The Role of Alveolar Epithelial Type II-Like Cells in Uptake of Structurally Different Antigens and in Polarisation of Local Immune Responses

Background Our previous studies on intranasal tolerance induction demonstrated reduction of allergic responses with different allergen constructs. The underlying mechanisms varied depending on their conformation or size. Objective The aim of the present study was to compare the uptake of two structurally different allergen molecules within the respiratory tract following intranasal application. Methods The three-dimensional Bet v 1 (Bv1-Protein) and the T cell epitope peptide of Bet v 1 (Bv1-Peptide) were labelled with 5,6-Carboxyfluorescein (FAM) and their uptake was investigated in lung cells and cells of the nasal associated lymphoid tissue from naive and sensitised BALB/c mice. Phenotypic characterisation of FAM+ lung cells after antigen incubation in vitro and after intranasal application was performed by flow cytometry. Impact of Bv1-Protein and Bv1-Peptide on cytokine profiles and gene expression in vivo or in an alveolar epithelial type II (ATII) cell line were assessed in mono- and co-cultures with monocytes using ELISA and quantitative real-time PCR. Results Both antigens were taken up preferably by ATII-like cells (ATII-LCs) in naive mice, and by macrophages in sensitised mice. After intranasal application, Bv1-Peptide was taken up faster and more efficiently than Bv1-Protein. In vivo and in vitro experiments revealed that Bv1-Protein induced the transcription of thymic stromal lymphopoietin mRNA while Bv1-Peptide induced the transcription of IL-10 and MCP1 mRNA in ATII-LCs. Conclusion and Clinical Relevance Both tested antigens were taken up by ATII-LCs under steady state conditions and induced different polarisation of the immune responses. These data may have an important impact for the generation of novel and more effective prophylactic or therapeutic tools targeting the respiratory mucosa.


Introduction
Type I allergic disorders such as allergic rhinitis, asthma, and atopic eczema are affecting~20% of westernised countries [1,2]. The initial contact site of inhaled allergens in the human body is the respiratory mucosa. In allergy-prone patients, this contact results in T helper type 2-skewed (T H 2) immune response leading to IgE-mediated clinical manifestations. The only immune modifying and potentially curative treatment for type I allergy is allergen-specific immunotherapy (SIT). The goal of SIT is to generate a switch from allergen-specific T cells to tolerant or anergic T cells and down regulation of IgE-mediated immune responses in allergic patients [3]. Nevertheless, the use of natural allergen extracts in SIT is associated with certain drawbacks, such as varying allergen concentrations, presence of non-allergic materials, and de novo sensitisation against other components within the allergen extracts [4]. In order to increase the safety of SIT, we and others have generated recombinant allergens, non-IgE binding T cell epitopes, polypeptide constructs, or fragments of the allergens without IgE reactivity [5][6][7][8] and tested their efficacy in different mouse models of type I allergy [4].
In a mouse model of birch pollen allergy, intranasal administration of major birch pollen allergen Bet v 1 led to suppression of allergic immune responses and airway inflammation in sensitised mice [5,6]. Furthermore, we have shown that it is possible to induce tolerance by mucosal application of different allergen-derived peptides [7,8].
The precise mechanism of interaction between structurally diverse antigens and the respiratory epithelium, which might lead to sensitization or tolerance, is still far from being elucidated. Only recently, there is increasing evidence stressing the role of epithelial cells in orchestrating immune responses to allergens [9,10].
In the present study, we aimed to investigate how differences in conformation of antigens (three-dimensional vs. linear) are influencing their recognition and uptake by innate immune cells and what is the consequence of this interaction in the context of steady state condition or allergy. As model antigens we used the three-dimensional recombinant Bet v 1 (Bv1-Protein) and the single T cell epitope peptide of Bet v 1 (Bv1-Peptide) and investigated their uptake by different cell populations in vitro and in vivo.

