C57BL/6 (B6) mice were purchased from Japan SLC Inc. Slpi^−/− and Fcer1g^−/− mice have been described previously (20, 26). Congenic B6-Ly5.1 (CD45.1) mice were purchased from Sankyo Labo Service. These mice were kept and bred in the animal unit at Kanazawa Medical University and Tohoku Medical and Pharmaceutical University, an environmentally controlled and specific pathogen-free facility, in accordance with the guidelines for experimental animals defined by the facilities. All of the studies were approved by the Animal Studies Committee at Kanazawa Medical University and Tohoku Medical and Pharmaceutical University.

Total RNA was extracted using ReliaPrep RNA Cell (Promega) or miReasy Mini kit (Qiagen), and cDNA was prepared using a ReverTra Ace qPCR RT kit (Toyobo) for reverse transcription. The gene transcript levels of mouse SLPI and MMP-9 and housekeeping S16 ribosomal protein (RPS16) were quantified by a real-time PCR using Go-Taq qPCR Master Mix (Promega) on a DNA Engine Opticon2 system (MJ Research). The relative amount of gene transcript was calculated and normalized by dividing the calculated value for the gene of interest by the housekeeping gene value. The PCR conditions for all genes were as follows: 95°C initial activation for 2 min, followed by 40 cycles of 95°C for 15 s, and 60°C for 60 s, and fluorescence determination at the melting temperature of the product. The primers were as follows: mouse RPS16 forward, 5′-GATATTCGG GTCCGTG TGA-3′, and reverse, 5′-TTGAGATG GACTGTCGGATG-3′, yielding a 69-bp product; mouse Slpi forward, 5′-AGCCACAATGCCGTACTGACT-3′, and revere, 5′-AGGCTTCCTCCAC ACTGGTT T-3′, yielding a 115-bp product; mouse Mmp9 forward, 5′-TGACTACGATAAGGACG GCAAA-3′, and reverse, 5′-GATGAACGGGAAC ACACAGG-3′, yielding a 100-bp product.

Basophils, eosinophils, and mast cells were derived from BM cells. The preparation of BM-derived basophils (BMBs) was carried out as described elsewhere (27). Briefly, BM cells were cultured with 5 ng/ml IL-3 (Pepro Tech) for 12 days. On Day 12, c-kit⁻ DX5⁺ cells were isolated as basophils using the MACS system with magnetic microbead-conjugated anti-DX5 antibody (Miltenyi Biotec). BM-derived eosinophils (BMEos) were induced from BM cells as previously described (28). Briefly, BM cells were cultured in the presence of 100 ng/ml stem cell factor (SCF) (Miltenyi Biotec) and 100 ng/ml FMS-like tyrosine kinase 3 ligand (Flt3L; Miltenyi Biotec) for 4 days. On Day 4, SCF and Flt3L were replaced with IL-5 (10 ng/ml; Peprotech). The cultured cells were collected and used as BMEos (purity, ≥95%) on Day 14. Mast cells were grown from BM cells as described (29). Mast cells were prepared by culturing BM cells in the presence of 5 ng/ml IL-3 for 8–12 weeks. Cellular proliferation was determined with propidium iodide staining using an ADAM-MC automatic cell counter (NanoEnTek Inc.).

Cells were fixed in 4% paraformaldehyde in PBS and then treated with permeabilizing buffer (BD Bioscience). The cells were blocked with blocking reagent (Toyobo) for 1 h at room temperature, incubated with biotinylated goat anti-mouse SLPI (R&D Systems) in Can get signal solution A (Toyobo) overnight at 4°C, and then incubated with allophycocyanin-labeled streptavidin in Can get signal solution B (Toyobo) for another 1 h at room temperature. The specimens were mounted with SlowFade Gold and the nuclei were stained with 4′,6-diamino-2-phenylindole (DAPI; Thermo Fisher Scientific). Images were obtained with a BZ-9000 fluorescence microscope (Keyence). For the TEM analysis, the samples were postfixed in 1% OsO₄ in 0.1 M sodium cacodylate buffer and embedded in Quetol 651 (Polysciences). Ultrathin sections (80 nm) of cells were cut on an Ultracut UCT ultramicrotome (Leica). Images were obtained with a TEM H-7650 (Hitachi High-Technologies).

