Demographic and clinical characteristics of the subjects
Demographic and clinic characteristics of all participants enrolled in this study are presented in the Additional file 1: Table S1. There was no significant difference with regard to age, gender and smoker status between the 3 groups. Peripheral blood eosinophils and total IgE were increased in both subtypes of CRSwNP patients compared to control subjects. Assessment for the significance of differences between CRSwNP+AS group and CRSwNP-alone group indicated that atopy and recurrence were significantly more frequent in patients with CRSwNP+AS than in patients with CRSwNP-alone (62.5% vs 28.6% and 66.7% vs 26.9%, respectively). Patients with CRSwNP+AS also had a significantly decreased forced exhalation volume in one second (FEV1) / forced vital capacity (FVC) ratio and an increased fractional exhaled nitric oxide (FeNO) compared to patients with CRSwNP-alone. Similarly, CRSwNP+AS patients had a significantly higher percentage of eosinophils (6.45%) and total IgE (143.00 kU/l) in peripheral blood compared to CRSwNP-alone patients (3.00% and 53.90 kU/l, respectively; Figs. 1a and b).
Local features of nasal tissue in different patient group
To identify the local inflammatory patterns of nasal tissue in CRSwNP patients with and without comorbid asthma, the infiltrating eosinophils, neutrophils, plasma cells, and lymphocytes were assessed. Tissue sections stained with H&E demonstrated that patients with CRSwNP+AS had significantly more eosinophils in nasal polyp tissues than patients with CRSwNP-alone (Figs. 1c, d and e).
Given the relationship between IgE and Type 2 immune response, we further examined the total IgE levels in nasal tissues and found that IgE levels in nasal tissues from CRSwNP+AS patients were significantly higher compared to IgE levels in nasal tissues of CRSwNP-alone patients and controls (Fig. 1f). Moreover, Spearman correlation test showed that the percentage of local eosinophils was positively correlated with the concentration of tissue IgE (Additional file 1: Figure S1).
Whole transcriptome profiling of nasal tissues from CRSwNP+AS and CRSwNP-alone
To identify the gene expression profiles of CRSwNP with and without comorbid asthma, RNA sequencing was performed on nasal tissue samples from control, CRSwNP-alone and CRSwNP+AS patients. Analysis of DE-mRNAs and DE-lncRNAs demonstrated that there were 5218 DE-mRNAs and 2949 DE-lncRNAs between CRSwNP-alone and control, and 2512 DE-mRNAs and 464 DE-lncRNAs between CRSwNP+AS and control (Additional file 1: Figure S2). Additionally, 968 DE-mRNAs and 312 DE-lncRNAs were identified between CRSwNP+AS and CRSwNP-alone. Hierarchical clustering of top 500 differentially expressed genes was shown in Additional file 1: Figure S3.
Common dysregulated genes shared by CRSwNP+AS and CRSwNP-alone
A total of 1988 common DE-mRNAs were shared by CRSwNP-alone and CRSwNP+AS respectively compared to control (Fig. 2a and Additional file 1: Table S2). Assessment of these common DE-mRNAs by pathway enrichment analysis using Enrichr Demonstrated that the top enriched KEGG pathways were associated with cytokine-cytokine receptor interaction, chemokine signalling pathway, staphylococcus aureus infection, asthma, and cell adhesion molecules (Fig. 2b). The top enriched BioCarta pathways were related to eicosanoid metabolism, the co-stimulatory signal during T-cell activation and IL-10 anti-inflammatory signalling pathway (Fig. 2b). Likewise, gene ontology enrichment analysis showed that common DE-mRNAs were mainly associated with regulation of immune system process, immune response, response to stimulus and cell-cell signalling (Additional file 1: Figure S4).
Similarly, a total of 176 common DE-lncRNAs were shared by CRSwNP+AS versus control and CRSwNP-alone versus control (Fig. 2c and Additional file 1: Table S2). WGCNA was applied to explore the potential functions of common DE-lncRNAs, and construction of a hierarchical clustering tree demonstrated 9 modules, each of which was labelled with a distinct colour (Fig. 2e). The size of these modules ranged from 41 to 539 genes; with the largest turquoise coloured module (Additional file 1: Figure S5) comprising 45 lncRNAs and 494 mRNAs, which were highly coexpressed. Pathway enrichment analysis showed that the genes of this turquoise module were mainly associated with cytokine-cytokine receptor interaction, asthma, T cell receptor signalling pathway, staphylococcus aureus infection, chemokine signalling pathway, and Th1, Th2, Th17 cell differentiation (Fig. 2d).
