Research ArticleALLERGY

Hookworm recombinant protein promotes regulatory T cell responses that suppress experimental asthma

See allHide authors and affiliations

Science Translational Medicine  26 Oct 2016:
Vol. 8, Issue 362, pp. 362ra143
DOI: 10.1126/scitranslmed.aaf8807

Airway allergy alleviated by hookworm protein

One reason for allergy prevalence in the developed world may be a lack of exposure to parasites, which can influence immune development and function. Because administering live parasites to people might pose safety issues, Navarro et al. tested the ability of the hookworm protein AIP-2 to treat airway allergic sensitization. Administration of AIP-2 could prevent or treat asthma symptoms in a mouse model, in a mechanism that was dependent on dendritic cells and regulatory T cells. Encouragingly, AIP-2 also reduced activation of human dendritic cells and T cells, indicating that these findings may readily translate to the clinic.

Abstract

In the developed world, declining prevalence of some parasitic infections correlates with increased incidence of allergic and autoimmune disorders. Moreover, experimental human infection with some parasitic worms confers protection against inflammatory diseases in phase 2 clinical trials. Parasitic worms manipulate the immune system by secreting immunoregulatory molecules that offer promise as a novel therapeutic modality for inflammatory diseases. We identify a protein secreted by hookworms, anti-inflammatory protein-2 (AIP-2), that suppressed airway inflammation in a mouse model of asthma, reduced expression of costimulatory markers on human dendritic cells (DCs), and suppressed proliferation ex vivo of T cells from human subjects with house dust mite allergy. In mice, AIP-2 was primarily captured by mesenteric CD103+ DCs and suppression of airway inflammation was dependent on both DCs and Foxp3+ regulatory T cells (Tregs) that originated in the mesenteric lymph nodes (MLNs) and accumulated in distant mucosal sites. Transplantation of MLNs from AIP-2–treated mice into naïve hosts revealed a lymphoid tissue conditioning that promoted Treg induction and long-term maintenance. Our findings indicate that recombinant AIP-2 could serve as a novel curative therapeutic for allergic asthma and potentially other inflammatory diseases.

INTRODUCTION

Immunoepidemiological observations have highlighted the protective effect of some anthropophilic helminth infections against allergies (13). One such study including almost 13,000 individuals in rural Ethiopia found that the risk of wheeze was independently reduced by infection with the hookworm Necator americanus but unrelated to viral exposure (4). Moreover, a compelling number of reports indicate that, although children infected with some intestinal helminths, and hookworms in particular, have a reduced incidence of allergic diseases, anthelmintic drug treatment provokes a significant increase in sensitization, wheezing risk, and atopic responses to allergens (59). It should be noted, however, that not all human helminths protect against lung inflammation; Ascaris infection is thought to promote asthma, possibly because of cross-reactivity between worm proteins and known allergens from other sources, such as house dust mites (HDMs) (10).

More recently, the efficacy of experimental infections with N. americanus or the pig whipworm Trichuris suis has been assessed in clinical trials for inflammatory bowel disease (1113), multiple sclerosis (14), asthma (15), rhinitis (2), and celiac disease (16). Although probiotic worm therapy has generated substantial interest in the treatment of chronic inflammatory conditions (17), concerns prevail around the implications of experimental human infection with a live pathogen (18) and its potential scalability as a therapeutic modality. Identifying protective worm molecules could circumvent this issue and promote the development of helminth-derived biologics for treating a range of disorders that result from a dysregulated immune system, such as asthma.

Allergic diseases affect up to 30% of the population worldwide, with asthma being one of the most common chronic diseases in which affected individuals may suffer considerable morbidity (1922). Although corticosteroids are effective at managing the disease, 10% of patients do not respond to the treatment, and they are associated with severe long-term side effects (2326). Whereas healthy individuals have more regulatory T cells (Tregs) than T helper 2 (TH2) effector cells, allergic patients have a defect in either the number of regulatory cells, their suppressive activity, or both (27, 28). Various approaches to redress the balance between regulatory and effector cells, such as allergen-specific immunotherapy (2931) and biologics targeting cytokines or immunoglobulin E (IgE) (27, 28), have been tested, but each has limitations.

Crude excretory/secretory (ES) products of numerous helminth parasites can suppress inducible lung inflammation in mouse models (2931). However, only a small number of purified ES immunomodulatory molecules have been characterized, and even fewer have been demonstrated to protect against allergic airway disease when produced in recombinant form. Among them, ES-62, a glycoprotein from a filarial nematode, protects against airway inflammation by polarizing the TH response via its phosphorylcholine moiety (32). AvCystatin, also secreted by filarial and other nematodes, is a recombinant cysteine protease inhibitor that ameliorates lung inflammation in mouse models of asthma by inducing interleukin-10 (IL-10)–producing macrophages (33, 34). Parasitic nematodes can also suppress lung inflammation in mice by secreting homologs of human cytokines such as macrophage migration inhibitory factor (35). Here, we show that hookworm ES products protect against experimental asthma and that this protection is likely due, at least in part, to the secretion of a protein that we have termed anti-inflammatory protein-2 (AIP-2). We show that recombinant AIP-2 induces the expansion of mouse mesenteric CD103+ dendritic cells (DCs), resulting in the generation of Tregs that home to the mucosa. Moreover, we show that AIP-2 generates a proregulatory imprint on mesenteric lymphoid tissues, promoting long-term specific protection against allergic asthma. Finally, we show that AIP-2 suppresses expression of costimulatory markers on human DCs and suppresses HDM-specific proliferation of peripheral blood mononuclear cells (PBMCs) from allergic asthma patients, highlighting the potential efficacy of helminth-derived molecules for treating a range of human inflammatory diseases.

RESULTS

Hookworm ES products and a recombinant TIMP-like ES protein protect against ovalbumin-induced airway inflammation

The anti-inflammatory properties of hookworm ES products have mostly been addressed using mouse models of colitis (3638) and exhibit a pro-TH2 phenotype characterized by eosinophilia, IL-4 production, induction of M2 macrophages, and CD4+ T cells secreting both IL-4 and IL-10 (37). To investigate whether purified hookworm ES products could suppress inflammatory responses generated in asthma, we sensitized mice to ovalbumin (OVA)/aluminum hydroxide (alum), treated with a daily intraperitoneal injection of ES products (AcES, 1.0 mg/kg) or vehicle for 5 days, and challenged with OVA or phosphate-buffered saline (PBS) (Fig. 1A). Mice treated with AcES displayed reduced peribronchial and perivascular cellular infiltration of the lungs and mucus hypersecretion (Fig. 1B). A differential cell count of bronchoalveolar lavage fluid (BALF) showed a significant decrease in eosinophil and lymphocyte infiltrates in these mice (Mann-Whitney, P = 0.0044; Fig. 1C). OVA-specific serum IgE levels were significantly reduced after AcES treatment (P = 0.0286, Fig. 1D). Quantities of both lung TH2 cytokines IL-5 and IL-13, as well as proinflammatory cytokines IL-6 and IL-17A, were significantly decreased in AcES-treated mice (Mann-Whitney, P < 0.0001, P = 0.0009, P = 0.0013, and P = 0.0007, respectively; fig. S1). Although these results indicate that AcES treatment is efficient at suppressing allergic responses, it presents substantial disadvantages as a therapeutic, given the requirement to reproducibly culture live parasites in a sterile environment. We therefore sought to identify the specific components of AcES that were responsible for protection against inflammatory airway disease.

