CCL17 Controls Mast Cells for the Defense against Filarial Larval Entry

Mast cells (MCs) are in close proximity to epithelial surfaces such as in the skin, respiratory and gastrointestinal systems, allowing prompt interaction with environmental agents. As a result of the observation that parasites can induce strong IgE-mediated MC activation, they have been implicated in the defense against parasitic infections, which is mediated by adaptive immune responses. However, the precise role of MCs in vivo is yet largely unclear. Increased numbers of mucosal MCs are often observed in tissues affected by gastrointestinal helminth infections. Indeed, functional MCs are required for the expulsion of Trichinella spiralis (1, 2), whereas defense against Heligmosomoides polygyrus (3) and Nippostrongylus brasiliensis (4) does not appear to be dependent on MCs (5). Different to parasites residing in the gut lumen, the function of MCs in response to tissue-dwelling filarial nematodes in primary infection has not been analyzed. Onchocerca volvulus, the filarial species that causes river blindness, affecting 37 million people in Africa, resides in s.c. nodules in humans and attracts MCs in high numbers. As opposed to live filariae within the onchocercomata, dead filariae induced significantly enhanced MC enrichment (6). The rare hyperreactive clinical manifestation of onchocerciasis (sowda), characterized by strong immune responses and low number of established worms, clearly correlates with strong mastocytosis (7). Therefore, in human onchocerciasis, the localization and frequency of MCs are inversely correlated to vitality and productivity of the worms and thus on the release of O. volvulus Ags. However, it has not yet been investigated whether MCs take part in the early defense against filariae, determining the fate of the parasite after entering its mammalian host. In the murine model of filariasis with Litomosoides sigmodontis infection, it is possible to follow the infection route from larval entry to full establishment. In this model, it has been shown that the extent of susceptibility is determined during the first days, with only a proportion of infectious L. sigmodontis stage 3 (L3) surviving and developing into adults (8). The worm load, which is maximal by day 10, stays constant for the subsequent 20 d (8). Innate mechanisms must therefore exist that allow only some of the incoming larvae to migrate and develop into adults.

Members of the chemokine superfamily are known to be involved in innate immunity. This also applies to the thymus and activation-regulated chemokine (TARC/CCL17), which is produced by dendritic cells (DCs) upon microbial challenge (9). It has previously been observed that CCL17 expression is high in human PBMC from filaria-infected individuals (10). However, localization and function of CCL17 in filarial infections remain unclear.

Using L. sigmodontis infection, we report that deficiency of CCL17 results in a higher worm load, accompanied by a higher number of skin MCs. MC function is essential for local vascular permeability, and blockade of MC granule release or histamine signaling prevents the more effective establishment of worms in the host seen in CCL17-deficient mice. A unique feature of many filarial species is the mutualistic symbiosis with Wolbachia endobacteria. These endobacteria can activate the TLR system (11–13) mainly through TLR2 and induce innate inflammatory responses. CCL17-mediated MC numbers and vascular permeability were indeed dependent on the presence of Wolbachia. Our study provides evidence for an essential role for CCL17 during innate immune response against infective L3 larvae to control MC-derived vascular permeability and thereby helminth infection.

Six- to 8-wk-old female C3H/HeN and background-matched CCL17-deficient (E/E) and heterozygous (E/+) mice (9) and TLR2^−/− (14) mice, were maintained under specific pathogen-free conditions, according to animal welfare guidelines. “E” denotes replacement of CCL17 by enhanced GFP (eGFP). As indicated, mice were daily treated i.p. (8 mg/kg body weight) with the MC stabilizer disodium cromoglycate (cromolyn) dissolved in 0.9% NaCl (Sigma-Aldrich) starting 6 d before infection until mice were sacrificed. In vivo treatments used anti-CCL17 Ab (MAB529, administered i.p. at 20 μg/mouse on days −1, 1, 3, and 5; R&D Systems). Anti–IL-10R Ab (clone 1B1.3a, provided by Dr. M. D. Taylor, University of Edinburgh) and anti-CD25 (clone PC61) were administered i.p at a concentration of 1 mg on days −1 and 3. For inhibition of histamine signaling, 400 μg pyrilamine (HR1 antagonist) and 800 μg cimetidine (HR2 antagonist) were administered 1 h prior to and 2 h postinfection (p.i.) as well as once per day for 5 consecutive days, as described previously (15). At least five mice were used for each group.