Flow cytometry was conducted using the following antibodies (All purchased from Biolegend unless stated otherwise): anti-IgE (RME-1), anti-TLR4 (SA15-21), anti-Siglec-F (E50-2440; BD Bioscience), anti-CCR3 (J073E5), anti-CD11b (M1/70; BD Bioscience), anti-ST2 (DIH9), anti-FcεRIα (MAR-1), anti-CD23 (B3B4), anti-CD123 (5B11), anti-CD49b (DX5), anti-CD117 (2B8), anti-CD45.1 (A20), and anti-CD45.2 (104). Fc-mediated nonspecific staining was blocked with anti-CD16/32 (2.4G2 hybridoma culture supernatant). Events were acquired using a FACSCanto II (BD Bioscience), and the data of 10,000–100,000 events were analyzed using the FACSDiva (BD Bioscience) or FlowJo software programs (FlowJo). The surface molecule expression was calculated by defining the delta mean fluorescence intensity between the specific antibody stain and the isotype-matched control antibody. Splenic basophils and eosinophils, and peritoneal eosinophils were isolated by using a cell sorting system (SH800; Sony Biotechnology). Separation of these cells were shown in Figure S1 in Supplementary Material. Briefly, CD4⁺CD8⁺B220⁺ cells were depleted from B6 splenocytes using magnetic separator (Miltenyi Biotec). Splenic basophils were sorted by gating on FcεRIα⁺ cells from CD4^–CD8^–B220⁻ spleen cells. Splenic eosinophils were also sorted by gating on Siglec-F⁺ cells from CD4^–CD8^–B220⁻ spleen cells. Peritoneal eosinophils were isolated by gating on Siglec-F⁺ cells. The purity of the sorted populations was consistently ≥90%, as determined by the FcεRIα⁺ DX5⁺ and Siglec-F⁺ phenotypes.

Bone marrow-derived basophils were incubated for 1 h at 37°C with 5 µg/ml anti-2,4,6-trinitrophenol (TNP)-IgE in 200 µl of culture medium and then stimulated for 12 h at 37°C with 1 µg/ml TNP-OVA in 200 µl of HEPES-tyrode’s buffer, pH 7.4. Eosinophils were incubated with LPS (0.1 µg/ml) for 12 h. The culture supernatants were collected, and the tryptase and chymase activities were detected by adding MeOSuc-AAPV-pNA and N-Suc-AAPF-pNA (chromogenic substrates; Sigma-Aldrich) at a final concentration of 1 mM, respectively. The absorbance was measured at 405 nm at 37°C.

β-Hexosaminidase activity was assayed as previously described (30). Briefly, 50 µl of the sample was incubated with 50 µl of 1 mM p-nitrophenyl-N-acetyl-β-d-glucosamide (Sigma-Aldrich) dissolved in 0.1 M citrate buffer (pH 5.0) in a 96-well microtiter plate at 37°C for 1.5 h. The reaction was stopped with 200 µl/well of 0.1 M NaOH/0.2 M glycine, pH 10.7, and measured at 405 nm in a plate reader. For the analysis of the total cell content of β-HEX, cells were lysed with 1.0% (vol/vol) Triton X-100 in HEPES-tyrode’s buffer. The percentage of degranulation was calculated as follows: the absorbance of culture supernatants at 405 nm/absorbance of the total cell lysate supernatants at 405 nm.

