Research ArticleFibrosis

Combinatorial targeting of TSLP, IL-25, and IL-33 in type 2 cytokine–driven inflammation and fibrosis

See allHide authors and affiliations

Science Translational Medicine  04 May 2016:
Vol. 8, Issue 337, pp. 337ra65
DOI: 10.1126/scitranslmed.aaf1938

Teaming up against inflammation and fibrosis

The cult of the individual extends to new therapies—new targets are identified and validated (or not) on a one-on-one basis. However, no protein is an island, and failure with a monotherapy does not invalidate a target. Now, Vannella et al. demonstrate that this is indeed the case for some types of progessive type 2 inflammation and fibrosis. Using a variety of models including helminth infection and allergic lung inflammation, the authors show that individual disruption of the type 2 inflammatory molecules thymic stromal lymphopoietin (TLSP), interleukin-25 (IL-25), and IL-33 had no effect on the progression of type 2–dependent inflammation or fibrosis. However, targeting all three simultaneously blocked disease development and progression.

Abstract

Thymic stromal lymphopoietin (TSLP), interleukin-25 (IL-25), and IL-33 are important initiators of type 2–associated mucosal inflammation and immunity. However, their role in the maintenance of progressive type 2 inflammation and fibrosis is much less clear. Using chronic models of helminth infection and allergic lung inflammation, we show that collective disruption of TSLP, IL-25, and IL-33 signaling suppresses chronic and progressive type 2 cytokine–driven inflammation and fibrosis. In a schistosome lung granuloma model or during chronic Schistosoma mansoni infection in the liver, individual ablation of TSLP, IL-25, or IL-33/ST2 had no impact on the development of IL-4/IL-13–dependent inflammation or fibrosis. However, significant reductions in granuloma-associated eosinophils, hepatic fibrosis, and IL-13–producing type 2 innate lymphoid cells (ILC2s) were observed when signaling of all three mediators was simultaneously disrupted. Combined blockade through monoclonal antibody (mAb) treatment also reduced IL-5 and IL-13 expression during primary and secondary granuloma formation in the lungs. In a model of chronic house dust mite–induced allergic lung inflammation, combined mAb treatment did not decrease established inflammation or fibrosis. TSLP/IL-33 double-knockout mice treated with anti–IL-25 mAb during priming, however, displayed decreased inflammation, mucus production, and lung remodeling in the chronic phase. Together, these studies reveal partially redundant roles for TSLP, IL-25, and IL-33 in the maintenance of type 2 pathology and suggest that in some settings, early combined targeting of these mediators is necessary to ameliorate progressive type 2–driven disease.

INTRODUCTION

Type 2 immunity is characterized by the production of the cytokines interleukin-4 (IL-4), IL-5, IL-9, and IL-13, which play diverse roles in the immune response (1). In addition to suppressing the proinflammatory activity of type 1 immune responses (2), type 2 immunity regulates wound healing (3), metabolic homeostasis (4), and immunity to several extracellular parasites (5). However, although the type 2 response exhibits many host-protective functions, should these responses persist or become dysregulated, they can contribute to the development of disease. Chronic type 2 cytokine production underlies diseases including allergic asthma, atopic dermatitis, allergic rhinitis, ulcerative colitis, and many chronic fibroproliferative disorders (69). Therefore, a better understanding of the mechanisms that regulate the initiation, maintenance, and resolution of type 2 immune responses could reveal novel approaches to treat a host of important human diseases.

Three predominantly epithelial cell–derived cytokines—TSLP ( thymic stromal lymphopoietin), IL-25, and IL-33—have emerged as important initiators of type 2 immunity in mammals, and their expression during type 2 disease in humans is widely documented (1015). These alarmins are released from the epithelium and other local stromal compartments when cells are damaged or stressed by allergens, pollutants, or pathogens and thereby trigger the production of the canonical type 2 cytokines IL-5, IL-9, and IL-13 by human and mouse cells of the innate and adaptive immune system (16, 17). TSLP targets dendritic cells (DCs), basophils, mast cells, monocytes, natural killer T cells, and type 2 innate lymphoid cells (ILC2s) (1821). In humans, TSLP has been shown to induce naïve human CD4+ T helper 2 (TH2) cell responses, but only in the presence of DCs (22). IL-25 and IL-33 exhibit similar TH2-inducing activity, but rather than targeting DCs, myeloid cells, and TH2 cells, they largely promote type 2 immunity by stimulating ILC2s as well as basophils, mast cells, and eosinophils. IL-33 will amplify antigen-dependent and antigen-independent effector responses from both human and mouse TH2 cells (16, 17). One recent study revealed that IL-33 can enhance TSLP- and DC-mediated human TH2 memory responses in vitro, suggesting the alarmins could play a role in maintaining immune responses (23). Although TSLP, IL-25, and IL-33 have all been shown to promote type 2 immunity when overexpressed in mice (1012), the requirement for these cytokines in the development of type 2 immunity in response to allergens and helminth parasites has been more variable, with some studies identifying little to no role for TSLP, IL-25, or IL-33 when targeted individually (2428). This variability has been attributed to the redundant and overlapping functional activities of these cytokines. IL-33 and IL-25 have both been shown to induce production of IL-13 by human ILCs in vitro [for example, (29)]. However, this theory has not been systematically investigated in vivo, nor have the combined roles of the three cytokines been dissected in models of chronic type 2–dependent disease.

Here, we used both genetic- and monoclonal antibody (mAb)–based strategies to investigate whether bifunctional or trifunctional targeting of TSLP-, IL-25–, and IL-33–dependent signaling more effectively controls pathogenic TH2 responses than disrupting any of the pathways individually. The roles of the three cytokines in the initiation and maintenance of primary and secondary type 2 immune responses were investigated in both acute and chronic models of lung inflammation and during chronic helminth infection. These models involve innate-initiated pathways as well as the development of antigen-specific T cell responses that influence outcomes at later stages. A major goal was to investigate whether type 2 cytokine–driven inflammation and fibrosis could be ameliorated more effectively if all three epithelial cytokines were targeted in combination. Moreover, in contrast to previous studies that have focused on their role in the “initiation” of type 2 immunity (30), our studies were also designed to investigate whether TSLP, IL-25, and IL-33, either alone or in combination, are required for the “maintenance” of established type 2–driven disease, because this is the stage where most therapeutic strategies are initiated.