Bone marrow cells were removed from femurs and tibias of CCL17^E/+ mice by flushing the bones with PBS and filtered through a nylon mesh. Cells were centrifuged and 5 × 10⁵/ml were plated out in RPMI 1640 medium containing 10% FCS, 1% l-glutamine, and 50 μg/ml gentamicin, supplemented with 20% supernatant of GM-CSF–transfected X63Ag8 cells. Medium was replaced 3 d after the start of the culture. At day 8, CD11c⁺ cells were isolated by magnetic bead separation with the MACS system (Miltenyi Biotec). The purity of DCs was typically >95%. Cells were washed and plated at a density of 2.5 × 10⁵/ well and stimulated with the respective Ags for 24 h.

Infective L3 larvae were transmitted through the bite of the vector mite Ornithonyssus bacoti (16), termed “natural infection” in the following sections. For a “localized” infection mites were collected in the lid of a 1.8-ml Eppendorf tube, which was fixed to the skin surface. On average, 10 infected mites could be observed to take their blood meal on a circle with a diameter of 0.5 cm. L3 without Wolbachia were harvested from mites that had acquired a blood meal from jirds treated with doxycycline for 4 wk, with the consequence that microfilariae, which the mites picked up were Wolbachia deficient (16). Parasites were harvested at 10, 20, and 30 d p.i. from the thoracic cavity and counted as described previously (17). For Ag preparation, worms were rinsed in sterile PBS before being mechanically minced. Insoluble material was removed by centrifugation of 300 × g for 10 min at 4°C. Protein concentrations of crude extracts were determined using the Advanced Protein Assay (Cytoskeleton). All procedures were conducted under sterile conditions.

Skin tissues were homogenized in TRIzol reagent 2 × 30 s at 6800 rpm in a Precellys 24 (PeqLab) using 2.8-mm steel beads. RNA was extracted with TRIzol reagent (Invitrogen). Genomic DNA was removed from RNA samples by DNase digestion. cDNA was generated using Omniscript-RT kit (Qiagen), following the manufacturer's protocol. Expression levels of IL-4, IL-5, IL-10, IL-12p35, IL-17A, CCL5, and thymic stromal lymphopoietin were quantified by real-time PCR using a Rotor-Gene RG-3000 (Corbett Research). Real-time PCR was performed in 20 μl consisting of 2 μl 10× PCR buffer, 0.2 mM 2′-deoxynucleoside 5′-triphosphates, 0.2 μl Sybr Green I (1/1000 in DMSO), 0.5 U Hotstar Taq polymerase (Qiagen), and 2 μl cDNA template. Primer sequences are given in Supplemental Table I. Reaction conditions were 15 min at 95°C, followed by 45 cycles of 15 s at 94°C, 20 s at 58°C, and 20 s at 72°C. Specific mRNA expression levels were normalized to β-actin mRNA expression levels.

Peripheral and mediastinal lymph nodes, lung, and skin of the lumbar area of mice were prepared for confocal microscopy of eGFP expression as described previously (9). Cryostat sections were analyzed at given time points after natural infection by confocal microscopy (×400) for eGFP expression of cells. For MC evaluation, the skin at the site of L3 inoculation was isolated 6 h p.i., and formalin-fixed paraffin sections were stained with toluidine blue. The total number of mast cells in the epidermis, dermis, and hypodermis was counted on 20 successive, complete fields of a light microscope (×1000) per section and were averaged from 10 sections of each skin sample, which were separate from each other by five 4-μm sections that were not analyzed. Mast cells were classified as non-degranulated or degranulating (18). The examiner (J.K.F.) was blinded with regard to the nature of the mouse groups.

