Paracoccidioides brasiliensis causes paracoccidioidomycosis, one of the most prevalent systemic mycosis in Latin America. Thus, understanding the characteristics of the protective immune response to P. brasiliensis is of interest, as it may reveal targets for disease control. The initiation of the immune response relies on the activation of pattern recognition receptors, among which are TLRs. Both TLR2 and TLR4 have been implicated in the recognition of P. brasiliensis and regulation of the immune response. However, the role of TLR9 during the infection by this fungus remains unclear.
We used in vitro and in vivo models of infection by P. brasiliensis, comparing wild type and TLR9 deficient (−/−) mice, to assess the contribution of TLR9 on cytokine induction, phagocytosis and outcome of infection. We show that TLR9 recognizes either the yeast form or DNA from P. brasiliensis by stimulating the expression/production of pro-inflammatory cytokines by bone marrow derived macrophages, also increasing their phagocytic ability. We further show that TLR9 plays a protective role early after intravenous infection with P. brasiliensis, as infected TLR9−/− mice died at higher rate during the first 48 hours post infection than wild type mice. Moreover, TLR9−/− mice presented tissue damage and increased expression of several cytokines, such as TNF-α and IL-6. The increased pattern of cytokine expression was also observed during intraperitoneal infection of TLR9−/− mice, with enhanced recruitment of neutrophils. The phenotype of TLR9−/− hosts observed during the early stages of P. brasiliensis infection was reverted upon a transient, 48 hours post-infection, neutrophil depletion.
Our results suggest that TLR9 activation plays an early protective role against P. brasiliensis, by avoiding a deregulated type of inflammatory response associated to neutrophils that may lead to tissue damage. Thus modulation of TLR9 may be of interest to potentiate the host response against this pathogen.
Paracoccidioides brasiliensis is an etiological agent of paracoccidioidomycosis, one of the most prevalent systemic mycosis in South America, affecting over 10 million people. One of the hallmark features of this fungus is its multinucleated nature, where a single P. brasiliensis cell can have from one to over 50 nuclei. Thus, fungal cell death may culminate in the release of large amounts of DNA and in the activation of the host innate immune system via toll like receptor (TLR) 9, a member of the TLR family. In this study, we evaluated the contribution of TLR9 activation during the infection by P. brasiliensis using an experimental model of P. brasiliensis infection of wild-type and TLR9−/− mice. We found that the lack of this receptor results in tissue damage and increased expression of cytokines, culminating in a higher death rate of TLR9−/− mice early post infection. This was mainly associated to neutrophils, as depletion of these cells, during the first 48 hours post-infection, converted the early response of TLR9−/− mice to infection to that of wild-type mice. Our work provides evidence for the protective effects of TLR9 activation during early stages of P. brasiliensis infection, highlighting the relevance of a balanced innate immune response to avoid tissue damage.
Citation: Menino JF, Saraiva M, Gomes-Alves AG, Lobo-Silva D, Sturme M, Gomes-Rezende J, et al. (2013) TLR9 Activation Dampens the Early Inflammatory Response to Paracoccidioides brasiliensis, Impacting Host Survival. PLoS Negl Trop Dis 7(7): e2317. doi:10.1371/journal.pntd.0002317
Editor: Joseph M. Vinetz, University of California San Diego School of Medicine, United States of America
Received: March 5, 2013; Accepted: June 5, 2013; Published: July 25, 2013
Copyright: © 2013 Menino et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: J.F. Menino was supported by a grant from Fundação para a Ciência e Tecnologia (FCT), Portugal (SFRH/BD/33446/2008). This work was supported by a grant from FCT (PTDC/BIA-MIC/108309/2008). M. Saraiva is a Ciência 2007 fellow and M. Sturme is a Ciência 2008 fellow. We would also like to thank FAPESP (Fundação para Amparo a Pesquisa do Estado de São Paulo) and CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) for financial support. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Paracoccidioides brasiliensis is a causative agent of paracoccidioidomycosis (PCM), one of the most prevalent systemic mycosis in Latin America . One of P. brasiliensis biological hallmarks is its particular temperature-dependent morphological dimorphism. This fungus switches from the environmental non-pathogenic mycelial/conidial form at ambient temperatures to the pathogenic multiple budding yeast form, with an high variability of cell sizes, when exposed to temperatures similar to those of the mammalian host , , . The mechanism of infection by P. brasiliensis entails the inhalation of airborne conidia that, when in the lung and exposed to host temperatures, undergo a complex morphological switch to the pathogenic yeast form , . PCM is divided in two different forms: the acute or sub-acute form and the chronic form, depending on the natural course of infection and clinical manifestations of the patient , . The clinical manifestations rely on the virulence of P. brasiliensis infecting strain, the degree and type of immune response triggered, the tissues infected, and importantly, on intrinsic characteristics of the host , .