Fig. 1. Ancylostoma caninum ES products and AIP-2 recombinant protein suppress OVA-induced airway hyperresponsiveness, lung inflammation, mucus production, and collagen deposition.

(A) Experimental procedure. Mice were sensitized with two intraperitoneal injections of OVA in alum and treated with intraperitoneal injection of AcES, AIP-2, or PBS (1 mg/kg) daily for 5 days. Mice were challenged with OVA aerosols from day 14 onward for 5 days. Mice were analyzed 2 days after the last aerosol. (B) Lung sections were stained with hematoxylin and eosin (H&E) or periodic acid–Schiff reagent (PAS). Micrographs depict representative tissues under 10-fold magnification. (C) Bronchoalveolar lavage cells were analyzed by flow cytometry. Data show the number of eosinophils (E) and lymphocytes (L) in the indicated groups. (D) Levels of OVA-specific serum IgE were assessed by enzyme-linked immunosorbent assay (ELISA). (E) Lung function was assessed by invasive plethysmography measuring resistance and compliance in response to increasing doses of methacholine. Results show the mean ± SEM from a representative experiment of three, with n = 5. *P ≤ 0.05; **P ≤ 0.01, ANOVA. (F) Concentrations of IL-5, IL-6, IL-13, and IL-17A in the lungs as assessed by cytometric bead array. (G) Lung sections were stained with Masson’s trichrome. Micrographs depict representative tissues under 10-fold magnification. Results represent individual mice or the mean ± SEM from either two (AcES) or five (AIP-2) independent experiments, each with n = 6 to 10. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001, Mann-Whitney.

We previously showed that protective moieties of AcES in a mouse model of colitis were proteins (36, 37). Characterization of the AcES proteome revealed that AIP-2 [formerly known as Ac-TMP-2; see (39)] was an abundant protein (40). To determine whether recombinant AIP-2 could protect against experimental asthma, mice were sensitized to OVA/alum as previously described, treated with recombinant AIP-2 expressed in Pichia pastoris or purified, heat/protease-denatured AIP-2 (dAIP-2), or vehicle, and challenged with aerosolized OVA or PBS (Fig. 1A and fig. S2). Mice treated with AIP-2 displayed reduced peribronchial and perivascular cellular infiltration of the lungs, mucus hypersecretion, and collagen deposition compared to those treated with vehicle (Fig. 1, B and G). Similarly, BALF differential cell count showed a significant decrease in eosinophil and lymphocyte infiltrates upon AIP-2 treatment (P < 0.0001, P < 0.0001, and P = 0.079, respectively; Fig. 1C and fig. S2B), as were OVA-specific IgE serum titers compared to controls (P < 0.0001, Fig. 1D). Lung resistance and compliance were significantly improved by AIP-2 treatment [two-way analysis of variance (ANOVA), PRes = 0.0128, 0.011, and 0.0018 and PComp = 0.0024, 0.0022, and 0.0024 for the 6, 12, and 25 mg/ml doses of methacholine, respectively; Fig. 1E and fig. S2A]. However, mice treated with dAIP-2 were not protected against airway inflammation (fig. S1, A and B), and cytokine levels were significantly lower in the AIP-2–treated group compared to vehicle and dAIP-2 control groups (Mann-Whitney, P < 0.0001, P < 0.0001, P = 0.0034, and P < 0.0001, respectively, Fig. 1F; P = 0.0076 and 0.0079, respectively, fig. S2C).

Recombinant AIP-2 is N-glycosylated by P. pastoris at Asn48 of the mature protein. To demonstrate that the N-glycans decorating AIP-2 were not responsible for its anti-inflammatory properties, we made a site-directed mutant of the protein where Asn48 was mutated to Gln and termed the protein AIP-2Q48. AIP-2Q48 conferred equivalent protection against OVA-induced asthma to glycosylated AIP-2 in terms of preventing cellular recruitment and cytokine production in the lungs (fig. S3). To understand the cellular mechanisms that govern AIP-2–mediated protection, we sought to identify whether Toll-like receptor (TLR) signaling was involved. MyD88−li/Trif−Tr mice were sensitized to OVA and subsequently treated with AIP-2 or vehicle alongside wild-type (WT) mice. BALF differential cell count and lung IL-5 and IL-13 concentrations were not significantly different between WT and MyD88−mi/Trif−Tr mice treated with AIP-2 or vehicle before challenge, indicating that TLR signaling was not involved in AIP-2–mediated protection against allergic inflammation (fig. S4).

Intraperitoneal injection of AcES leads to peritoneal eosinophilia (37). To investigate whether AIP-2 induces a similar effect, peritoneal lavages from mice treated with AIP-2 were collected and found to contain the same numbers of eosinophils, lymphocytes, neutrophils, and macrophages as controls (fig. S5), suggesting that, unlike AcES, AIP-2 is nonimmunogenic. Furthermore, we investigated whether a course of treatment with AIP-2 induced specific antibody and T cell responses (fig. S6). Mice were treated daily with either vehicle, AIP-2, or OVA for 5 days and compared with control mice that were immunized with two intraperitoneal injections of AIP-2 formulated with alum administered 1 week apart. Anti–AIP-2 antibody levels of mice treated with either AIP-2 or OVA were negligible compared to control mice that were immunized with adjuvanted AIP-2 (fig. S6A). In addition, there was no induction of proliferation of T cells stimulated with AIP-2 in the groups treated with either AIP-2 or OVA (fig. S6B).

On the basis of various epidemiological, clinical, and experimental studies, there is no concrete evidence that hookworms induce global immunosuppression that impairs the host’s ability to fight bacterial or viral infections (17, 41). However, because of the potent suppression of inflammation observed with AIP-2 in our model, we sought to verify whether AIP-2 had general immunosuppressive abilities by treating mice with AIP-2 or vehicle, as previously described, and then injecting them with OVA formulated with Freund’s complete and incomplete adjuvants in the hocks. Popliteal lymph nodes were collected, and cells were counted and restimulated in vitro with OVA (fig. S7A). Cellularity and levels of interferon-γ (IFN-γ), IL-6, and tumor necrosis factor–α were similar in both AIP-2–treated and vehicle control mice, suggesting that AIP-2 had no effect on TH1-mediated vaccination-like responses (fig. S7, B and C). Together, these results suggest that AIP-2 is highly efficient at suppressing OVA-induced inflammatory responses and that the protein integrity is essential for the induction of protective processes.

Preventative or therapeutic mucosal delivery of AIP-2 protects against airway inflammation

Administration of human biologics is typically via the parenteral route, but less invasive delivery systems are desirable (42). To determine whether AIP-2 could prevent airway inflammation when delivered directly to the mucosa, we administered it intranasally as a bolus into the lungs (Fig. 2A), whereupon it induced significant protection against eosinophil recruitment to and TH2 cytokine production in the lungs (Mann-Whitney, PEos = 0.0022, PLymphos = 0.0022, PIL-5 = 0.0004, and PIL-13 = 0.0002; Fig. 2, B and C), as well as significant reductions in OVA-specific serum IgE titers (P < 0.0001, Fig. 2D). Moreover, to determine whether AIP-2 could suppress OVA-induced airway inflammation at the time of onset, mice were injected with AIP-2 daily for 4 days, commencing 2 days after the first OVA aerosol administration (Fig. 2A). Therapeutic delivery of AIP-2 during aerosol challenges induced significant protection against eosinophil recruitment to and TH2 cytokine production in the lungs and OVA-specific serum IgE (Mann-Whitney, PEos = 0.0152, PLymphos = 0.0022, PIL-5 = 0.0023, PIL-13 = 0.001, and PIgE < 0.0001; Fig. 2, B to D). To address the efficacy of a single AIP-2 treatment followed by two distinct phases of aerosol challenge (3 weeks apart), mice were sensitized to OVA, treated with AIP-2 or vehicle from days 12 to 16, and challenged with OVA aerosols from days 14 to 18. Mice were then left to rest for 3 weeks and were challenged a second time from days 36 to 40 without a repeat of treatment with AIP-2. We observed a twofold decrease in airway eosinophilia and a significant decrease in IL-5 and IL-13 production (Mann-Whitney, PEos = 0.0317, PLymphos = 0.0286, PIL-5 = 0.0159, and PIL-13 = 0.0079; Fig. 2, E and F).

