Fig 1.
Repeatability of nasal curettage sampling.
Nasal cells were collected from the left (L) and right (R) nostril of three volunteers, processed independently and their composition was assessed by flow cytometry. After excluding debris and doublets, epithelial cells were identified by Epcam expression. A viability dye and CD45 were used to identify live immune cells. Among those cells, side scatter, CD66b, CD14 and CD3 were used to identify granulocytes, monocytes and T cells respectively.
Fig 2.
Nasal curettage yields reproducible and consistent results over time.
(A) The percentage of granulocytes (closed circles) and T cells (open circles) in 218 nasal cell samples collected over a five month period (n = 117 volunteers, sampled up to five times). Individual samples and loess curves are depicted for both populations. (B, C) The correlation for individuals in four repeated measurements over a 33-day period for (B) granulocytes and (C) T cells.
Fig 3.
Comparison of samples collected by nasal wash and nasal curette.
(A) Epithelial (open circles) and immune (closed circles) and cell yields were compared between nasal wash pellets and nasal curette samples. Individuals samples and median and interquartile range are shown. (B) Median proportions of granulocytes, T cells, monocytes, lineage- HLA-DR+ and uncharacterized cells among immune cells in nasal curette (n = 139 individuals) and nasal wash (n = 8) samples. *p < 0.05, ***p < 0.001, ****p < 0.0001 Mann-Whitney test.
Fig 4.
Comparison of samples from nasal mucosa and blood.
(A) Median proportions of granulocytes, T cells, monocytes, lineage- HLA-DR+ and uncharacterized cells among immune cells in blood (n = 10) and nasal curette (n = 139). **** p < 0.0001 Mann-Whitney test. (B) The percentage of HLA-DR+ T cells in blood and nasal curette samples and mean fluorescent intensity (MFI) of HLA-DR and CD66b on granulocytes was measured for blood, nasal curette and nasal wash (n = 8) samples. Median and interquartile range are shown. *p < 0.05, ***p < 0.001 Kruskal-Wallis, followed by Dunn’s Multiple Comparison Test. (C) Multi-dimensional scaling analysis shows the clustering of samples from blood (grey circles), nasal curette (open squares, 11 randomly selected) and nasal wash (black triangles). The epithelial cell yield, activation state of granulocytes and composition of the immune cells were taken into account. Kruskal stress = 5.8% and Analysis of Similarity ANOSIM p-value = 0.001.
Fig 5.
Comparison of nasal curettes with visible red blood cell contamination (n = 55) or not (n = 62).
(A) Median proportions of granulocytes, T cells, monocytes, lineage- HLA-DR+ and uncharacterized cells among immune cells. (B) The percentage of HLA-DR+ T cells (left axis) and mean fluorescent intensity (MFI) of HLA-DR and CD66b on granulocytes (right axis) of nasal curette samples visibly contaminated with erythrocytes (open squares) or not (open circles). Individuals samples and median and interquartile range are shown.
Fig 6.
Comparison of cytokine levels in samples collected by nasal wash and using nasosorption devices.
(A) Levels of IL-8, IP-10 and MCP1 were measured by Luminex in nasosorption devices and by ELISA from paired nasal washes (n = 41). The r2 and p indicate goodness-of-fit and p-value calculated by linear regression analysis (B) A heat map depicts log-transformed cytokine concentrations, with white and black indicating low and high levels, respectively. Each of the columns corresponds to one sample (n = 6 nasal wash and nasosorption) and each of the rows to one cytokine. A legend assigning color gradient to log-transformed cytokine levels is displayed. (C) The ratio of cytokine concentrations measured in paired nasosorption and nasal wash (median and IQR are shown, n = 6).
Fig 7.
Tolerability of novel nasal sampling methods.
The percentage of volunteers rating (A) nasal curettage and (B) nasosorption on discomfort, pain and lacrimation.