Despite the fact that a large number of individuals are exposed to the fungus, only a minority develops the disease, suggesting that, for the majority of the population, a protective immune response is developed , . Therefore, the understanding of the protective characteristics of the immune response to P. brasiliensis is of interest as it may reveal targets for disease control. The initiation of the immune response relies on the activation of the innate immune system upon recognition of pathogen-associated molecular patterns (PAMPs) . This recognition is mediated by the family of pattern recognition receptors (PRRs) that is composed by a large number of receptors in immune cells , . Activation of PRRs culminates with the expression of several immune mediators, including pro- and anti-inflammatory cytokines and also with the activation of a series of microbicidal mechanisms that aim at eliminating the pathogen . The most widely recognized type of PRRs are the toll-like receptors (TLRs) . Over the past few years, several studies have demonstrated a relevant role for TLRs in the recognition of fungal pathogens, such as P. brasiliensis, Candida albicans, Aspergillus fumigatus, and Cryptococcus neoformans , , , , , , . The role of MyD88, an adaptor protein used by all TLRs (with the exception of TLR3), during P. brasiliensis infection remains controversial, with some authors reporting that this protein is not essential for an effective response against P. brasiliensis , and others claiming that MyD88 is important for the activation of innate fungicidal mechanisms and for the induction of the effector and regulatory cells of the adaptive immune response . A role for both TLR2 and TLR4 in the recognition and internalization of P. brasiliensis has been reported in human monocytes and neutrophils . In a model of experimental PCM, TLR2 deficiency leads to increased Th17 immunity associated with diminished expansion of regulatory T cells and increased lung pathology due to unrestrained inflammatory reactions . In contrast, P. brasiliensis recognition by TLR4 leads to an increased production of Th17 cytokines, enhanced pro-inflammatory immunity, and impaired expansion of regulatory T cells, resulting in a more severe form of infection . However, the involvement of TLR9 in P. brasiliensis infection has not yet been addressed. Several lines of evidence suggest that TLR9 may play a role in infection by P. brasiliensis, similarly to what is already described for other pathogenic fungi, such as A. fumigatus, C. albicans and C. neoformans ,,,,,. Firstly, P. brasiliensis DNA is known to have large numbers of CpG motifs , the natural ligand to TLR9 . Secondly, due to the fungus multinucleated nature , , a high amount of DNA is expected to be released upon cell death during infection. Thirdly, previous studies show that, in in vitro models, P. brasiliensis DNA increases the phagocytic index of macrophages, whereas in in vivo models of P. brasiliensis infection, TLR9 activation may act as a Th1-promoting adjuvant in a time/concentration dependent-manner , .
In this study, we investigated the role of TLR9 in the recognition of P. brasiliensis, and its influence on the infective process and evolution of the disease. Our results show that TLR9 recognizes P. brasiliensis, playing a major regulatory role during early times of in vivo infection with its absence making the host more prone to increased liver pathology and premature death, mainly mediated by neutrophils.
Materials and Methods
Microorganisms and culture media
The strain ATCC 60855 of P. brasiliensis registered at the American Type Culture Collection (Rockville, MD) was used throughout this experiment. Yeast cells were maintained at 37°C by subculturing in brain heart infusion (BHI) (Duchefa) solid media supplemented with 1% glucose and gentamicin (50 µg/mL). For both in vitro and in vivo assays, yeast cells were grown in BHI liquid medium supplemented with 1% glucose and gentamicin (50 µg/mL) at 37°C with aeration on a mechanical shaker (220 rpm). Cell growth was monitored for 148 h by microscopic counting using a Neubauer's Chamber and cells were collected during the exponential growth phase (72 h of growth, 1.65±0.8×107 cells/mL) for all the experimental assays.
