Nasal epithelium: new insights and differences of the cytokine profile between normal subjects and subjects with allergic rhinitis*

Background: The role of the nasal epithelium in the induction of a proper cytokine response in normal subjects and subjects with allergic rhinitis is still not completely elucidated. Methodology: We aimed to compare nasal epithelial immune responses in allergic rhinitis patients of different ages compared to healthy volunteers. Primary nasal epithelial cells from 47 subjects (33 normal and 17 with allergic rhinitis) were collected and cultured. Their unstimulated supernatants were analysed for 21 cytokines and chemokines. Statistical analysis was performed with the R statistical software and the RStudio interface. Results: Differences of the spontaneous release of epithelial cytokines and chemokines were noticed between the two study groups. The levels of GMCSF, MIP1A, MIP1B, IL28A, TNFA, CCL5 were significantly lower in the allergic rhinitis group compared to healthy volunteers’ group, independent of age. Most differences were noticed in the younger allergic rhinitis group (0-12 years old). Conclusions: Despite the cross-sectional nature of the study and the limited number of subjects, allergic rhinitis appears to be associated with dysfunction of cytokine and chemokine spontaneous release from nasal epithelial cells which may represent an abnormal innate immunity maturation pattern.


Introduction
Despite the dramatic increase of allergic rhinitis (AR), affecting 20%-30% of the global population (1) , and the intense study of the role of allergy and innate immunity underpinning its pathophysiology, there are still critical knowledge gaps (2) .
Several studies, mostly on allergic asthma, have explored the significance of the airway epithelium on the induction of an appropriate immune response towards potentially harmful inhaled substances and microbial pathogens or driving the downstream innate responses. However, there are limited studies focusing on the nasal epithelium, which represent the initial defense barrier, particularly in the context of AR.
It is well known that central to the inflammatory cascade in AR, is an immunoglobulin E (IgE)-dependent response to extrinsic protein antigens orchestrated by eosinophils, T cells, mast cells and basophils, which release several mediators such as chemokines and cytokines (3,4) . These mediators are released locally and maintain a persistent inflammation of the nasal mucosa by attracting inflammatory cells and interfering with the mucosal barrier. Recently, studies have indicated that a defective epithelial barrier is a key part of the underlying pathophysiology of diseases such as atopic dermatitis (5) , asthma (6) , chronic rhinosi-nusitis (7) , otitis media (8) and AR (9,10) .
Although the nasal epithelial cells are the building blocks of the mucosa, their contribution to innate and adaptive immunity only recently started to be understood, particularly in relation to age-dependent maturation (11) . The development of the human immune system is a continuous process, while the majority of immune mechanisms are not fully developed from birth (12) . In literature, several studies have been conducted in peripheral blood (13) , there is very little evidence in epithelial responses at different ages.
This cross-sectional study aimed to evaluate nasal epithelial responses in allergic rhinitis patients of different ages compared to healthy volunteers and characterize the molecular pathways involved.

Eligibility criteria and sample collection
This cross-sectional study enrolled allergic and non-allergic children (0 -18 years) and adults (18-45 years whereas histamine (10 mg/ ml) and saline were used as a positive and negative control, respectively (15) . Wheel diameter of 3 mm or greater was considered as positive (16) . When SPT could not be performed, the sensitization of participants was evaluated by sIgE measurement using IMMULITE® 2000 immunoassay system (Siemens, Germany) or ImmunoCap allergen-specific IgE blood test (Phadia AB, Uppsala). Values >0.35 kU/L were considered positive. Clinical data were recorded in a standardized manner.

Protein expression
Protein expression of multiple cytokines was measured in

Statistical analysis
Initially, all continuous variables (demographic and cytokines) were tested for normality with the Shapiro-Wilk test, and they were all found to be non-normal. Based on that, non-parametric tests were performed on them, and their descriptives were provided in the form of "median (25th-75th percentile)". In order to perform cytokine comparisons among two groups (control vs AR), Wilcoxon's rank-sum test (equivalent to Mann-Whitney test) was applied. Pearson's chi-squared test of independence was used to recognize any dependency between categorical variables. Linear regression analysis was performed using all the patients (both control and AR groups), associating jointly the age and the AR status with the levels of the cytokines' expression in order to evaluate their combined effect.
All statistical tests were considered as two-sided and statistical significance was taken at the level of 5%. Analysis was performed with the R statistical software and the RStudio interface, both open-source products.

Bioinformatic pathway analysis
We performed pathway analysis using the Functional Annotation Tool DAVID Bioinformatics Resources 6.8, NIAID/NIH, considering proteins statistically significantly different between AR patients and controls.

Subjects' characteristics and PNECs cultures
A total of 59 subjects were included (39 normal subjects and 20 subjects with allergic rhinitis -(AR)). The PNECs from 47 subjects (79,66%) were successfully cultured, and their unstimulated Also, a marginal difference (p=0.057) was detected between AR patients and control group according to parental smoking cigarettes per day: parents of AR patients were smoking more (>20) cigarettes per day than parents of controls (<20).

