To investigate whether GAG can exert immunomodulatory effects in humans, we tested whether GAG induces the production of pro- and/or anti-inflammatory cytokines in human PBMCs. GAG did not induce the proinflammatory cytokines TNFα, IL-6, IL-8, IFN-γ, IL-17, IL-5 and IL-9 (Figure 1A), neither did it induce the anti-inflammatory cytokine IL-10 (Figure 1A). To determine whether GAG modulates Aspergillus-induced innate monocyte-derived cytokines, PBMCs were stimulated for 24 hours with Aspergillus conidia (these morphotypes of A. fumigatus were selected because they do not contain GAG that would interfere with the study of the immunological function of GAG) in the presence or absence of GAG. The presence of GAG did not have a significant effect on the production of the innate cytokines TNFα and IL-6, or the anti-inflammatory cytokine IL-10 (Figure 1B). However, when the production of the characteristic T-helper cytokines IL-17, IL-22 and IFN-γ induced by A. fumigatus was investigated, the IL-17 and IL-22 responses were significantly reduced in the presence of GAG (Figure 1C). To determine whether the effects of GAG are specific for Aspergillus-driven T-helper (Th) responses, or whether GAG has a general ability to modulate human Th responses, the effects of GAG on cytokine-driven Th responses were studied. GAG significantly decreased the proinflammatory Th cytokine production induced by the cytokine combinations IL-1β/IL-23 and IL-12/IL-18 that induce IL-17/IL-22 and IFN-γ, respectively (Figure 1E). Thus, GAG can inhibit human proinflammatory Th cytokine production induced by Aspergillus and cytokine cocktails.

Human Th cytokine responses such as IL-17 and IL-22 production are highly dependent on the IL-1 receptor pathway [11], [12]. To investigate whether the observed modulation of Th cytokines by GAG was due to an interaction of GAG with the IL-1 pathway, we determined the capacity of GAG conditioned medium (culture supernatants of PBMCs that were exposed to 10 µg/ml GAG for 24 hours) to reduce IL-1β bioactivity. Indeed it was shown that GAG significantly reduced the bioactivity of IL-1β while culture supernatants of unstimulated PBMCs did not (Figure 2A). Bioactivity of the IL-1 signalling pathway is dependent on IL-1 receptor agonists (IL-1α and IL-1β) and IL-1 receptor antagonists [13]. One of the natural inhibitors of the IL-1 signalling is the interleukin-1 receptor antagonist (IL-1Ra); therefore the ability of GAG to induce IL-1Ra was tested. IL-1Ra concentrations in the supernatant of the cells stimulated with GAG were significantly increased, whereas GAG induced none of the IL-1 receptor agonists, IL-1α or IL-1β (Figure 2B), showing that GAG has the capacity to modulate immune responses by blocking the IL-1 receptor pathway.

To demonstrate that IL-17, IL-22, and IFN-γ production by human PBMCs is indeed dependent on the IL-1 receptor pathway and that IL-1Ra can inhibit the production of these Th cytokines, Th1 and Th17 inducing stimuli were studied in the presence or absence of IL-1Ra. Addition of IL-1Ra reduced IL-17, IL-22, and IFN-γ induction by both Aspergillus conidia and by IL-1/IL-23 and IL-12/IL-18 cytokine combinations (Figure 3A). To determine whether the immunosuppressive effect of GAG was mediated through the induction of IL-1Ra, PBMCs were stimulated with IL-1β/IL-23 and GAG in the presence or absence of neutralizing anti-IL-1Ra antibodies. GAG reduced IL-17 and IL-22 levels significantly, which was not observed in the presence of neutralizing anti-IL-1Ra antibodies, demonstrating that the inhibitory effects of GAG on Th cytokine production are dependent on IL-1Ra (Figure 3B).

The in vitro stimulations described above suggest that the immunomodulatory effects of GAG are due to inhibition of IL-1 bioactivity by inducing IL-1Ra. To assess whether this has relevant consequences in vivo, we measured IL-1Ra transcription in the lungs of mice infected with Aspergillus with or without the administration of GAG. Induction of Il1ra was increased in the presence of GAG (Figure 4A). To determine which cells are responsible for the induction of Il1ra, we isolated macrophages, neutrophils and epithelial cells from the lungs of naïve mice. Macrophages and neutrophils, but not epithelial cells, expressed Il1ra after stimulation with Aspergillus in the presence of GAG (Figure 4B). Interestingly, not all microbiological stimuli can prime for increased GAG-induced Il1ra, since pre-stimulation with LPS did not increase Il1ra induction by GAG (Figure 4B).