RESULTS

Function of IL-25 during the initiation and maintenance of type 2 inflammation

We have previously shown that TSLP is not required for type 2–driven granuloma formation and fibrosis induced by the eggs of the helminth parasite Schistosoma mansoni (26). Another group has demonstrated that many helminths could bypass the need for TSLP in the development of type 2 responses by directly modulating DC function (28). However, the relative importance of IL-25 and IL-33 to the maintenance of established type 2–driven disease and the potential redundancy of these mediators have not been assessed. Therefore, we began by exploring the contribution of IL-25 in type 2–dependent inflammation and fibrosis by overexpressing IL-25 in mice that were injected intravenously with live S. mansoni eggs. Hydrodynamic delivery of an IL-25–expressing plasmid to naïve mice boosted IL-25 mRNA expression more than 1000-fold in the liver (Fig. 1A). As observed in previous studies (10), corresponding increases in IL-4, IL-5, and IL-13 expression were observed in both the liver and the lungs (Fig. 1A). When the IL-25–expressing plasmid was delivered 24 hours before exposure to S. mansoni eggs, the resulting granulomatous response to the eggs in the lungs was exacerbated (Fig. 1B). Granuloma volume more than doubled in the IL-25–pretreated mice, and their lesions contained many more eosinophils than control mice, which was likely due to type 2 cytokine induction in the lungs (Fig. 1C). Goblet cell hyperplasia and mucus production were also augmented in the lungs of mice treated with the IL-25 plasmid. The effects of IL-25 plasmid administration were reduced in IL-13Rα1−/− mice, demonstrating that the IL-25–mediated increase in type 2–associated pathology was dependent on IL-4/IL-13–mediated signaling through the type 2 IL-4 receptor (IL-4R) complex (Fig. 1B). Eosinophils accumulated after plasmid administration, however, which is likely explained by IL-25–driven IL-5 expression (Fig. 1C).

Fig. 1. Ablating IL-25 offers no protection against type 2–mediated pathology.

(A) Quantitative polymerase chain reaction (qPCR) analysis of gene expression in lung and liver tissue from wild-type (WT) C57BL/6 mice 7 days after hydrodynamic injection of IL-25 (n = 5 mice) or phosphate-buffered saline (PBS) (n = 2). D0, day 0. (B) Histopathology analysis of lungs from WT and IL-13Rα1−/− mice 7 days after S. mansoni egg exposure and 8 days after hydrodynamic injection of IL-25 or PBS (n = 12 to 15 per group; pooled from two independent experiments). Scale bars, 50 μm. AB-PAS, Alcian blue–periodic acid–Schiff. (C) Cytokine quantification from bronchoalveolar lavage fluid (BALF) of mice in (B) (n = 4 to 5 per group). (D) Histopathology analysis of lungs from IL-25−/− mice and littermate controls 7 days after challenge with S. mansoni eggs with (2°) or without priming (1°), with 5000 S. mansoni eggs 14 days before challenge (granuloma volume 1°, n = 18 to 23 per genotype pooled from three experiments; granuloma volume 2°, n = 9 to 10 per genotype pooled from two experiments; eosinophils, n = 5 per genotype). (E) Histopathology analysis and fibrosis quantification of livers of IL-25−/− mice and littermate controls 12 weeks after infection with S. mansoni cercariae (n = 9 per genotype). Student’s t test was used to measure all P values, and P > 0.05 except where reported. Error bars represent SEM, and each data point represents a value for an individual mouse. Data are representative of two independent experiments unless otherwise noted. Wk0, week 0.

Although these studies established that IL-25 could exacerbate type 2 cytokine–driven pathology, they did not reveal whether endogenously expressed IL-25 was critical to the development of granulomatous inflammation and fibrosis. To clarify the role of IL-25 in both the initiation and maintenance of type 2–driven fibrosis, we used IL-25−/− mice in both primary (Fig. 1B) and secondary intravenous S. mansoni egg challenge models (31). In these experiments, naïve or egg-sensitized IL-25−/− mice and wild-type littermates were challenged intravenously with live S. mansoni eggs, and granuloma formation was quantified on day 7 after challenge. Neither primary nor secondary granuloma formation was significantly reduced in the absence of IL-25 (Fig. 1D). The number of granulomatous eosinophils in each group was also indistinguishable during both primary and secondary challenges (Fig. 1D, right panel). Finally, to evaluate the requirement for IL-25 in a more chronic type 2 disease setting, we exposed wild-type and IL-25−/− mice to S. mansoni cercariae and quantified granuloma volume, tissue eosinophilia, and fibrosis in the liver after 12 weeks of infection. Similar to the results in the lungs (Fig. 1D), no significant change in type 2–dependent pathology was observed in the livers of IL-25−/− mice compared with those of the wild-type mice when chronically infected with S. mansoni (Fig. 1E).

Role of IL-33 in type 2 inflammation and fibrosis

Given that the inflammation and fibrosis induced by S. mansoni eggs in both lungs and liver were IL-4–, IL-13–, and IL-13Rα1–dependent but did not require IL-25 or TSLP (26), we next examined whether IL-33/ST2 receptor signaling was required in this setting. As observed in IL-25−/− mice, mice deficient in IL-33 showed no significant reduction in either primary (Fig. 2A) or secondary granuloma formation (Fig. 2B) when challenged intravenously with live S. mansoni eggs. In both models, type 2–driven fibrosis and eosinophilia were similar in wild-type and IL-33−/− mice. We also infected wild-type and IL-33−/− mice with S. mansoni cercariae and examined the development of type 2–dependent pathology in the liver at acute (week 9) and chronic (week 12) phases of infection. Although recent studies using hepatotoxic chemicals or schistosome egg–driven models have suggested that IL-33 expression is critical to the development of fibrosis in the liver (32), we observed no reduction in hepatic fibrosis in IL-33−/− mice at either time point (Fig. 2C) by measuring hydroxyproline or staining liver sections with picrosirius red (Fig. 2D). The number of eosinophils in the lesions and the diameter of granulomas were also similar in the absence of IL-33, confirming unimpaired type 2–driven inflammation (Fig. 2E). The marked type 2 cytokine response that normally develops in the livers of infected wild-type mice was also similarly observed in IL-33−/− mice, and in the case of IL-4 expression was even slightly increased (Fig. 2F), further suggesting that IL-33 signaling is dispensable for the development of type 2 cytokine–driven pathology during both acute and chronic S. mansoni infection.