Thirty-five L3 or 10 μg soluble L. sigmodontis adult worm Ag (LsAg) or LsAg depleted of Wolbachia (16) were inoculated s.c. into the lumbar area of the mice or localized natural infection was performed. After 6 h, Evans blue (30 mg/kg body weight; Sigma-Aldrich) was injected into the tail vein of the mice (19). After 1 h, mice were sacrificed, and the skin at the site of inoculation was removed and incubated in 3 ml formamide (Sigma-Aldrich) at 60°C for 36 h. Vascular leakage into the tissue was determined by quantifying the amount of skin-extracted Evans blue at OD 620 nm, in relation to the weight of the extracted tissue and an Evans blue standard curve.

The following mAbs against mouse Ags were used as PE or allophycocyanin conjugates for analysis of BMDCs or cell composition in the thoracic cavity: CD40 (3/23), CD80 (16-10A1), CD4 (RM4), CD8a (53-6.7), CD25 (PC61), CD45R/B220 (RA3-6B2), CD11b (M1/70), or CD11c (HL3). Surface staining with mAb, acquisition, and analyses were performed as described previously (20, 21). For eGFP expression of reporter mice, organs were collagenase/DNaseI digested and CD11c⁺ cells sorted using MACS (Miltenyi Biotec). For surface staining PE-conjugated MHC class II (M5/114.15.2) and PE-Cy7–conjugated CD11c (HL3) Abs were used. Flow cytometric analysis was performed using FACSCanto running Diva software (BD Biosciences). Cytokines in the thoracic cavity fluid were quantified by ELISA, using standard protocols (DuoSet; R&D Systems).

Comparison between the groups of single experiments in case of parametrical data was done using the Student t test or ANOVA and Tukey’s multiple comparison tests. In case of nonparametrical data, Kruskal–Wallis or Mann–Whitney U test was used. Data in all graphs are presented as mean ± SD or median and range. Statistics were performed with GraphPad Prism 5 software or SPSS. Significant values are indicated as follows: *p < 0.05, **p < 0.01, and ***p < 0.001.

DCs are among the first immune cells that filarial parasites encounter when entering the skin. Their key role is to ingest, sense, and process Ag to efficiently activate adaptive immunity, but also to recruit immune cells to the site of Ag encounter (e.g., by secretion of chemokines). We found that after stimulation with filarial extract (LsAg), bone marrow-derived DCs (BMDCs) upregulated the activation markers CD40 and CD80 compared with unstimulated controls and secreted significant amounts of IL-6 and CCL17 upon stimulation with LsAg (Fig. 1).

To investigate the in vivo expression of CCL17, mice were infected with the rodent filaria L. sigmodontis, an established model for filarial diseases (8, 17, 22). Using a fluorescence (eGFP)-based in vivo reporter system (9), in which an eGFP cassette was inserted into the endogenous murine ccl17 locus, we show that CCL17 is expressed in murine skin (Fig. 2A) as early as 6 h post filarial infection, but not in naive mice. Also, in the draining lymph node and lung CCL17 expression was increased over time. These eGFP-producing cells have been previously identified as CD11c⁺ DCs (9). Consistently, quantification of CCL17⁺CD11c⁺ cells revealed their presence at higher percentages in the draining lymph nodes at 6 h p.i. as compared with naive mice, which peaked at day 5 p.i. (Fig. 2B). Lungs contained these DCs at enhanced numbers at 1 d p.i. compared with the uninfected stage or 6 h p.i. (Fig. 2C). These organs are along the migratory pathway of infective larvae, their final destination being the thoracic cavity. CCL17⁺ DCs were not present in the spleen (data not shown), according to the lack of CCL17 expression in the spleen upon bacterial challenge (9).

To analyze the relevance of CCL17 as functional player in filariasis, we observed parasitological and immunological consequences during the course of a natural infection with L. sigmodontis in CCL17^E/+ heterozygous and CCL17^E/E-deficient mice (“E” denotes replacement of CCL17 by eGFP). At day 10 p.i., when the filariae have reached the thoracic cavity, an up to 4 times higher parasite load in CCL17^E/E mice compared with CCL17^E/+ mice was observed (Fig. 3A, Supplemental Fig. 1). This difference was maintained throughout day 30 p.i., being in line with several earlier observations of L. sigmodontis infection that the worm levels, which are established by day 10, do not change over the next 20 d (8). Furthermore, anti-CCL17 Ab treatment in heterozygous mice was reminiscent of CCL17-deficient mouse phenotype. Anti-CCL17 treatment of CCL17^E/+ mice led to increased parasite numbers 10 d p.i. equivalent to CCL17^E/E mice (Fig. 3B), suggesting that the phenotype is due to the absence of CCL17 rather than indirect ontogenetic side effects of the genetic knockout. Our results indicate early CCL17-dependent mechanisms being responsible for a significant proportion of host defense against incoming larvae. Homozygous (+/+) and heterozygous (E/+) control mice did not differ with regard to parasite loads (Fig. 4D).