Fig. 2. Therapeutic or preventative mucosal delivery of Ac-AIP-2 protects against allergic airway inflammation in mice.

(A) Experimental procedure. Mice were sensitized with two intraperitoneal injections of OVA in alum. Mice were treated with 30 μl of intranasal administration of AIP-2 (1 mg/kg) or PBS 5 days after sensitization (preventative group, pAIP-2) or 3 days after the beginning of OVA challenges (therapeutic group, thAIP-2). Analyses were performed 24 hours after the last aerosol. (B) Bronchoalveolar lavage cells were analyzed by flow cytometry. (C) Concentrations of IL-5 and IL-13 in the lungs were analyzed by cytometric bead array. (D) Levels of OVA-specific serum IgE were analyzed by ELISA. Data represent the mean ± SEM from two independent experiments, each with n = 6 to 8. (E and F) Mice were sensitized to OVA, treated with AIP-2 (1 mg/kg) or PBS from days 12 to 16 as described previously, and challenged with OVA aerosols from days 14 to 18. Mice were rested for 21 days after the last aerosol and challenged again from days 36 to 40. Analyses were performed 24 hours after the last aerosol. (E) Numbers of eosinophils (E) and lymphocytes (L) in the BALF were assessed by flow cytometry. Vcl, vehicle. (F) Concentrations of IL-5 and IL-13 in the lungs were analyzed by cytometric bead array. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001, Mann-Whitney.

AIP-2 specifically expands mesenteric CD11c+CD11bCD103+ DCs that are important for mediating protection against OVA-induced inflammation

To further explore AIP-2–induced anti-inflammatory mechanisms, we sought to identify the specific tissue targeted by the treatment when administered systemically. Naïve mice were injected daily for 5 days with AIP-2 or vehicle. Multiple tissues were collected, and cells were enumerated, revealing that mesenteric lymph node (MLN) cells were significantly expanded in relative comparison to the brachial, popliteal, and inguinal lymph nodes, spleen, Peyer’s patches, small intestine epithelial cells, and lamina propria (Mann-Whitney, P = 0.0003; Fig. 3A, left). Whereas both DCs, based on CD11c+ expression, and B cells were significantly expanded upon treatment (P = 0.0079 and 0.0082, respectively; Fig. 3A, right), injection of Alexa Fluor 647 (AF647)–labeled AIP-2 revealed that CD11c+ cells were the main population to capture AIP-2 (fig. S8A). Most AIP-2+ CD11c+ cells coexpressed CD103 but not CD11b (Fig. 3B). AIP-2 administration induced a significant decrease in the expression of major histocompatibility complex class II (MHC II) (P = 0.0079, 0.0076, and 0.0081, respectively; Fig. 3C), findings in agreement with a previous report on a related hookworm tissue inhibitor of metalloproteinase (TIMP)–like protein, Ac-TMP-1 (43). The CD11c+ population was selectively expanded after AIP-2 treatment (Fig. 3D, left and middle). Phenotypic characterization of these cells showed no difference in expression of CD80 and CD86 despite the significant increase of CD11c+CD11b and CD11c+CD11b+ DC subpopulations (Mann-Whitney, P = 0.0198 and 0.0381, respectively; Fig. 3D, left, and fig. S8, B and C). To verify that the expansion of MLN CD103+ DC was specific to Ac-AIP-2, naïve mice were injected in parallel with endotoxin-free OVA, as described above (Fig. 3, D and E). The expansion of MLN CD103+ DC observed with AIP-2 administration was not detected with OVA or vehicle, suggesting that the effects of AIP-2 were specific (Fig. 3D, right). Aldehyde dehydrogenase (ALDH1A2) activity, which is responsible for the conversion of retinol to retinoic acid, was also significantly enhanced in these cells upon AIP-2 administration when compared with OVA or vehicle (Mann-Whitney, P = 0.0027; Fig. 3E). CD103+ DCs have been widely described for their highly specific tolerogenic function in the mesenteries (4448). Therefore, to understand their role in mediating AIP-2–induced protection against airway inflammation, CD11c.DTR mice were sensitized and treated with AIP-2 in the presence or absence of diphtheria toxin (DT) to selectively deplete all CD11c-expressing cells (Fig. 3, F to H). Administration of DT during AIP-2 treatment abolished the suppression of airway infiltration of eosinophils and lymphocytes (Fig. 3G) and inhibition of IL-5 and IL-13 production in the lungs (Fig. 3H). Together, these data suggest that DCs are critical for the mediation of AIP-2–induced protection against airway inflammation and, more specifically, that AIP-2 expands MLN CD103+ DCs and enhances ALDH1A2 activity.

Fig. 3. AIP-2 treatment induces selective expansion of CD11c+CD11bCD103+ DCs in the mesenteric lymph nodes that are essential for protection against airway inflammation.

Mice were treated with intraperitoneal injections of AF647-labeled Ac-AIP-2 or PBS for 5 days. (A) Indicated tissues (PLN, peripheral lymph nodes; SPL, spleen; PP, Peyer’s patch; IEC, intestinal epithelial cells; siLP, small intestine lamina propria) were collected, and cell suspensions were prepared and counted (left). MLN cells were stained with major hematopoietic markers and analyzed by flow cytometry (right). Data are expressed as mean ± SEM from four independent experiments, each with n = 6 to 8. (B) MLN cells were stained with indicated surface markers and analyzed for the expression of AIP-2–AF647. Frequencies are expressed as mean ± SEM from three independent experiments, n = 5 to 8. (C) Mean fluorescence intensity (MFI) of MHC II expression on DC subtypes. Results represent the mean of three independent experiments, each with n = 5 to 8. (D) Mice were treated with Ac-AIP-2, OVA, or vehicle intraperitoneally for 5 days. MLNs were collected 24 hours after the last injection, and cells were analyzed by flow cytometry. Flow cytometric frequencies are expressed as individual mice with mean ± SEM from three independent experiments, each with n = 5 to 10. pDC, tolerogenic plasmacytoid DC. (E) MLN cells were incubated with Aldefluor. Representative histograms of ALDH1A2 activity assessed by Aldefluor staining of CD11c+CD11bCD103+ cells from mice treated with vehicle, OVA, or AIP-2 (left). Data represent the mean ± SEM for each group from two experiments, each with n = 5 (right). (F to H) CD11c.DTR mice were sensitized to OVA/alum and treated with DT or mock injection every third day during treatment with vehicle or AIP-2. Mice were then challenged with OVA or PBS, and inflammation was assessed 24 hours after the last aerosol. (F) Schematic representation of the experimental procedure. (G) Numbers of eosinophils (E) and lymphocytes (L) in the BALF were assessed by flow cytometry. (H) Concentrations of IL-5 and IL-13 in the lungs were assessed by ELISA. Results represent individual mice from a representative experiment or the mean ± SEM from two independent experiments, each with n = 5. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001, Mann-Whitney.

