Background: Eosinophilia is a hallmark of helminth infections and eosinophils are essential in the protective immune responses against helminths. Nevertheless, the distinct role of eosinophils during parasitic filarial infection, allergy and autoimmune disease-driven pathology is still not sufficiently understood. In this study, we established a mouse model for microfilariae-induced eosinophilic lung disease (ELD), a manifestation caused by eosinophil hyper-responsiveness within the lung.

Methods: Wild-type (WT) BALB/c mice were sensitized with dead microfilariae (MF) of the rodent filarial nematode Litomosoides sigmodontis three times at weekly intervals and subsequently challenged with viable MF to induce ELD. The resulting immune response was compared to non-sensitized WT mice as well as sensitized eosinophil-deficient dblGATA mice using flow cytometry, lung histology and ELISA. Additionally, the impact of IL-33 signaling on ELD development was investigated using the IL-33 antagonist HpARI2.

Results: ELD-induced WT mice displayed an increased type 2 immune response in the lung with increased frequencies of eosinophils, alternatively activated macrophages and group 2 innate lymphoid cells, as well as higher peripheral blood IgE, IL-5 and IL-33 levels in comparison to mice challenged only with viable MF or PBS. ELD mice had an increased MF retention in lung tissue, which was in line with an enhanced MF clearance from peripheral blood. Using eosinophil-deficient dblGATA mice we demonstrate that eosinophils are essentially involved in driving the type 2 immune response and retention of MF in the lung of ELD mice. Furthermore, we demonstrate that IL-33 drives eosinophil activation in vitro and inhibition of IL-33 signaling during ELD induction reduces pulmonary type 2 immune responses, eosinophil activation and alleviates lung lacunarity.

In conclusion, we demonstrate that IL-33 signaling is essentially involved in MF-induced ELD development.

Microfilariae purification

MF were purified as described previously [20]. In short, cotton rats (Sigmodon hispidus) naturally infected with L. sigmodontis were bled into EDTA tubes (Sarstedt AG & Co. KG, Nümbrecht, Germany). The blood was diluted 1:2 in pre-warmed RPMI-1640 media (Gibco, ThermoFisher Scientific, California, USA) and layered on top of a 25% and 30% sucrose gradient (Carl Roth, Karlsruhe, Germany) of 0.25 M in iso-osmotic percoll (Merck, Darmstadt, Germany). The gradient was centrifuged at 300g for 30 min without breaks. The white layer containing the MF was collected and washed 3 times with pre-warmed RPMI-1640. MF were counted in a Neubauer chamber (Laboroptik GmbH, Friedrichsdorf, Germany) and adjusted to 1x 10⁵ in 150 µl PBS (Dulbecco´s PBS, Gibco, ThermoFisher Scientific, California, USA). The MF were checked for viability via microscopy.

Bone marrow-derived eosinophil culture

Bone marrow was isolated from hind legs of WT BALB/c mice using a 20-gauge needle and 5 ml RPMI-1640 for each bone. Isolated cells were centrifuged (400g, 10 min, 4°C) and red blood cells were lysed by adding 1 ml of red blood cell lysis buffer (eBioscience, ThermoFisher Scientific, California, USA) for 5 min at room temperature (RT). Cells were washed with 9 ml PBS (Gibco, ThermoFisher Scientific, California, USA) and centrifuged at 400g at room temperature for 10 min, adjusted to 1x 10⁶/ml in RPMI-1640 with the addition of 20% fetal calf serum, penicillin (10000 U/ml), streptomycin (100 mg/ml), glutaMAX, gentamycin (50 mg/ml) (Gibco, ThermoFisher Scientific, California, USA for all contents). Cells were cultured with 100 ng/ml FMS-like tyrosine kinase 3 ligand (FLT3L) and 100 ng/ml stem cell factor (SCF) (PeproTech Inc., ThermoFisher Scientific, California, USA) for 4 days and subsequently with 20 ng/ml IL-5 (PeproTech Inc., ThermoFisher Scientific, California, USA) until day 12. Half of the medium was replaced on day 2, 6, 10 and the whole medium was replaced on day 4 and 8. Cells were adjusted using CASY automated cell counter (OLS OMNI Life Science GmbH & Co. KG, Bremen, Germany). Cells were centrifuged and adjusted to 1x 10⁶/ml in the aforementioned culture media with 10% fetal calf serum. Cell purity was determined using flow cytometry and cell viability was assessed using the Annexin-V-PI staining kit (ThermoFisher Scientiffic, California, USA) and was above 80% purity and 55-85% viability (S8A, B Fig).