To investigate whether there are different filaria-driven immune responses in the presence or absence of CCL17, the cellular composition and cytokine responses in the thoracic cavity were analyzed at day 10, 20, and 30 p.i. Of the cell types analyzed, no major differences were observed, except for slight but significant elevation of CD4⁺CD25⁺ T cells. Numbers of other cells such as CD4⁺ and CD8⁺ T cells and DCs were unaltered in both groups (Supplemental Table II). Analysis of concentrations of the cytokines IL-5, IL-10, IL-12, and IFN-γ in the thoracic cavity fluid did not reveal Th1/Th2 imbalance, except a small increase in IL-10 (Supplemental Table III). In the mediastinal lymph nodes, there was no alteration in the cellular composition between CCL17-deficient and control mice at day 10 p.i. (Supplemental Table IV). Upon Ag specific restimulation of mediastinal lymph node cells, cytokine production was at the limit of detection and did not differ between CCL17-deficient and control mice (data not shown).

To unravel a potential role of IL-10 or CD25⁺ cells at this stage of infection, we administered neutralizing or blocking Abs to CD25 and IL-10 receptor in vivo. However, in vivo the parasitological outcome in CCL17^E/E mice was not altered (Supplemental Fig. 3A, 3B).

Because CCL17 expression was observed as early as 6 h p.i., we also analyzed gene expression in the skin by quantitative PCR. IL-12p35 and CCL5 were upregulated already 6 h after transmission of the larvae irrespective of the genetic status of the mouse (Fig. 3C, 3D), however, key components of adaptive immunity, such as IL-4, IL-5, IL-10, IL-17A, and thymic stromal lymphopoietin could not be detected (data not shown).

Within the repertoire of host cells remaining in proximity to epithelial surfaces, such as the skin, MCs are particularly frequent. However, the role of MCs in primary filarial infection is unclear. We analyzed the number and status of skin MCs in CCL17^E/E mice and heterozygous littermates after either natural infection with L3-containing mites or a mock infection using mites without L3, the latter being used to control for the vector bite itself. MCs were grouped as either nondegranulating or degranulating (18). As shown in Fig. 5A, a slight increase in the number of total skin MCs could be observed following exposure of mice to mites without L3 in both CCL17^E/E and CCL17^E/+ mice. In contrast, a significant increase was observed after exposure to L3-containing mites in CCL17^E/E mice but not in heterozygous controls. The number of degranulating MCs was also significantly higher in CCL17^E/E mice after exposure to L3-containing mites (Fig. 5C). This difference was more profound after a localized natural infection where mites were allowed to bite only on a circular field with a diameter of 0.5 cm (Fig. 6C, left pair of bars, see also below).

Equivalent results were also observed in heterozygous mice treated with anti-CCL17 Abs. Ab treatment resulted in increased numbers of total MCs (Fig. 5D) and degranulating MCs (Fig. 5F) similar to those seen in CCL17-deficient mice after natural challenge with infective larvae (Fig. 5C).

MCs reside in close association with blood vessels and nerves at the sites that are most likely to be exposed to pathogens. Many of the mediators that are released by MCs cause alterations in selective cellular recruitment and vascular function (5).

We investigated whether deficiency of CCL17 could modify the permeability of endothelial cells. After s.c. injection of medium or LsAg (Fig. 7A), mice were inoculated with Evans blue and extravasation into the surrounding tissue was measured (19). Aside from the expected change in OD when injecting Evans blue compared with noninjected mice, we found that vascular permeability was significantly enhanced in CCL17^E/E as well as in anti–CCL17-treated CCL17^E/+ mice compared with untreated CCL17^E/+ controls. Changes in vascular permeability were also induced in CCL17-deficient mice by infective larvae, either when injected (Fig. 7B) or when transmitted by their natural vector in a localized infection (Fig. 7C).