To investigate the significance of IL-1Ra in vivo and to determine whether the effects induced by GAG are dependent on IL-1Ra, we studied the effects of GAG in wild type (WT) and Il1ra^−/− mice with invasive aspergillosis. Il1ra^−/− mice were highly resistant to invasive aspergillosis, as indicated by long-term survival (Figure 4C) and reduced fungal burden (Figure 4D). Administration of GAG resulted in increased protein levels of IL-1Ra in the lungs of wild-type mice during infection (Figure 4E). In line with previous observations, GAG increased the susceptibility to invasive aspergillosis in WT mice but not in Il1ra^−/− mice (Figure 4D). Il1ra^−/− mice had increased Mpo expression (Figure 4F) and PMN influx in their respiratory tract (Fig. 4G). As expected, administration of GAG reduced inflammatory PMN recruitment in WT but not in Il1ra^−/− mice (Figure 4G). These data demonstrate that IL-1Ra has an important role in invasive aspergillosis, and support the concept that the induction of IL-1Ra by GAG may have important clinical consequences.

In ABPA, GAG administration decreased PMN recruitment, but not eosinophilic infiltration in the BAL and lung of allergic mice (Figure 4H), a finding consistent with decreased Th17 but not Th2 cell responses in the draining lymph nodes (Figure 4I). To address whether GAG would have similar effects on human Th2 responses, PBMCs isolated form healthy subjects were pre-incubated for 1 h with GAG and subsequently stimulated with Aspergillus conidia for 7 days. Similar to mice, IL-17 production decreased in the presence of GAG, but not Th2 cytokines such as IL-5 and IL-13 (Figure 4J). Thus, GAG has the potential to ameliorate Th17-dependent immunopathology in ABPA.

Since IL-1Ra treatment can be beneficial for autoinflammatory diseases such as chronic granulomatous disease CGD colitis in humans [14], we investigated whether GAG could be beneficial in experimental DSS-induced colitis in mice with chronic granulomatous disease (CGD). The administration of GAG resulted in the amelioration of clinical signs of colitis (weight loss and stool consistency) (Figure 5A-B) and of inflammatory lesions (Figure 5C) in both wild-type and CGD mice. However the protective effects of GAG were most apparent in CGD mice. GAG induced IL-1Ra and, consistently, reduced IL-1β and IL-17 (Figure 5D). Concomitantly, there was an increased production of IL-10, an anti-inflammatory cytokine that plays an important role in the protection of colitis [15], [16]. The effects of GAG on CGD colitis were similar to those of IL-1Ra administration (unpublished data).

All studies with human blood samples were conducted in the Radboud University Nijmegen Medical Centre and the use of healthy volunteers was approved by the institutional ethics review board. Peripheral venous blood samples from healthy volunteers were obtained after written informed consent was provided.

All animal studies were conducted within the University of Perugia and were performed according to the Italian Approved Animal Welfare Assurance A–3143–01. Legislative decree 245/2011-B regarding the animal license was obtained by the Italian Ministry of Health lasting for three years (2011–2014). Infections were performed under avertin anesthesia and all efforts were made to minimize suffering.

Galactosaminogalactan (GAG) was isolated from A. fumigatus culture supernatant and purified from the urea-soluble fraction as previously described [7]. Lyophilized GAG was resuspended in 10 mM HCl at 2 mg/ml and used in a final concentration of 10 µg/ml. Before using GAG in stimulation experiments it was incubated with polymixin B to neutralize potential contamination of lipopolysaccharide.

A clinical isolate of Aspergillus fumigatus V05-27, which has been previously characterized was used for stimulations [27]. Conidia and hyphae were prepared and heat-killed as previously described [6]. A concentration of 1×10⁷/ml was used in the experiments, unless otherwise indicated. Recombinant human IL-1β, IL-23, IL-12 and IL-18 were purchased from R&D Systems (Minneapolis, MN, USA) and were used in end concentrations of 100 ng/ml, 50 ng/ml, 10 ng/ml and 50 ng/ml respectively. Recombinant human (rh) IL-1Ra (Amgen, Inc., Thousand Oaks, CA, USA) was used to antagonize IL-1β signalling at a final concentration of 10 ng/ml. Anti-humanIL-1Ra (R&D) was used to block IL-1Ra in a final concentration of 10 µg/ml, and was compared to isotype control.

PBMCs were isolated as described previously [28]. The cells were counted using a particle counter (Beckmann Coulter, Woerden, The Netherlands) and the cell number was adjusted to 5×10⁶/ml. PBMCs were plated in 96-well round-bottom plates (Corning, NY, USA) at a final concentration of 2,5×10⁶/ml and in a total volume of 200 μl. Cells were pre-stimulated for 1 hour with medium or 10 µg/ml GAG. Following prestimulation, the PBMCs were stimulated with culture medium, heat killed A. fumigatus conidia (1×10⁷/ml), IL-1β/IL-23 (100 and 50 ng/ml respectively) or IL-12/IL-18 (10 and 50 ng/ml respectively). These experiments were also performed in the presence of 10 μg/ml anti-IL-1Ra antibody or isotype control. Plates were incubated at 37°C, 5% CO₂ for 24 hours, 48 hours or 7 days. 7 day cultures were supplemented with 10% human serum. In this serum we detected anti-GAG antibodies as described previously [7]. After incubation, culture supernatants were collected and stored at -20°C until cytokine measurements were performed.