Fig. 2. Ablating IL-33 offers no protection against type 2–mediated pathology.

(A) Fibrosis quantification and histopathology analysis of lungs from WT C57BL/6 and IL-33−/− mice 7 days after challenge with S. mansoni eggs (n = 7 to 10 per genotype). (B) Fibrosis quantification and histopathology analysis of lungs of the same mouse strains 21 days after priming with S. mansoni eggs and 7 days after challenge with eggs (n = 10 per genotype). (C) Fibrosis quantification of livers from the same mouse strains infected with S. mansoni cercariae (n = 7 to 10 per genotype). (D) Micrographs of representative liver tissue sections of mice in (C) collected 9 weeks after infection and stained with picrosirius red. Scale bars, 100 μm. KO, knockout. (E) Histopathology analysis of livers from the mice in (C) (n = 7 to 10 per genotype). (F) Intracellular cytokine analysis of lymphocytes isolated from livers of mice in (C) 9 weeks after infection was measured by flow cytometry (n = 8 per genotype). Student’s t test was used to measure all P values, and P > 0.05 except where reported. Data are representative of two independent experiments for each of the models.

Disrupting TSLP, IL-25, and IL-33 signaling during S. mansoni infection

To investigate whether TSLP, IL-25, and IL-33 were playing redundant roles in the maintenance of type 2 cytokine–dependent granuloma formation and fibrosis, we developed strategies to simultaneously disrupt all three cytokine pathways. In initial studies, TSLP−/− mice were crossed with IL-33−/− mice to generate a double-knockout (DKO) mouse, and a highly effective neutralizing mAb was introduced to the DKO mice to block IL-25. C57BL/6 mice were infected with S. mansoni cercariae, and the response in the absence of TSLP, IL-25, and IL-33 signaling was evaluated at acute (week 9) and chronic (week 12) phases of infection. We first measured TSLP, IL-25, and IL-33 gene expression and found that each gene is constitutively expressed in whole liver tissue at detectable levels (fig. S1). These levels of expression do not change significantly during S. mansoni infection on a whole-tissue level. In contrast to the studies in which individual cytokines were targeted, we observed a small yet significant decrease in granuloma volume in the triple-deficient mice in the acute phase (Fig. 3A). This was also accompanied by a 25 to 30% decrease in hepatic fibrosis (Fig. 3B) and a small yet significant decrease in the number of granuloma-associated eosinophils (Fig. 3C). The decrease in pathology observed at week 9 was associated with a significant decrease in the frequency of IL-13–producing ILC2s in the mesenteric lymph nodes (MLNs) (Fig. 3D), which is consistent with the ILC2-promoting activity of IL-25 and IL-33 (33). The frequency of ILC2s in the liver, however, was not significantly different between the two groups (Fig. 3D, right panel). Total leukocyte numbers were similar in the liver tissue and MLNs of both cohorts.

Fig. 3. Simultaneous disruption of all three mediators has a transient effect on TH2 pathology driven by S. mansoni.

(A) Granuloma measurement (n = 14 to 19 per group pooled from two independent experiments) from livers of S. mansoni–infected WT C57BL/6 mice administered isotype control antibody and IL-33/TSLP DKO mice administered anti–IL-25. (B) Fibrosis quantification from livers of infected mice (n = 7 to 10 per group). (C) Quantification of granuloma eosinophils from livers of infected mice (n = 7 to 9 per group). (D) Quantification of CD4IL-13ST2+ICOS+ leukocytes from MLNs (n = 7 per group) and livers (week 9, n = 7 to 8 per group; week 12, n = 14 to 15 per group pooled from two independent experiments) of infected mice by flow cytometry. ICOS, inducible T cell costimulator; IP, intraperitoneal. (E) Micrographs of representative liver tissue sections of mice 12 weeks after infection and stained with picrosirius red. Scale bars, 500 μm. (F) Intracellular cytokine analysis of liver lymphocytes of infected mice by flow cytometry (n = 14 to 17 per genotype pooled from two experiments). Student’s t test was used to measure all P values, and P > 0.05 except where reported. Data are representative of two independent experiments.

By 12 weeks after infection, the decrease in IL-13–producing ILC2s in MLNs observed at week 9 was no longer significant (Fig. 3D, left panel), and although there was a modest but consistent decrease in pathology at week 9, granuloma volume and fibrosis became indistinguishable between wild-type and DKO + αIL-25–treated mice (Fig. 3, A to C). Both groups of mice displayed a marked increase in IL-13–dependent fibrosis by week 12 as determined by both hydroxyproline assay (Fig. 3B) and picrosirius red staining (Fig. 3E). In addition, whereas the frequency of IL-13–producing ILC2s was lower in the MLN at week 9 (Fig. 3D), a marked increase in IL-4– and IL-13–producing CD4+ T cells was observed at the same time point in the granulomatous livers of the DKO + αIL-25–treated mice (Fig. 3F). Antigen-specific CD4+ TH2 cell cytokine production likely compensated for the transient decrease in ILC2s, thus explaining the unimpaired development of IL-13–dependent fibrosis in triple-deficient mice by week 12.

Disrupting TSLP, IL-25, and IL-33 signaling during acute granuloma formation

After considering the transient nature of immune control affected by disrupting the three mediators during S. mansoni infection, we hypothesized that the effect of blocking all three cytokines would be more apparent when applied to a more acute model where the cytokines are blocked from the onset of injury. Primary and secondary lung granuloma models were used for these studies because they provide simple and short-term systems to dissect the importance of TSLP, IL-25, and IL-33 during both the sensitization and maintenance phases of a type 2 cytokine–driven inflammatory response (31). Groups were treated with either isotype control antibodies or with αTSLP, αIL-25, and αST2 (IL-33R receptor) monoclonal neutralizing antibodies for the entire length of the experiments. The pathological effects of TSLP, IL-25, and IL-33 have been directly linked to the enhanced production of IL-4, IL-5, and IL-13 by downstream target cells such as CD4+ TH2 cells, ILC2s, and other innate lymphocytes (33, 34), and much of the pathology that results from the persistent activation of type 2 immunity has been attributed to IL-4/IL-13–mediated signaling through IL-4R (35). Therefore, we used IL-4Rα–deficient mice as positive controls.

