Novel approach to inhibit asthma-mediated lung inflammation using anti-CD147 intervention

Abstract

Extracellular cyclophilins have been well described as chemotactic factors for various leukocyte subsets. This chemotactic capacity is dependent upon interaction of cyclophilins with the cell surface signaling receptor CD147. Elevated levels of extracellular cyclophilins have been documented in several inflammatory diseases. We propose that extracellular cyclophilins, via interaction with CD147, may contribute to the recruitment of leukocytes from the periphery into tissues during inflammatory responses. In this study, we examined whether extracellular cyclophilin-CD147 interactions might influence leukocyte recruitment in the inflammatory disease allergic asthma. Using a mouse model of asthmatic inflammation, we show that 1) extracellular cyclophilins are elevated in the airways of asthmatic mice; 2) mouse eosinophils and CD4+ T cells express CD147, which is up-regulated on CD4+ T cells upon activation; 3) cyclophilins induce CD147-dependent chemotaxis of activated CD4+ T cells in vitro; 4) in vivo treatment with anti-CD147 mAb significantly reduces (by up to 50%) the accumulation of eosinophils and effector/memory CD4+ T lymphocytes, as well as Ag-specific Th2 cytokine secretion, in lung tissues; and 5) anti-CD147 treatment significantly reduces airway epithelial mucin production and bronchial hyperreactivity to methacholine challenge. These findings provide a novel mechanism whereby asthmatic lung inflammation may be reduced by targeting cyclophilin-CD147 interactions.

FIGURE 1

OVA mice exhibit asthma-like Th2-mediated airway inflammation. Female BALB/c mice were injected i.p. with OVA/alum (OVA) or alum only (control) on day 0. OVA and control mice were then challenged i.n. with OVA on days 7–10. BAL fluid was collected and perfused lungs were harvested on day 12 after sacrifice, or bronchial hyperreactivity was measured. A, Mean (±SE) numbers of neutrophils (NEUTs), lymphocytes (LYMPHs), and eosinophils (EOs) isolated from BAL fluid of naive, control, and OVA mice, with n = 3 mice per group. Leukocyte differentials were obtained from Wright-Giemsa-stained cytospins. B, Lung cells from each group of mice were stained with Alexa Fluor 647 anti-CCR3 or CyChrome anti-CD4 in combination with FITC anti-CD62L and analyzed by FACS. Bar graphs, Mean (±SE) numbers of CCR3⁺ granulocytes (eosinophils) and CD4⁺CD62L⁻ lymphocytes (effector/memory CD4⁺ T cells), with n = 3 mice per group. C, Bronchial hyperreactivity was measured using a Buxco apparatus. Data show mean (±SE) ratio of Penh after escalating doses of inhaled methacholine:PBS, with n = 5 mice per group. ***, p < 0.001; **, p < 0.01. D, Pooled lung cells from control and OVA mice were restimulated with OVA protein and T cell-depleted splenocytes for 4 days in culture to induce cytokine secretion by Ag-specific T cells. Bar graphs, Mean levels of IL-5 and IL-13 in culture supernatants as measured by ELISA, with n = 5 mice per group. <L.D., Below the limit of detection of the assay.

FIGURE 2

CypA and CypB are elevated in BAL fluid from OVA mice. BAL fluid collected from individual mice on day 12 of the OVA regimen was cleared of cells by centrifugation, and equivalent volumes (20 µl) were assayed by Western blot analysis for the presence of CypA and CypB using anti-CypA or anti-CypB Ab followed by HRP-conjugated secondary Ab. A, These panels show the detection of CypA and CypB in naive (untreated) mice (lanes 1–2) and OVA mice (lanes 3–7). Recombinant CypA (1.9 µg) and CypB (3 ng) were used as positive controls (lane 8). B, Densitometric analysis was performed on the protein bands in A. Bar graphs, Mean (±SE) band density of untreated vs OVA groups for CypA and CypB blots. **, p < 0.01; *, p < 0.05. C, Panel shows the detection of CypA in BAL fluid from naive (untreated), alum control, and OVA mice. These data are representative of a minimum of two separate experiments.

FIGURE 3

Mouse eosinophils and CD4⁺ T cells express cell surface CD147. Peripheral blood leukocytes and lung cells from OVA mice were collected on day 12 and stained with anti-CD147 (□) or isotype control (■) mAb followed by FITC-conjugated secondary Ab and then Alexa Fluor 647 anti-CCR3 or CyChrome anti-CD4. Histograms show expression of CD147 gated on CCR3⁺ granulocytes (eosinophils) and CD4⁺ T lymphocytes from peripheral blood (A) and lung tissues (B). These data are representative of a minimum of two separate experiments.

