils in cystic fibrosis

Abstract

Cystic fibrosis is a life-shortening genetic disorder, caused by mutations in the gene that encodes cystic fibrosis transmembrane-conductance regulator, a cAMP-activated chloride and bicarbonate channel. Persistent neutrophilic inflammation is a major contributor to cystic fibrosis lung disease. However, how cystic fibrosis transmembrane-conductance regulator loss of function leads to excessive inflammation and its clinical sequela remains incompletely understood. In this study, neutrophils from F508del-CF and healthy control participants were compared for gene transcription. We found that cystic fibrosis circulating neutrophils have a prematurely primed basal state with significantly higher scores for activation, chemotaxis, immune signaling, and pattern recognition. Such an irregular basal state appeared not related to the blood environment and was also observed in neutrophils derived from the F508del-CF HL-60 cell line, indicating an innate characteristic of the phenotype. Lipopolysaccharides (LPS) stimulation drastically shifted the transcriptional landscape of healthy control neutrophils toward a robust immune response; however, cystic fibrosis neutrophils were immune-exhausted, reflected by abnormal cell aging and fate determination in gene programming. Moreover, cystic fibrosis sputum neutrophils differed significantly from cystic fibrosis circulating neutrophils in gene transcription with increased inflammatory response, aging, apoptosis, and necrosis, suggesting additional environmental influences on the neutrophils in cystic fibrosis lungs. Taken together, our data indicate that loss of cystic fibrosis transmembrane-conductance regulator function has intrinsic effects on neutrophil immune programming, leading to premature priming and dysregulated response to challenge.

Keywords: cystic fibrosis; f508del-CFTR mutation; peripheral blood neutrophils; single-cell RNA sequencing; sputum neutrophils; transcriptomic analysis.

Fig. 1.

Atlas and transcriptional signatures of peripheral blood neutrophils from patients with CF and HCs. (A) UMAP of peripheral blood neutrophils from their basal state and LPS-stimulated activating state. There are 4 groups of cells marked by corresponding colors: (i) HC; (ii) HC_LPS; (iii) CF; and (iv) CF_LPS. (B) UMAP of neutrophils from HC and HC_LPS groups. (C) UMAP of neutrophils from CF and CF_LPS groups. (D) Venn diagram showing similarities of top expressed genes among the 4 groups of neutrophils. (E) Heatmap showing the top 30 DEGs from each group. Selected immune related genes are marked. (F) Dotplot showing GO enrichment of DEGs from 4 groups.

Fig. 2.

Gene scores for immune properties of CF and HC neutrophils at rest and upon LPS stimulation. Boxplots showing neutrophil immune scores of HC and CF neutrophils in resting (− LPS) and LPS stimulated (+ LPS) conditions. (A) Neutrophil activation score (GO:0042119). (B) Neutrophil chemotaxis score (GO:0030593). (C) Score for cytokine-mediated signaling pathway (GO:0019221). (D) Score for neutrophil migration (GO:1990266). (E) Score for immune response-regulating signaling pathway (GO:0002764). (F) Score for Toll-like receptor signaling pathway (GO:0002224). (G) Score for pattern recognition receptor signaling pathway (GO:0002221). (H) Score for response to LPS (GO:0032496). ANOVA test was used to judge statistical significance between 2 comparing groups. **P < 0.01, ***P < 0.001. Red asterisks indicate that CF neutrophils have a significantly higher score and blue asterisks a significantly lower score in each comparison with HC counterparts. ns, no significant difference.

Fig. 3.

Cell state transition of CF and HC neutrophils in response to LPS. (A) Monocle trajectory of HC neutrophils colored by states (basal, activating, and aging). (B) Monocle trajectory of HC neutrophils colored by groups (HC and HC_LPS). (C) UMAP of the neutrophils from HC and HC_LPS groups, colored by states. (D) GO enrichment of DEGs in each state of HC neutrophils. (E) Monocle trajectory of CF neutrophils, colored by states. (F) Monocle trajectory of CF neutrophils, colored by groups. (G) UMAP of the neutrophils from CF and CF_LPS groups, colored by states. (H) Proportions of the activating state and aging state in the population of HC_LPS or CF_LPS cells. (I, J) Venn diagrams showing shared top expressed genes between HC neutrophils and CF neutrophils in the activating state (I) or aging state (J).

Fig. 4.

Cell fate decision of CF and HC neutrophils in response to LPS. (A) RNA velocity plot revealing the origin and interrelationship of 3 states of HC neutrophils. (B) UMAP-embedding root cells and end point of HC neutrophils. (C) RNA velocity plot showing the origin and interrelationship of 3 states of CF neutrophils. (D) UMAP-embedding root cell and end point of CF neutrophils. (E) Heatmap showing upregulated genes (red) and downregulated genes (blue) in CF samples in the activating state or aging state, as compared to HC. (F–H) Comparisons of neutrophil property and function scores between HC and CF neutrophils in the activating state or aging state. (F) Neutrophil activation scores. (G) Differentiation and aging scores. (H) Granule formation scores. Significance was determined by 2-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001. Red asterisks indicate that CF neutrophils have significantly higher scores, and blue asterisks significantly lower scores, as compared to HC neutrophils. ns, no significant difference.

Fig. 5.

Preactivation of neutrophils derived from CF HL-60 cells. (A) UMAP of WT differentiated HL-60 cells (WT_dHL-60). (B) UMAP of CF differentiated HL-60 cells (CF_dHL-60). (C) Dotplot showing GO enrichment of DEGs in 4 groups. (D–K) Boxplots showing activation scores of neutrophils derived from WT and CF HL-60 cells with or without LPS stimulation. The significance was determined by 2-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001. Red asterisks indicate that the derived CF neutrophils have significantly higher scores, and blue asterisks significantly lower scores, as compared to the derived WT neutrophils. ns, no significant difference.

Fig. 6.

Gene reprogramming in CF sputum neutrophils as compared to CF blood neutrophils. (A) UMAP of 3 groups of CF PMNs: (i) CF blood PMNs in their native state, (ii) CF blood PMNs stimulated with LPS, and (iii) CF sputum PMNs. (B) Heatmap showing the top 20 DEGs from each group of CF neutrophils. Selected immune-related genes are marked. (C) GO enrichment comparisons of DEGs among the 3 groups of CF neutrophils. (D) KEGG pathway analysis of DEGs from each group of CF neutrophils.

Fig. 7.

Comparisons of immune-related properties between CF blood neutrophils and CF sputum neutrophils. (A–D) Comparisons of neutrophil property and function scores among CF blood PMNs, CF blood PMN_LPS, and CF sputum PMNs. (A) Differentiation and aging scores. (B) Granule formation scores. (C) Neutrophil activation scores. (D) Inflammation scores. Significance was determined by 1-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001. Red asterisks indicate that the latter group has significantly higher scores, and blue asterisks significantly lower scores, as compared to the former group in each comparison. Gray lines in each violin plot represent the median.