Sensitization and induction of ELD mice

The sensitization protocol was adjusted from Egwang and Kazura, 1990 [7]. Mice were injected subcutaneously into the neck fold with 1x 10⁵ dead (killed by freezing) MF in 100 µl PBS once a week for three weeks. Two weeks later, mice were injected with 1x 10⁵ freshly isolated, living motile MF in 150 µl PBS (Dulbecco´s PBS, Gibco, ThermoFisher Scientific, California, USA) intravenously. Control mice received 150 µl PBS.

HpARI2 treatment of ELD mice

Full length HpARI2 with a 6-His tag was expressed in Expi293 cells, and purified by nickel affinity chromatography, as previously described [57, 58]. HpARI2 was used to inhibit IL-33 signaling during ELD development. Therefore, 10 µg HpARI2 in 10 µl PBS (Gibco, ThermoFisher Scientific, California, USA) was administered intranasally under anesthesia induced with 2% isoflurane (Piramal Critical Care, New-Delhi, India). For ELD induction, mice were sensitized as usual once a week for three weeks total with dead MF and two weeks later, starting 1 h before the final MF challenge, mice were treated every three days with HpARI2. Control mice received 10 µl PBS intranasally.

Broncho-alveolar lavage

After euthanasia of the animals, the thorax was opened carefully to allow the lungs to dilate completely. The throat of the animals was cut open to expose the trachea. A 20G Vasofix (B. Braun Melsungen, Melsungen, Germany) intravenous vein catheter was inserted into the trachea and the lung was flushed with 1 ml PBS. The first ml was centrifuged at 400g for 10 min at 4°C and the supernatant was used for the analysis of cytokines. The lung was flushed with additional 4 ml PBS (Gibco, ThermoFisher Scientific, California, USA). Finally, 1 ml of PBS was injected into the lung to dilate the lobes before paraffin (Carl Roth, Karlsruhe, Germany) fixation. Flushed out BAL cells were centrifuged at 400g for 10 min at 4°C and used for flow cytometry.

Lung tissue dissociation

The lung tissue was dissociated using the gentleMACS in combination with the tissue dissociation kit (Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s protocol. Briefly, the lung was cut into the respective lobes and added into the dissociation tube with 2.4 mL of 1× buffer S, 100 μL of enzyme D, and 15 μL of enzyme A. Lung lobes were dissociated at 37°C for 30 min under rotation of the tube (program: 37C_m_LDK_1) in the gentleMACS (Miltenyi Biotec, Bergisch Gladbach, Germany). Isolated cells were collected by centrifugation (400g, 10 min at 4°C) and red blood cells were lysed for 5 min at RT, washed and resuspended in culture media (eBioscience, ThermoFisher Scientific, California, USA). Finally, cells were filtered through a 70 µm MACS SmartStrainer (Miltenyi Biotec, Bergisch Gladbach, Germany). Cell viability was assessed using DAPI staining after lung dissociation (S8C Fig).

Purification of native proteins from the lung tissue

Lung tissue was placed into RIPA-buffer (eBioscience, ThermoFisher Scientific, California, USA) and lysed in a Precellys24 (VWR international GmbH, Langenfeld, Germany). Cells were incubated for 1 h at 4°C and centrifuged for 30 min at 16,000 g at 4°C. The supernatant was collected and adjusted to 100 µg/ml using Bradford protein quantification. Samples were diluted 1:100 in Bradford reagent (Sigma-Aldrich, MERCK, Darmstadt, Germany) and compared to a BSA standard.

Detection of proteins from lysed lung tissue

10 µg lung protein extraction were used for the respective ELISA kits. CCL5, CCL11 (R &D Systems, Minnesota, USA), Amphiregulin (ThermoFisher Scientific, California, USA), major basic protein (Antikörper-Online, Aachen, Germany), MMP activity assay (Abcam, Cambridge, United Kingdom) and eosinophil peroxidase (Lifespan Biosciences, Washington, USA) ELISA measurements were performed according to the manufacturers’ protocols.