Bone marrow-derived eosinophils were incubated with LPS (Sigma-Aldrich), IL-33, and Phorbol 12-myristate 13-acetate (PMA)-ionomycin (Biolegend) for 12 h at the indicated concentration. The EPO activity was measured by the spectrophotometric method (31). Briefly, 100 µl of culture supernatant from each sample was placed in a 96-well plate, and 100 µl of substrate solution containing 0.1 µM o-phenylenediamine-dihydrochloride, 0.1% Triton X-100, and 1 µM hydrogen peroxide (Sigma-Aldrich) was added in each well. After incubation for 30 min at 37°C, the enzymatic reaction was stopped by adding 50 µl of 4 M sulfuric acid. Absorbance was measured at 492 nm using a Multiscan JX system (Thermo Fisher Scientific).

Bone marrow-derived basophils and BMEos were incubated with the indicated stimulators. The levels of cytokines and histamine in the culture supernatants were determined using ELISA kits, in accordance with the manufacturers’ instructions. The mouse IL-4, IL-6 ELISA MAX Standard kit (Biolegend), the IL-13 Ready-Set-Go ELISA set (Thermo Fisher Scientific), a Histamine ELISA test kit (Neogen), and a Cysteinyl leukotrine ELISA kit (Cayman Chemical) were used.

Chemotaxis of BMEos was performed in 24-well chemotaxis chambers containing polycarbonate filters (pore size: 5 µm, Kurabo, Osaka, Japan). The wells of the lower chamber were filled with 5% bovine serum albumin (Wako)-RPMI1640 medium (Sigma-Aldrich) with LPS (Escherichia coli O111:B4; Sigma-Aldrich) and CCL11 (R&D Systems) and incubated for 2 h at 37°C. Eosinophils were then applied to the wells of the upper chambers and incubated for a further 2.5 h. The number of cells migrating from the upper chamber to the lower chamber was counted via the trypan blue-exclusion test. An invasion assay was performed with a BioCoat invasion system (BD Bioscience) in accordance with the manufacturer’s instructions. The insert membrane was coated with a basement membrane extract. Eosinophils (5 × 10⁵ cells) were applied to the wells of the upper chambers. The lower chambers of the 24-well plate were filled with 750 µl of serum-free RPMI1640 medium with 1 µg/ml LPS and 10 nM CCL11 and then incubated for 22 h at 37°C. The number of cells migrating from the upper chamber to the lower chamber was again counted using the trypan blue-exclusion test. The percent of invasion was calculated as the number of cells invading through the Matrigel insert membrane divided by the number of cells migrating through the control insert membrane.

Cells were solubilized in lysis buffer (1.0% NP-40, 50 mM HEPES, pH7.4, 150 mM NaCl) containing protease and phosphatase inhibitor (Thermo fisher scientific). Cell lysates were separated by SDS-PAGE, transferred to a Polycinylidene difluoride (PVDF) membrane, and detected with the following antibodies using ECL substrate (Bio-Rad): rabbit anti-PLCγ2, antiphospho-PLCγ2, anti-Erk1/2, antiphospho-Erk1/2, antiphospho-JNK1, anti-Elk-1, antiphospho-Elk-1(Ser383), anti-NF-κBp65, antiphospho-NF-κBp65, mouse anti-IκBα (Cell Signaling Technology), rabbit antiphospho Elk-1 (Thr417) (Thermo Fisher Scientific), rabbit anti-NF-κBp65 mouse anti-IκBβ (Santa Cruz Biotechnology), goat antimouse SLPI, rabbit antimouse MMP-9 (R&D Systems), rabbit anti-JNK1, rat anti-mouse MCP8, rat antimouse MCP11 (Biolegend), horseradish peroxidase (HRP)-conjugated goat (Santa Cruz Biotechnology), mouse (Cell Signaling Technology), or rabbit IgG antibodies (Cell Signaling Technology). For the immunoprecipitaion analysis, cell lysates (5 × 10⁶ cells) were precleaned by Dynabeads protein G (VERITAS), and were sequentially incubated with mouse anti-JIP3 (F-6; Santa Cruz Biotechnology) or mouse IgG1κ isotype control (MG1-45; Biolegend) and Dynabeads protein G. The immunoprecipitates were detected with the following antibodies using SuperSignal™ West Pico Chemiluminescent Substrate (Thermo Fisher Scientific): mouse-anti JIP3 (F-6) or biotinylated goat-anti mouse SLPI (R&D Systems), and HRP-conjugated goat IgG antibodies (Santa Cruz Biotechnology) or streptavidin (Biolegend). Digital images were obtained using an ImageQuant LAS4000 mini instrument (GE Healthcare). Densitometry was performed on scanned blots using the ImageQuant TL software program (GE Healthcare).

