Immunoproteasome Inhibition Reduces the T Helper 2 Response in Mouse Models of Allergic Airway Inflammation

Abstract

Background: Allergic asthma is a chronic disease and medical treatment often fails to fully control the disease in the long term, leading to a great need for new therapeutic approaches. Immunoproteasome inhibition impairs T helper cell function and is effective in many (auto-) inflammatory settings but its effect on allergic airway inflammation is unknown.

Methods: Immunoproteasome expression was analyzed in in vitro polarized T helper cell subsets. To study Th2 cells in vivo acute allergic airway inflammation was induced in GATIR (GATA-3-vYFP reporter) mice using ovalbumin and house dust mite extract. Mice were treated with the immunoproteasome inhibitor ONX 0914 or vehicle during the challenge phase and the induction of airway inflammation was analyzed.

Results: In vitro polarized T helper cell subsets (Th1, Th2, Th17, and Treg) express high levels of immunoproteasome subunits. GATIR mice proved to be a useful tool for identification of Th2 cells. Immunoproteasome inhibition reduced the Th2 response in both airway inflammation models. Furthermore, T cell activation and antigen-specific cytokine secretion was impaired and a reduced infiltration of eosinophils and professional antigen-presenting cells into the lung and the bronchoalveolar space was observed in the ovalbumin model.

Conclusion: These results show the importance of the immunoproteasome in Th2 cells and airway inflammation. Our data provides first insight into the potential of using immunoproteasome inhibition to target the aberrant Th2 response, e.g. in allergic airway inflammation.

Keywords: GATA-3 reporter mice; Th2 cells; allergic airway inflammation; eosinophilia; immunoproteasome inhibition.

Figure 1

The immunoproteasome is strongly expressed in T helper cell subsets. Naïve CD4⁺ T cells were magnetically sorted from spleens of GATIR and IL-17A-GFP mice and cultured with polarizing cytokine/antibody cocktails for 7 days. (A) Flow cytometric analysis of polarization on day 7. Intracellular staining for IFN-γ and FoxP3 was performed for Th1 and Tregs, respectively. GATA-3 and IL-17A expression was analyzed by measuring the respective reporter fluorochrome vYFP or GFP. Histogram overlays show fluorescence levels in the different Th populations with naïve CD4⁺ T cells on day 0 as controls (grey). For each parameter, one other population is shown to demonstrate the specificity of the signal (Th1 – black, Th2 – red, Th17 – petrol, Treg – purple). (B) Polarized cells were harvested and lysed on day 7. Lysates were analyzed for LMP7, LMP2, MECL-1, β5c, β2c and β1c by immunoblotting. γ-tubulin was used as loading control. Lanes represent extracts from three different mice. Whole spleen lysates of wildtype (WT) and LMP7-MECL-1-double-deficient mice were used as controls. Note the additional band of LMP2 in the LMP7-MECL-1-double-deficient mice which likely represents the accumulated precursor that is not incorporated into the mature proteasome in the absence of LMP7 and MECL-1.

Figure 2

GATA-3-vYFP expression signal correlates with Th2 cytokine expression. GATIR mice were sensitized with OVA/Alum by two intraperitoneal (i.p.) injections on day 0 and 7. On day 14, 15 and 16 they were challenged with aerosolized OVA for 20 min and analyzed on day 17. (A) Experimental setup. (B) GATA-3-vYFP expression levels in indicated CD4⁺ T cell subsets in the lung compared to bulk CD4⁺ T cells (grey filling) and (C) quantification of the mean fluorescence intensity (MFI). Data is shown as mean ± SD, n=3; statistical differences are not indicated for better visibility.

Figure 3

Immunoproteasome inhibition reduces the Th2 response in OVA-induced airway inflammation. GATIR mice were sensitized with OVA/Alum by two intraperitoneal (i.p.) injections on day 0 and 7. On day 14, 15 and 16 they were challenged with aerosolized OVA for 20 min. Mice received subcutaneous injections of 10 mg/kg ONX 0914 or vehicle on day 12, 14, 15 and 16 and were analyzed on day 17. Naïve mice served as controls. (A) Experimental setup. (B) Lysates of spleen, lung and lymph nodes (LN) were analyzed by immunoblotting against the indicated proteins. The shift of electrophoretic mobility of LMP7 results from covalent modification with ONX 0914. β-actin served as a loading control. (C) Gating strategy for the identification of Th2 cells in the lung as GATA-3^hi and GATA-3^hi co-expressing ST2 after doublet and dead cell exclusion and pre-gating on CD3⁺ CD4⁺. (D) Quantification of the frequencies. (E) Frequency of total CD3⁺ CD4⁺ T cells in the lung. Frequency (F, H) and absolute count (G, I) of IL-4 and IL-13 expressing CD4⁺ T cells in the lung. (J) Cells from spleen and lymph nodes were restimulated with ovalbumin for 4 days in vitro to measure the frequency of antigen-specific IL-4⁺ CD4⁺ cells. Concentrations of anti-OVA IgG1 (K) and IgE (L) in the serum. Data is shown as mean ± SD and statistical significance was determined depending on the data structure. (D, E) Pooled data from 5 independent experiments (naïve: n=7; vehicle/ONX 0914 n=13-14) was analyzed by one-way ANOVA with Holm-Sidak test for multiple comparison. (F-I) Representative result of two independent experiments, n=3, data was analyzed by two-tailed t test. (J) Representative result of two independent experiments, (n=4) data was analyzed by two-way ANOVA with Holm-Sidak test for multiple comparison. (K) Pooled data of three independent experiments (naïve: n=4; vehicle: n=8; ONX 0914: n=9) was analyzed by one-way ANOVA and Holm-Sidak test. (L) Pooled data of three independent experiments (naïve: n=4; vehicle: n=11; ONX 0914: n=10) was analyzed by Kruskal-Wallis test (not normally distributed). *p <0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 4

