Fig 2: Extended predictive value of serum CXCL17 levels in influenza patients and cellular sources. (A) Multiple correlation analysis of the clinical variables and serum CXCL17 levels of influenza patients (n = 68). The heat color map gradient was constructed using Spearman R values. (B) Serum CXCL17 levels in influenza patients according to the severity of their acute respiratory distress syndrome (ARDS) on admission (mild, PaO2/FiO2 >200, n = 3; moderate, PaO2/FiO2 100–200, n = 27; severe, PaO2/FiO2 <100, n = 38; Kruskal–Wallis test and post hoc Dunn test). (C) Comparison of the levels of CXCL17 between COVID-19 patients that survived (n = 14) or died (n = 10). Mann–Whitney U test. (D) Serum levels of CXCL17 were compared between influenza patients that developed acute kindly injury (AKIN, n = 19), and individuals with normal renal function (n = 49). (E) Differences in CXCL17 levels between influenza patients according to their need for renal replacement therapy (RRT; n = 52 vs 16). (B–E) Plots display medians and interquartile ranges. Mann–Whitney U test. (F) Receiver operating characteristic (ROC) curve of the CXCL17 levels in influenza patients that developed AKIN and those that maintained normal renal function. The graph display area under the curve (AUC) and 95% CI interval values. (G) ROC curve analysis of the serum CXCL17 levels in patients requiring RRT. (H) Human A549 alveolar epithelial cells and (I) peripheral blood-monocyte derived macrophages were infected with a clinical isolate of the influenza A(H1N1) pdm09 virus for 24, 48, and 72 h n = 6–7 per group per time point). Levels of CXCL17 in supernatants from infected and uninfected cells were compared using the unpaired Student’s t test (n = 5 to 9 per group) at each time point. Values of p were corrected for multiple comparisons using the Holm–Sidak method. The data represent mean ( ± SE) values from 2 to 3 independent experiments per time point and experimental condition. *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001.

Fig 3: CXCL17 in the lung and serum of influenza, COVID-19, and PTB patients. (A) Morphological changes observed in the lung of influenza and COVID-19 patients (n = 2 per group). H&E, ×400. The expression of CXCL17 was analyzed in lung autopsy specimens from influenza (left lower panel) and COVID-19 (right lower panel). IHQ, ×400. (B) Serum CXCL17 levels in healthy donors (HD; n = 30) patients with influenza (n = 68), COVID-19 subjects (n = 24), and PTB patients (n = 30). Kruskal–Wallis test and post hoc Dunn test (plots display medians with interquartile ranges; ****p = 0.0001). (C) K-means clustering analysis of the clinical characteristics of influenza and COVID-19 patients. (D) Random forest algorithm showing the most important factors that differentiate influenza from COVID-19. The points represent the mean decrease Gini values, indicative of the importance of each variable with respect to the mean importance of the model (discontinuous vertical line). (E) Receiver operating characteristic (ROC) curve of the levels of CXCL17 in influenza and COVID-19 subjects. The graph displays area under the curve (AUC) and 95% CI interval values.