The murine cell line NOB-1 responds to both human or mouse IL-1 by production of IL-2, furthermore these cells are unresponsive to other cytokines like tumor necrosis factor (TNF), colony stimulating factors (CSFS), IL-3, IL-5, IL-6 and IFN-γ [29]. NOB-1 cells were plated in 96 well flat-bottom plates at a final density of 1×10⁶ cells/ml and were stimulated for 24 hour using culture supernatants of unstimulated PBMCs or PBMCs stimulated in presence of GAG (GAG conditioned medium). After 24 hours of incubation at 37°C, 5% CO₂ the culture supernatants of the NOB-1 cells were collected and IL-2 production by the NOB-1 cells was measured by ELISA (R&D systems).

Cytokines were measured using commercially available ELISAs (R&D Systems)(Biolegend, San Diego, CA, USA) (Sanquin, Amsterdam, The Netherlands) according to the protocols supplied by the manufacturer. IL-1α, IL-1β, TNF-α, IL-6, IL-8, IL-1Ra and IL-10 were measured in culture supernatants of 24 hour cultures, and IL-5, IL-9, IL-13, IL-17, IL-22 and IFN-γ were measured in culture supernatants of 7 day cultures.

Female, 8- to 10-weeks old, BALB/c (wild-type, WT) mice were purchased from Charles River (Calco, Italy). Breeding pairs of homozygous Il1ra^−/− mice on the BALB/c background, were kept under specific-pathogen free conditions at the breeding facilities of the University of Perugia, Perugia, Italy. Experiments were performed according to the Italian Approved Animal Welfare Assurance 229-2011-B.

Viable conidia from the A. fumigatus Af293 strain were obtained as described [30]. For infection mice were anesthetized in a small plastic cage, containing 3% Isofluoran (Isofluran Forene Abbot Scandinavia AB, Solna) before intranasal (i.n.) instillation of a suspension of 2×10⁷ resting conidia/20 µl saline. Mice were treated with 250 μg/kg i.n. of GAG the day of infection and on days 1 to 3 post infection. Mice were monitored for survival, fungal growth (colony forming unit/organ, mean ± SEM), as described [31], histopathology, myeloperoxidase (Mpo) and Il1ra mRNA expression in lung cells and IL-1Ra production. For histology, sections (3–4 ìm) of paraffin-embedded tissues were stained with periodic acid-Schiff (PAS) reagent. For allergy, mice received an i.p. and s.c. injection of 100 µg of A. fumigatus culture filtrate extract (CCFA) dissolved in incomplete Freund's adjuvant (Sigma) followed by two consecutive intranasal injections (a week apart) of 20 µg CCFA. A week after the last intranasal challenge, mice received 10⁷ Aspergillus resting conidia and evaluated a week later (16424201). GAG (250 μg/kg i.n.) or vehicle alone was administered daily, for a week, in concomitance with the Aspergillus infection.

Lungs were filled thoroughly with 1 ml aliquots of pyrogen-free saline through a 22-gauge bead-tipped feeding needle introduced into the trachea. The lavage fluid was collected in a plastic tube on ice and centrifuged at 400 g, 4°C, for 5 min. For differential BAL cell counts, cytospin preparations were made and stained with May- Grünwald Giemsa reagents (Sigma-Aldrich). At least 200 cells per cytospin preparation were counted and the absolute number of each cell type was calculated. Photographs were observed using a BX51 microscope (Olympus, Milan, Italy) and images were captured using a high-resolution DP71 camera (Olympus).

Mice received either regular drinking water (control) or 2.5% dextran sulfate sodium (DSS) in drinking water for 7 days and then allowed to recover by drinking water alone for an additional 7 days. GAG was given intraperitoneally (1 mg/kg) daily for a week. Weight changes were recorded daily, and the day after the 7-days of rest mice were killed and tissues were collected for histology and cytokine analysis. Colonic sections were stained with H&E [32]. To assess colitis severity, stool and histological scores were used that recently were introduced and proven sensitive to experimental therapy [33].

Purified peritoneal CD11b⁺ Gr-1⁺ polymorphonuclear neutrophils (PMNs) (>98% pure on FACS analysis) were obtained as described [34]. Lung epithelial cells were isolated as described [35] Murine macrophages were isolated from total lung cells after 2 hours plastic adherence at 37°C. PMNs, epithelial cells and macrophages were exposed to unopsonized Aspergillus conidia at the ratio of 1∶1 or LPS (10 ng/ml) at 37°C for 1 hour in the presence of different concentrations (1 or 20 µg/ml) of GAG for 18 hours before the assessment of Il1ra mRNA expression.

The differences between the various stimulations were analyzed with the Wilcoxon signed rank test (p-value of <0.05 was considered statistically significant). All experiments were performed at least twice and data represent cumulative results of all experiments performed and are presented as mean +/− standard error of the mean (SEM) unless otherwise indicated. Data was analyzed using GraphPad Prism v5.0.