As observed in previous lung granuloma studies where TSLP, IL-25, and IL-33 were targeted individually, the combined blockade of all three cytokines had no significant impact on the volume of the lesions in mice undergoing either primary (Fig. 4A) or secondary (Fig. 4B) granuloma formation. In marked contrast, the lesions in IL-4Rα−/− mice were about 50% smaller than those in isotype control treated mice (Fig. 4, A and B). The triple blockade did lead to >80% reduction in the number of granuloma-associated eosinophils during primary granuloma formation (Fig. 4A, right panel and tissue sections). Macrophages and primarily lymphocytes comprised the granulomas in the absence of eosinophils. Nevertheless, the eosinophil deficit in the triple blockade mice was completely corrected when the mice were undergoing a secondary challenge (Fig. 4B, right panel and tissue sections). IL-4Rα−/− mice, in contrast, displayed a near complete absence of eosinophils after both primary and secondary challenges.

Fig. 4. Combined TSLP, IL-25, and ST2 mAb blockade during granuloma generation diminishes type 2 immunity but not pathology.

(A) Histopathology analysis of WT BALB/c and IL-4R−/− mice 7 days after injection with S. mansoni eggs. WT egg–injected mice were intraperitoneally administered either anti-ST2, anti-TSLP, and anti–IL-25, or corresponding isotype control antibodies (n = 8 to 9 per group). Micrographs are of representative lung sections stained with Masson’s trichrome. Scale bars, 50 μm. (B) Histopathology analysis of lungs from WT BALB/c and IL-4R−/− mice 7 days after injection with S. mansoni eggs and 21 days after priming with S. mansoni eggs (isotypes IP and triple block IP, n = 8 to 9 per group; IL-4R−/−, n = 5). WT egg–injected mice were intraperitoneally administered either anti-ST2, anti-TSLP, and anti–IL-25, or corresponding isotype control antibodies for all 3 weeks. Micrographs are of representative lung sections stained with Masson’s trichrome. Scale bars, 50 μm. (C) Quantification of gene expression in lung tissue from mice in (A) assayed by qPCR and shown relative to expression in lungs of naïve BALB/c mice (n = 3). (D) Quantification of gene expression in lung tissue from mice in (B) assayed by qPCR and compared to a different group of naïve BALB/c controls (n = 3). Student’s t test was used to measure all P values, and P > 0.05 except where reported. Data are representative of two independent experiments.

Although the effects of the triple blockade on egg-induced pathology were minimal, there were notable changes in cytokine expression in the lungs. Triple blockade mice displayed significant reductions in IL-4, IL-5, and IL-13 expression in the lungs during primary granuloma formation (Fig. 4C) and in IL-5 and IL-13 during secondary granuloma formation (Fig. 4D). It is worth noting that although these measurements imply a significant reduction in the type 2 cytokines after triple blockade, they were expressed at significantly higher levels than those in IL-4Rα–deficient mice. Changes in expression of two eosinophilic chemokines, Ccl5 and Ccl11, do not explain the eosinophil phenotype we observed. Ccl5 and Ccl11 were not affected by the triple blockade during the primary response (Fig. 4C), although both chemokines were reduced in the triple blockade mice during a secondary response (Fig. 4D). Rather, the pattern of Il5 gene expression likely explains why granuloma eosinophilia is reduced by the triple blockade during primary granuloma formation and is restored during secondary granuloma formation. The reduced Il5 expression in triple blockade mice during primary granuloma formation was on par with the expression observed in IL-4Rα–deficient mice. During secondary granuloma formation, the Il5 expression was reduced by the triple blockade, but it was still expressed at significantly higher levels than in IL-4Rα–deficient mice.

As seen in many type 2 cytokine–driven diseases, we observed increased Il33, Tslp, and Il25 gene expression in the lungs of wild-type mice in the primary granuloma model (Fig. 4C). Whereas gene expression of Il33 and Tslp increased in the lungs of mice undergoing secondary granuloma formation, Il25 was expressed at baseline levels during the secondary response (Fig. 4D). The increase was IL-4Rα–dependent as Il33 and Tslp expression diminished to baseline levels in IL-4Rα−/− mice. We hypothesize that the low alarmin expression in IL-4Rα−/− mice is due to decreased inflammation–driven injury in these mice.

Together, our studies with S. mansoni demonstrated that TSLP, IL-25, and IL-33 play redundant roles in the maintenance of chronic type 2 immunity. Targeting all three cytokines simultaneously from the initiation of primary or secondary granuloma formation reduced type 2 cytokine production but offered little protection from egg-induced pathology.

Efficacy of TSLP, IL-25, and ST2 blockade on established chronic allergy

With evidence that the combined blockade of TSLP, IL-25, and IL-33 signaling had a significant impact on type 2 cytokine expression, we hypothesized that the triple blockade might ameliorate type 2–mediated pathology in a different disease model. We sought to investigate a model that primarily targets epithelial cells, the predominant source of TSLP, IL-25, and IL-33, to determine whether the maintenance of type 2 immunity induced via mucosal epithelial injury was more dependent on the targeted cytokines. We chose to test the effects of administering single, double, and triple mAb blockades to a model of house dust mite (HDM)–induced allergic inflammation entering its chronic stage. Genes for all three alarmins are expressed at steady state in the lungs, and HDM induces expression of each of the alarmins with complementary kinetics (fig. S2). Il33 was up-regulated acutely and at chronic stages of allergic disease. Tslp was only up-regulated in the initial hours after first HDM exposure, and Il25 was up-regulated only at chronic time points. BALB/c mice were chronically challenged intranasally with HDM on days 0, 7, and 14 and then received eight additional doses spread over a total of 45 days. Beginning 3 weeks after the initiation of the allergic response, separate groups of HDM-treated mice were administered doses of anti-ST2, anti-TSLP, anti–IL-25 every 3 to 4 days in various combinations to achieve single, double, or triple blockades. Additional control groups received either saline or isotype control antibodies with or without HDM. On day 46, all mice were analyzed. As expected, in the lungs of isotype-treated control mice, chronic HDM exposure resulted in a marked increase in inflammatory cells in the lungs (Fig. 5A) and nearly a twofold increase in collagen content (Fig. 5B), confirming extensive lung remodeling and fibrosis. However, none of the single, double, or triple blockade combinations led to a significant decrease in inflammation or fibrosis in the sensitized mice. When the triple blockade mice were analyzed more closely, we also observed little to no change in the type 2 cytokine response in the lungs (Fig. 5C), and the total number of leukocytes in the bronchoalveolar lavage (BAL) and lungs appeared indistinguishable between the triple blockade and isotype control treated mice (Fig. 5D). We did, however, observe a significant decrease in the percentage of eosinophils in the lungs but not in the BAL (Fig. 5E).