These data suggest a new mechanism whereby L3 induces vascular permeability, probably to facilitate their migration by entering the lumen of vessels situated along this pathway.

To prove that the increased parasite load observed in CCL17^E/E mice is essentially dependent on the increased number of degranulating MCs and the ensuing elevated vascular permeability either in the skin or at other sites along the migratory pathway of L3 toward the pleural space, we inhibited MC degranulation by injection of disodium cromoglycate (cromolyn). It is hypothesized that cromolyn induces conformational changes of moesin that is involved in the regulation of a functional association between the cell surface and the cytoskeleton in MCs, thereby inhibiting the secretion of granules (23).

Fig. 4A shows that cromolyn treatment reduced elevated numbers of degranulating MCs in CCL17-deficient CCL17^E/E littermates. In accordance to their role as inducers of vascular permeability, CCL17^E/E skin tissues showed reduced uptake of Evans blue following cromolyn treatment (Fig. 4B), arguing that inhibition of MC degranulation by cromolyn reduced vascular permeability to the level in CCL17^E/+ mice. Even more important, cromolyn treated CCL17- deficient mice no longer displayed increased numbers of parasites in the thoracic cavity (Fig. 4C), thus showing reversion to the phenotype of heterozygous CCL17^E/+ mice. Treatment of CCL17^E/+ mice with cromolyn did not alter parasite loads (Supplemental Fig. 2A).

To further underscore the role of MCs for parasite establishment in CCL17-deficient mice, we blocked the activity of histamine, the main MC product released during degranulation, in vivo. Administration of antagonists for histamine receptors 1 (pyrilamine) and 2 (cimetidine) reverted the increased parasite loads of CCL17-deficient mice to the levels of both homozygous (+/+) and heterozygous (E/+) control mice (Fig. 4D, Supplemental Fig. 2B).

These data conclusively suggest that the increase in MCs in CCL17-deficient mice is of functional relevance for vascular permeability and parasite establishment.

Most filarial nematode species have the unique feature that they contain endosymbiotic bacteria of the genus Wolbachia, which are essential for development of L3 into adult worms and for fertility and survival of the adults (11, 24, 25). Therefore, we investigated whether Wolbachia endosymbionts are involved in MC activation and consequently increased vascular permeability. First, we injected LsAg or LsAg devoid of Wolbachia into the skin, followed by injection of Evans blue. Only the extract containing Wolbachia Ags was able to increase vascular permeability in CCL17^E/E mice (Fig. 6A). Next, it was of interest whether this would also hold true for infective L3. Therefore, the effect of normal, Wolbachia-containing L3 on the induction of MC degranulation was compared with that of L3 generated in vectors that had ingested MF from animals treated with the Wolbachia-depleting drug doxycycline, as described previously (Ref. 16, see also 1Materials and Methods). Infective L3 were either injected (Fig. 6B) or transferred by mites in a localized infection (Fig. 6C). These data reveal that in CCL17^E/E mice, the number of degranulating MCs was reduced upon challenge with Wolbachia-depleted L3 in comparison with Wolbachia-containing L3, whereas in heterozygous mice, MC degranulation was not significantly affected by the presence or absence of Wolbachia (compare the white bars in Fig. 6B, 6C). Taken together, these results indicate that Wolbachia are involved in MC degranulation and increased vascular permeability.

Because several publications could link activation of the immune system by Wolbachia with TLR2 (12, 13, 26), we set out to test whether the observed dependency on Wolbachia for the induction of vascular permeability was due to interaction with TLR2. Therefore, we compared the effects of anti-CCL17 Ab treatment on CCL17^E/+ mice, where this treatment results in elevated parasite loads (Figs. 3B, 6F) and increased vascular permeability (Fig. 6D), to TLR2-deficient mice on the same background (C₃H).