The protocol by which IgE-mediated chronic allergic inflammation (IgE-CAI) was previously described (27). DX5⁺ cells containing basophils were fractionated with ant-DX5 magnetic beads from B6 and Slpi^−/− mice. DX5⁺ BM cells were transferred to Fcer1g^−/− mice 4 days after 5-fluorouracil (5-FU) (Sigma-Aldrich) marrow suppression. Two days after the adaptive cell transfer, the recipient mice were intravenously injected with 2 µg of DNP-IgE (SPE-7; Sigma-Aldrich). Antigen challenge was performed in the right ear by applying 0.6% DNFB acetone-olive oil solution (20 µl). Simultaneously, an equal amount of acetone-olive oil solution was administered to the left ear. The ear thickness was measured using a dial thickness gage (Mitsutoyo). For the histological analysis, both ears were removed from euthanized mice on Day 6 after the antigen challenge and stained with HE.

Mouse IgE anti-TNP mAbs (C38-2) (BD Bioscience) were administered intravenously through the tail vein (volume, 100 µl/mouse). Mice were injected intravenously with 1.0 mg of TNP-OVA in saline 24 h after the injection of IgE (150 µg/mouse). Changes in the core body temperature associated with systemic anaphylaxis were monitored by measuring changes in the rectal temperature using a rectal probe coupled to a digital thermometer (Natsume Seisakusyo).

A plasmid containing full-length mouse SLPI cDNA was purchased from DNAFORM. The open reading frame corresponding to SLPI with a fused 6× His-Tag at the C-terminal was amplified from a plasmid and cloned into pcDNA3.1 (Thermo Fisher Scientific). The primers were as follows: mouse SLPI forward, 5′-CCCCCGAATTCGAGAGCTCC-3′, and reverse, 5′-CACCGAGCATCTA GACTAGTGGTGATGGTGATGGTGATGATGACGACCTTCGATCATCGGGGGCA-3′. Plasmid transfection was performed using ScreenFect A (Wako), in accordance with the manufacturer’s instructions. Briefly, cells were seeded at 2 × 10⁶ cells/ml and transfected of Slpi plasmid DNA (0.3 µg) with ScreenFect A. The mRNA of Slpi and Mmp9 were analyzed 1 day after transfection.

B6 and Slpi^−/− BMEOs were stimulated with/without LPS (1 µg/ml) for 3 h. Total RNA was purified with an miRNeasy Mini kit (Qiagen) in accordance with the manufacturer’s instructions. The purified RNA was converted to sense-strand cDNA using an Ambion WT Expression Kit (Thermo Fisher Scientific) and then labeled using an Affymetrix GeneChip WT Terminal Labeling and Controls Kit. Labeled cDNAs were hybridized onto the Affymetrix GeneChip Mouse Gene 1.0 ST Array using an Affymetrix GeneChip Hybridization, Wash, and Stain Kit, according to the manufacture’s protocols. The signals were quantified using an Affymetrix CeneChip Scanner 3000, and raw data were obtained using the Affymetrix GeneChip Command Console Software program (version 1.2.1.20). The data were normalized using the Robust Multichip Average algorithm in the “affy” package (32) of the Bioconductor project software program,¹ after which transcript signals were calculated by log2-transformation of the normalized data. Further analyses were performed using the R software program.² The data were archived in the NCBI Gene Expression Omnibus (accession number, GSE87638).³

The model of HDM (Greer Laboratories)-induced asthma was developed as reported previously, with slight modification (33). Briefly, mice were anesthetized with isoflurane and intranasally sensitized with 1 µg of HDM on Day 0. Seven days later, they were challenged with 1 µg of HDM for 5 consecutive days. The mice were sacrificed, and their organs were dissected for a histological analysis on Day 14. BALF cells were obtained by an intratracheal injection of EDTA-containing PBS.

