A possible role for neutrophils in allergic rhinitis revealed after cellular subclassification

A re-examination of former concepts is required to meet today’s medical challenges in allergic rhinitis. Previously, neutrophils have been treated as a relatively homogenous cell population found in the nose both when the patient is suffering at the height of the allergic season as well as when the patient report no symptoms. However, new data indicates that neutrophils can be divided into different subsets with diverse roles in inflammation. We showed increased levels of neutrophils in peripheral blood, nasal biopsies and nasal lavage fluid (NAL) from allergic patients during the pollen season compared to healthy controls. A closer examination revealed that the activated subset of neutrophils, CD16high CD62Ldim, outweighed the normal form CD16high CD62Lhigh in nasal tissue among these patients. This skewed distribution was not seen in controls. The normal subset prevailed in peripheral blood from patients as well as controls, whereas CD16high CD62Ldim and CD16dim CD62Ldim subsets, the latter considered “end state” neutrophils before apoptosis, were elevated in NAL. Functional in vitro experiments revealed that activated neutrophils exhibit a T cell priming capacity and an ability to enhance eosinophil migration. Activated neutrophils may thus contribute to allergic inflammation seen in allergic rhinitis by priming T cells and attracting eosinophils.

Allergic rhinitis (AR) is an inflammatory disease of the nasal mucosa, induced by a reaction to a normally harmless antigen. It is characterized by troublesome local nasal symptoms as well as general fatigue, negatively impacting quality of life. In addition, the significant prevalence of AR is associated with staggering indirect costs for the Western societies [1][2][3] . In order to develop novel strategies for further treatment, traditional concepts of allergic inflammation have to be re-examined 4 .
Eosinophils are well described markers of the allergic reaction, known to play an important role in Th2-mediated immune responses [5][6][7][8][9] . Neutrophils have evoked less interest, as they are found in the nose both when the patient is suffering at the height of the allergy season as well as when no symptoms can be reported. Nevertheless the number of neutrophils increases in the nose in symptomatic individuals during the allergy season and their large absolute cell number in comparison with corresponding eosinophils deserves attention 10-12 . Neutrophils have long been considered short-lived, terminally differentiated cells with a well-established role in acute bacterial infections 13,14 . However, information has recently emerged indicating that neutrophils can be divided into different subsets and that the various subsets might have diverse roles in inflammatory diseases. The new subsets are defined by differences in expression of Fcγ RIII (CD16) and L-selectin (CD62L). Three different subsets have recently been identified: CD16 dim CD62L high (immature), CD16 high CD62L high (mature) and CD16 high CD62L dim (activated and CD11b bright as well as CD11c bright ) 15 . The same authors reported that the activated subset of mature human neutrophils causes certain immunological responses. Differentiated, CD11b + neutrophils have further been suggested to be reduced at the site of C. perfringens infections through systemic reduction of mature neutrophils, possibly explaining polymicrobial infections seen in these patients 16 .
We now suggest that some of these neutrophil subsets might affect the course of inflammation seen in AR. Hence, the present study is designed to identify these subsets as well as determine their ability to traffic between different compartments i.e. peripheral blood, nasal mucosa and nasal lavage fluid (NAL) during the allergic Activated neutrophils can help activate T cells. Functional in vitro experiments were performed to evaluate the immunological importance of the activated neutrophil subsets. Mature neutrophils, CD16 high CD62L high (Fig. 3(a)), were isolated from peripheral blood from healthy controls and activated in vitro with LPS, TNF-α and IL-8. Flow cytometric analysis verified an increase of the activated neutrophil subset (CD16 high CD62L dim ) after activation ( Fig. 3(b)). The expression of the activation marker CD66b increased simultaneously on neutrophils, supporting their state of activation (see Supplementary Fig. S1). A schematic figure illustrates the phenotypic maturation of neutrophils in the activation process ( Fig. 3(c)). After thoroughly washing the neutrophils, isolated autologous T cells from peripheral blood were added in a co-culture. There was an increase in both the fraction of CD69 + /CD4 + T cells (p < 0.0007) ( Fig. 4(a)) and in the expression levels of CD69 on CD4 + T cells (p < 0.0051) ( Fig. 4(b)) when T cells had been primed with activated neutrophils prior to CD3 stimulation. The same findings were made when using blood from patients with AR outside the pollen season (see Supplementary Fig. S2(a,b)). The increased activation of CD4 + T cells was confined to CD45RO − T cells indicating that neutrophils specifically affected naïve CD4 + T cells (see Supplementary Fig. S3). The T cell activation  level in control experiments with naïve neutrophils was comparable with controls using only T cells and CD3, ruling out a T cell priming capacity for naïve neutrophils. Experiments with transwell cell cultures were set up to determine how T cell priming was mediated by activated neutrophils. When using transwell plates, no increase in CD69 + /CD4 + T cells was seen upon priming with activated neutrophils (Fig. 4(c,d)) indicating that the enhanced T cell activation of neutrophils was cell-cell contact or close contact dependent. As a control, monocytes were analysed and no increased activation was seen, making it unlikely that the facilitated T cell activation was mediated by impurities from activated monocytes (see Supplementary Fig. S4). Representative dot plots illustrates the gating procedures ( Fig. 4(e)).

