Apolipoprotein A‐IV acts as an endogenous anti‐inflammatory protein and is reduced in treatment‐naïve allergic patients and allergen‐challenged mice

Abstract Background Recent studies pointed to a crucial role for apolipoproteins in the pathogenesis of inflammatory diseases. However, the role of apolipoprotein‐IV (ApoA‐IV) in allergic inflammation has not been addressed thoroughly thus far. Objective Here, we explored the anti‐inflammatory effects and underlying signaling pathways of ApoA‐IV on eosinophil effector function in vitro and in vivo. Methods Migratory responsiveness, Ca2+‐flux and apoptosis of human peripheral blood eosinophils were assessed in vitro. Allergen‐driven airway inflammation was assessed in a mouse model of acute house dust mite‐induced asthma. ApoA‐IV serum levels were determined by ELISA. Results Recombinant ApoA‐IV potently inhibited eosinophil responsiveness in vitro as measured by Ca2+‐flux, shape change, integrin (CD11b) expression, and chemotaxis. The underlying molecular mechanism involved the activation of Rev‐ErbA‐α and induced a PI3K/PDK1/PKA‐dependent signaling cascade. Systemic application of ApoA‐IV prevented airway hyperresponsiveness (AHR) and airway eosinophilia in mice following allergen challenge. ApoA‐IV levels were decreased in serum from allergic patients compared to healthy controls. Conclusion Our data suggest that ApoA‐IV is an endogenous anti‐inflammatory protein that potently suppresses effector cell functions in eosinophils. Thus, exogenously applied ApoA‐IV may represent a novel pharmacological approach for the treatment of allergic inflammation and other eosinophil‐driven disorders.

cytokines, 2 and thereby modulate the immune microenvironment and promote several immunoregulatory functions. Eosinophils are involved in antigen-presentation 3 and T-cell activation, 4 interact with and activate other immunocompetent cells such as dendritic cells, 5 mast cells, 6 macrophages, 7 and neutrophils. 8 Moreover, activated eosinophils signal to and activate resident tissue cells such as epithelial cells, 9 endothelial cells, 10 goblet cells, 11 smooth muscle cells, 12 fibroblasts, 13 and neurons, 14 overall leading to the progression of inflammation, mucus secretion, tissue remodeling, and angiogenesis. 15 Thus, eosinophils are potent effectors and modulators of various diseases ranging from bronchial asthma 16 and atopic dermatitis 17 to eosinophilic esophagitis, 18 colitis ulcerosa 19 and hypereosinophilic syndrome. 20 In asthmatics, levels of eosinophil granule proteins such as eosinophil cationic protein (ECP) or eosinophil peroxidase (EPO) largely correlate with disease severity. 21 Moreover, patients who receive treatment based on eosinophil counts in sputum have significantly fewer exacerbations than patients treated according to standard therapy. 22 Of note, eosinophilic inflammation of the upper airways may also occur independent of allergy as observed in chronic rhinosinusitis (CRS) patients. 23 Similar to allergies, CRS causes not only physical suffering, but also impacts psychological well-being and daily functioning. Patients with eosinophilic CRS represent a unique subtype and remain largely resistant to medical and surgical interventions. Thus, therapies that specifically target eosinophilic expansion and effector functions are urgently needed.
The apolipoprotein ApoA-IV is-to some extent-found on chylomicrons and HDL in plasma; however, its lipid-free form is predominant in circulation, 24 where it is presumed to play anti-inflammatory roles. In fact, the expression of human ApoA-IV in ApoE−/− mice protected them from oxidative stress, decreased the secretion of proinflammatory cytokines after LPS administration and reduced the formation of atherosclerotic lesions. 25,26 Furthermore, in an experimental model of DSS-induced colitis, ApoA-IV inhibited leukocyte and platelet adhesive interactions and suppressed the upregulation of P-selectin on colonic endothelium. 27 In humans, ApoA-IV was found to inhibit histamine release from basophils in vitro, 28 and, interestingly, ApoA-IV levels increased in the blood of allergic rhinitis patients following sublingual immunotherapy and were inversely correlated with symptom-medication scores. 28 In this study, we set out to explore whether the anti-inflammatory properties of ApoA-IV also extend to eosinophil effector function in vitro and in a mouse model of allergen-induced pulmonary inflammation. Moreover, we assessed ApoA-IV levels in patients with eosinophil-driven diseases such as allergy and chronic rhinosinusitis. Our data clearly indicate that ApoA-IV is a potential resolution factor in eosinophilic inflammation and might have beneficial effects on eosinophil-driven diseases.