Fig. 2f shows the top 50 hub genes with high connectivity and edge weigh identified and visualized by cytoscape network. Of these top hub genes, only LINC01146 was significantly up-regulated in both CRSwNP-alone and CRSwNP+AS. Pathway analysis of the coexpressed mRNAs of LINC01146 demonstrated that LINC01146 was mostly associated with T cell receptor signalling pathway, natural killer cell mediated cytotoxicity, Fc gamma R-mediated phagocytosis, and Th1 and Th2 cell differentiation (Additional file 1: Figure S6).
Distinct transcriptome signatures in nasal tissue of CRSwNP+AS
Overall, 212 mRNAs were down-regulated and 756 mRNAs up-regulated in nasal tissues of CRSwNP+AS patients, compared to nasal tissues of CRSwNP-alone patients (Fig. 3a). The 50 most significant DE-mRNAs are shown in Additional file 1: Table S3. Enrichr pathway analysis showed that arachidonic acid metabolism, ECM-receptor interaction, IL-17 signalling pathway, GATA3 participate in activating Th2 cytokine genes, and fibrinolysis pathway were the top significant pathways enriched by DE-mRNAs (Fig. 3b).
Detailed examination of the expression of genes involved in arachidonic acid metabolism demonstrated that 14 of 16 DE-mRNAs related to arachidonic acid metabolism; including PTGS1, TBXAS1, ALOX15B, PLA2G4D, LTC4S, GGT1, GGT5, GPX6, PLA2G4B, PLA2G6, CYP4F8, JMJD7-PLA2G4B, PTGIS and GPX3; were up-regulated in CRSwNP+AS compared to CRSwNP-alone (Fig. 3c).
Severe type 2 inflammation in nasal tissues of CRSwNP+AS
As indicated by above transcriptome data, type 1, type 2 and type 3 related signalling pathways might be differentiated between CRSwNP-alone and CRSwNP+AS. Thus, we investigated the mRNA expression of critical cytokines and their receptors, which indicate the activities of different inflammatory endotypes; namely (1) IFNG and IFNG receptor (IFNGR1 and IFNGR2) for type 1 inflammation; (2) IL5, IL4, IL13 and their receptors (IL5RA, IL5RB, IL4R, IL13RA1 and IL13RA2) for type 2 inflammation; (3) IL17A and IL17A receptor (IL17RA and IL17RC) for type 3 inflammation; as well as (4) TNF, IL1B, IL6 and their receptors (TNFRSF1A, TNFRSF1B, IL1R1, IL1R2 and IL6R) for proinflammatory activity.
We found that the expression of cytokines IL5 and IL13, and receptors IL5RA and IL5RB, indicating augmented type 2 inflammation, was significantly enhanced in both CRSwNP-alone and CRSwNP+AS compared to control (Fig. 3d and Additional file 1: Table S4). Furthermore, nasal tissue from CRSwNP+AS demonstrated significantly higher expression of IL4, IL5 and IL13; and conversely significantly lower expression of IL17A than nasal tissue of CRSwNP-alone.
Assessment of correlations between the significant cytokines and the other significant inflammatory indicators determined above, demonstrated that the expression of both IL5 and IL13 was positively correlated with the percentage of local eosinophils, concentration of tissue IgE and the expression of LTC4S, which reflected an imbalanced arachidonic acid metabolism (Figs. 4 a-f). In contrast, the expression of IL17A was negatively correlated with tissue IgE and the expression of LTC4S (Figs. 4 g-i).
LncRNA signatures in nasal tissue of CRSwNP+AS
We identified 229 up-regulated and 83 down-regulated DE-lncRNAs in nasal tissues of CRSwNP+AS patients compared to nasal tissues of CRSwNP-alone patients (Fig. 5a). The top 50 significant DE-lncRNAs are shown in Additional file 1: Table S5. A coexpression network constructed based on the expression of DE-lncRNAs and DE-mRNAs using WGCNA demonstrated 7 colour-coded modules underneath the cluster tree (Additional file 1: Figure S7). Assessment of the relationship between each module and tissue type (CRSwNP+AS), estimated using the module eigengenes, demonstrated that all 7 modules were significantly correlated with tissue type changes (Fig. 5b). Of these, the blue module was most negatively correlated with the phenotypic traits of CRSwNP+AS (r = −0.73, P = 2x10-4), and the yellow module the most positively correlated (r = 0.66, P = 0.001).
Overall, there were 31 lncRNAs and 94 mRNAs in the blue module, and only one lncRNA (HK3-006) and 63 mRNAs in the yellow module. Pathway enrichment analysis indicated that the blue module was mainly associated with IL-17 signalling pathway and cytokine-cytokine receptor interaction (Fig. 5c), whereas the yellow module was mainly related to asthma, arachidonic acid metabolism, and signalling pathways such as calcium and JAK-STAT signalling pathway (Fig. 5d). Top 50 hub genes identified from the blue and yellow modules, respectively, are shown in Figs 5 e and f.