Eosinophilic lung inflammation was established with an intranasal challenge of chitin. Chitin (Sigma Aldrich) was prepared as previously described (34). Congenic B6-CD45.1 mice were intravenously transferred with eosinophils from B6 or Slpi^−/− mice (CD45.2) in accordance with the methods of a previous study (28). One hour after the eosinophil transfer, chitin (10⁵ beads) was intranasally administered to the recipient mice. BALF cells were obtained 24 h after the antigen challenge. The donor cells were distinguished from the recipient cells by anti-CD45.1/CD45.2 antibody staining.

The statistical significance of differences was determined using a paired Student’s t-test. P values of <0.05 were considered to indicate statistical significance.

We first examined the expression of Slpi transcripts and proteins in mast cells, basophils, and eosinophils derived from BM cells. As shown in Figure 1A, murine SLPI-encoding Slpi transcripts were found to be abundant in BMBs. BMEos expressed a similar level of SLPI mRNA to BM cells, but SLPI mRNA was barely detected in mast cells. In addition, Slpi transcripts were detected in basophils and eosinophils sorted from spleen cells. SLPI mRNA was also observed in the peritoneal fluid eosinophils (Figure 1B), suggesting that SLPI is expressed in residential basophils and eosinophils.

Because these allergic effector cells are induced from BM cells by their specific growth factors for the lineage commitment, we next examined the implication of SLPI in cellular differentiation using Slpi^−/− mice (20). As shown in Figures S2A, S3A, and S4A in Supplementary Material, the proliferation curves of B6 and Slpi^−/− mice were comparable during the terminal differentiation of BM cells into mast cells, basophils, and eosinophils. There were no significant differences in the population of terminally differentiated cells between B6 and Slpi^−/− mice (Figures S2B, S3B, and S4B in Supplementary Material), suggesting that SLPI does not affect the growth of these effector cells. As shown Figure 1C, immunoblotting also demonstrated that BMBs express a large amount of SLPI protein and that the SLPI expression of eosinophils is similar to that of BM cells; however, no SLPI expression was observed in mast cells, suggesting that SLPI is expressed in basophils and eosinophils, but not in mast cells. We found that the degranulation and cytokine production in response to IgE or LPS did not differ between B6 and Slpi^−/− mast cells (Figures S2C,D in Supplementary Material). Moreover, the absence of SLPI did not affect mast cell-dependent IgE-mediated systemic anaphylaxis (Figure S2E in Supplementary Material), showing that SLPI is dispensable for the activation of mast cells.

We thus focused on the morphological analyses of BMBs and BMEos. As shown in Figure 1D, immunostaining revealed that basophils express SLPI. There were no marked differences in the Diff-Quick, or alcian blue staining patterns between B6 and Slpi^−/− BMBs. TEM demonstrated that Slpi^−/− BMBs resemble B6 BMBs morphologically, with both sharing lobulated nuclei and granules. Both B6 and Slpi^−/− BMBs showed similar expression levels of FcεRI, IL-3Rα, and ST2 (IL-33 receptor) (Figure S3C in Supplementary Material). The amount of mast cell protease (MCP)-8 and 11—granule serine proteases that are known as specific basophil markers—and the amount of lysosomal enzyme β-hexosaminidase (HEX) did not differ between B6 and Slpi^−/− BMBs (Figures S3D,E in Supplementary Material). We next explored the morphology, cell surface receptors, and granule contents in BMEos. As shown in Figure 1E, BMEos expressed SLPI at low levels, results that were consistent with the mRNA and immunoblotting data. The Diff-Quick and TEM images of B6 and Slpi^−/− BMEos were comparable. Both B6 and Slpi^−/− BMEos displayed similar expression levels of TLR4, Siglec-F, C-C chemokine receptor (CCR) 3, CD11b, and ST2 (Figure S4C in Supplementary Material). The granular enzyme, EPO, was also detected in Slpi^−/− BMEos at almost the same level as in B6 BMEos (Figure S4D in Supplementary Material). Collectively, these data showed that the basophils and eosinophils of mice express SLPI and suggested that SLPI deficiency did not affect the proliferation or the morphology of BMBs or BMEos.