| ME THODS
Detailed description of patient cohorts, ethical permits, materials, and procedures is provided in the Methods section in this article's Online Repository (Appendix S1).

| Isolation peripheral blood eosinophils
Human peripheral blood eosinophils were isolated from citrated whole blood from allergic or healthy donors. In brief, erythrocytes were removed by dextran sedimentation and polymorphonuclear leukocytes (pellet) were separated from mononuclear cells (buffy coat) by density gradient centrifugation using Histopaque 1077. Eosinophils were

G R A P H I C A L A B S T R A C T
ApoA-IV is decreased in serum from untreated allergic patients compared to healthy controls. ApoA-IV potently inhibits eosinophil function in vitro. Systemic application of ApoA-IV reduces airway hyperreactivity and airway eosinophilia in a murine model of allergic asthma. ApoA-IV, apolipoprotein A-IV; HDM, house dust mite separated from neutrophils in the polymorphonuclear leukocyte fraction by negative magnetic selection using the MACS cell separation system (Eosinophil Isolation Kit; Miltenyi Biotec, Bergisch Gladbach, Germany) with a resulting purity of typically ≥ 98%. 29

| Shape change assay
Eosinophil shape change was assessed in polymorphonuclear leukocyte (PMNL) preparations and monitored by flow cytometry as an increase in the forward scatter signal. 30

| Calcium flux
Intracellular Ca 2+ release from purified human eosinophils was detected by flow cytometry using the Ca 2+ sensitive dye Fluo 3-AM. 29

| Chemotaxis
Eosinophil chemotaxis experiments were done with purified eosinophils, whereas neutrophil chemotaxis was performed in separate experiments with PMNL preparations. Chemotaxis assays were performed in a 48-well micro chemotaxis chamber using PVP-free polycarbonate filters with a pore size of 5 µm. Migrated cells were enumerated by flow cytometry. 32 Therefore, eosinophils and neutrophils were gated by their forward and side scatter properties and by autofluorescence.

| Cholesterol-rich microdomain (lipid raft) assessment
Lipid raft abundance was quantified by flow cytometry in purified eosinophils stained with FITC-cholera toxin B. 33

| Apoptosis assay
Purified eosinophils were stained with FITC-annexin-V/PI and analyzed by flow cytometry.

| House dust mite-induced allergic lung inflammation
The HDM model was performed as described by Plantinga et al 34 In brief, Balb/c mice were sensitized i.n. with 1 μg HDM extract on day 1 and were challenged intranasally with 10 μg HDM per day from day 7 to day 11. On day 15, lung function testing was performed or BAL fluid, bone marrow, and spleens were taken. Leukocytes were analyzed by flow cytometry.

| Statistical analysis
Data are shown as mean + or ± SEM for n observations, where n denotes independent experiments with cells from different donors. Comparisons of groups were performed as appropriate using Student's t test or Mann-Whitney test, 1-way ANOVA followed by Dunnett's or Tukey's post hoc test or 2-way ANOVA for repeated measurements followed by Bonferroni's post hoc test to determine the levels of significance for each group. Probability values of P < .05 were considered as statistically significant.