Incubation of lung cells with antigens in vitro
Lungs from naive BALB/c mice were digested in RPMI containing 0.4 mg/ml collagenase II (Sigma-Aldrich, St. Louis, MO, USA) and 0.04 mg/ml DNase I (Sigma-Aldrich) for 1 hour at 37°C. Tissue was passed through a 70 μm cell strainer. Lung cells were stimulated with 5 μg/ml of Bv1-Protein-FAM or Bv1-Peptide-FAM for 0.5, 1, 6, 24, and 48 hours. FAM + cells were detected by flow cytometry.
THP1 monocyte/macrophage cell line (ATCC) was cultured in RPMI supplemented with 10% heat-inactivated FCS and 10 mM β-2 Mercaptoethanol (Carl Roth GmbH, Karlsruhe, Germany). Both cell lines were maintained at 37°C under a 5% CO 2 atmosphere. For co-culture, A549 cells were plated in 48 wells flat-bottom plates (5x10 5 cell/well) and incubated until confluence. THP1 cells were added in a concentration of 5x10 5 cells/well. For transwell co-culture, the cells were separated with 0.4-μm-pore filters (Corning, NY, USA). The co-cultures were incubated with 5 μg/ml of the Bv1-Protein or the Bv1-Peptide for 6 days. Cells were harvested separately to evaluate gene expression and cell free supernatants were stores at -20°C for further analysis.

ELISA
Levels of IL-6 were measured by commercially available ELISA (Diaclone, Besancon, Cedex, France). IL-8 levels were measured in 96-well plates coated with 2 μg/ml of mouse anti-human IL-8 mAb (Thermo Scientific). 50 μl of a 1:10 dilution of the samples were investigated in triplicates. Thereafter 0.2 μg/ml of biotin-labelled mouse anti-human IL-8 mAb (Thermo Scientific) was applied followed by horseradish peroxidase-conjugated streptavidin. With tetramethylbenzidin substrate (Millipore, Billerica, MA, USA) colour development was achieved and the absorbance at 450 nm was measured.

Isolation of ATII-LCs
Primary ATII-LCs were isolated as described by Driscoll B. et al. 2012 [22]. Briefly, mice were euthanized and left kidney was excised. Lungs were perfused with 5 ml of PBS through the right ventricle of the heart. One ml collagenase (Sigma-Aldrich) was instilled into the lungs via tracheal catheter followed by 0.5 ml 1% low-melt agarose (Sigma-Aldrich) warmed to 45°C. Lungs were immediately covered with ice for 2 minutes. Lungs were then dissected, placed in a culture tube containing 1 ml of collagenase (Sigma-Aldrich), and incubated for 45 minutes at room temperature. Lung tissue was teased apart, transferred to a culture dish containing 7 ml of DMEM and 100 μl of DNase (Sigma-Aldrich) and then filtered through 100-, 40-and 25 μm nylon meshes. Single cells were stained with a mixture of APC labelled anti-mouse antibodies (CD11b, CD11c, CD16/32, CD19, CD31, CD45, and F4/80). ATII-LCs were isolated by negative selection with anti-APC coated magnetic beads using the Magnetic Activated Cell Sorting (MACS) system (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer's instructions.

Statistics
Statistical analysis was performed with GraphPad Prism 5 software (San Diego, CA, USA). Data were assessed for significance using the Student's t test (for comparison of two groups) or ANOVA (for multiple group comparison). For 2 −ΔΔCT method values were log transformed to obtain a symmetric distribution. These values were analysed by a General Linear Model with Bv1-Protein/Bv1-Peptide as the main factor of interest and mice nested within the experiment factor. Mice (3 per group) and experiment (1/2) were defined as random factors. By this procedure potential differences between experiments (different RT-PCR plates) and differences between animals for which triplicate measurements were available are accounted for in the analysis of the difference between reaction to Bv1-Protein and Bv1-Peptide. Homogeneity of variances was examined by Levene's tests and normality of residuals was checked by Kolmogorov-Smirnov tests with Lilliefors' corrected p-values. A similar approach was used for the experiments with cell lines. Experiments were repeated three times in this case and the analysis was performed applying a model with a random factor (experiments 1 to 3) and a fixed factor (Bv1-Protein/Bv1-Peptide).