We next examined the cytokine production and serine protease activity in B6 and Slpi^−/− BMBs upon IgE stimulation. As shown in Figure 2A, Slpi^−/− BMBs produced more IL-4, 6, and 13 than B6 BMBs. Slpi^−/− BMBs also showed higher tryptase activity than B6 BMBs (Figure 2B). In contrast, there was no significant difference in the chymase activity of B6 and Slpi^−/− BMBs. As shown in Figure 2C, the release of β-HEX upon stimulation with IgE or compound 48/80 (an IgE-independent degranulator) did not differ between B6 and Slpi^−/− basophils, suggesting that SLPI does not affect basophil degranulation. Furthermore, B6 and Slpi^−/− BMBs secreted comparable levels of IgE-induced chemical mediators, histamine and cysteinyl leukotrienes (CysLT) (Figure 2D). We further investigated FcεR downstream signaling in B6 and Slpi^−/− BMBs. As shown in Figure 2E (left panel), the phosphorylation of phospholipase (PLC)-γ2 and extracellular signal-regulated kinase (Erk) 1/2 were equivalently increased in B6 and Slpi^−/− BMBs. We did not detect any elevation of NF-κB p65 phosphorylation or degradation of IκB α/β in B6 or Slpi^−/− BMBs (Figure 2E, right panel). Electrophoretic mobility shift assays (EMSAs) showed that the DNA binding activity of NF-κB was not altered upon stimulation (data not shown), suggesting that SLPI represses pathways other than NF-κB in basophils. Collectively, these results showed that SLPI inhibits IgE-mediated cytokine production and tryptase activity in basophils.

The use of the DX5⁺ BM cell adaptive transfer system showed that basophils are indispensable for the development of the IgE-mediated delayed-onset cutaneous anaphylaxis reaction (IgE-CAI) in Fcer1g^−/− mice (27). Only basophils express FcεRI in the fraction of DX5⁺ BM cells (27). Thus, to investigate whether SLPI regulates basophil activation in vivo, we induced IgE-CAI in 5-FU-treated Fcer1g^−/− mice reconstituted with DX5⁺ BM cells, including basophils, selected from B6 or Slpi^−/− BM cells (Figure 3A). We confirmed that the population of FcεRI⁺ cells did not differ between the B6 and Slpi^−/− DX5⁺ BM cells (data not shown). As shown in Figure 3B, ear swelling in mice transferred with Slpi^−/− basophils was significantly increased in comparison to mice transferred with B6 basophils. A histopathological examination at six days after antigen challenge also demonstrated that recipients transferred with Slpi^−/− basophils showed augmented inflammatory cellular infiltration (Figure 3C). These findings demonstrated that basophil SLPI regulates the IgE-mediated allergic inflammatory responses.

We next investigated the cytokine production and protease activities of B6 and Slpi^−/− BMEos. IL-33, unlike LPS, is known to be a strong inducer of the Th2 cytokines IL-4 and 13 (35). We thus measured the IL-4, 6, and 13 levels after LPS or IL-33 treatment. As shown in Figure 4A, Slpi^−/− BMEos produced more IL-6 than B6 BMEos upon LPS stimulation, whereas B6 and Slpi^−/− BMEos did not secrete IL-4 or IL-13 (Figure 4B). When stimulated with IL-33, B6 and Slpi^−/− BMEos produced any or all of IL-4, 6, and 13, but there were no differences in the amounts of cytokines between B6 and Slpi^−/− BMEos (Figure 4B). In addition, the enzymatic activities of tryptase, chymase, and EPO were not increased in B6 or Slpi^−/− BMEos upon LPS or IL-33 stimulation (Figures 4C,D), showing that SLPI deficiency does not affect the serine protease activities or the degranulation responses in eosinophils, and that SLPI inhibits the IL-6 secretion induced by LPS but not that induced by IL-33.