Infiltration of inflammatory cells is reduced upon ONX 0914 treatment in OVA-induced airway inflammation. GATIR mice were sensitized with OVA/Alum by two intraperitoneal (i.p.) injections on day 0 and 7. On day 14, 15 and 16 they were challenged with aerosolized OVA for 20 min. Mice received subcutaneous injections of 10 mg/kg ONX 0914 or vehicle on day 12, 14, 15 and 16 and were analyzed on d17. Naïve mice served as controls. (A, B) Infiltrating cells in the broncho-alveolar lavage fluid (BALF) were analyzed by flow cytometry. (A) Absolute cell count and (B) relative frequency of the different cell populations identified as eosinophils (CD45⁺ CD11c^- CD11b⁺ Ly6G^- Siglec-F⁺), neutrophils (CD45⁺ CD11c^- CD11b⁺ Ly6G⁺), CD11c⁺ and lymphocytes (CD45⁺ CD11b^- SSC-A^low) (naïve: n=3; vehicle/ONX 0914 n=7). Single cell suspensions from lungs were analyzed by flow cytometry and cells were identified as (C) eosinophils (CD45⁺ CD11c^- CD11b⁺ Ly6G^- Siglec-F⁺) and neutrophils (CD45⁺ CD11c^- CD11b⁺ Ly6G⁺) (naïve: n=3; vehicle/ONX 0914 n=9) as well as (D, E) different CD11c⁺ subpopulations (naïve: n=3; vehicle/ONX 0914 n=7). (F, G) Formalin-fixed sections were stained with CongoRed and the positive area was quantified by color deconvolution. Arrows indicate CongoRed-positive cells, representing eosinophils. Scale bar indicates the distance of 50 µm. Data is shown as a representative of two independent experiments with similar results (n=3). (H, I) Hematoxilin-eosin-staining of formalin-fixed samples, scale bar indicates the distance of 100 µm (naïve: n=3; vehicle/ONX 0914 n=7). All data is shown as mean ± SD and was pooled from 2-3 independent experiments (except G). Significance of differences was analyzed by two-way (A-C) or one-way (D-I) ANOVA with Holm-Sidak test for multiple comparison. *p <0.05, **p < 0.01, and ****p < 0.0001. For better visibility, only the results for the comparison Vehicle vs. ONX 0914 are indicated.

Figure 5

ONX 0914 reduces the Th2 response in house dust mite-induced airway inflammation. (A-H) GATIR mice received intranasal applications of 50 µg house dust mite (HDM) extract on day 0, 7, 14 and 21. One hour before the last immunization, mice were treated with 10 mg/kg ONX 0914 or vehicle subcutaneously. Inflammatory infiltration was analyzed by flow cytometry on day 23. Naïve mice served as controls. (A) Schematic setup of experiments (B-H). Frequency of GATA-3^hi (B), GATA-3^hi ST2⁺ (C) and IL-13⁺ T cells (D) in the lung. Frequencies of (E) CD11c⁺ MHC-II^hi cells as well as (F) granulocytes in the lung. (G) Absolute cell counts and (H) relative frequencies of cell populations in the BALF, identified as eosinophils (CD45⁺ CD11c^- CD11b⁺ Ly6G^- Siglec-F^-), neutrophils (CD45⁺ CD11c^- CD11b⁺ Ly6G⁺; left), CD11c⁺ and lymphocytes (CD45⁺ CD11b^- SSC-A^low). (I-K) Mice of day 23 (d23 (Vehicle)) were immunized and treated as described above. For analysis of the inflammation on day 21 (d21), mice received only three immunizations on day 0, 7 and 14 and were analyzed without treatment and the last immunization. (I) Experimental setup for experiments (J, K). (J) Infiltration of myeloid cells in the BALF and (K) frequency of GATA-3^hi ST2⁺ Th2 cells in the lung on day 21 and 23. Data is shown as mean ± SD (naïve: n=3, vehicle/ONX 0914: n=6-7, d21: n=5). *p <0.05 and **p < 0.01. For better visibility, only the results for the comparison vehicle vs. ONX 0914 are indicated. Groups d21 and d23 were not significantly changed.

Figure 6

ONX 0914 treatment reduces T cell activation but does not affect the frequency of proliferating cells. GATIR mice were sensitized with OVA/Alum by two intraperitoneal (i.p.) injections on day 0 and 7. On day 14, 15 and 16 they were challenged with aerosolized OVA for 20 min. Mice received subcutaneous injections of 10 mg/kg ONX 0914 or vehicle on day 12, 14, 15 and 16 and were analyzed on day 17. (A) Experimental setup. CD44 was analyzed on CD4⁺ T cells in the spleen. Gating (B) and frequency (C) of CD44⁺ populations of CD4⁺ T cells in the spleen as well as mean fluorescence intensity (MFI) of CD44⁺ on CD4⁺ T cells (right) (n=4). (D) Gating of Ki-67 positive populations among all CD4⁺ T cells (left) and CD4⁺ GATA-3^hi cells (right) in the lung. (E) Quantification of the respective frequency (n=3). Data is shown as mean ± SD as a representative from 2 independent experiments. Significance of differences was analyzed by two-tailed t test (C) and two-way ANOVA (E). *p < 0.05.