Spleen dissociation

Spleens were isolated from the euthanized mice and dissociated using a syringe stamp and pressing the organ through a 70 µm MACS SmartStrainer (Miltenyi Biotec, Bergisch Gladbach, Germany). Splenocytes were collected by centrifugation at 400g at RT for 5 min and red blood cells were lysed (eBioscience,Thermofisher Scientific, California, USA). Cells were washed with RPMI1640 media (Gibco, ThermoFisher Scientific, California), counted and used for flow cytometry.

Microfilariae quantification from lung

Lung and heart were isolated from the mice and placed into a glass petri dish. 20 G needles (B. Braun, Melsungen, Germany) were placed into the left and right atrium. An additional needle was placed into the right atrium and attached to a 10 ml syringe (B. Braun, Melsungen, Germany) containing PBS (Gibco, ThermoFisher Scientific, California, USA). The lungs were flushed with 10 ml PBS. The flow through was collected in a 15 min falcon tube and centrifuged for 10 min at 400g at RT. The supernatant was discarded and 1 ml of 1x red blood cell lysis buffer (ThermoFisher Scientific, California, USA) was added to the pellet. After 5 min at RT, the samples were centrifuged again at 400g for 10 min at RT. 950 µl of the supernatant was carefully removed. The pellet was resuspended in the remaining 50 µl and MF were counted via microscopy.

Microfilariae quantification from peripheral blood

50 µl of blood was collected from mice in EDTA Tubes (Sarstedt AG & Co. KG, Nümbrecht, Germany) and transferred into 1.5 ml Eppendorf tubes (Eppendorf SE, Hamburg, Germany). 1 ml of red blood cell lysis (ThermoFisher Scientific, California, USA) was added and incubated for 5 min at RT. Following centrifugation at 400g for 10 min at RT, 1 ml of supernatant was removed, the pellet resuspended and all MF contained in the remaining 50 µl pellet were counted via microscopy.

Flow cytometry

10⁶ cells were blocked in 1 µg/ml rat IgG/PBS (Sigma-Aldrich, MERCK, Darmstadt, Germany) for 1 h at 4°C. The blocking buffer was removed by centrifugation (400g, 8 min, 4°C) and the cell pellet was dissolved in 100 µl fixation buffer (BioLegend, California, USA). Cells were fixated at RT for 15 min and washed with PBS (Gibco, ThermoFisher Scientific, California, USA), centrifuged at 400g for 10 min at RT and stained in 20 µl surface marker antibody mix (S1 Table) for 30 min at 4°C. Stained cells were washed twice with PBS and centrifuged at 400g for 10 min at RT. For additional intracellular stainings, cells were washed twice with 100 µl permeabilization buffer (BioLegend, California, USA) and subsequently stained for 30 min at 4°C. Finally, cells were washed twice with PBS and measured using a CytoflexS flow cytometer (Beckman Coulter, California, USA). ILC2s were gated as CD45⁺, linage^-, TCRb^-, CD90.2⁺, ST2⁺, GATA3⁺, RELM-α positive macrophages as CD45⁺, CD206⁺, Siglec-F⁺, RELM-α⁺ and eosinophils as CD45⁺, CD11c^-, Siglec-F⁺, CD11b⁺. Further, neutrophils were gated as CD45⁺, Ly6G⁺, alveolar macrophages as CD45⁺, CD206⁺, Siglec-F⁺. The gating strategy is shown in S9 Fig

Enzyme linked immunosorbent assay

Invitrogen ELISA (IFNγ, IL-4, IL-5, IL-33,) (Invitrogen, ThermoFisher Scientific, California, USA) were performed according to the manufacturer’s protocol from serum, bronchoalveolar lavage (BAL), pleura wash or lung homogenates. Plates were coated with 1x coating buffer. 50 µl serum was diluted 1:2 in assay buffer, 100 µl of either BAL, pleura wash or lung homogenates was transferred onto the plates alongside the respective standard samples and incubated for 2 h at RT and 200 rpm. Samples were blocked in assay diluent. Plates were washed and incubated with the detection antibody for 1 h at RT and 200 rpm. Plates were washed and horseradish peroxidase--conjugate was added for 30 min at RT and 200 rpm. Colorimetric detection was performed by adding 50 µl 3,3´5,5`-Tetramethylbenzimidin into each well. The reaction was stopped using 50 µl of 1 M H₂SO₄ (Carl Roth, Karlsruhe, Germany) to each well. Optical density of the plates was measured in a SpectraMax Molecular Devices LLC, California, USA) at 450 and 570 nm wave length. Concentrations were calculated using the Softmax Pro software (Molecular Devices LLC, California, USA).