| ApoA-IV impairs eosinophil responsiveness
Since the effects of ApoA-IV on eosinophils have not been deciphered so far, we first explored the anti-inflammatory capacity of ApoA-IV in assays of eosinophil shape change, integrin upregulation, and intracellular Ca 2+ mobilization. When encountering a chemotactic factor, such as CCL11, eosinophils immediately prepare for diapedesis through the endothelium by rearranging their cytoskeleton.
Such morphological changes can be detected by flow cytometry as increases in the forward scatter properties of the cells. We studied the effects of ApoA-IV on eosinophil shape change in PMNL samples from healthy nonallergic donors. We pretreated samples with recombinant ApoA-IV or vehicle for 30 minutes, followed by stimulation with serial dilutions of CCL11, and shape change was monitored by flow cytometry. Of particular interest, already very low concentrations of ApoA-IV (1 µg/mL) led to a statistically significant decrease of eosinophil shape change, as the responsiveness to CCL11 was decreased by 50% ( Figure S2A). Besides shape change, upregulation of adhesion molecules such as α m β 2 integrins (CD11b/ CD18; Mac-1) is another precondition for eosinophil migration. To measure the impact of ApoA-IV on integrin mobilization, we pretreated human eosinophils in PMNL fractions with ApoA-IV (1 µg/ mL) or vehicle and stimulated again with CCL11. Coinciding with the effect on shape change, ApoA-IV clearly reduced the presence of CD11b molecules on the cell surfaces by 30% ( Figure S2B). Beside morphological changes and integrin upregulation, CCL11 induces a rapid and transient rise in intracellular Ca 2+ ions. Similar to shape change and CD11b, we found that ApoA-IV reduced this CCL11-induced Ca 2+ mobilization in a concentration-dependent fashion. As presented in Figure S2C, Ca 2+ flux was diminished by 30% and 45% in the presence of 1 and 3 µg/mL of ApoA-IV, respectively.

| ApoA-IV inhibits eosinophil chemotaxis
Having established that ApoA-IV affects cellular responsiveness of eosinophils, we next investigated the direct impact of ApoA-IV-in comparison to ApoA-I or isolated HDL-on eosinophil migration. Chemotaxis assays were carried out in a modified Boyden chamber using a 48-well microchemotaxis assembly. As displayed in Figure 1A, ApoA-IV (1 µg/ mL) not only inhibited eosinophil chemotaxis toward CCL11 (3 nmol/L), but also toward prostaglandin D 2 (30 nmol/L) by 56% and 70%, respectively. Similarly, ApoA-IV even abolished migration toward house dust mite extract (HDM; 100 µg/mL) which has been shown to directly activate and mobilize eosinophils. 35 For neutrophil chemotaxis toward IL-8, a reduction of 41% from CI = 3.2 to CI = 2.3 was observed ( Figure 2C).

