The helminth product, ES-62, protects against airway inflammation by resetting the Th cell phenotype

Abstract

We previously demonstrated inhibition of ovalbumin-induced allergic airway hyper-responsiveness in the mouse using ES-62, a phosphorylcholine-containing glycoprotein secreted by the filarial nematode, Acanthocheilonema viteae. This inhibition correlated with ES-62-induced mast cell desensitisation, although the degree to which this reflected direct targeting of mast cells remained unclear as suppression of the Th2 phenotype of the inflammatory response, as measured by eosinophilia and IL-4 levels in the lungs, was also observed. We now show that inhibition of the lung Th2 phenotype is reflected in ex vivo analyses of draining lymph node recall cultures and accompanied by a decrease in the serum levels of total and ovalbumin-specific IgE. Moreover, ES-62 also suppresses the lung infiltration by neutrophils that is associated with severe asthma and is generally refractory to conventional anti-inflammatory therapies, including steroids. Protection against Th2-associated airway inflammation does not reflect induction of regulatory T cell responses (there is no increased IL-10 or Foxp3 expression) but rather a switch in polarisation towards increased Tbet expression and IFNγ production. This ES-62-driven switch in the Th1/Th2 balance is accompanied by decreased IL-17 responses, a finding in line with reports that IFNγ and IL-17 are counter-regulatory. Consistent with ES-62 mediating its effects via IFNγ-mediated suppression of pathogenic Th2/Th17 responses, we found that neutralising anti-IFNγ antibodies blocked protection against airway inflammation in terms of pro-inflammatory cell infiltration, particularly by neutrophils, and lung pathology. Collectively, these studies indicate that ES-62, or more likely small molecule analogues, could have therapeutic potential in asthma, in particular for those subtypes of patients (e.g. smokers, steroid-resistant) who are refractory to current treatments.

Graphical abstract

Fig. 1

The phosphorylcholine-containing glycoprotein, ES-62, inhibits eosinophil, neutrophil and lymphocyte infiltration of the lungs in the murine ovalbumin (OVA)-induced airway inflammation model. Differential eosinophil (A), neutrophil (B), lymphocyte (C), macrophage (D) and total (E) bronchoalveolar lavage fluid (BALF) cell counts in individual mice from each of the four treatment groups where, for A–D, n = 32 for PBS; n = 22 for ES-62; n = 51 for OVA and n = 38 for OVA + ES-62 groups and for E, n = 42 for PBS; n = 24 for ES-62; n = 59 for OVA and n = 44 for OVA + ES-62 groups. The bar represents the median value of the group and ^∗P ⩽ 0.05 or ^∗∗P ⩽ 0.01 for OVA compared with OVA + ES-62 groups.

Fig. 2

In vivo treatment with the phosphorylcholine-containing glycoprotein, ES-62, inhibits draining lymph node (DLN) cell antigen-specific Th2 cytokine production but not antigen-specific proliferation ex vivo. Lung DLN (thoracic) cells from individual mice within each in vivo treatment group were pooled and cultured with medium alone or ovalbumin (OVA) (500 μg/ml) for 72 h. Cell proliferation was measured by [³H] thymidine uptake and data (A) are expressed as mean ± S.D. (n = 3 replicate cultures) and are from one experiment representative of two. Culture supernatants IL-4 (B), IL-5 (C), IL-13 (D), IL-10 (E) and IFNγ (F) were measured for each group and data are expressed as mean concentrations ± S.D., n = 3 replicate cultures, where ^∗∗∗P ⩽ 0.001, apart from IL-4 and IFNγ which were measured as duplicate samples by Luminex, and represent single data sets representative of at least two independent experiments apart from IL-13 which was only measured in a single experiment to corroborate the decrease in the very low levels of IL-4 observed in that model. (G) The mean values of triplicate serum (obtained day 28) values from individual mice (n = 6) are shown where the bar represents the mean of the group and the OVA, but not OVA + ES-62, group shows significantly elevated levels of IL-5 (^∗P ⩽ 0.05) relative to the PBS and ES-62 control groups.