After injection of LsAg followed by Evans blue, we found that only genetically deficient CCL17^E/E and anti–CCL17-treated CCL17^E/+ mice displayed increased vascular permeability. In contrast, anti–CCL17-treated TLR2^−/− mice did not show permeability changes (Fig. 6D). The number of degranulating MCs was not increased in these mice after natural infection with L3 by the insect vector (Fig. 6E). Importantly, these results were associated with the parasitological outcome: as expected, CCL17^E/E and anti–CCL17-treated CCL17^E/+ mice contained significantly higher numbers of parasites 10 d p.i. However, TLR2^−/− mice when given anti-CCL17 Abs, did not display elevated parasite loads (Fig. 6F), indicating that signaling via TLR2 is indispensible for the increased MC function after CCL17 blockade. These data show that control of vascular permeability and parasite loads by CCL17 acts in a TLR2 dependent manner.

Early defense mechanisms against entry of filarial parasites are poorly documented. In particular, the role of chemokines has not been addressed. The data presented in this study show that CCL17 is involved in early immune responses limiting filarial parasite invasion into the host. Genetic deficiency of CCL17 or, equivalently, administration of anti-CCL17 Abs, results in MC accumulation, their degranulation, and increased vascular permeability. This mechanism is dependent on the presence of Wolbachia endosymbionts in infective L3 larvae and promoted by TLR2-dependent signaling. To our knowledge, they link CCL17, MCs, histamine, and control of filarial infection for the first time.

It is known that L3, after transmission into the skin of mice, migrate via the lymphatics (27–29) into the pleural space. Accordingly, CCL17 expression was detected not only upon artificial stimulation of BMDCs but also in CD11⁺ DCs in the skin, in the draining lymph nodes and the lung after natural filarial infection. Genetic deficiency or Ab depletion of CCL17 resulted in increased parasite loads, indicating a major role for CCL17 in parasite containment.

CCL17 is known to be produced by DCs to recruit activated and memory T cells to the site of infection. However, we did not observe differences in effector cell composition at day 10 p.i. neither in the thoracic cavity nor in the draining lymph node. This also applied for cytokine production, except for slight differences in CD4⁺CD25⁺ T numbers and IL-10 production in the thoracic cavity. In murine filariasis, regulatory T cells expand in numbers at the site of infection and are observed in the skin draining lymph nodes or coelomic cavities as early as day 7 p.i. (21, 30). However, in CCL17^E/E mice, neither IL-10 neutralization nor depletion of CD25⁺ T cells prior to infection altered the parasitological outcome, in accordance with previous results showing that anti-CD25 Abs altered the parasite loads only later during infection, but did not affect initial larval establishment (21). Together with the findings that neither cellular composition nor production of Th1/Th2 cytokines differed between CCL17^E/+ and CCL17^E/E mice, our data suggest that induction of effector or regulatory cells is not the major mode of action of CCL17 in early murine filarial infections.

This is also in accordance with earlier observations that differences in parasite survival were obvious already at day 10, and it has been shown that the proportion of worms that establish themselves in the host following the entry of infective L3 is determined during the first days [i.e., at times before adaptive immune responses can control the immune reaction (29, 31, 32)].

MCs are frequently located in proximity to epithelial surfaces, such as the skin or the lung, ideally placed to carry out their function in host defense. After natural transmission of L3 by the vector, but also after injection of parasite extract (LsAg), we detected MCs at higher numbers per se and among those higher proportions of degranulating MCs in the skin of CCL17^E/E mice. Because bites from mites that did not carry L3 failed to induce MC accumulation and degranulation, this rules out that vector saliva plays a major role in the activation of MCs. A similar inhibitory effect of CCL17 on human cord blood MC migration has been reported: CCL5-induced migration of human cord blood-derived MCs was suppressed by CCL17 (33). Conversely, in transgenic mice overexpressing CCL17, lower numbers of MCs as well as of lymphocytes and neutrophils were found in the skin of mice after a single challenge with oxazolone (34). Considering CCL5 as a possible chemoattractant for MCs, CCL5 gene expression was equivalently elevated in both CCL17^E/+ and CCL17^E/E mice. This suggests that CCL17 restricts CCL5-induced MC accumulation in the skin following exposure to L3, thus leading to the elevated numbers of MCs and facilitation of degranulation in the absence of CCL17. These data add to the recognition of CCL17 not only as chemoattractant but also as limiting factor for cell migration to maintain the homeostasis at pathogen-exposed interfaces.