AIP-2 induces Tregs that accumulate in the mucosa and are critical for protection against OVA-induced inflammation

Both MLN CD103+ DCs and parasitic helminths are noted for their abilities to induce and maintain tolerance via induction of regulatory immune cells (41, 44, 45). Therefore, we examined the impact of AIP-2 treatment on CD4+Foxp3+ Treg populations. We observed a twofold increase in Tregs in the airway and intestinal mucosa expressing the gut-homing chemokine receptor CCR9 upon AIP-2 treatment in comparison to vehicle (Fig. 4, A and B, table S1, and fig. S9). The coculture of purified MLN CD103+ DCs from AIP-2–treated mice with CD4+ T cells from DO11.10 donors in the presence of OVA revealed a significantly enhanced capacity of these cells to induce de novo Tregs in vitro in comparison to the yeast-derived control recombinant protein, human albumin (HA) (P = 0.0118, Fig. 4C). These findings agree with the increased ALDH1A2 activity that we observed. To assess the role of DCs in the accumulation of Tregs in the mucosa, CD11c.DTR mice were injected with DT every third day starting 24 hours before the 5-day AIP-2 or vehicle treatment regimen described previously. In the absence of DCs, the frequencies and numbers of Tregs in the trachea and small intestine lamina propria in the DT-treated AIP-2 group were comparable to the DT-treated vehicle control, indicating that DCs are critical for AIP-2–induced Treg accumulation (Fig. 4D). Notably, DT had no side effects, as demonstrated by DT administration in WT mice on either the frequency or number of Tregs (fig. S9). To investigate the importance of Tregs in the protection induced by AIP-2 against airway inflammation, we used Foxp3DTR (DEREG) mice, which allow for green fluorescent protein (GFP) marking of and selective and efficient depletion of Foxp3+ Tregs (49). DEREG mice were sensitized to OVA, exposed to DT or not, treated with AIP-2 or vehicle, and challenged with aerosolized OVA or PBS (Fig. 4F). AIP-2–treated DEREG mice that did not receive DT injections were protected against airway inflammation, but specific deletion of Tregs with DT completely abrogated the protection in these mice against eosinophilia and lymphocyte infiltration into the airways (Fig. 4F), production of IL-5 and IL-13 in the lungs (Fig. 4G), and OVA-specific serum IgE production (Fig. 4H). Together, these data indicate that Tregs are essential for AIP-2–mediated protection against OVA-induced airway inflammation.

Fig. 4. AIP-2 induces CCR9+ Tregs critical for the suppression of OVA-induced inflammation.

Mice were treated with intraperitoneal injections of AIP-2 or PBS daily for 5 days. MLNs, spleen, peripheral lymph nodes (inguinal, popliteal, brachial), trachea (Tr), and small intestine lamina propria were collected and analyzed by flow cytometry. (A) Frequency of Foxp3+ cells among CD3+CD4+ T cells represented as individual mice with mean ± SEM from three independent experiments, each with n = 5 to 8. (B) Frequency of CCR9-expressing cells among the CD4+CD25+Foxp3+ Treg population expressed as mean ± SEM from three independent experiments, with n = 5 to 8. (C) MLN CD11c+CD11bCD103+ DCs were purified from mice treated with intraperitoneal injections of AIP-2 or HA (1 mg/kg) as a control protein for 5 days by flow cytometry. Cells (1 × 105) were cocultured with 2 × 105 carboxyfluorescein diacetate succinimidyl ester (CFSE)–labeled CD4+ T cells purified from the spleen of a DO11.10 mouse in the presence of OVA (0.01 mg/ml) for 4 days. Cells were analyzed by flow cytometry 3 days later. Data show representative differentiation profiles after gating on CD4+ T cells (left). Summary graph indicates the mean of triplicates ± SEM from two experiments, each with n = 30 mice per group (right). (D) CD11c.DTR mice received DT injections intraperitoneally starting 24 hours before vehicle or AIP-2 (1 mg/kg) and every third day during the 5-day treatment course with AIP-2. Data show the frequency (left) and number (right) of CD4+Foxp3+ cells in the indicated tissues expressed as the mean ± SEM of two independent experiments, each with n = 3 to 6. (E) Experimental procedure. DEREG mice were sensitized with two intraperitoneal injections of OVA in alum and treated with daily intraperitoneal injections of Ac-AIP-2 or PBS for 5 days. Indicated mice intraperitoneally received DT injections or PBS every third day starting 24 hours before the first AIP-2 injection. Mice were challenged with OVA aerosols from day 14 onward for 5 days. Mice were analyzed 2 days after the last aerosol. (F) Numbers of eosinophils (E) and lymphocytes (L) in the BALF were assessed by flow cytometry. (G) Concentrations of IL-5 and IL-13 in the lungs were assessed by ELISA. (H) Levels of OVA-specific IgE in the serum were assessed by ELISA. Results represent the mean ± SEM of three independent experiments, each with n = 5. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001, Mann-Whitney.

Complete MLN tissue resection results in the absence of mucosal Treg accumulation after AIP-2 treatment

MLNs have specific properties that allow for the conversion of naïve T cells into Tregs upon interaction with ALDH1A2-expressing CD103+ DCs, at which time they acquire CCR9 expression (44, 50). To further demonstrate the origin of the immune response generated by AIP-2, MLNs were surgically resected, as previously described (50). After a recovery period of 6 weeks, mice were treated with AIP-2 daily for 5 days. Although no major difference in the frequency of CD4+ Foxp3+ cells was observed in the periphery between WT and MLN resection (MLNrx) mice, Tregs were significantly decreased in the airway and intestinal mucosa of MLNrx mice (Mann-Whitney, PTrachea = 0.0022 and PLamina Propria = 0.0024; Fig. 5A, left). The remaining Tregs at these mucosal sites were likely the resident cells before surgical removal of MLNs. Furthermore, a significant decrease in CCR9- and cytotoxic T lymphocyte–associated protein 4 (CTLA-4)–expressing Treg in the small intestine lamina propria supports the idea that AIP-2 induces de novo Tregs that originate from the mesenteries (P = 0.0043 and 0.0022, respectively; Fig. 5A, right). To validate the critical role of the mesenteries in the protection against airway inflammation, mice underwent MLNrx or not, were subsequently sensitized to OVA, were treated with AIP-2 or vehicle, and were challenged with OVA or PBS aerosols once the recovery period was complete (Fig. 5, B to F). AIP-2 had no impact on airway hyperresponsiveness, airway infiltration of eosinophils, lung production of IL-5 and IL-13, and OVA-specific serum IgE in MLNrx mice (Fig. 5, C to F). Together, these data identify an essential role for the mesenteries in the capture and presentation of AIP-2, which results in the induction of Tregs that are capable of exiting the tissue to accumulate in the airway and intestinal mucosa.

Fig. 5. Resection of MLN tissue prevents the accumulation of Tregs in the mucosa.