Fig. 3

Exposure to the phosphorylcholine-containing glycoprotein, ES-62, in vivo does not modulate ovalbumin (OVA)-induced elevation of IgG1 and IgG2a but inhibits IgE and suppresses Th2 cells. Serum samples from each mouse were obtained on day 28. Samples were analysed by ELISA for OVA-specific IgE (A) and total IgE (B), at 1/80 and 1/200 dilutions, respectively, and OVA-specific IgG1 (C) and IgG2a (D) over the full titration range where data for each group are presented as the mean ± S.E.M. (n = 6 mice/group) and ^∗∗∗P ⩽ 0.001, OVA + ES-62 compared with the OVA group. The data are from one experiment representative of two. In E, draining lymph nodes (DLN) cells from each of the four treatment groups were stained for expression of IgE. (F) DLN sections from OVA and OVA + ES-62 groups were stained for B220⁺ B cells, CD3⁺ T cells and for the indicated transcription factor (either GATA3, Foxp3 or Tbet) and the transcription factor positive cells were quantified within the T cell regions. Exemplar tissue maps from a single experiment of DLNs from OVA and OVA + ES-62 groups displaying the relevant distribution of such transcription factor positive cells (red) within the T cell paracortical region (white) surrounded by B220⁺ B cells (blue) and their quantitation of numbers of the relevant transcription factor positive T cells by laser scanning cytometry (LSC) are shown, together with an exemplar relocated image of a portion of the section from the OVA group where Foxp3 is stained in red and CD3, green (F). In an independent experiment, the proportions of Foxp3-expressing unstimulated CD4⁺ T cells from the DLN of pooled mice from the four treatment groups are shown, with the percentage of cells in the Foxp3⁺CD4⁺ quadrant annotated to the right of the plots (G). Pooled DLN cells from the four treatment groups were cultured with medium (I) or phorbol 12-myristate 13-acetate (PMA) plus ionomycin (IONO) (H, J) in the presence of Brefeldin A before staining for intracellular expression of IL-10 (H) or IFNγ (I, J) prior to FACS analysis: cytokine expression (y-axis) was plotted against side scatter (x-axis) and the percentage of DLN cells in the cytokine positive gate is shown and annotated to the right of the plots. (J) The Mean Fluorescence Intensity (MFI) values for IFNγ production for the four treatment groups are PBS: 24,701; ES-62: 21,915; OVA: 23,439 and OVA + ES-62: 27,596. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Exposure to the phosphorylcholine-containing glycoprotein, ES-62, in vivo increases CD4 and CD8 IFNγ responses during murine ovalbumin (OVA)-induced airway inflammation. (A) The proportion of draining lymph node (DLN) cells expressing IFNγ following stimulation with phorbol 12-myristate 13-acetate (PMA) plus ionomycin in the indicated treatment groups by individual mice (PBS, n = 6; OVA + IgG, n = 9; OVA + ES-62 + IgG, n = 10; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 9; pooled from two independent experiments). The bar represents the mean of the group and the OVA + ES-62 + IgG, but not OVA + IgG, group is significantly different from the PBS group (grey ^∗∗P ⩽ 0.01). In an independent experiment, the cytokine expression (y-axis) by cells from pooled DLN from the indicated groups was analysed and the proportions of IFNγ-producing cells in the lymphocyte gate of DLN cells from mice from the OVA and OVA + ES-62 groups that were CD4⁺ (B, C) or CD8⁺ (D, E) following stimulation with medium (B, D) or PMA plus ionomycin (C, E) are shown. The proportions (F) and numbers (G) of PMA plus ionomycin-stimulated CD4⁻ cells expressing IFNγ in the DLN of individual mice from the indicated groups (PBS, n = 6; OVA + IgG, n = 9; OVA + ES-62 + IgG, n = 10; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 9; pooled from two independent experiments) are shown where the bar represents the mean of the group and the OVA + ES-62 + IgG group demonstrates significantly elevated levels of IFNγ production relative to the PBS control group (where ^∗P ⩽ 0.05). (H) The proportions of PMA plus ionomycin-stimulated CD4⁺ cells expressing IFNγ in the DLN of individual mice from the indicated groups (PBS, n = 6; OVA + IgG, n = 9; OVA + ES-62 + IgG, n = 10; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 9; pooled from two independent experiments) are shown where the bar represents the mean of the group and no significant differences are detected.

Fig. 5

Neutralising anti-IFNγ antibodies prevent ES-62-mediated protection against cellular infiltration in murine ovalbumin (OVA)-induced airway inflammation. Differential neutrophil (A), macrophage (B), lymphocyte (C) and eosinophil (D) bronchoalveolar lavage fluid (BALF) cell counts for individual mice in each of the indicated treatment groups (PBS, n = 8; OVA + IgG, n = 10; OVA + ES-62 + IgG, n = 8; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 10; pooled from two independent experiments) are shown where the bar represents the median of the group and black ^∗P ⩽ 0.05 and ^∗∗P ⩽ 0.01 are for OVA + ES-62 + IgG compared with OVA + ES-62 + anti-IFNγ groups whilst grey ^∗∗P ⩽ 0.01 and ^∗∗∗P ⩽ 0.001 are for PBS compared with the indicated treatment group. (E) Representative H & E sections from the OVA + ES-62 + IgG (n = 3 mice) and the OVA + ES-62 + anti-IFNγ (n = 3 mice) groups. Images were captured using an Olympus BX41 microscope mounted with an Olympus U-CMAD3 camera and Cell-B2.0 software. Images are at 10× magnification and include automatic yellow scale bar (200 μm) at the bottom right corner of image.