Fig. 5. Efficacy of TSLP, IL-25, and ST2 mAb blockade on established chronic allergy.

WT BALB/c mice were sensitized and challenged intranasally (IN) with HDM, and starting on day 21, anti-ST2, anti-TSLP, and/or anti–IL-25 was administered in various combinations to different groups to achieve single, double, or triple blocks. Additional control groups received only isotype control antibodies with or without HDM. To properly control for the triple blockade group, groups administered single and double blocks also received immunoglobulin G1 (IgG1) in the absence of anti-ST2 or anti–IL-25, and rat IgG1 in the absence of anti-TSLP. All mice were analyzed on day 46. (A) Histopathology analysis of lung sections stained with Masson’s trichrome and scored for peribronchial and perivascular inflammation (n = 6 to 10 per group pooled from two experiments). (B) Quantification of fibrosis from lung tissue. (C) Quantification of gene expression from lung tissue measured by qPCR. (D) Quantification of leukocytes in the BALF and lung tissue. (E) Quantification of eosinophils shown as a percentage of total inflammatory cells in BALF and lung tissue. Student’s t test was used to measure all P values, and P > 0.05 except where reported.

Disrupting TSLP, IL-33, and IL-25 signaling during initiation and maintenance of type 2–driven chronic allergy

The failure of the triple blockade to protect against type 2–driven pathology when applied to established allergy further suggested that TSLP, IL-25, and IL-33 are not critical for the maintenance of chronic type 2–driven allergic lung inflammation. To test whether disrupting the signaling of the three cytokines during the initiation of type 2 cytokine–driven allergic lung inflammation provides a benefit, in a final series of experiments, IL-33/TSLP DKO mice were treated with anti–IL-25 during the entire course of chronic HDM exposure. Here, the deficient mice displayed marked and significant decreases in fibrosis when compared with control HDM mice on day 46 (Fig. 6A). Although peribronchial and perivascular inflammation in the lungs was similar in both groups, we observed a marked decrease in endarteritis and mucus staining in the lumen of the deficient mice (Fig. 6B). In addition, the total number of BAL cells (Fig. 6C) and the number of eosinophils in the BAL (Fig. 6D) and lungs (Fig. 6E) were reduced. The decrease in inflammatory eosinophils was also accompanied by a marked and highly significant reduction in IL-4, IL-5, and IL-13 production in the lungs (Fig. 6F), and IL-13 and IL-5 were also significantly decreased in the BAL (Fig. 6G). We observed similar results using anti-ST2, anti-TSLP, anti–IL-25 neutralizing antibodies in wild-type mice during the entire course of chronic HDM exposure (fig. S3).

Fig. 6. Disruption of all three mediators during initiation and maintenance of type 2–driven chronic allergy reduces inflammation and fibrosis.

WT C57BL/6 and IL-33/TSLP DKO mice were sensitized and challenged intranasally with HDM over 45 days. DKO mice were intraperitoneally administered αIL-25 (DKO + αIL-25/HDM), and HDM-treated WT C57BL/6 mice were intraperitoneally administered an IgG1 isotype control (isotype/HDM). A control group of C57BL/6 mice received intranasal saline instead of HDM and the isotype (isotype/saline). All mice were analyzed on day 46. (A) Quantification of fibrosis from lung tissue (isotype/saline, n = 5; isotype/HDM, n = 9; triple block/saline, n = 8). (B) Histopathology analysis of lung sections stained with Masson’s trichrome for scoring of inflammation and AB-PAS for mucus scoring. Micrographs are of representative lung sections stained with Masson’s trichrome. Scale bars, 50 μm. (C) Quantification of leukocytes in BALF. (D) BALF leukocyte differential. (E) Quantification of eosinophils in lung tissue. (F) Intracellular cytokine quantification of lung tissue lymphocytes by flow cytometry. (G) Intracellular cytokine quantification of BALF lymphocytes by flow cytometry. Student’s t test was used to measure all P values, and P > 0.05 except where reported. Data are representative of two independent experiments.

DISCUSSION

Although TSLP, IL-25, and IL-33 each has been identified as important initiators of type 2 immunity, their role in the maintenance of progressive type 2 disease was much less clear. Here, using chronic models of helminth infection and type 2 cytokine–driven lung inflammation, we found that trifunctional targeting of TSLP, IL-25, and IL-33 was more efficacious than blocking any of the mediators alone. This conclusion is strengthened because we made the observations using mice on both C57BL/6 and BALB/c backgrounds. In a schistosome lung granuloma model or during chronic S. mansoni infection in the liver, selective ablation of TSLP, IL-25, or IL-33/ST2 had little to no impact on the development of IL-4/IL-13–dependent inflammation or fibrosis. Nevertheless, we observed the modest albeit significant reductions in egg-induced inflammation in the liver when signaling of all three mediators was disrupted simultaneously. The reduction in inflammation in the schistosome infection model was also accompanied by a small yet significant decrease in the number of granuloma-associated eosinophils, a 25 to 30% decrease in hepatic fibrosis, and a significant reduction in the number of IL-13–producing ILC2s in the MLNs. The deficient mice also displayed reduced expression of IL-5 and IL-13 during primary and secondary granuloma formation in the lungs. Furthermore, when signaling of all three mediators was disrupted in a model of chronic HDM-induced allergic lung inflammation, inflammation, mucus production, and lung remodeling were decreased. Together, these studies revealed redundant roles for TSLP, IL-25, and IL-33 in the maintenance of these type 2–associated pathologies and suggest that aggressive trifunctional targeting of these mediators may more effectively ameliorate progressive type 2–driven disease.