(A) Three-week-old mice were anesthetized, and the MLNs were surgically removed. After a 6-week recovery period, mice were treated with intraperitoneal injections of AIP-2 for 5 days. Tissues were collected 24 hours after the last injection, and the cells were analyzed by flow cytometry. Data represent mean frequencies ± SEM obtained by flow cytometry analysis from two independent experiments, with n = 8. (B to F) Induction of OVA-specific airway inflammation in MLNrx mice. (B) Experimental approach. Three-week-old mice underwent surgical resection of MLNs. After recovery, they were sensitized to OVA, treated with AIP-2 or vehicle, and challenged with OVA aerosols or PBS, as described previously. (C) Lung function assessed by invasive plethysmography measuring resistance (left) and compliance (right) in response to increasing doses of methacholine. Results show the mean ± SEM from two representative experiments, each with n = 8. *P ≤ 0.05; **P ≤ 0.01, ANOVA showing significance between the WT, PBS, OVA and WT, AIP-2, OVA groups. (D) Numbers of eosinophils (E) and lymphocytes (L) in the BALF were assessed by flow cytometry. (E) Concentrations of IL-5, IL-6, IL-13, and IL-17A in the lungs were assessed by cytometric bead array. (F) Levels of OVA-specific serum IgE were assessed by ELISA. Results represent the mean ± SEM of two independent experiments, each with n = 8. *P ≤ 0.05; **P ≤ 0.01, Mann-Whitney. n.d., not detected.

AIP-2 induces a long-term “imprint” on MLN cells that can transfer protection via MLN transplant

We demonstrated that AIP-2 is captured by MLN CD103+ DC, resulting in the significant amplification of this population. We have also shown upon AIP-2 treatment that Tregs, critical for the suppression of OVA-specific airway inflammation, are expanded. The process of tolerance induction via CD103+ DC priming is not unique. Moreover, the adoptive transfer of mucosal or MLN Tregs into sensitized animals has been widely used to investigate the multiple mechanisms of suppression of allergic responses (47, 51, 52). To verify that similar observations could be made with a hookworm recombinant protein, MLN Tregs from vehicle- or AIP-2–treated animals were purified and adoptively transferred into OVA-sensitized hosts (Fig. 6, A and B). As expected, whereas the transfer of non-Foxp3 CD4 T cells did not modify the outcome of the aerosol challenges, recipients injected with Foxp3-expressing CD4 T cells from AIP-2– or vehicle-treated donors were equally protected (Fig. 6, A and B). To examine whether AIP-2 modifies the MLN microenvironment, mice were treated with AIP-2 or vehicle, as described above, and MLNs were surgically removed and transplanted into naïve recipients after removal of their own mesenteric tissue (Fig. 6A). After recovery, mice were sensitized to OVA and challenged with OVA or PBS aerosols. Nine weeks after the transplant, mice that received MLN transplant (MLNtx) from AIP-2–treated donors were still protected against airway hyperresponsiveness, lung eosinophilia, BALF lymphocyte infiltration, and lung IL-5 and IL-13 production (Fig. 6, C to F).

Fig. 6. Transplant of AIP-2–treated MLN tissue into MLN-deficient mice protects against OVA-induced airway inflammation.

(A and B) DEREG mice were treated intraperitoneally with AIP-2 (1 mg/kg) or vehicle for 5 days, and MLN CD4+ cells were enriched with magnetic beads and sorted based on the expression or not of GFP by fluorescence-activated cell sorting (FACS). Age-matched OVA-sensitized C57Bl/6 recipients received 0.2 × 106 cells from each donor group. Mice were then challenged daily with 2% OVA aerosols for 5 days. (A) Eosinophils (E) and lymphocytes (L) were analyzed from the bronchoalveolar lavage by flow cytometry. (B) Lung IL-5 and IL-13 cytokines were measured by cytometric bead array. Data show the mean ± SEM from one experiment, with n = 5. **P ≤ 0.001, Mann-Whitney. (C) Schematic representation of the MLNtx experimental procedure. Three-week-old mice were treated with intraperitoneal injections of AIP-2 or PBS. MLNs were collected 24 hours after the last injection and transplanted into age-matched naïve recipients. After a 6-week recovery period, recipients were sensitized and challenged with OVA. Analyses were performed 24 hours after the last aerosol administration. (D) Resistance and compliance in response to increasing doses of methacholine were analyzed by invasive plethysmography. Results show the mean ± SEM from two representative experiments, each with n = 8. **P ≤ 0.01; ***P ≤ 0.001, ANOVA showing significance between the WT, PBS, OVA and WT, AIP-2, OVA groups. (E) Numbers of eosinophils (E) and lymphocytes (L) in the BALF were analyzed by flow cytometry. (F) Concentration of IL-5 and IL-13 in the lungs were quantified by cytometric bead array. Results represent the mean ± SEM from two independent experiments, each with n = 8. *P ≤ 0.05; **P ≤ 0.01, Mann-Whitney.

AIP-2 controls the proliferation of mouse and human effector T cells in vitro

To investigate whether the tolerogenic environment generated by AIP-2 can alter the proliferation of effector responses in vitro, total MLN cells from mice treated with intraperitoneal injections of AIP-2 or OVA were cultured ex vivo in the presence of anti-CD3 stimulation and analyzed after 3 and 5 days of incubation (Fig. 7, A to C). A high frequency of CD4+Foxp3+ Tregs was observed on both days 3 and 5 in the cultures originating from AIP-2–treated mice. AIP-2 cultures showed a lower frequency of CFSE-diluted CD4+Foxp3 T cells (Teff) on both days 3 and 5, with a marked difference in proliferation in comparison to the cells from OVA-treated mice on day 5 (Fig. 7, A and B). Together, AIP-2 seemed to generate a high frequency of Tregs in vitro and a decreased frequency of proliferative Teff, which resulted in a significantly higher Treg/Teff ratio in comparison to OVA-treated mice (PDay3 = 0.0079 and PDay5 = 0.0006; Fig. 7C). Finally, to determine whether AIP-2 had immunoregulatory properties that acted upon human cells, PBMCs were isolated from seven healthy donors, and the cells were incubated with fluorescently labeled AIP-2 or OVA (AF647) or vehicle for 24 hours (Fig. 7, D and E, and fig. S10). Although all DCs seemed to comparably capture AIP-2 or OVA in vitro, AIP-2+CD11c+ DCs displayed a significant decrease in the expression of the CD80 activation molecule and a down-regulation in the expression of human leukocyte antigen (HLA)–DR, albeit nonsignificant, to levels comparable to the vehicle control (PCD80 = 0.0411; Fig. 7, D and E, and fig. S10). This marked effect of AIP-2 on the most efficient antigen-presenting cells suggests that AIP-2 affects human DC function. To verify this hypothesis in an allergic context, five human subjects with clinically characterized HDM allergy were recruited. PBMCs were isolated as described previously, and the cells were labeled with CFSE and preincubated with AIP-2 or vehicle for 2 hours before stimulation with HDM extract for 5 days. Similar to our previous observations with murine cells, HDM stimulation of PBMCs resulted in a specific proliferative response that was significantly reduced in the presence of AIP-2 (P = 0.0003, Fig. 7F).

Fig. 7. AIP-2 induces Tregs in vitro and inhibits Teff proliferation.