Detection of IgE

Parasite-specific serum IgE was detected by coating ELISA plates (Sigma-Aldrich, MERCK, Darmstadt, Germany) with 20 µg/ml female L. sigmodontis adult worm crude extract diluted in PBS (Gibco, ThermoFisher Scientific, California, USA) for 24 h at 4°C. The supernatant was discarded and plates were washed and blocked for 1 h with 5% BSA/PBS (MERCK, Darmstadt, Germany) at RT. Serum samples were diluted 1:50, 50 µl were added onto the plate and incubated for 2 h at RT. After the incubation, the samples were discarded and the plates were washed three times. The biotinylated anti-mouse-IgE antibody was diluted 1:200 (2µg/ml) in PBS, added to each well and incubated for 1 h at RT, at 200 rpm. Following a washing step, the plates were incubated with the conjugated horseradish peroxidase for 30 min at RT, shaking at 200 rpm. Colorimetric detection was performed by adding 50 µl 3,3´5,5`-Tetramethylbenzimidin into each well. The reaction was stopped using 100 µl of 1 M H₂SO₄ H₂SO₄ (Carl Roth, Karlsruhe, Germany) to each well. Optical density of the plates was measured in a SpectraMax (Molecular Devices LLC, California, USA) at 450 and 570 nm wave length.

DNA quant

DNA quantification was performed as described previously [20]. In brief, the Invitrogen Quant-iT dsDNA Assay Kit, high sensitivity (ThermoFisher Scientific, California, USA), was used for DNA quantification of supernatant of stimulated bone marrow-derived eosinophils. Eosinophils were cultured in 96-well plates with medium containing murine IL-5 (20ng/ml PeproTech inc., ThermoFisher Scientific, California, USA). After 24 h, 2.5 U/well of the micrococcus exonuclease (PeproTech inc., ThermoFisher Scientific, California, USA) was added to all samples at 37°C for 15 min to disassociate DNA bound to the wells. The micrococcal nuclease reaction was stopped by adding 1 mM EDTA. Culture plates were subsequently centrifuged at 400g for 8 min at 4°C and the supernatant was removed and transferred to a new 96-well plate (Sigma-Aldrich, MERCK, Darmstadt, Germany). The quantification was done with 100 μl Quant-iT™ dsDNA HS buffer after a 1:200 dilution. The diluted DNA Quant-iT solution was added to 20 μl of the supernatant. For quantification, the provided standard ranging from 0-10 ng/ml was used. The absorption of sample and standard was measured using the Tecan infinity M200 (Tecan Group, Männedorf, Switzerland) at 485/525 nm.

Lung histology

Mice were sacrificed and the lung lobes were inflated with 1 ml PBS (Gibco, ThermoFisher Scientific, California, USA) via the trachea using a 20G Vasofix intravenous vein catheter (B. Braun, Meisungen, Germany). To maintain the inflated status of the lung tissue, the trachea was closed with a yarn tied into a knot. The lung was removed together with the heart and placed into 2% methanol free formalin (ThermoFisher Scientific, California, USA) for 24 h at RT. The lung tissue was washed twice with PBS (Gibco, ThermoFisher Scientific, California, USA) and dehydrated by being submerged in ethanol baths of ascending concentration (50, 70, 80, 96, 100%) and finally 100% xylene for 30 min (STP-120 automated sample processor, Especialidades Médicas Myr, S.L., Tarragona, Spain). The tissue was embedded in paraffin (Merck, Darmstadt, Germany) at 56°C (modular tissue embedding center, Especialidades Médicas Myr, S.L., Tarragona, Spain). Hardened paraffin blocks were cut into 4 µm sections, paraffin was removed with alcohol baths, sections were stained with Mayer´s hematoxylin and eosin (Merck, Darmstadt, Germany) and mounted with entellan (Merck, Darmstadt, Germany). Stitched histology images were taken on the Axio-observer 5 (Zeiss, Oberkochen, Germany) using Zen V3.6 (Zen Blue edition V3.6, Zeiss, Oberkochen, Germany). Fifty 10x10 µm Random regions of interest were generated using KNIME (https://www.knime.com/ [59]) and lacunarity analysis was performed using the ImageJ [60] plugin FracLac (Karperien, A., FracLac for ImageJ. http://rsb.info.nih.gov/ij/plugins/fraclac/FLHelp/Introduction.htm). Protocol adapted from Chenery et al., 2019 [61].