Previous studies identified critical roles for TSLP, IL-25, and IL-33 in type 2 immunity to some helminth parasites (3643). However, most of these studies have focused on Nippostrongylus brasiliensis infection, in which expulsion of the nematode parasite is delayed or accelerated by relatively minor changes in type 2 immunity. Our initial studies focused on the schistosome lung granuloma and S. mansoni infection models because these models provide robust systems to dissect the role of TSLP, IL-25, and IL-33 during both the initiation and maintenance phases of type 2–driven inflammation (31). As reported previously with TSLP (26), we observed little to no role for IL-25 or IL-33 in IL-4/IL-13–dependent granuloma formation in the lungs. A recent study found modest decreases in acute inflammation in the absence of IL-25 (44), but in our studies, IL-25 or IL-33 deficiency alone had no discernable impact on the development of type 2 immunity or type 2–dependent pathology, even during the initiation of a primary granulomatous response. A similar outcome was observed in the liver after acute and chronic infection with S. mansoni, suggesting that TSLP, IL-25, and IL-33 were either not required or possibly playing redundant roles (26, 28, 38, 45). Although we found little to no role for TSLP, IL-25, or IL-33/ST2 when each mediator was ablated individually, we observed significant reductions in type 2 inflammation and fibrosis in the liver when all three mediators were targeted simultaneously, confirming their overlapping activities in response to significant damage during acute schistosomiasis. It is possible the degree of damage from parasites and other environmental triggers may impact the redundancy of the alarmins. Also, schistosome egg antigens have been identified that are capable of directly activating type 2 responses by modulating DC function (46, 47). Basophil- and autocrine T cell–derived IL-4 may also be sufficient to initiate and maintain type 2 responses (48, 49). Therefore, alarmins may not be critical to the activation or maintenance of all type 2 cytokine–driven inflammatory responses.

The type 2 response is a critical driver of wound repair pathways (1). However, when type 2 cytokine production persists or becomes dysregulated, it can lead to the development of pathological fibrosis (3). Consequently, because of their type 2–inducing activity, there has been a great deal of interest in understanding the roles of TSLP, IL-25, and IL-33 in progressive fibrosis, with numerous studies identifying increased production of these cytokines in various fibrotic diseases (5054). Many recent studies have shown that when overexpressed in mice, TSLP, IL-25, and IL-33 induce fibrosis in multiple tissues. For example, IL-25 was shown to promote lung remodeling in a model of HDM-induced allergic airway disease and indirectly induced pulmonary fibrosis by stimulating the production of IL-13 from ILC2s (44, 54). Transgenic overexpression of IL-33 has also been shown to promote IL-13–dependent cutaneous fibrosis (55), ILC2-mediated hepatic fibrosis (32), and bleomycin-induced pulmonary fibrosis in mice (56). Transgene-induced expression of TSLP has also been shown to induce pulmonary fibrosis in the lungs by up-regulating type 2 cytokine expression (52). Nevertheless, evidence that these epithelial-derived alarmins are critical to the development of TH2-associated fibrosis in a natural model of fibrosis was lacking before this study. Our studies with the schistosome lung granuloma and infection models show unequivocally that IL-13–dependent fibrosis can develop in the lungs and liver independently of TSLP, IL-25, and IL-33. We did, however, observe a significant decrease in fibrosis when all three mediators were targeted simultaneously, with the reduction in fibrosis associated with a significant decrease in IL-13–producing ILC2s. At more chronic time points after infection with S. mansoni, the early reduction in fibrosis and ILC2 activity appeared to be compensated for by an increased CD4+ T cell–derived IL-13 response, suggesting that TSLP, IL-25, IL-33, and ILC2s may not be critical to the maintenance of established and progressive fibrosis once the adaptive immune response has taken over. The relative involvement of an adaptive antigen-specific response may therefore be important in determining the relative contribution of these innate pathways to chronic disease. Regardless, these data further emphasize the potential benefit of early combinatorial targeting of TSLP, IL-25, and IL-33 in the treatment of type 2–driven disease.

Because epithelial cells are a major source of TSLP, IL-25, and IL-33 and schistosome eggs primarily damage the endothelium, it is possible that these cytokines are less important to the development of type 2 pathology in schistosomiasis. Therefore, in a final series of experiments, we used a chronic model of HDM-induced allergic lung inflammation to explore the combined roles of TSLP, IL-25, and IL-33 in a disease where the epithelium is the primary target. Here, in contrast to the lung granuloma studies, in which a mAb triple blockade administered from initial egg challenge had little impact on type 2 pathology, disrupting TSLP, IL-25, and IL-33 signaling from first allergen exposure had a significant suppressive effect on the development of fibrosis, endarteritis, and mucus deposition in the lumen. The number of inflammatory cells in the BAL was also reduced, as were the number of eosinophils in the BAL and lungs, with the reduction in eosinophils consistent with a recent study exploring the roles of TSLP, IL-25, and IL-33 in a model of chitin-induced lung inflammation (57). We also observed marked and highly significant reductions in IL-5 and IL-13 production in the lung and BAL fluid. When the combined mAb blockade of TSLP, IL-25, and IL-33 was applied to a model of established allergic lung inflammation, the marked protective effects were almost completely lost, however, suggesting either that TSLP, IL-33, and IL-25 are not required for the maintenance of an established antigen-specific type 2 response or that earlier intervention with TSLP, IL-33, and IL-25 antagonists is needed.

Although TSLP, IL-33, and IL-25 were all initially identified as critical drivers of type 2 immunity (10, 12, 50), several subsequent studies have illustrated that type 2 immunity can develop independently of these cytokines (2426, 28). The results from our experiments suggest that much of the data in the latter studies are likely explained by the overlapping activities of TSLP, IL-33, and IL-25. Our data also suggest that the three alarmins may be dispensable for the maintenance of type 2 immunity and chronic type 2–associated pathology because continued exposure to complex antigens like schistosome eggs or HDM allergen generates a potent and sustained adaptive CD4+ type 2 response that can supplant the requirement for alarmins and innate lymphocytes. A recent double-blind, placebo-controlled study of AMG 157, a neutralizing anti-human TSLP mAb, showed that TSLP blockade could reduce allergen-induced bronchoconstriction and eosinophilia (58). Whether targeting TSLP alone would show clinical benefit in moderate to severe asthma, however, could not be discerned from this small study tested on allergic individuals with near-normal baseline lung function.