The influx of MCs detected by histology appears to be rather minor; however, this could be due to the failure to detect MCs that are already fully degranulated, or to the nature of a natural infection. We have established a way to localize the bite of mites to a limited area of the skin, which resulted in a higher ratio of degranulating MCs between infected CCL17-deficient and control mice in a localized compared with a dispersed natural infection (2.9 versus 1.6; compare Figs. 6C and 5C). The parasite infection differs considerably from skin inflammation models in that the parasite trespasses the skin within one day following infection, trying to avoid skin irritation, whereas in other models inflammatory triggers are administered over days or weeks (35, 36).

In addition to MCs located in the skin, these cells could also play an important role for the facilitation of larval migration at other sites (e.g., lymphatics or lungs), which the parasites encounter on their route toward the pleural space, where the juvenile and adult worms reside. To address the function of those MC comprehensively, we carried out functional blockade of MCs and the histamine signaling pathway.

As a consequence of recruitment and activation, degranulating MCs release a magnitude of substances, among them many molecules that are involved in vascular permeability, such as histamine and granzymes. To identify MC function in the context of early filarial infection, we pharmacologically blocked MC function by administration of the MC stabilizator cromolyn. Indeed, cromolyn treatment reversed all phenotypes investigated (i.e., MC degranulation, vascular permeability, and parasite loads). MC inhibition had no effect on parasite recovery in CCL17^E/+ mice, suggesting that there is a threshold of degranulation below which filarial recovery is not modified (Supplemental Fig. 2).

Additional studies using histamine receptor antagonists in vivo point toward histamine, the main vasodilatory molecule secreted by MCs, because inhibition of the mediator of the effects seen in CCL17-deficient mice, was able to revert the parasitological phenotype (Fig. 4D). Vascular permeability is a common feature of inflammation, which is associated with leukocyte extravasation, increased interstitial fluid, and consequently, increased lymphatic drainage; this elevated flow into the lymphatic vessels may explain the greater entry of larvae inside the lymphatic vessels.

Prototypic MC activation is induced by cross-linking of FcR-bound IgE. Recent studies show that activation of murine MCs by TLR2 or TLR4 ligands acts synergistically with IgE-mediated activation, because administration of TLR ligands can augment airway sensitization in mice. It has been suggested that this process involves TLR-dependent activation of DCs (reviewed in Ref. 37). It is however unclear, whether degranulation itself can be induced via TLR directly. Some reports showed that TLR2 ligands are able to induce degranulation directly (38, 39), but a clear mechanism cannot be defined yet, because such responses have not been observed consistently in all studies (40, 41). At this point it should be noted that MCs exhibit some variation in phenotype according to anatomical location, species and culture conditions.

Wolbachia endosymbionts present in most filarial species (42) and also in L. sigmodontis provide ligands for TLR2 (and TLR6). We show that vascular permeability after injection of Ag and MC numbers after transmission of L3 via the natural vector is markedly reduced in the absence of Wolbachia in a TLR2 dependent manner.

Our data demonstrate that CCL17 inhibits filaria-induced MC activation/function in genetically modified or anti–CCL17-treated, mice. In particular, higher numbers of MCs in CCL17-deficient mice were accompanied by more degranulating MCs. This was associated with increased vascular permeability in CCL17-deficient mice. Both MCs and vascular permeability could be causatively linked to the parasitological phenotype seen in CCL17-deficient mice, as inhibition of MC degranulation by cromolyn or administration of histamine-receptor antagonists reverted worm establishment. These observations were monitored within the first barrier that the parasite has to encounter, suggesting the skin to be a site for MC action. However, the parasite will also trespass other organs such as the lung on its way to the pleural space. This was not differentiated in this study and constitutes interesting further studies.

In conclusion, this study suggests a new molecular mechanism for filarial worm migration and establishment by MC-mediated increased vascular permeability. This is counterregulated by CCL17 expression of the host, implicating an important role for CCL17 in early responses against a skin-invading pathogen.

The authors have no financial conflicts of interest.