| ApoA-IV acts NR1D1-dependently and signals via PI3K/PDK1 and PKA
It is assumed that ApoA-I attenuates neutrophil function via the ATPbinding cassette transporter AI (ABCAI), whereas anti-inflammatory effects of HDL are mediated via scavenger receptor BI (SRBI). 36 Thus, we next scrutinized whether ApoA-IV also signals via ABCAI or SRBI binding. As shown in Figure 2, ABCAI blocking averted the effect of ApoA-I ( Figure 2A) and the SRBI antibody impeded the HDL-mediated decrease of eosinophil chemotaxis ( Figure 2B). Notably, neither ABCAI nor SRBI blocking could prevent the inhibitory effect of ApoA-IV ( Figure 2C).
To further elucidate the downstream components of the ApoA-IV pathway in eosinophils, cells were incubated with protein kinase inhibitors. As illustrated in Figure 3A, blocking PI3K and PDK1 prevented the ApoA-IV induced inhibition of eosinophil chemotaxis.
Moreover, eosinophils that were pretreated with the PKA inhibitor H89 (1 µmol/L) even reached 112% of the CCL11-induced chemotaxis ( Figure 3B). In contrast, the adenylyl cyclase inhibitor SQ22536 (10 µmol/L) showed no significant effect. Hence, the anti-inflammatory activity of ApoA-IV seems to require PI3K and PDK1 as well as cAMP-independent activation of PKA. Consistently, PI3K activation has already been associated with other responses to ApoA-IV. 39,40
Since the ApoA-IV-related apolipoprotein ApoE was found to modulate the expression of proinflammatory molecules such as the CCL11 receptor CCR3 on activated microglia, 41 we investigated whether eosinophil CCR3 surface expression is altered in response to ApoA-IV (3 µg/mL). As depicted in Figure S3B, ApoA-IV did not reduce CCR3 staining after a 60-minutes treatment.  Figure S4A). Moreover, ApoA-IV specifically increased the percentage of apoptotic (PI-negative and positive) cells from 43 ± 5.2% (vehicle treatment) to 60 ± 2.6% in allergic donors, while ApoA-I was less effective (increase of apoptotic cells to 48 ± 5.6%) ( Figure S4B). No significant differences were observed for early apoptotic ( Figure S4C), late apoptotic ( Figure S4D), and necrotic cells ( Figure S4E).  as reflected by a ~42% reduction in eosinophil counts in the BAL fluid of ApoA-IV-treated mice compared to vehicle-treated controls ( Figure 4C). ApoA-IV also tended to reduce the numbers of alveolar macrophages in the BAL fluid; however, this difference did not reach significance ( Figure S6A). Similarly, counts of lymphocytes, monocytes, and neutrophils in the BAL fluid remained unchanged ( Figure   S6B-D). Of note, ApoA-IV supplementation also protected from systemic eosinophilia as reflected by a ~60% reduction of eosinophil counts in spleen tissue ( Figure S7A) and in bone marrow ( Figure S7B).

| ApoA-IV is decreased in serum of allergic patients and accumulates in mucus during chronic rhinosinusitis
ApoA-IV levels have been shown to increase under immunotherapy in patients with allergic rhinitis. 28 However, ApoA-IV serum levels in allergic patients and healthy controls have not been compared yet. Hence, in this study we evaluated ApoA-IV serum levels in 17 nonallergic healthy subjects and 49 untreated patients with respiratory allergic symptoms to aeroallergens (mainly grass pollen) (for further details please refer to Appendix S1, table E1). As presented in Figure 5A, ELISA analysis revealed consistently reduced ApoA-IV concentrations in serum of allergic patients, with a mean value of 428.8 ± 31.02 µg/mL, whereas serum ApoA-IV was 810.7 ± 120.1 µg/mL in healthy nonallergic controls. Albeit, no correlation between ApoA-IV serum levels and laboratory parameters such as sIgE was found (data not shown).
Similar to allergy, chronic rhinosinusitis (CRS) is characterized by a pronounced eosinophilic inflammation of the lining of the nose and paranasal sinuses. Thus, we assessed ApoA-IV levels in the mucus of CRS patients (for further details please refer to Appendix S1, table E2). Interestingly, ApoA-IV levels correlated with their histology scores: Low mean ApoA-IV levels of 0.47 ± 0.1 µg/mL and 0.55 ± 0.2 µg/mL were found in healthy controls and patients with low or medium clinical scores, respectively, whereas a mean mucus level of 5.28 ± 1.8 µg/mL was observed in patients with high histology scores or patients suffering from nasal polyps (CRSwNP) ( Figure 5B). Moreover, a significant positive correlation was found between ApoA-IV mucus levels and radiologic Lund-Mackay 42 scores (r = 0.5039; P = .033) ( Figure 5C).
Thus, our data indicate that ApoA-IV expression and/or metabolism is altered in allergic patients. Moreover, ApoA-IV associates with the severity of inflammation in mucus of CRS patients.