Differences in perturbations of epithelium and other stromal cells may dictate the relative contribution of the three alarmins, and further studies with different animal models of allergy (for example, allergen dosing, variety, airway hypersensitivity) will be important before large-scale human studies are considered. The cost and time required for chronic models prevented us from testing all combinations of single, double, and triple blockades in every model. Notably, the triple blockade with mAbs from the start of allergic disease is effective, but its impact was not identical to congenital knockouts by all measures. Although all three antibodies were confirmed to exhibit highly effective neutralizing activity, it is possible that incomplete target coverage with the antibodies might in part explain these differences as well as the minimal efficacy of treating mice with established allergic disease. It is also possible that intracrine alarmin signaling such as IL-33–mediated activation of nuclear factor κB contributes to these small differences. In any case, antibody target coverage should be carefully evaluated in any future study in humans. Chronic human disease is likely maintained by a complex assortment of signals combined with sporadic exposure to specific antigen, and a better understanding of the hierarchy of these cues will help to clarify the relative contributions of TSLP, IL-33, and IL-25, as well as ILC2s. Our data suggest that a strategy that simultaneously suppresses more than one of these alarmins from the early phase of the disease may be required to effectively target type 2 cytokine–driven disease.

MATERIALS AND METHODS

Study design

Our primary objective was to investigate the effects of ablating IL-33, TSLP, and IL-25 signaling on chronic type 2 inflammation and fibrosis. To do this, we developed strategies to disrupt the signaling of the cytokines in mouse models of progressive type 2 immune-related pathology. No statistical methods were used to predetermine sample size. Group sample size was chosen using records of variance in past experiments, and variance is similar between groups being statistically compared. Samples or data points were excluded only in the case of a technical equipment or human error that caused a sample to be poorly controlled for. Mice or samples were randomly assigned to experimental groups or processing orders. Group allocation was blinded for all mouse work, when possible (such as administration of proteins, schistosomes, or allergens, sample quantification and analysis, and pathology scoring). The ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines in the EQUATOR (Enhancing the Quality and Transparency of Health Research) Network library were followed for this report.

Animals

The National Institute of Allergy and Infectious Diseases Division of Intramural Research Animal Care and Use Program, as part of the National Institutes of Health Intramural Research Program, approved all of the experimental procedures (protocol “LPD 16E”). The program complies with all applicable provisions of the Animal Welfare Act (www.aphis.usda.gov/animal_welfare/downloads/awa/awa.pdf) and other federal statutes and regulations relating to animals. IL-33−/− and IL-33/TSLP DKO mice on a C57BL/6 background were provided by Amgen Inc. C57BL/6, BALB/c, and IL-4Rα−/− mice were obtained from Taconic Farms Inc. IL-25−/− mice were obtained from Regeneron Pharmaceuticals Inc. Male and female mice between the ages of 6 and 12 weeks were used randomly to begin experimental models because of limited availability, and no sex-specific differences were observed. Groups in individual experiments were sex- and age-matched. All animals were housed under specific pathogen–free conditions at the National Institutes of Health in an American Association for the Accreditation of Laboratory Animal Care–approved facility.

Parasite infection

Mice were infected percutaneously via the tail with 35 cercariae from a Puerto Rican strain of S. mansoni (Naval Medical Research Institute) obtained from infected Biomphalaria glabrata snails (Biomedical Research Institute). Infection with 35 cercariae in wild-type mice leads to substantial disease and liver fibrosis but low mortality through the chronic phase of infection. Mice were perfused at the time of euthanasia to determine worm and tissue egg burdens as described previously (59).

Chronic HDM-induced allergy

Mice anesthetized with isoflurane were challenged intranasally with 200 μg of HDM in 30 μl of saline on days 0, 7, and 14 followed by eight additional 50-μg doses in 30 μl of saline spread over a total of 45 days. Lungs were harvested on day 46.

Schistosome egg–induced lung granuloma models

For the primary lung granuloma model, 5000 live S. mansoni eggs (Biomedical Research Institute) in saline were injected intravenously into mice on day 0. Lungs were harvested on day 7 for analysis. For the secondary lung granuloma model, 5000 S. mansoni eggs were also injected intraperitoneally on day 0. Mice were injected intravenously with 5000 live eggs containing mature embryos again on day 14 before lungs were harvested on day 21.

Hydrodynamic delivery of IL-25

Mice were injected intravenously with 10 μg of a mammalian expression plasmid coding for murine IL-25 in 2 ml of warm saline (60).

Triple block of IL-33, TSLP, and IL-25 with mAbs

Anti-mouse ST2 (61), anti-mouse TSLP (38), and anti-mouse IL-25 (62) mAbs were generated and selected by Amgen Inc. after extensive in vitro and in vivo testing. Previously unpublished tests for the efficacy of anti-TSLP included a bone marrow–derived DC bioassay measuring the inhibition of TSLP-induced CCL17/TARC (thymus and activation regulated chemokine) production and an assay measuring the inhibition of TSLP-induced proliferation of a pro–B cell line stably transduced with murine TSLP receptor. Neutralization of IL-33, TSLP, and IL-25 signaling was achieved by administering 250 μg of these antibodies, respectively, via intraperitoneal injection twice weekly. To properly control for the neutralizing antibodies, groups administered with single and double blocks also received 250 μg of mouse IgG1 in the absence of anti-ST2 or anti–IL-25, and 250 μg of rat IgG1 in the absence of anti-TSLP.

Histopathology

Liver or lung tissue was fixed in Bouin-Hollande solution, embedded in paraffin for sectioning, and stained (Histo-Path of America) with Wright’s Giemsa (S. mansoni models), hematoxylin and eosin, or Masson’s trichrome (allergy model) for analysis of inflammation, picrosirius red or Masson’s trichrome for fibrosis analysis, or Periodic acid–Schiff stain for analysis of mucus production. A scale of 1 to 4 (4 being the highest) was used for scoring. A blinded pathologist measured the diameter of about 30 granulomas and quantified the granulomatous eosinophils in Giemsa-stained sections of each sample with granulomatous pathology. Images were scanned with an Aperio ScanScope (Leica Biosystems).