Activated neutrophils mediate eosinophil migration.
To further analyse the importance of neutrophils in allergic inflammation, experiments were set up to study the impact of activated neutrophils on eosinophils. Isolated neutrophils activated in vitro with LPS, TNF-α and IL-8 upregulated eosinophilic migration after 3 h of incubation in the transwell system ( Fig. 5(a)). Naïve neutrophils had no impact on eosinophil migration. Thus, we concluded that activated neutrophils can increase eosinophil migration, potentially accounting for the local eosinophil infiltration seen in allergy. Representative dot plots illustrates the gating procedures ( Fig. 5(b)).

Discussion
Patients with AR exhibited increased numbers of neutrophils in peripheral blood, nasal biopsies and NAL during the pollen season. When studying this amplification on a subset-level, it was evident that the mature neutrophil subset CD16 high CD62L high dominated in peripheral blood while the subsets CD16 high CD62L high together with the activated form CD16 high CD62L dim were elevated in nasal biopsies. The activated subset CD16 high CD62L dim fraction was significantly higher than the normal fraction in nasal tissue from AR-patients. This was not convincingly seen in the healthy controls. Finally, the subsets CD16 high CD62L dim and CD16 dim CD62L dim , considered the end state before apoptosis, were elevated in NAL. A T cell priming capacity of activated neutrophils was demonstrated in vitro by co-culturing activated neutrophils and CD4 + T cells. Since no such priming was seen using a transwell system, this interaction was likely mediated by close or direct cell-cell contact. Further, activated neutrophils were found to upregulate eosinophil migration.
Special attention has lately been paid to the findings that neutrophil subsets can be defined by differences in CD16 and CD62L expression 15,23 . The CD16 high CD62L dim neutrophil subset has previously been suggested to cause T cell inhibition through Mac-1 resulting in immunosuppression 15 . However, the way in which different neutrophil subsets influence various diseases has scarcely been investigated. We have previously demonstrated that activated neutrophils seem to have anti-tumorigenic properties 24 . Moreover, we have suggested that the enhanced systemic adaptive immune response seen among patients with AR might protect against head and neck squamous cell carcinoma, a phenomenon likely mediated by peripheral blood mononuclear cells 25 . The same subtype has been demonstrated in other cancer forms 26 . Neutrophils from allergic patients have been shown to downregulate surface expression of CD62L upon allergen stimulation 27 . The distribution of different neutrophils based on their expression of CD16 and CD62L has to our knowledge however never been studied in allergy.
Both neutrophils and eosinophils are known to increase upon allergen exposure. Even though the neutrophils far outnumber the amount of eosinophils found both locally and systemically during symptomatic AR, the former has been given much less attention as potential players in the allergic reaction (Fig. 1). This is probably due to the fact that neutrophils can be found in the nose in symptomatic patients at the height of the allergic season as well as in the asymptomatic phase. However, when examining changes in neutrophil subsets in different compartments, a different picture emerges (Fig. 2). It becomes clear that activated and differentiated neutrophil subsets accumulate at the site of allergic provocation, in this case in the nose. Furthermore, we demonstrated that these activated neutrophils have the ability to prime CD4 + T cells, a phenomenon known to be of great importance for the start of allergic inflammation at the site of antigen exposure 28,29 . This notion is supported by a recent publication demonstrating that timothy grass pollen can stimulate neutrophil immune responses through the secretion of IL-8 11 . In addition, the T cells in our experiment were CD45RO − negative, indicating that the activated neutrophil subset could have a role in the allergic sensitisation process, by affecting naïve antigen-specific CD4 + T cells 30 . Hence, locally activated neutrophils in the upper airways seem to have the ability to mediate the inflammatory process in allergy. Further, neutrophils have recently been shown to be components of the response to, immunisation accumulating in lymph nodes and forming an adaptive immune response 31 . Neutrophil subsets could even partly explain the airway epithelial injury seen in asthma 32,33 . Altogether, these reports suggest that neutrophils are more complex than previously thought.
T cell activation and proliferation is central in allergic immunity. However, it is still not completely understood which cells are responsible for local mucosal T cell activation. It has previously been shown that it takes several days to recruit conventional dendritic cells to the nasal mucosa upon allergen challenge 34 . We have earlier shown the immunological importance of nasal epithelial cells activating T cell at the site of allergen exposure, i.e. the nose 35 . In addition, previous publications in mice have showed that neutrophils can influence CD8 + T cell response and possibly also direct or prime Th1 T cell responses [36][37][38] . The current experiments demonstrate that activated human neutrophils can prime T cells, thus facilitating CD3 activation. No such priming was seen when naïve neutrophils were used or when T cells and CD3 were used alone speaking against potential contamination from monocytes. Separate analyses of monocytes in wells containing activated neutrophils demonstrated no monocyte activation, definitively ruling out a role for monocytes in facilitating T cell activation (see Supplementary Fig. S4). These assays identify that activated neutrophils have stimulatory properties on T cells, further suggesting the possible functional impact of locally activated neutrophil subsets in allergic patients.
The presence of both circulating and local eosinophils during ongoing allergic inflammation is well established 39 . In contrast, the detailing of their pharmacological suppression and the resulting clinical improvement has over the years been limited 40,41 . Interestingly, a significant decrease of not only eosinophils but also neutrophils in the nasal mucosa was seen upon sublingual immunotherapy (IT) to Parietaria species 42 . Neutrophils were also shown to be decreased in a skin chamber model upon AR IT 43,44 . Further, IL-9 positive neutrophils increased in the nasal mucosa during pollen season, something that could be successfully inhibited by IT 45 .
The present data on neutrophil subsets might be a next step in understanding the great infiltration of neutrophils in the nasal mucosa in allergy. We demonstrate the ability for activated subsets to directly prime T cells and enhance migration of eosinophils 46,47 . This is in line with previous publications suggesting a more complex role for neutrophils 28,29,31,46 . Further experimental studies depleting or deactivating the activated neutrophils in an allergic setting, potentially affecting the course of the inflammation, are required. This could open up new therapeutic possibilities for AR.