FIGURE 4

Anti-CD147 mAb treatment in vivo reduces numbers of eosinophils and effector/memory CD4⁺ T cells in lung tissues. OVA mice were injected i.p. with isotype control or anti-CD147 mAb (10 µg) on days 6–11 of the OVA regimen. Lung cells were collected on day 12 and stained with Alexa Fluor 647 anti-CCR3 or CyChrome anti-CD4 in combination with FITC anti-CD62L for FACS analysis. Bar graphs, mean (±SE) numbers of CCR3⁺ granulocytes (eosinophils) and CD4⁺CD62L⁻ lymphocytes (effector/memory CD4⁺ T cells) in lung tissues, with n = 10 mice per group. **, p < 0.01; *, p < 0.05.

FIGURE 5

Anti-CD147 mAb treatment in vivo results in decreased Ag-specific Th2 cytokine secretion. OVA mice were injected i.p. with isotype control or anti-CD147 mAb (10 µg) on days 6–11 of the OVA regimen. Pooled lung cells were collected on day 12 and restimulated for 4 days in culture with OVA protein and T cell-depleted splenocytes to induce cytokine secretion by OVA-specific T cells. Bar graphs, mean (±SE) levels of IL-5 and IL-13 in culture supernatants as measured by ELISA, with n = 10 mice per group. An OVA restimulation dose of 20 µg/ml is shown. *, p < 0.05.

FIGURE 6

Anti-CD147 mAb treatment reduces the percentage of PAS⁺ airways and bronchial hyperreactivity. OVA mice were injected i.p. with isotype control or anti-CD147 mAb (10 µg) on days 6–11 of the OVA regimen. Mice were sacrificed on day 12, and lungs were fixed in 10% formalin. Fixed lungs were embedded in paraffin and 5-µm sections were cut and stained with PAS reagent to identify changes in airway epithelial mucin. A, PAS⁺ and PAS⁻ airways present in each stained section were counted blind by two independent investigators. The bar graph indicates the mean (±SE) percentage of airways that were PAS⁺ in isotype- vs anti-CD147-treated mice, with n = 5 mice per group. ***, p < 0.001. B, Bronchial hyperreactivity was measured on day 12 using a Buxco apparatus. Data show the mean (±SE) ratio of Penh after escalating doses of inhaled methacholine:PBS in isotype- vs anti-CD147-treated vs naive (untreated) mice, with n = 10 mice per group. Baseline PenhPBS values were subtracted from each PenhMch value. *, p < 0.05.

FIGURE 7

Mouse eosinophils are not chemotactic to CypA and CypB. A, Chemotaxis of eosinophils purified from peripheral blood of IL-5 Tg (BALB/c) mice to different doses of CypA or CypB was assessed using a Transwell assay. Recombinant eotaxin-1 (270 ng/ml) was used as a positive control. Bar graph, mean (±SD) chemotactic indices for CypA and CypB and eotaxin-1, with n = 3 wells per group. B, ERK activation was assayed by Western blot analysis after stimulation of eosinophils with CypA (400 ng/ml). Eotaxin-1 (270 ng/ml) was used as positive control for ERK signaling in eosinophils. Upper and lower panels, The detection of phosphorylated and total ERK, respectively. C, Eosinophils (from IL-5 Tg (BALB/c) mice) were stained with anti-CD147 (□) or isotype control (■) mAb followed by FITC-conjugated secondary Ab and then Fluor 647 anti-CCR3. Histogram, Expression of CD147 gated on CCR3⁺ granulocytes (eosinophils). These data are representative of a minimum of two separate experiments.

FIGURE 8

Activated mouse CD4⁺ T cells are chemotactic to CypA and CypB. CD4⁺ T cells were purified from naive BALB/c splenocytes and from splenocytes, which had been activated overnight with Con A (1 µg/ml). A and B, Chemotaxis of naive and activated CD4⁺ T cells to RANTES (10 ng/ml), CypA or CypB (400 ng/ml) was assessed using modified Boyden chamber assays. In some wells, anti-CD147 or isotype control mAb was included in the top and bottom chambers. Bar graphs, Mean (±SE) chemotactic indices for each group, with n = 6 wells per group. Based on prior optimization of this assay, a chemotactic index of 1.3 (− − −) was chosen as a cut-off point for a positive chemotactic response. ***, p < 0.001. C, ERK activation was assayed by Western blot analysis after stimulation of naive or activated CD4⁺ T cells with CypA (400 ng/ml). RANTES (1 ng/ml) was used as positive control for ERK signaling in CD4⁺ T cells. Upper and lower panels, The detection of phosphorylated and total ERK, respectively. D, Naive and activated CD4⁺ T cells were stained with anti-CD147 (□) or isotype control (■) mAb followed by FITC-conjugated secondary Ab, and then CyChrome anti-CD4. Histograms show expression of CD147 gated on CD4⁺ T lymphocytes. These data are representative of a minimum of two separate experiments.