ATII-LCs from lungs of naive mice are the major population internalising Bv1-Protein and Bv1-Peptide in vitro
Lung cells derived from naive BALB/c mice were incubated with 5 μg/ml of FAM-labelled Bv1-Protein or Bv1-Peptide for different time points (0.5-48 hours). Antigen internalisation reached the maximum after 6 hours of incubation ( Fig 1A) and slightly decreased after 48 hours. In order to characterise the cell populations internalising Bv1-Protein and Bv1-Peptide, naive lung cell cultures were incubated for 6 hours with either of the antigens in vitro. ATII-LCs reveal to be the most efficient population in uptake of Bv1-Protein and Bv1-Peptide followed by macrophages, B cells, and DCs ( Fig 1B). ATII-LCs in naive but not in sensitised mice are the major population taking up antigens Bv1-Peptide-FAM and Bv1-Protein-FAM were intranasally applied to either naive or Bet v 1-sensitised and birch pollen-challenged mice (Fig 2A). One hour after Bv1-Peptide application or six hours after Bv1-Protein application, phenotypic characterisation of FAM + cells in lungs was performed (Fig 2B and 2C). ATII-LCs were the major cell population in uptake of both antigens in naive mice. However, in sensitised and challenged mice, macrophages displayed the dominant cell population internalising either Bv1-Protein or Bv1-Peptide.
After intranasal application, Bv1-Peptide is taken up at an earlier time point and more efficiently by cells of the NALT and lungs, compared to Bv1-Protein In order to investigate antigen uptake in vivo, Bv1-Protein and Bv1-Peptide were intranasally applied to naive BALB/c mice and cells of the NALT and lungs were collected. The number of Bv1-Peptide-FAM + cells reached highest levels already after 1 hour after in vivo application in both NALT and lungs. In contrast, Bv1-Protein internalisation peaked 6 hours after intranasal application (Fig 3A and 3B). After 24 hours none of the antigens were detectable in NALT or lung cells. Phenotypic characterisation of the lung cells revealed that antigens were taken up preferably by ATII-LCs, followed by macrophages, DCs and B cells (Fig 3C). In NALT, majority of both antigens was taken up by B cells (S1 Fig). There were no FAM-positive DCs and macrophages in NALT (S1 Fig). Precise gating strategy is displayed in S1 and S2 Figs.
A549 cells internalised Bv1-Peptide earlier and more efficiently than Bv1-Protein in vitro Already 30 minutes after incubation of the human ATII cell line, A549, with Bv1-Peptide-FAM, 70% of all cells were FAM + . Even after 48 hours, 50% of cells remained positive. In contrast, after 30 minutes of incubation with Bv1-Protein-FAM, only 20% were FAM + . This number did not significantly increase over the experimental period of 48 hours (Fig 4A). Internalisation of both antigens after 4 hours was confirmed by confocal microscopy (Fig 4B). Incubation of A549 cells with Bv1-Protein led to increased levels of TSLP mRNA (Fig 5A), while Bv1-Peptide induced the transcription of IL-10 mRNA (Fig 5B). Levels of MCP1 mRNA remained unchanged after 24 hour incubation with Bv1-Protein or Bv1-Peptide (Fig 5C).
In a transwell co-culture system, A549 and THP1 cells were stimulated with Bv1-Protein and Bv1-Peptide. The gene expression of both cell types was evaluated separately after 6 days. A549 show significant elevated levels of MCP1 and IL-10 upon stimulation with Bv1-Peptide, whereas incubation with Bv1-Protein increases mRNA expression of TSLP (Fig 5D-5F). No induction of IL-10, TSLP or MCP1 mRNA was found in THP1 cells after incubation with both antigens (data not shown).
Primary murine ATII-LCs, which were isolated from mice after intranasal application of Bv1-Peptide, reveal significant levels of IL-10 and MCP1 mRNA (Fig 5H and 5I), whereas Bv1-Protein induced transcription of TSLP mRNA (Fig 5G) in these cells. These data are in accordance with data obtained from transwell co-culture experimentswith A549 and THP1 cell lines (Fig 5D-5F).  In vitro incubation of A549/THP1 co-culture with Bv1-Peptide induced significantly higher levels of IL-6 and IL-8 than incubation with Bv1-Protein Without any stimulus, A549 cells constitutively secrete IL-8. Production of the cytokine was increased by Pam3-Cys, but was not altered by addition of Bv1-Protein, Bv1-Peptide, or LPS (Fig 6A). In THP1 cells, both LPS and Pam3-Cys induced high levels of IL-8, but level of this cytokine was under detection limit in cultures incubated with Bv1-Protein and Bv1-Peptide ( Fig 6A). Furthermore, neither A549 nor THP1 cells produce IL-6 after incubation with Bv1-Protein or Bv1-Peptide (Fig 6B). In contrast, direct co-culture of these two cell lines led to increased levels of IL-6 and IL-8 upon addition of Bv1-Protein or Bv1-Peptide (Fig 6). Moreover, Bv1-Peptide induced significantly higher levels of both cytokines compared to Bv1-Protein. This cytokine production was depended on direct cell contact, as no differences in cytokine production were detected in transwell culture system (Fig 6).