Fig. 6

Neutralising anti-IFNγ antibodies modulate IL-4 responses in ES-62-treated mice undergoing ovalbumin (OVA)-induced airway inflammation. Draining lymph node (DLN) cells from the four indicated treatment groups were cultured with phorbol 12-myristate 13-acetate (PMA) plus ionomycin in the presence of Brefeldin A before staining for intracellular expression of IL-4 prior to FACS analysis: cytokine expression (y-axis) was plotted against side scatter (x-axis) and the percentage of DLN cells in the cytokine positive gate shown and annotated to the right of the plot labels (A). Data pooled from two further independent experiments show the numbers (B–F) and Mean Fluorescence Intensity (MFI) (G) of spontaneously- (B, D, G) and PMA plus ionomycin-stimulated (C, E, F) IL-4-producing DLN (B, C), CD4⁻ DLN (D, E) and CD4⁺ DLN (F, G) cells. The bar represents the mean value of the group where black ^∗P ⩽ 0.05 represents the difference between the OVA + ES-62 + IgG compared with the OVA + ES-62 + anti-IFNγ group whilst grey ^∗P ⩽ 0.05, ^∗∗P ⩽ 0.01 and ^∗∗∗P ⩽ 0.001 are for PBS compared with indicated treatment group (PBS, n = 6; OVA + IgG, n = 9; OVA + ES-62 + IgG, n = 10; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 9; pooled from two independent experiments).

Fig. 7

The phosphorylcholine-containing glycoprotein, ES-62, suppresses IL-17 responses in murine ovalbumin (OVA)-induced airway inflammation. Pooled draining lymph node (DLN) cells from the four indicated treatment groups were cultured with medium (A) or phorbol 12-myristate 13-acetate (PMA) plus ionomycin (B) in the presence of Brefeldin A before staining for intracellular expression of IL-17 prior to FACS analysis: cytokine expression (y-axis) was plotted against side scatter (x-axis) and the percentage of DLN in the cytokine positive gate shown and annotated to the right of the plots. (C) RORγt expressing unstimulated CD4⁺ T cells from the DLN of pooled mice from the four treatment groups are shown. The numbers (D) of spontaneously (upper; (PBS, n = 6; OVA + IgG, n = 9; OVA + ES-62 + IgG, n = 10; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 9; pooled from two independent experiments)) and PMA plus ionomycin-stimulated (lower; (PBS, n = 6; OVA + IgG, n = 9; OVA + ES-62 + IgG, n = 9; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 9; pooled from two independent experiments)) IL-17-producing DLN cells and the numbers of IL-17-producing CD4⁻ DLN cells (E) and proportions of IL-17-producing CD4⁺ cells (F) from PMA plus ionomycin-stimulated DLN cells from individual mice (PBS, n = 6; OVA + IgG, n = 9; OVA + ES-62 + IgG, n = 9; OVA + anti-IFNγ, n = 10 and OVA + ES-62 + anti-IFNγ, n = 9; pooled from two independent experiments) are presented; the bar represents the mean value of the group and grey ^∗P ⩽ 0.05 shows the difference between the PBS and indicated treatment group. In an independent experiment (G), the levels of IL-17 mRNA (against a glyceraldehyde 3-phosphate dehydrogenase (GAPDH) reference) in the pooled DLN cells from the four indicated treatment groups are shown, normalised to the levels found in the control PBS group. (H) The levels of IL-17 detected by Luminex following stimulation of pooled DLN cells from mice in each group with medium or OVA (500 μg/ml) in a further independent experiment are shown and where ^∗∗∗P ⩽ 0.001 for triplicate samples. In (I), pooled bone marrow-derived dendritic cells (bmDCs) from OVA and OVA + ES-62 mice were pulsed with OVA peptide on day 7 and cocultured with DO11.10 Tg CD4⁺CD62L⁺ T cells for 72 h before analysing the levels of IL-17 production in culture supernatants by ELISA. The data are presented as the mean ± S.D. where ^∗∗P ⩽ 0.01, OVA compared with OVA + ES-62 at 30 nM of OVA peptide and ^∗∗∗P ⩽ 0.001, OVA compared with OVA + ES-62 at 300 nm of OVA peptide in one experiment. In a further independent experiment, (J), the levels of IL-4, IFNγ and IL-17 mRNA (against a GAPDH reference) in the lungs from the OVA + ES-62 group (grey bars) are shown, normalised to the levels found in the OVA group (black bars).