Upon stimulation, eosinophils rapidly migrate to the inflammatory sites, and across the epithelia into tissue (1, 7–9). We therefore conducted a chemotaxis assay after stimulation with eosinophil chemotactic factors LTB4, CCL2, and CCL11. As shown in Figure 4E, the chemotactic activities in response to these chemoattractants were comparable between B6 and Slpi^−/− BMEos, indicating that SLPI does not affect the cellular migration induced by chemokines alone. We therefore performed a Matrigel invasion assay after costimulation with LPS and CCL11. As shown in Figure 4F, the invasion activity in Slpi^−/− BMEos was increased in comparison to that in B6 eosinophils. Collectively, these results suggest that SLPI regulates the LPS-mediated IL-6 production and invasion activity in eosinophils.

We investigated the gene alteration in B6 and Slpi^−/− BMEos before and after administration with LPS using DNA microarray analyses. Surprisingly, the expression of Mmp9 in Slpi^−/− BMEos was markedly higher than that in B6 BMEos (Figure 5A). A quantitative RT-PCR confirmed that the expression of Mmp9 transcripts in Slpi^−/−BMEos was significantly higher than that in B6 BMEos before and after LPS stimulation (Figure 5B). To clarify whether the genetic disruption of SLPI affected the Mmp9 gene in eosinophils, we introduced an Slpi plasmid into Slpi^−/− BMEos. As shown in Figure 5C, the Mmp9 transcripts in the Slpi plasmid-transfected Slpi^−/− BMEos were decreased to half the level observed in the mock-transfected Slpi^−/− BMEo. Furthermore, the absence of SLPI did not alter the expression of MMP-9 protein in BMBs or BM cells (Figure 5D), suggesting that the MMP-9 expression is not impaired by the SLPI gene disruption itself, and that SLPI transcriptionally represses MMP-9 in eosinophils. On the other hand, immunoblotting showed that the expression of MMP-9 proteins was remarkably increased in Slpi^−/− BMEos at the steady state; however, the MMP-9 expression in Slpi^−/− BMEos was not enhanced after LPS treatment (Figure 5E). In addition, MMP-9 was not increased in B6 or Slpi^−/− BMEos after IL-5 stimulation (Figure 5E), inferring that IL-5 does not affect the MMP-9 expression in terminally differentiated eosinophils. It was shown that MMP-9 is critical for the migration of eosinophils though the basement membrane components (36). Since MMP-9 is increased in Slpi^−/− BMEos in the steady state, our data suggested that SLPI represses excessive eosinophil migration in part by MMP-9 expression.

The expression of the Mmp9 gene is induced upon the phosphorylation of several transcription factors, including NF-κB and Elk-1 (37, 38). Because SLPI is shown to regulate NF-κB activation in macrophages and neutrophils upon LPS stimulation (17, 20, 21), we investigated the TLR4-downstream signaling in eosinophils (Figure 6A). In contrast to the previous results, the degradation of IκBα/β was not clearly observed in B6 or Slpi^−/− BMEos though the degradation of IκBα was slightly but not significantly increased in Slpi^−/− BMEos in 60 min after stimulation (Figure 6A, lanes 1 and 2). The phosphorylation of NF-κB p65 in B6 and Slpi^−/− BMEos was comparable (Figure 6A, lanes 3 and 4). In addition, we were unable to detect the activation of NF-κB or CCAAT enhancer binding proteins (C/EBPs) in an EMSA (data not shown).