Discussion
In the present study we sought to investigate, whether the conformation/size of molecules used for intranasal tolerance induction impacts on uptake by host cells and polarisation of the immune responses. Therefore, internalisation of two different allergen molecules, the three-dimensional Bv1-Protein and the linear Bv1-Peptide, was investigated in NALT and lungs, as primary sites of antigen uptake in the respiratory mucosa.
In vitro studies with lung cells revealed that both Bv1-Peptide and Bv1-Protein were taken up with the same intensity at all investigated time points (maximum uptake was detected after 6 hours, Fig 1), where around 15-20% of all cells were FAM + . In contrast, investigation of antigen uptake in vivo, after intranasal application, showed that Bv1-Peptide was internalised earlier and more efficiently than Bv1-Protein in NALT and lungs (Fig 3), however, only around 1-2% of all cells were FAM + in the case of Bv1-Peptide and only around 0.5% in the case of Bv1-Protein. Moreover, Bv1-Peptide was internalised faster and at an earlier time point compared to Bv1-Protein. These data indicated that the size of the antigens might be an important factor for the effectivity and kinetic of the uptake. Accordingly, several studies investigated the relationship between increasing molecular weight and the effectivity of transport across the mucosal membrane in the respiratory tract and concluded an inverse relationship between molecule size and amount of antigen uptake [23,24]. Moreover, we observed differences in intensity of uptake between in vitro and in vivo conditions. The nasal respiratory epithelia is covered with a constantly regenerating mucus, responsible for the fact that nasally-applied molecules are cleared from the mucosa within a short period (i.e. around 20 minutes) [23]. Thus, there is a narrow "window of opportunity" for antigens to be available for mucosal cells in comparison to in vitro conditions, where antigens are in the direct contact with cell populations of the respiratory tract during the incubation period (i.e. 0.5-48 hours).
Based on our kinetic studies on antigen uptake in vitro (Fig 1), phenotypic characterisation of lung cells taking up either Bv1-Protein or Bv1-Peptide was performed. Here we describe that in naive mice, ATII-LCs are the major population internalising both Bv1-Protein and   (Figs 1 and 3). Apart from the fact that dendritic cells are described as a prominent population responsible for antigen uptake in airways [25,26], our data on ATII-LCs are supported also by other studies [27,28]. Moreover, these authors showed that small and larger molecules (<2 kDa and >60 kDa) diffuse across airways and are taken up and captured for several hours in alveolar epithelial cells [27,28].