| D ISCUSS I ON
In the present study, we demonstrate through several lines of evidence that apolipoprotein A-IV bears potent anti-allergic properties and thereby reveal a hitherto unknown anti-inflammatory (C) Pearson correlation of ApoA-IV mucus levels and CT scores was statistically significant. Data are shown as individual values and in A and B as mean ± SEM; *P < .05, ***P < .001, Student's t test or 1-way ANOVA mechanism: First, we found that recombinant ApoA-IV inhibits eosinophil responses to chemoattractants in assays of Ca 2+ mobilization, shape change, integrin (CD11b) surface upregulation, and chemotaxis. The underlying molecular mechanism appears distinct from ApoA-I and HDL-induced signaling cascades as it occurs independently from ABCAI and SRBI binding, but is mediated through a novel pathway involving nuclear receptor NR1D1 (Rev-ErbA-α) and the protein kinases PI3K, PDK1, and PKA. Second, we established that ApoA-IV specifically enhances apoptosis in eosinophils from allergic individuals but not healthy volunteers. Third, and in line with these in vitro data, systemic administration of ApoA-IV prevented pulmonary eosinophilia and markedly improved airway hyperresponsiveness in a mouse model of HDM-induced airway inflammation. And finally, we found that therapy-naïve allergic patients have noticeably lower ApoA-IV serum levels compared to healthy individuals. Moreover, we show that ApoA-IV is present in mucus from CRS patients, where it might act in an anti-inflammatory manner.
Up to now, low ApoA-IV levels have been associated with serious conditions such as cardiovascular disorders [43][44][45] and cancer. [46][47][48] For instance, ApoA-IV has been identified as a reliable biomarker in ovarian cancer. 49 ApoA-I were found to be positively correlated with FEV1 in subjects with allergic asthma. 62 Conversely, a study in schoolchildren showed that high ApoA-I is associated with the manifestation of asthma and atopy. 63 Of interest, increased serum levels of ApoA-IV were previously reported in patients with allergic rhinitis under allergen-specific immunotherapy. 28 We made the surprising observation that ApoA-IV serum levels are noticeably reduced in therapy-naïve allergic patients. However, the reasons for decreased ApoA-IV levels under inflammatory and allergic conditions are a matter of speculation. Previous studies revealed an association of APOA-IV gene variants with ApoA-IV levels and increased risks for certain diseases such as coronary heart disease, 68 renal diseases, 69 depression, 70 and obesity. 71 For instance, Ser347 homozygotes have clearly lower ApoA-IV plasma levels compared with carriers of the Thr347 allele and show a significantly increased risk of coronary heart diseases. 68 Moreover, it was demonstrated that individuals who are homozygous for the Ser347 allele have higher BMI and percentage body fat compared with individuals homozygous for Thr347. 71 Up to-date no data are available whether APOA-IV gene variants are also associated with a higher risk for chronic atopic diseases such as allergic asthma or rhinitis. In nonallergic patients with chronic rhinosinusitis, we provide evidence that anti-inflammatory ApoA-IV is not only present in nasal mucus but it is also correlated with the extent of inflammation.
We assume that ApoA-IV accumulates in the paranasal sinuses due to increased vascular permeability. However, it has been proposed that monocytes 72 and dendritic cells 73 are able to express ApoA-IV, thus we cannot exclude that ApoA-IV is also released locally by infiltrating inflammatory cells.