(A to C) Mice were treated with intraperitoneal injections of OVA or AIP-2 (1 mg/kg) for 5 days. MLNs were collected, and the cells were stained with CFSE and stimulated with CD3 (1 μg/ml) for 5 days. (A) On days 3 and 5, cells were analyzed by flow cytometry. Data show representative FACS plots with mean ± SEM of two experiments, each with n = 6. (B) Proliferation of Foxp3CD4+ T cells. Data show representative proliferation profiles (left) and summary graph (right) on both days 3 and 5 with mean ± SEM of two experiments, each with n = 6. (C) Treg/Teff ratio calculated on both days 3 and 5. Data show the mean ± SEM from two experiments, with n = 6. (D) PBMCs from seven healthy individuals were prepared and incubated at a density of 1 × 106 cells/ml for 24 hours with AF647-labeled AIP-2 or OVA (20 μg/ml) or vehicle at 37°C. Cells were then collected, washed, and stained for flow cytometry analysis. Data represent capture of AF647-labeled proteins by CD11c+HLA-DR+CD56CD14CD1a cells (D), and the expression of CD80 and HLA-DR on AF647+CD11c+ cells (E) expressed as mean ± SEM. (F) PBMCs from five individual HDM-sensitized patients were prepared and stained with CFSE. Cells were pretreated with AIP-2 or PBS (20 μg/ml) for 3 hours before adding HDM extract (20 μg/ml) and incubating for 5 days. Cells were stained with anti-CD3 and anti-CD4 and analyzed for CFSE dilution by flow cytometry. Representative histograms (left) and summary graph (right) with the mean ± SEM are shown for PBS- and AIP-2–treated cells stimulated with HDM. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001, Mann-Whitney.

DISCUSSION

Human infections with helminths, including hookworms, are universally associated with TH2 responses characterized by IL-4, IL-5, and IL-13, high levels of IgE, eosinophilia, and mastocytosis (53). This same immune signature is seen in allergic diseases, including asthma. Paradoxically, however, the vast majority of people infected with parasitic helminths do not succumb to hypersensitivity responses, nor do they develop potentially fatal allergies to their parasites. There are different interpretations on the mechanisms involved, and these are well summarized by the “hygiene hypothesis” and the “old friends hypothesis” (54, 55). Human hookworms induce a robust antibody (including IgE) response to a plethora of antigens (56), but this vigorous response is spectacularly unsuccessful at inducing protective immunity in hookworm-endemic populations. Whereas the explanation for the observed failure of humans to mount protective anti-hookworm responses is multifactorial, one clear mechanism is through the production of ES proteins with immunoregulatory properties (38, 43, 5760).

Here, we have demonstrated that hookworm ES products can significantly suppress OVA-induced airway inflammation. We have shown that the TIMP-like protein Ac-AIP-2 restores lung function and suppresses eosinophil and lymphocyte infiltration into the airways, mucus hyperplasia, and collagen deposition in the lungs upon systemic or mucosal delivery. This protection was independent of endotoxin contamination because MyD88−/−Trif−/− mice treated with Ac-AIP-2 were still protected from allergic asthma. AIP-2 did not appear to be particularly immunogenic because mice did not produce a detectable specific antibody response after treatment. Treatment with AIP-2 did not affect the establishment of TH1 immunity in response to an adjuvanted antigen. However, in an allergic disease model, lung TH2 cytokine production was significantly decreased upon AIP-2 administration, as was the production of proinflammatory cytokines IL-6 and IL-17A. We have yet to demonstrate whether AIP-2–induced Tregs are generated from an expansion of the naturally occurring subtype or the differentiation of naïve CD4+ T cells; however, we have shown here that they express the suppressive molecules CTLA-4 and CCR9, specifically imprinted in the mesenteries. The total resection of mesenteric lymphoid tissue allowed us to unequivocally demonstrate the MLN origin of AIP-2 Tregs, and furthermore, AIP-2 treatment induced a selective expansion of MLN CD11c+CD11bCD103-expressing DCs that were the predominant antigen-presenting cells to capture the protein. Selective depletion of DCs during the course of AIP-2 treatment demonstrated that these cells are critical for the establishment of protection against airway inflammation. In addition, AIP-2 seemed to promote the production of retinoic acid by these regulatory MLN CD103+ DCs, which translated into an enhanced ability to induce Tregs in vitro, explaining the potent regulatory response induced upon treatment. Furthermore, total MLN transplant from AIP-2–treated mice into naïve donors showed that the effects of AIP-2 on the resident populations within the tissue were significant and long lasting.

Cording and colleagues published recently that de novo Tregs induction likely depends on synergistic contributions of DCs and stromal cells, influenced themselves by the intestinal microbiome (61). It is possible that the long-lasting protection induced by AIP-2 originates from direct modifications on the numbers and tolerogenic function of MLN resident CD103 DCs, potentially involving the stromal cells and/or the gut microbiome.

In vitro restimulation of MLN cells showed a higher Treg/Teff ratio in mice previously treated with AIP-2 when compared to a control protein (OVA), resulting in significant suppression of proliferation of Foxp3CD4+ T cells upon anti-CD3 monoclonal antibody (mAb) activation. It is likely that the Tregs generated in vitro could control the proliferation of effector responses. In vitro stimulation with HDM extract of human PBMCs from confirmed allergic patients resulted in a similar suppression of effector responses in the presence of AIP-2 in comparison with vehicle. Incubation of PBMCs from healthy human donors with labeled AIP-2 confirmed that the hookworm protein seemed to be preferentially captured by CD11c-expressing cells, resulting in the down-regulation of activation molecules, as seen in murine DCs. Together, these data suggest that AIP-2 creates a protolerogenic environment that is potent enough to induce Tregs and suppress Teff responses in both mice and humans. In addition, the fact that PBMCs respond in a similar way to AIP-2 as do MLN cells in vitro suggests that our observations in mice may also be applicable in humans.

AIP-2 offers potent protection against allergic airway inflammation via the activation and expansion of Tregs and CD103+ DCs in the MLNs. AIP-2–induced Tregs might therefore control mucosal TH2 inflammation, possibly via direct effects on the composition of commensal microbial communities in the lung or the gut (6264). In contrast with other helminth recombinant proteins that suppress inflammation via type 2 macrophages or the modulation of voltage-gated potassium channels on memory T cells (33, 59, 65), AIP-2 seems to mainly affect MLN tolerogenic DCs and Tregs because depletion of either cell type ablated AIP-2–driven protection against OVA-induced airway pathology.

Our findings show an important and long-lasting effect of AIP-2, promoting the generation of regulatory responses without inducing general immunosuppression. By acting upstream of inflammatory cascades, AIP-2 appears to reset the balance of Teff-to-Treg responses by generating a long-term imprint on lymphoid tissues. Furthermore, the potential translatability of this particular hookworm protein into a human therapeutic highlights the utility of helminth recombinant proteins as a new generation of biologics. Our study, however, presents certain limitations because most of the protective responses induced by AIP-2 were localized to the mesenteries in mice, which could not be directly translated clinically. Although we showed protolerogenic effects on DCs from healthy human volunteers and inhibition of PBMC proliferation ex vivo in asthmatic human patients, experimental results have yet to be confirmed with human secondary lymphoid tissues. The notion of a recombinant protein drug is substantially more palatable and easily regulated as a therapeutic modality than experimental helminth infection. This class of biologics has appeal for a broad range of inflammatory diseases that are driven by an imbalance between pro- and anti-inflammatory T cell responses and have reached alarming prevalence in industrialized nations.

MATERIALS AND METHODS

Study design

The overall goal of the study was to determine whether a hookworm-derived protein produced in recombinant form could suppress allergic asthma. An investigation of the mechanism of action in mice was performed in combination with humans using ex vivo tissue culture–based approaches with cells derived from both healthy volunteers and asthmatic patients to demonstrate putative efficacy. To evaluate the therapeutic potential of our recombinant protein, we used different modes of delivery, genetically modified animals for conditional depletion, and surgical transplant techniques to address the suppressive mechanisms. Sample sizes were determined on the basis of previous experience and statistical analyses. Standard measurements were used in the OVA model of allergic asthma to determine efficacy, including lung function by plethysmography, airway remodeling and collagen deposition, cellular infiltration, and cytokine production. For the human component of our study, seven healthy volunteers were recruited, including men and women of different ethnicities. Peripheral blood DCs were cultured with AIP-2, and modulation of expression of costimulatory molecules was determined by flow cytometry. Peripheral blood was obtained from five asthmatic patients with clinical history of asthma to HDM and cultured with AIP-2 in the presence of HDM to assess allergen-specific proliferation. Sample sizes, replicates, and statistical measurements are included in the figures and legends and in the text where appropriate.