Fibrosis assay

Hydroxyproline was measured as a surrogate for collagen content. A known weight of liver or lung tissue was hydrolyzed in 6 N HCl at 110°C for 18 hours and then neutralized in 10 N NaOH before colorization. A standard curve composed of dilutions of 1 mM hydroxyproline (Sigma-Aldrich) (63).

Leukocyte isolation for intracellular cytokine staining and eosinophil identification

About 200 mg of lung or liver tissue was ground into a single-cell suspension through a 100-μm nylon mesh. Leukocytes were separated on a 40% Percoll (Sigma-Aldrich) gradient (2000 rpm for 15 min) and treated for 2 min with 1 ml of ACK (ammonium chloride–potassium bicarbonate) lysis buffer to lyse erythrocytes. After 3 hours of stimulation with phorbol 12-myristate 13-acetate (10 ng/ml), ionomycin (1 μg/ml), and brefeldin A (10 μg/ml), leukocytes were fixed and permeabilized for 30 min (Cytofix/Cytoperm buffer, BD Biosciences) and then stained for 30 min with antibodies for CD4 (clone RM4-5, eBioscience), inteferon-γ (XMG1.2, eBioscience), IL-4 (11B11, eBioscience), IL-5 (TRFK5, BD Pharmingen), and IL-13 (eBio13A, eBioscience) diluted in the Perm/Wash buffer (BD Biosciences). Unstimulated lung leukocyte aliquots were set aside and stained for 30 min with anti–SiglecF (sialic acid–binding immunoglobulin-like lectin F). Positive SiglecF staining and scatter profiling were used to identify eosinophils by flow cytometry. Leukocytes collected from BAL were isolated with ACK lysis buffer, stimulated, fixed, permeabilized, and stained as above. Expression of CD4, SiglecF, and the intracellular cytokines was analyzed with a BD FACSCanto II flow cytometer and FlowJo v.7.6 software (Tree Star).

Leukocyte isolation from liver and MLN for ILC2 staining

Liver or lymph node tissue was ground into a single-cell suspension through a 100-μm nylon mesh, and hepatic leukocytes required further separation using a 40% Percoll gradient and ACK lysis as described above. Leukocyte samples from both tissues were stimulated, fixed, and permeabilized as described above. Then they were stained for 30 min with antibodies for CD16/32 (clone 2.4G2, BD Biosciences), CD4 (RM4-5, eBioscience), IL-13 (eBio13A, eBioscience), ST2 (DJ8, MD Biosciences), and ICOS (C398.4A, BioLegend) diluted in Perm/Wash buffer (BD Biosciences). Expression of the surface markers and intracellular IL-13 was analyzed with a BD FACSCanto II flow cytometer and FlowJo v.7.6 software (Tree Star).

RNA isolation and real-time qPCR

Lung or liver tissue was homogenized in TRIzol Reagent (Life Technologies) using Precellys 24 (Bertin Technologies). Total RNA was extracted from the homogenate by addition of chloroform followed by the recommendations of the MagMAX-96 Total RNA Isolation Kit (Life Technologies). RNA was then reverse-transcribed using SuperScript II Reverse Transcriptase (Life Technologies). Real-time reverse transcription PCR was performed on an ABI PRISM 7900HT Sequence Detection System (Applied Biosystems). Quantities of mRNA expressed by a particular gene were determined using Power SYBR Green PCR Master Mix (Applied Biosystems), normalized to ribosomal protein, large, P2 mRNA levels in each sample, and then articulated as a relative increase or decrease compared with mRNA levels expressed by the same gene in naïve controls. Primers were designed using Primer Express Software (version 2.0, Applied Biosystems). Forward and reverse primer sequences are listed in table S1.

BAL, cell differential determination, and enzyme-linked immunosorbent assay

Ice-cold PBS (1 ml) supplemented with 5 mM EDTA was injected through the trachea into the lungs and aspirated using a syringe. Cells (~1 × 105) were spun for 5 min with a Shandon Cytospin 3 centrifuge (Thermo Scientific) onto a slide before being fixed with methanol and stained with Diff-Quik (Boehringer) to identify leukocyte cell types. Levels of IL-4, IL-5, and IL-13 in the undiluted BAL were quantified using a Luminex-based multiplex assay according to the manufacturer’s protocol (EMD Millipore).

Statistical analysis

All data were analyzed with Prism (version 5; GraphPad). Data sets were compared with a two-tailed t test, and differences were considered significant if P values were less than 0.05. A Welch’s correction was used when an F test comparing variances had a P value of less than 0.05.

SUPPLEMENTARY MATERIALS

www.sciencetranslationalmedicine.org/cgi/content/full/8/337/337ra65/DC1

Fig. S1. Alarmin gene expression in the liver.

Fig. S2. Kinetics of alarmin gene expression in chronic HDM model.

Fig. S3. Neutralizing all three alarmins with mAbs during initiation and maintenance of type 2–driven allergy reduces inflammation and fibrosis.

Table S1. qPCR primer sequences.

Source data

REFERENCES AND NOTES

Acknowledgments: We thank the NIH animal facilities for their conscientious care of mice, and Amgen Inc. and Regeneron Pharmaceuticals Inc. for providing mice. Funding: This research was supported by the Intramural Research Program of NIH, National Institute of Allergy and Infectious Diseases. Author contributions: K.M.V., T.R.R., M.R.C., D.E.S., and T.A.W. conceived and designed the experiments; K.M.V., T.R.R., A.W.C., L.A.B., K.M.H., R.W.T., and S.W. performed the experiments; K.M.V., T.R.R., L.A.B., L.B., K.M.H., K.N.K., M.R.C., D.E.S., and T.A.W. analyzed the data; K.M.H. analyzed the statistics; A.W.C., A.L.B., M.R.C., and D.E.S. contributed reagents/materials/analysis tools; and K.M.V. and T.A.W. wrote the paper. Competing interests: A.L.B., M.R.C., and D.E.S. work for a for-profit company. The other authors declare that they have no competing interests. Data and materials availability: Genes of interest can be accessed in NCBI’s (National Center for Biotechnology Information’s)GenBank with the following codes: Rplp2: NM_026020, Il4: NM_021283, Il5: NM_010558, Il13: NM_008355, Ifnγ: NM_008337, Il25: NM_080729, Il13: NM_008355, Ccl5: NM_013653, Ccl11: NM_011330, Il33: NM_001164724, Tslp: NM_021367.
View Abstract

Navigate This Article