METHODS
Human subjects and samples. Data were collected from 11 2-year-old children at stable clinic visits. Subjects with CF were enrolled in the IMPEDE-CF study at Emory University and Children’s Healthcare of Atlanta (Atlanta, GA, USA). All aspects of subject enrollment and sample collection were approved by the Emory University Institutional Review Board (IRB00097352). Consent for sample collection from subjects was obtained from parents on the day of the clinic visit. Subject demographics are summarized in Table 1 . Collection of IS, blood, chest CT scan and BAL from the RML and LIN, were completed in that order within the same 4-hour visit. Samples were stored on ice following collection and delivered to the laboratory for immediate processing. Details of sample collection procedures are provided in online supplement and Figure S1 , with sample yields indicated in Table S1 .
Chest CT imaging and PRAGMA scoring. Chest CT scans were obtained under general anesthesia at Children’s Healthcare of Atlanta [22] with full inspiratory and expiratory breath hold maneuvers, and scored using the validated PRAGMA-CF method [23]. Scoring of mucus plugging (%MP), bronchiectasis (%Bx) and abnormal airways (%AA) after excluding areas of atelectasis were done with inspiratory scans, and scoring of trapped air (%TA) used expiratory scans. Total Disease (%Dis) score was calculated by combining all abnormal areas as a percentage of total scored CT sections after excluding areas of atelectasis.
Total protein quantification. Total protein levels were quantified in IS and BAL using the Pierce copper sulfate/bicinchoninic acid (BCA) assay (ThermoFisher), with bovine serum albumin used for calibration.
Soluble immune mediator quantification. A custom assortment of 20 soluble immune mediators was measured in IS and BAL using a highly sensitive chemiluminescent assay (U-PLEX; Meso Scale Diagnostics), according to the manufacturer’s protocol. To enable comparison between IS and BAL samples, mediator concentrations were normalized to total protein concentration (measured using BCA assay), as illustrated in Figures 1 and 2 .
Flow cytometry. Multicolor flow cytometry was performed on cells from blood, BAL, and IS using previously described methods [11]. A gating strategy was developed to identify all major leukocyte populations in blood and airway fluids, as detailed inFigure S4 . Additional experimental details are presented in Supplemental Methods. Consistent fluorescence output was achieved through stringent cytometer calibration techniques, as detailed before [24].
Extracellular NE activity. Extracellular NE activities in IS, RML BAL and LIN BAL were measured with a Förster resonance energy transfer (FRET) assay using the NEmo-1 probe (Sirius Fine Chemicals SiChem GmbH), as previously described [25-27]. NE concentration was normalized to total protein concentration (measured using BCA assay) in the fluid.
Statistical analysis. Data were analyzed in Prism (GraphPad, version 8) using nonparametric statistics due to the limited number of samples, including the Wilcoxon matched-pairs signed rank test and Spearman test for correlation. Specifically for cytokines, a value of half the lower limit of detection or twice the upper limit of detection was assigned for data points which fell outside the limits of detection. Those values are clearly labeled as closed symbols in related figures. To avoid over-representation of imputed values, statistical comparisons of mediator concentration were not performed where more than half the data points of a sample group were imputed values. To conduct correlations, imputed values were removed, and correlations were conducted using R only for mediators with at least 6 non-imputed data points.