Mice

BALB/c.ARC and C57Bl/6 mice (3 to 12 weeks old) were purchased from the Animal Resources Centre (Perth, Western Australia, Australia) and housed under specific pathogen–free conditions according to the Australian animal rights and regulation standards. B6.MyD88-TRIFF knockout mice were provided by S. Akira (Osaka University, Japan) via L. Schofield from Walter and Eliza Hall Institute (Melbourne, Victoria, Australia). DEREG mice (49) were obtained from T. Sparwasser and provided by C. R. Engwerda along with the CD11c.DTR strain. All mice were bred in the animal facility at James Cook University (JCU) (Cairns, Queensland, Australia). All experimental protocols were approved by the JCU Animal Ethics Committee under projects A1771, A2172, and A1913.

Reagents

A. caninum adult hookworms were cultured as previously described (37). The supernatant (AcES) was collected, and endotoxin was removed using one of two methods: EndoTrap blue (Profos), according to the manufacturer’s instructions, or Triton X-114 (Sigma-Aldrich), as previously described, with some minor changes (37). Recombinant Ac-AIP-2 and the glycosylation mutant Ac-AIP-2Q48 were expressed as secreted proteins in the yeast P. pastoris using methods described elsewhere (66). Mutation of Asn48 to Gln was achieved using polymerase chain reaction, as described elsewhere (67). The complementary DNAs encoding the mature sequences of Ac-AIP-2 (amino acids 17 to 244) and Ac-AIP-2Q48 were cloned in frame into pPICZαA (Invitrogen) using Eco RI and Xba I restriction sites. The recombinant plasmids were linearized by Sac I digestion and transformed into P. pastoris strain X-33 by electroporation according to the manufacturer’s instructions (Invitrogen). The transformants were selected on yeast extract–peptone–dextrose plates containing Zeocin and assessed for the expression of recombinant protein via Western blot. A Western blot–positive clone for each protein was grown in a shaker flask, and the expression of the recombinant 6×His-tagged Ac-AIP-2 and Ac-AIP-2Q48 were induced with methanol, as per the manufacturer’s instructions (Invitrogen). The recombinant fusion proteins were purified with a nickel affinity column, and eluates containing Ac-AIP-2 and Ac-AIP-2Q48 were concentrated using Amicon Ultra Centrifugal concentrators and buffer-exchanged into PBS (pH 7.4). Lipopolysaccharide contents in AcES and Ac-AIP-2 were below 5 ng/mg, as determined using the Limulus amoebocyte lysate assay (Pierce Thermo Fisher Scientific). AIP-2 was denatured by trypsin digestion (1 μg/ml) (Sigma-Aldrich) and heat denaturation, as described previously (37). AIP-2 was labeled with AF647 using a protein labeling kit (Life Technologies) following the manufacturer’s instructions. mAbs to mouse CD3, CD4, CD25, CD19, B220, 120G8, TCR-β, Foxp3, CCR9, CD103, CTLA-4, F4/80, Siglec-F, Gr-1, CD11c, CD11b, IA/IE, CD80, CD86, IL-4, IL-5, IL-10, and IL-17A and to human CD3, CD4, CD56, CD14, CD1a, CD11c, HLA-DR, and CD80 were purchased from eBioscience and BD Biosciences.

AcES or AIP-2 treatment and induction of allergic asthma

Sensitization was performed by two intraperitoneal injections of 20 μg of endotoxin-free OVA (Hyglos GmbH) in 2 mg of alum (Pierce) at days 0 and 7. On days 12 to 16, mice were treated with AcES or AIP-2 (1 mg/kg) in PBS (Life Technologies). From days 14 to 18, mice were exposed to either a daily aerosol of OVA (0.2%) (Sigma-Aldrich) or PBS for 20 min using an ultrasonic nebulizer (Ventalair Max, Allersearch). On day 20, mice were analyzed for lung function and the hallmarks of allergic airway disease.

Airway hyperresponsiveness

Invasive measurements of dynamic lung resistance and compliance were performed 2 days after the last aerosol challenge using flexiVent (SCIREQ, Emka Technologies), as previously described (47). Briefly, mice were anesthetized [ketamine (50 mg/kg) and xylazine (5 mg/kg)], tracheotomized, and immediately intubated with an 18-gauge catheter, followed by mechanical ventilation. Respiratory frequency was set at 150 breaths/min with a tidal volume of 0.2 ml, and a positive-end expiratory pressure of 2 ml of H2O was applied. Increasing concentrations of methacholine (0 to 50 mg/ml) were administered at a rate of 20 puffs per 10 s, with each puff of aerosol delivery lasting 10 ms, via a nebulizer aerosol system with a 2.5- to 4.0-μm aerosol particle size generated by a nebulizer head (Aeroneb, Aerogen). Baseline resistance was restored before administering subsequent doses of methacholine.

Analysis of BALF cells

Mice were bled by severing the caudal vena cava, and a cannula was inserted into the trachea. Lungs were washed three times with 1 ml of warmed PBS. For differential BALF cell counts, cells were stained with anti–Siglec-F, anti–Gr-1, anti-CD3, and anti-CD19 mAbs (BD Biosciences) and analyzed by flow cytometry using a FACSCanto II flow cytometer and FACSDiva software. Eosinophils were defined as non-autofluorescent Siglec-F+CD3CD19, and lymphocytes were defined as CD3+ or CD19+.

Serum antibody measurements

Serum OVA-specific IgE was measured by ELISA. Antigen-coated MaxiSorp plates (Nunc, Sigma-Aldrich) were incubated with serial dilutions of sera and biotinylated anti-IgE mAbs (BD Biosciences). Horseradish peroxidase–conjugated streptavidin (BD Biosciences) and tetramethylbenzidine (KPL) were used for detection.

Cytokine assays

Lung samples were homogenized in calcium- and magnesium-free Hanks’ balanced salt solution and phosphatase and protease inhibitor cocktail (Roche). Multiplex IL-5, IL-13, IL-17A, and IFN-γ analyses were performed with cytokine bead arrays using a FACSCanto II flow cytometer (BD Biosciences).

Tissue processing

Trachea, lungs, MLNs, peripheral lymph nodes (brachial, inguinal, and popliteal), or spleens were processed in RPMI 1640 medium containing 2% fetal bovine serum (FBS), 400 U of type I collagenase, and DNase I (1 mg/ml) (Life Technologies) using Miltenyi gentleMACS (Miltenyi Biotec) and incubated for 15 min at 37°C. Cells were strained through a 70-μm cell strainer (BD Biosciences). Erythrocytes were lysed with ACK lysis buffer. Small intestine lamina propria were obtained after digestion in RPMI containing 5% fetal calf serum (FCS), 5 mM EDTA, and 2 mM dithiothreitol, as described previously (68). Briefly, Peyer’s patches were removed, and tissue pieces were incubated under agitation for 30 min. Intestinal epithelial lymphocytes were discarded by filtration, and the remainder was further incubated in RPMI 1640 containing 5% FCS, 400 U of type I collagenase, and DNase I (1 mg/ml) for 30 min at 37°C. Small intestine lamina propria cells were filtered and stained for flow cytometry.