Lipopolysaccharide also activates MAP kinase signaling, Erk1/2 and JNK1. Both pErk1/2 and pJNK1 phosphorylate the transcriptional factor Elk-1, which has multiple serine (Ser) and threonine (Thr) phosphorylation sites (39, 40). Although pSer384 Elk-1 mainly contributes to the transcriptional activation of Elk-1 (39–41), pThr418 Elk-1 counteracts the transcriptional activation of Elk-1 itself (41). We therefore examined the phosphorylation of Erk1/2, JNK1, and Elk-1 (murine Ser383 and Thr417) upon LPS stimulation (Figure 6A). While Erk1/2 phosphorylation was comparably increased in B6 and Slpi^−/− BMEos, JNK1 was significantly phosphorylated in Slpi^−/− BMEos without stimulation (Figure 6A, lanes 7 and 8 and Figure 6B, left panel). Surprisingly, pSer383 Elk-1 was increased in Slpi^−/− BMEos, whereas Elk-1 (Ser383) was barely phosphorylated in B6 BMEos, even after LPS stimulation (Figure 6A, lanes 9 and 11 and Figure 6B, right panel). Conversely, Thr418 Elk-1 was equivalently phosphorylated in both B6 and Slpi^−/− BMEos (Figure 6A, lanes 10 and 11), inferring that SLPI represses pSer383 Elk-1 via association with JNK1 and/or Elk-1. We therefore performed an immunoprecipitation assay using anti-SLPI and JNK1 or Elk-1 antibodies; however, we were unable to detect any interaction between these molecules (data not shown). It was reported that the JIP family proteins function as specific scaffold proteins for JNK signaling (24, 25). Although our microarray data showed that B6 and Slpi^−/− BMEos express JIP1 and 3 genes (GSE87638), we found that JIP3, but not JIP1, is present in both eosinophils at the protein level (data not shown). JIP3 is shown to bind to the cytoplasmic domains of TLR4 with or without LPS stimulation and regulates JNK1 signaling (42). In addition, JIP3 promotes the retrograde transportation of JNK1 and lysosome (43). Because SLPI is secreted in response to various stimuli (13, 15) (Figure 5E), we examined the interaction between JIP3 and SLPI at the steady state. An immunoprecipitation assay using mouse anti-JIP3 antibody showed that JIP3 associates with SLPI (Figure 6C). These results suggested that SLPI is associated with the JIP3 scaffold protein and represses Elk-1 activation in eosinophils.

To investigate whether or not SLPI contributes to eosinophil-mediated inflammatory responses in vivo, we examined HDM (house dust mite)-induced airway inflammation because TLR4 has an essential role in the HDM model (44). Mice were intranasally sensitized with small amounts of HDM (1 µg) on Day 0 and subsequently challenged on Days 7–11. After HDM, the frequency and numbers of eosinophils in the bronchoalveolar lavage fluid (BALF) cells of Slpi^−/− mice were significantly higher exposure than those in B6 mice (Figures 7A,B). A histopathological examination of a lung specimen showed that the cellular infiltration in Slpi^−/− mice was markedly greater than that in B6 mice (Figure 7C). In addition, as shown in Figure S5 in Supplementary Material, the expression of MMP-9 in the BALF were augmented in Slpi^−/− mice in comparison to B6 mice. The results were also consistent with the eosinophil numbers in the BALF.

Fungal chitin was found to induce acute eosinophilic allergic inflammation in a Myd88-dependent manner (34). We therefore used the adaptive transfer of eosinophils to examine chitin-induced airway inflammation (28). BMEos (CD45.2) were adoptively transferred in CD45.1 congenic B6 mice. Simultaneously, chitin was intranasally administered to the transferred mice. One day after the antigen challenge, we evaluated the fraction of donor (CD45.2⁺) and recipient (CD45.1⁺) Siglec-F⁺ eosinophils among BALF cells. As shown in Figures 7D,E, the numbers of the recipient eosinophils were comparable between the transferred mice, whereas the population of CD45.2⁺ donor eosinophils in mice transferred with Slpi^−/− BMEos was significantly higher than that in those that received B6 eosinophils. These results showed that SLPI regulates eosinophil-mediated airway inflammation.