Bv1-Peptide in vitro and in vivo
Surprisingly, we noted a marked difference in cell types internalising antigens in allergic versus naive mice, where macrophages, but not ATII-LCs, were the dominant population internalising Bv1-Protein and Bv1-Peptide (>50% of FAM + cells were macrophages in allergic mice, while >50% of FAM + cells were ATII-LCs in naive mice). In a mouse model of birch pollen allergy, we previously described that airway challenge with birch pollen leads to increased recruitment of macrophages into the alveolar space [5]. Accordingly, the important role of macrophages for allergen uptake in an allergic status has been demonstrated in a mouse model of OVA-induced airway inflammation [29]. An increased allergen uptake was associated with increased secretion of surfactant proteins by ATII cells [29]. Thus, we suggest that the immune status of the host might be another important factor influencing the process of antigen uptake by distinct cell populations after intranasal application.
As our data pointed out the importance of ATII-LCs in the antigen uptake process, further studies on the functionality of these cells were performed. Previous studies using isolated murine ATII cells showed their trans-differentiation towards ATI cells within days during in vitro cultivation [30,31], indicating that they are not suitable for long term experiments in vitro. We therefore used the human ATII cell line A549 for further experiments. When stimulating A549 cells alone or in a transwell co-culture with the monocyte cell line THP1 with Bv1-Protein, enhanced TSLP mRNA levels, known to initiate or enhance T H 2 responses at the lung epithelial cell surface [32], were measured in A549 cells, whereas Bv1-Peptide induced the transcription of IL-10 mRNA (Fig 5). Our data indicate a distinct polarisation pattern of A549 cells depending on the encountered antigen. Similar to our observations, it has been shown that stimulation with correctly folded allergens, such as phospholipase A 2 or OVA, induces T H 2 immune responses, while structurally modified allergens (e.g. heat-treated OVA) rather induce T H 1 responses [33,34]. The capacity of peptides to induce IL-10 as described in our study is also supported by other investigators, suggesting allergen peptides as suitable candidates for tolerance induction (33).
Beside their barrier function, ATII cells are able to produce several chemokines, important for the recruitment of monocytes and lymphocytes, of which MCP1 is a representative candidate [35,36]. While stimulation of A549 cells alone with allergen molecules did not induce upregulation of this chemokine, A549 cells in a transwell co-culture with monocytes (THP1) led to increased MCP1 mRNA transcription upon Bv1-Peptide, but not Bv1-Protein, stimulation ( Fig 5). These data indicate that A549 cells are able to crosstalk with other neighbouring cells, such as monocytes/macrophages, in an antigen-specific but cell-cell contact-independent manner.
Importantly, the analysis of TSLP, IL-10 and MCP1 mRNA transcription in primary ATII-LCs isolated from naive mice 6 hours after the intranasal application of Bv1-Protein or one hour after the intranasal application of Bv1-Peptide revealed similar results compared to the A549 and THP1 co-culture (Fig 5). When performing direct co-culture of A549 cells with THP1 cells, we measured increased levels of IL-6 and IL-8 after incubation with Bv1-Protein and Bv1-Peptide (Fig 6), similarly as previously described by Striz et al. 2001 [37]. In order to investigate whether cell-cell contact between A549 and THP1 cells is important for cytokine production, transwell co-culture was performed (Fig 6), showing that indeed direct cell-cell contact is necessary for production of these cytokines.
In conclusion, our findings emphasise the importance of ATII-LCs in the uptake of antigens entering via the respiratory mucosa. This study further indicates that ATII-LCs may contribute to polarisation processes of the immune system depending on the structure and size of the encountered antigen during the non-inflamed steady state situation. Further investigations on the role and function of these cells during health and allergic disease may help to improve prophylactic as well as therapeutic strategies against asthma and allergic diseases.