Intestinal surgery

Three-week-old BALB/c.ARC mice were anesthetized with ketamine (50 mg/kg) and xylazine (5 mg/kg), and the MLNs of the small and large intestine were removed, as described previously (50). For MLNrx experiments, intestines were placed back into the abdomen after MLN removal. For MLNtx experiments, MLNs were isolated from AIP-2– or PBS-treated donors. The MLNs of the small and large intestine of the host were removed, and donor MLNs were transplanted into this region. Mice were allowed to recover for 6 weeks before subsequent experiments.

Cell cultures

For the in vitro control of proliferation experiments, mice were treated with intraperitoneal injections of OVA or AIP-2 (1 mg/kg) for 5 days. MLN cells were prepared and stained with CFSE, as described previously (69). Cells were seeded at a density of 0.3 × 106 cells per well in the presence of anti-CD3 mAb (1 μg/ml) or were left unstimulated. For the induction of Tregs in vitro, mice were treated with intraperitoneal injections of HA or AIP-2 (1 mg/kg) for 5 days. MLN cells were collected, and CD11c cells were enriched using the EasySep positive selection method (STEMCELL). Cells were further sorted with a FACSAria III cell sorter based on their expression of CD103 to a purity of >95%. CD103+CD11bCD11c+ DCs (1 × 105 cells) were cocultured with 2 × 105 CFSE-labeled CD4 T cells from a DO11.10 mouse. The spleen of the DO11.10 donor was processed, and the CD4 T cells were enriched using negative selection (Miltenyi Biotec). Cells were stimulated with endograde OVA (0.1 mg/ml) for 4 days. The cells were washed, the medium was refreshed, and the cells were incubated for another 3 days until analysis after staining with anti-CD4 and anti-Foxp3 mAbs. For the Treg transfer experiment, DEREG mice were treated with AIP-2 or vehicle (1 mg/kg) for 5 days, and the MLNs were harvested and prepared for CD4 T cell enrichment using magnetic beads (Miltenyi Biotec). Cells were sorted with a FACSAria III cell sorter based on GFP expression to a purity of >95%.

Human PBMCs

Seven healthy individuals (mean age, 38 years; two females and five males) and five allergic patients (mean age, 36 years; four females and one male) with clinical symptoms and histories of HDM allergies, verified for allergen-specific IgE responses by skin prick test with wheal sizes ≥2 mm classified as positive, were examined. All patients gave their written informed consent. The study was approved by the local Ethics Committee (JCU) and under institutional review board approval number H5869. Blood was collected into lithium heparin Vacutainer tubes (BD Biosciences). PBMCs were prepared from whole blood by diluting with an equal volume of RPMI 1640 medium and layered over a Lymphoprep density gradient using SepMate separation method (STEMCELL Technologies). After centrifugation (800g for 20 min at room temperature), cells were collected and washed once at 600g and twice at 250g for 10 min at 4°C. For the capture assay, isolated and washed PBMCs were resuspended at 1 × 106 cells/ml in RPMI 1640 culture medium containing 10% FBS, penicillin (100 U/ml), streptomycin (100 mg/ml), 20 mM Hepes buffer, and 1× β-mercaptoethanol and incubated with PBS, AF647-labeled AIP-2, or OVA (20 μg/ml) at 37°C for 24 hours. Cells were then collected, washed, and blocked with rat serum (1:50), and AF647+CD11c+ DCs were defined as CD3CD56CD14CD1aHLA-DR+ cells and then analyzed for the expression of CD80 and HLA-DR by flow cytometry. For the proliferation assay, PBMCs were stained with CFSE and resuspended at 5 × 106 cells/ml in culture medium. Cells were then preincubated with AIP-2 (20 μg/ml) or vehicle for 2 hours and stimulated with HDM extract (20 μg/ml; Stallergenes) for 5 days.

Statistical analyses

ANOVA for repeated measures was used to determine the levels of difference between groups of mice for plethysmography measurements. Comparisons for all pairs were performed by unpaired two-tailed Mann-Whitney U test. Significance levels were set at P = 0.05.

SUPPLEMENTARY MATERIALS

www.sciencetranslationalmedicine.org/cgi/content/full/8/362/362ra143/DC1

Fig. S1. Lung cytokine profile of mice treated with AcES.

Fig. S2. Denaturation of Ac-AIP-2 (dAIP-2) restores OVA-induced airway inflammation.

Fig. S3. Protection against OVA-induced airway inflammation with nonglycosylated recombinant AIP-2Q48.

Fig. S4. Ac-AIP-2–induced protection does not require the activation of TLRs.

Fig. S5. Intraperitoneal administration of Ac-AIP-2 does not induce cellular infiltration at the site of injection.

Fig. S6. Unadjuvanted Ac-AIP-2 treatment does not induce specific antibody production or in vitro T cell proliferation.

Fig. S7. Ac-AIP-2 does not impair vaccination (adjuvanted)–induced TH1-type inflammation.

Fig. S8. Ac-AIP-2 treatment modulates the expression of activation markers on the surface of MLN DCs.

Fig. S9. Ac-AIP-2 increases the frequency and number of mucosal Tregs.

Fig. S10. Ac-AIP-2 is captured by human DCs and decreases the expression of activation molecules.

Table S1. Source data (Excel).

REFERENCES

  1. Acknowledgments: We thank U. Bode, M. Buettner, and O. Pabst from Hannover Medical School (Germany) for sharing the transplant model. We thank A. Susianto for animal husbandry and C. Winterford and M. Christensen from the Histotechnology Unit at QIMR Berghofer Medical Research Institute (QIMRB) for their help with histology. We thank M. Montes de Oca from QIMRB for helpful discussions and provision of reagents. Funding: This work was supported by the National Health and Medical Research Council of Australia (NHMRC) program (grant 1037034 to A. Loukas and C.R.E.), Janssen R&D, NHMRC Principal Research Fellowship (to A. Loukas) and Senior Research Fellowship (to C.R.E.), and NHMRC Overseas Biomedical Fellowship (613718 to P.R.G.). S.N. was supported by an Australian Society of Parasitology researcher exchange travel award and a research grant from the Faculty of Medicine, Health and Molecular Sciences, JCU. Additional financial support was provided by the Australian and Queensland governments via the establishment of the Australian Institute of Tropical Health and Medicine at JCU. Author contributions: S.N. conceived the study, designed and performed the experiments, and wrote the manuscript. D.A.P. and S.T. expressed the recombinant proteins. I.B.F., D.A.P., L.J., and S.R. performed the experiments and proofread the manuscript. P.R.G., V.J., and C.R.E. provided reagents, gave advice on experimental design, and critiqued the manuscript. P.J.H., T.L., and B.Z. provided reagents and proofread the manuscript. R.P., J.C., A. Leech, and J.W.U. provided helpful discussions on experimental design and proofread the manuscript. T.S. provided reagents and proofread the manuscript. A. Loukas conceived the study, helped design the experiments, and wrote the manuscript. Competing interests: The authors declare that they have no competing interests. A. Loukas and S.N. have jointly invented “Method for treating inflammation,” as described in the following patent applications: Australian Patent Application No. 2012900999, filed on 13 March 2012; Patent Cooperation Treaty Patent Application No. PCT/AU2013/000247; U.S. Patent Application No. 14/384,681; WO2013/138822; WO2015/039189.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article