Methods
Patients. Eight non-smoking AR patients with allergy to birch pollen (n = 3), grass pollen (n = 1) or both (n = 4) were included in the study (mean age 28 years; M:F 2:6). All patients were healthy with the exception of their allergy. Their allergy was diagnosed on the basis of clinical history and positive skin prick test (SPT) and/or the radioallergosorbent test (RAST) for allergen-specific IgE. Samples were taken in-season and all patients had nasal symptoms of allergy at the time of participation. Seven patients had been off antihistamine treatment for more than 48 h, one had been off antihistamine treatment for more than 24 h. Four patients reported no use of topical nasal steroids for at least a month. Four patients had been off topical nasal steroids for more than three days. No patient had been on inhaled steroids for at least one month. In addition, six healthy non-smoking controls were included (mean age 21 years; M:F 2:4). RAST for allergen-specific IgE were negative for all healthy controls included in the study. The patients and healthy controls reported no history of infections in the airways or in the rest of the body three weeks prior to participating. The study was approved by the local ethical committee in Stockholm, Karolinska Institutet, and all patients and healthy controls gave their written informed consent. All methods were performed in accordance with the relevant guidelines and regulations.
Human nasal biopsies. Nasal biopsies were taken from inferior turbinates of healthy non-smoking controls under local anaesthesia as previously described 48  Recovery of nasal lavage fluid. Nasal lavage fluid was collected as previously described 48 . In brief, after cleaning excess mucous by forceful exsufflation, 8-10 ml of sterile saline solution (0.9% NaCl) at RT was aerosolised into the nostrils and passively collected from the nostrils. When 7 ml of fluid were recovered into a graded test tube, it was centrifuged for 10 min. The pellet was resuspended in PBS containing 2% FBS before analysis with flow cytometry.

T cell stimulation.
For all in vitro experiments, healthy controls with no known allergies were recruited. For the T cell stimulation experiments, patients with birch and/or grass pollen allergy were also recruited. All participants were non-smokers and all patients were healthy with the exception of their allergy. Blood was collected in heparin tubes. T cells and neutrophils were isolated from peripheral blood using ficoll-paque (Sigma-Aldrich, St. Louis, MO, USA) according to the manufacturer's instructions. The erythrocytes in the granulocyte rich pellet were lysed (0.8% NH 4 Cl, 10 mM KHCO 3 and 0.1 mM EDTA). Neutrophils were further purified with CD15 microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions. The neutrophils were diluted to a concentration of 2 × 10 6 cells/ml in TexMACS Medium (Miltenyi Biotec) containing 10% autologous plasma, 100 U/mL penicillin and 100 μ g/mL streptomycin (Life Technologies, Eugene, OR, USA). Thereafter, the neutrophils were activated with 1 μ g/ml LPS (product no: L2654, Sigma-Aldrich, St. Louis, MO, USA), 5 ng/ml TNF-α and 10 ng/ml IL-8 (R&D System, Minneapolis, MN, USA) for 15 min and then washed with PBS twice. The lymphocyte interface from the ficoll-paque isolation step was collected and washed with PBS. The cells were diluted to a concentration of 2 × 10 6 cells/ml in the same medium as the neutrophils.

Statistics.
Statistical differences between patients with allergy and healthy controls were performed using unpaired t-tests (Mann-Whitney). For more than two sets of matched data, a two-way ANOVA with Bonferroni post-tests was performed. For the in vitro experiments, a non-parametric (Friedman test) one-way ANOVA was used together with Dunn's post-test, except for the eosinophil migration experiments where Tukey's multiple comparisons test was used. A p-value of 0.05 or less was considered statistically significant (*p < 0.05, **p < 0.01, ***p < 0.001). Statistical analyses were performed using GraphPad Prism software (version 6.0, GraphPad Software, La Jolla, CA). All data are shown as mean ± S.E.M. For human data, n equals the number of patients.