Differentially expressed proteins between PNECs from AR patients and control individuals
Comparing the protein expression levels between control and AR PNECs, several mediators appeared to be decreased in subjects with allergic rhinitis ( Table 2). Six of them (GMCSF, MIP1A, MIP1B, IL28A, TNFA, CCL5) ( Figure 1) were statistically significantly lower in the AR subjects than the control group. Further, we analyzed the data according to age, comparing either two groups (adults vs children <18)) or three groups: children (0 -12yrs), adolescence (>12 -18yrs) and adults (Table S1).
As shown in Table S1, three of the cytokines (GMCSF, IL28A, TNFA) were significantly lower in children with allergic rhinitis than controls (p<0.05). Only one chemokine was differentially expressed in adults, MIP1B. MIP1B is lower in the adult AR group than adults' controls (p<0.05).
When comparing children (<12yrs) and adolescents (>12-18yrs), the most prominent differences in protein expression were detected in the age group of <12 years old, as seen in Table S1.  . We also performed linear regression analysis separately for controls and AR using IL28A levels and age and we did not find any difference. It seems that the age is the main factor of IL28A differentiation, and this may influence the results in the group of children. In addition, removing from the analysis the subjects under 3 years old, the statistical difference of IL28A between AR and controls was p=0.07, still remaining a trend of difference ( Figure S1).

Pathway analysis in differentially expressed proteins between PNECs from AR patients and healthy individuals
We performed pathway analysis using the Functional Annotation Tool DAVID Bioinformatics Resources 6.8 in statistically significantly different proteins between allergic rhinitis patients and controls. The pathways identified include key inflammatory and innate immune response pathways. Cytokine-cytokine receptor interaction, TNF signaling pathway, NOD-like receptor signaling pathway, Influenza A, Chemokine signaling pathway, and Toll-like receptor signaling pathway were found significantly downregulated in allergic rhinitis (p<0.05) ( Table 3).

Discussion
In the last decades, the airway epithelium, which is in the interface between the external and internal environment, has gained much attention as its dysfunction may cause the pathogenesis of many diseases. More specifically, multiple studies highlight the significant role of the airway epithelium for an appropri-  (17) . Further investigations (18)   The question of whether reduced cytokine release from PNECs in subjects with allergic rhinitis is a result of environmental stress or inflammation or genetic defect is still unclear (19) . Nevertheless, any reduction in mediator production may result in a suboptimal response of the epithelium and potential dysfunction of the epithelial barrier. It is well accepted that cytokines not only activate inflammatory cells but also regulate the epithelial barrier in allergic disease (20,21) . Cytokines, such as IL13, may alter tight junctions (TJs) of the mucosa, which play a key role in the mucosal barrier (22,23) . Our findings of reduced cytokine release from the nasal epithelial cell in the allergic population may suggest a distraction of the whole cytokine cascade and/or other sources of cytokines, such as neutrophils and macrophages, may replace the epithelial cell gap (18) .
It is well accepted that the development of the human immune system is a continuous process. On the other hand, it has been suggested that the reduced cytokine production from nasal epithelial cells of subjects with allergic rhinitis compared to normal subjects may represent in vitro differences of the cultured cells.
Cytokine release differences may represent an in vitro different state of differentiation of the epithelial cells from the allergic population (18) . This may also explain the difference between the two groups in the success rate of the epithelial cells cultures in our study. In addition, these differences may reflect an in vivo altered differentiation pattern or may be reflect a delayed maturation process. Based on the above, our findings may support the theory that the allergic nasal epithelium cannot undergo standard mechanisms of repair and differentiation.
Finally, our results suggest that in children, allergy may compromise the mechanical and immunological defense function of the nasal mucosa against viruses (24) . All the inflammatory mediators differentially expressed in the allergic population of our study are essential to the host response infection. The cytokines MIP1A, MIP1B and CCL5/RANTES have strong chemotactic effects on many inflammatory cells such as monocytes, T lymphocytes and neutrophils (25,26) . The GMCSF, released by the airway epithelium, along with IL-4 and tumor necrosis factor-α (TNF-α) cause DC maturation which then these cells are actively involved in the development and maintenance of an inflammatory response (27) . TNF-α is an inflammatory cytokine produced by macrophages/monocytes during acute inflammation and is responsible for a diverse range of signaling events within cells, leading to necrosis or apoptosis. It is also crucial for resistance to infection and cancers (28) . IL28A is essential for maintaining the integrity and homeostasis of epithelial tissue layers. It promotes innate immune responses from tissue epithelia to limit the damage caused by viral and bacterial infections (29) . IL28A was highlighted as the most age-related cytokine, and this may influence the results in the group of children. The fact that AR is an allergic disease that is present in older ages (>3 years), may influence the significancy of the results. Although, all the literature about the interferon deficiency in allergy upon virus stimulation supports that is due to allergic status without taking into consideration the age. In agreement with our results, Degir- Table 3. Pathway analysis of differential expressed genes between allergic rhinitis and healthy PNECs.  (30) .
Our study has several limitations. The nature of our data is crosssectional, and as a result, inter-individual variability could play a significant role in the observed differences in the cytokine profiles. Every effort was made to eliminate factors that affect cytokines release by applying strict inclusion criteria. The number of subjects in each group is low due to the significant cost and effort required for primary cell cultures. However, this is the first study of age-stratified comparison of mediators of nasal epithelial cells in allergic rhinitis and control subjects to the best of our knowledge.

Conclusion
Based on the above, we suggest that patients with allergic rhinitis, especially the young ones, have an altered epithelial cytokine profile, possibly related to abnormal immune maturation, supporting disease predisposition. These observations highlight the need to investigate further inflammatory pathways in the pathogenesis of the allergic response, which may lead to more effective early intervention in allergic rhinitis.

Authorship contribution
SS did the sampling, cell culture and wrote the paper. PNG conceived of the presented idea. LI performed statistical analysis.
GA, MP and X did the clinical evaluation of the cohort. AE verified the analytical methods. PE supervised the findings of this work. LE did the contributed to writing and editing the paper.