In our present work, we observed that ApoA-IV potently affected effector cells of allergic inflammation such as eosinophils and neutrophils. Pretreatment of eosinophils with recombinant ApoA-IV decreased their responses to chemoattractants by means of Ca 2+ flux, shape change and integrin surface expression. Moreover, ApoA-IV reduced eosinophil migration to baseline levels involving a signaling cascade mediated by Rev-ErbA-α, the NR1D1 (nuclear receptor subfamily 1, group D, member 1) gene product, which is a dominant transcriptional silencer that represses the expression of genes involved in numerous physiological functions, including circadian rhythm and metabolism, 74 and plays a crucial role in maintaining immune functions. 75,76 For instance, inflammatory stimuli were shown to promote Rev-ErbA-α degradation in mice, and complete lack of Rev-ErbA-α further enhanced inflammation in the lungs following inflammatory challenge. 76 In macrophages, Rev-ErbA-α decreased integrin expression and adhesion. 77 In addition, it was recently shown that pharmacological activation of Rev-ErbA-α reduced lipopolysaccharide (LPS)-induced neuro-inflammation in mouse microglia in vitro and in vivo. 78 In further experiments, we demonstrated that the ApoA-IV-induced signaling cascade involves the activity of PI3K, PDK1, and PKA. PI3Ks are a family of enzymes involved in cellular functions such as cell growth, proliferation, differentiation, motility, and migration. 79 In numerous cell types, PI3K acts in a heterodimeric form consisting of one 85-kDa regulatory and one 110-kDa catalytic subunit. In previous work, we already elucidated the critical role for the PI3K/PDK1 cascade in transducing inhibitory signals on eosinophil effector function mediated by the prostaglandin E 2 receptor EP4. 80 In another study, ApoA-IV has been identified to regulate food intake by acting as a satiation factor, which is released by, and is acting in, the hypothalamus. 39 In this context, ApoA-IV triggered the activation of the PI3K cascade in cultured primary hypothalamic neurons, and inhibition of PI3K signaling in rat brain noticeably decreased the potency of ApoA-IV to reduce food intake. 39 Moreover, cell culture experiments showed that ApoA-IV improved glucose uptake in adipocytes by upregulating GLUT4 translocation in a PI3Kdependent manner. 40 These results further support our observation that ApoA-IV engages with the PI3K signaling pathway to promote its anti-inflammatory actions.
To confirm the in vivo relevance of the observed anti-inflammatory activities of ApoA-IV, we performed a well-established mouse model of HDM-induced airway inflammation. First, we revealed that HDM-induced allergic inflammation in mice is accompanied by a significant drop in ApoA-IV serum levels compared to healthy control mice. However, whether this effect is due to reduced ApoA-IV synthesis from epithelial cells in the small intestine or due to increased ApoA-IV degradation by the kidneys needs to be clarified in further studies. Of note, we could show that daily systemic treatment with ApoA-IV for several days not only improved lung parameters, but also reduced eosinophil counts in the airways, spleen, and bone marrow of HDM-challenged mice, suggesting that the ability of ApoA-IV to inhibit air- Moreover, previous work provided further evidence that ApoA-IV is capable of inhibiting eosinophil-driven inflammatory processes other than asthma. ApoA-IV knockout mice exhibited a significantly greater inflammatory response in DSS-induced colitis than did their wild type littermates. This greater susceptibility to DSS-induced inflammation was reversed upon exogenous administration of ApoA-IV. The authors proposed that ApoA-IV is an endogenous antiinflammatory protein that acts via inhibition of P-selectin-mediated leukocyte and platelet adhesive interactions. 27 In conclusion, our results unequivocally demonstrate the antiinflammatory properties of ApoA-IV on effector cells of allergic inflammation. Further, we provide novel evidence that systemic elevation of ApoA-IV protects against airway hyperresponsiveness, leukocyte infiltration into the airways and reduces eosinophil count in the circulation. Moreover, ApoA-IV serum levels are significantly decreased in allergic patients and in HDM-exposed mice. Thus, the present data collectively suggest that ApoA-IV has promising diagnostic and therapeutic potential for allergic and inflammatory conditions, particularly those involving eosinophil effector functions.

ACK N OWLED G M ENTS
We thank Birgit Brodacz, Iris Red, Kathrin Rohrer and Ilse Lanz for their skilled technical assistance.

CO N FLI C T S O F I NTE R E S T
AH received consultancy fees from AstraZeneca. All other authors declare no conflicts of interest.