Identification of potential biomarkers and immune infiltration characteristics in severe asthma

Abstract

Objectives: We hope to identify key molecules that can be used as markers of asthma severity and investigate their correlation with immune cell infiltration in severe asthma.

Methods: An asthma dataset was downloaded from the Gene Expression Omnibus database and then processed by R software to obtain differentially expressed genes (DEGs). First, multiple enrichment platforms were applied to analyze crucial biological processes and pathways and protein-protein interaction networks related to the DEGs. We next combined least absolute shrinkage and selection operator logistic regression and the support vector machine-recursive feature elimination algorithms to screen diagnostic markers of severe asthma. Then, a local cohort consisting of 40 asthmatic subjects (24 with moderate asthma and 16 with severe asthma) was used for biomarker validation. Finally, infiltration of immune cells in asthma bronchoalveolar lavage fluid and their correlation with the screened markers was evaluated by CIBERSORT.

Results: A total of 97 DEGs were identified in this study. Most of these genes are enriched in T cell activation and immune response in the asthma biological process. CC-chemokine receptor 7 (CCR7) and natural killer cell protein 7(NKG7) were identified as markers of severe asthma. The highest area under the ROC curve (AUC) was from a new indicator combining CCR7 and NKG7 (AUC = 0.851, adj. p < 0.05). Resting and activated memory CD4 T cells, activated NK cells, and CD8 T cells were found to be significantly higher in the severe asthma group (adj. p < 0.01). CCR7 and NKG7 were significantly correlated with these infiltrated cells that showed differences between the two groups. In addition, CCR7 was found to be significantly positively correlated with eosinophils (r = 0.38, adj. p < 0.05) infiltrated in bronchoalveolar lavage fluid.

Conclusion: CCR7 and NKG7 might be used as potential markers for asthma severity, and their expression may be associated with differences in immune cell infiltration in the moderate and severe asthma groups.

Keywords: asthma treatment; eosinophil asthma; eosinophils; glucocorticoid; immune infiltration.

Figure 1.

Volcano plot of differentially expressed genes in MA and SA BALF samples. Blue dots represent significantly downregulated genes in SA samples, and red dots represent significantly upregulated genes. MA, moderate asthma; SA, severe asthma; BALF, bronchoalveolar lavage fluid.

Figure 2.

Gene set enrichment analysis of whole gene values between MA and SA using the h.all.v 6.2.symbols.gmt [Hallmarks] gene set database. NES, normalized enrichment score. Significant gene sets were cut off by FDR < 0.25 and p value < 0.05.

Figure 3.

Enrichment of the 97 differentially expressed genes in the asthma samples. (a), Bar chart of the top five biological processes sorted using DAVID GO terms with respect to the value of adj.p to which they correspond. (b), GO terms visualized using the ClueGO/CluePedia plug-in from Cytoscape. The values of p ≤ 0.05 indicate the node size and the node color code corresponding to the specific functional class in which they participate. DAVID, Database for Annotation, Visualization and Integrated Discovery.

Figure 4.

Functional enrichment analysis tool screened six biological pathways with a p value < 0.05.

Figure 5.

Construction of a PPI network of DEGs by STRING analysis. (a), Construction of a PPI network of DEGs by STRING analysis and molecules related to the same signaling are colored the same. (b), STRING analysis shows the interaction between genes in cluster 1; molecules related to immune response are colored in red, molecules related to leukocyte activation are colored in green, molecules related to regulation of immune system processes are colored in purple.

Figure 6.

Screening and validation of biomarkers for asthma severity from genes in cluster 1. (a), Least absolute shrinkage and selection operator (LASSO) logistic regression algorithm to screen diagnostic markers. Each color represents an individual gene. (b), Diagnostic marker screened by the support vector machine-recursive feature elimination (SVM-RFE) algorithm. (c), Venn diagram illustrating the intersection of diagnostic markers obtained by the two algorithms. (d), ROC curves of CCR7 and NKG7 in asthma severity prediction. CCR7 (red line), NKG7 (yellow line), and index of the two genes combined (dark line). (e), CCR7 and NKG7 expression of PBMC from healthy volunteers (n = 6) with vehicle or with 10⁻⁵ M dexamethasone for 12 and 24 h.

Figure 7.

Immune cell infiltration in asthma BALF. (a), Principal component analysis (PCA) plot of immune cell infiltration between moderate asthma cases and severe asthma cases. (b), Correlation heatmap of infiltrated immune cells. The area size of squares indicates the strength of the correlation, with red representing a positive correlation and blue representing a negative correlation. (c), Violin diagram of the proportion of immune cells in the two asthma groups. (d), Correlation of CCR7 and NKG7 with infiltrated immune cells in asthmatic BALF. Blank squares indicate a nonsignificant correlation, red colors indicate a positive correlation, and blue colors indicate a negative correlation. p < 0.05 was considered statistically significant.