This study was approved by the Portuguese national authority for animal experimentation Direção Geral de Veterinária (ID: DGV 594 from 1st June 2010). Animals were kept and handled in accordance with the guidelines for the care and handling of laboratory animals in the Directive 2010/63/EU of the European Parliament and of the Council.
Eight-week-old C57BL/6 mice were obtained from Charles River (Barcelona, Spain) and eight-week-old TLR9−/− (generated in a C57BL/6 background) were kindly provided by P. Vieira (Pasteur Institute of Paris, France). Mice were housed under specific-pathogen-free conditions with food and water ad libitum.
In vivo infection.
C57BL/6 WT or TLR9−/− mice were infected intravenously (i.v.) with 1×106 P. brasiliensis yeast cells grown to the exponential phase in BHI liquid medium. Prior to infection, cells were washed 3 times with lipopolysaccharide (LPS)-free phosphate-buffered saline (PBS) (Gibco), passed through a syringe to eliminate cell clumps, and submitted to Neubauer counting procedures (each mother and bud cells were considered as individual counts). Mice were monitored daily for 32 days. Mice showing severely impaired mobility in the first 48 hours of infection were sacrificed and liver and blood were collected for later analysis.
For the peritoneal cavity model of infection, mice were injected intraperitoneally (i.p.) with 1×106 P. brasiliensis yeast cells grown and treated as described above. Mice were sacrificed 6 hours post-infection and, after peritoneal lavage with 4 ml of PBS, total leukocyte number was determined.
Liver from dying mice were harvested in the time-period comprehending 48 h post-infection, fixed in 3.8% phosphate-buffered formalin and embedded in paraffin. Light-microscopy studies were performed on tissue sections stained with hematoxylin and eosin (HE) as previously described . The histological analysis was performed by the presence of necrotic areas and the type of inflammatory infiltrate (when present) in each field of 10× objective.
C57BL/6 WT and TLR9−/− mice were made neutropenic by treatment with the monoclonal antibody (MAb) RB6-8C5, as previously described , . Briefly, mice were injected i.v. in the lateral tail vein with 200 mg of MAb RB6-8C5, and 6 h later infected i.v. with 1×106 P. brasiliensis yeast cells grown to the exponential phase in BHI liquid medium and treated as indicated before. Mice were monitored as indicated before during the first 2 days of infection.
Preparation of fungal DNA.
P. brasiliensis DNA was extracted from exponentially growing cells. Briefly, cells were disrupted using a mix of lysis buffer (1 mM EDTA, 10 mM Tris–HCl, 1% SDS, 100 mM NaCl) and phenol/chloroform/isoamylalcohol (25∶24∶1). Heat-shock treatment (45 min at 65°C) was performed and the aqueous phase was treated using a column-based method (in-column proteinase K and RNAse treatments were performed). To elute DNA, LPS-free water was used.
In vitro infection
Bone marrow-derived macrophages (BMDMs) from C57BL/6 WT and TLR9−/− mice were prepared as described previously , , . BMDMs were seeded in 24-well plates at 5×105 cells/well and kept at 37°C and 5% CO2 atmosphere. Cells were challenged with 5 µg of P. brasiliensis DNA extracted from exponentially growing cells. In other experiments, BMDMs from C57BL/6 WT and TLR9−/− mice seeded as described above were challenged with P. brasiliensis yeast cells grown to the exponential phase, with late-stationary growing cultures or with P. brasiliensis lysed cells, using a 2∶1 multiplicity of infection (MOI; yeast/BMDMs ratio) for 24 h. Prior to infection, fungal cells were washed 3 times with LPS-free PBS, passed through a syringe to eliminate cell clumps, and submitted to Neubauer counting procedures (mother and bud cells were considered as individual counts). Supernatants from stimulated BMDMs were collected 24 h post-infection and stored at −80°C for later cytokine analysis.
P. brasiliensis phagocytosis assays
BMDMs from C57BL/6 WT and TLR9−/− prepared as previously described  were seeded in 24-well plates at 5×105 cells/well, kept at 37°C and 5% CO2 atmosphere and infected with P. brasiliensis yeast cells grown to the exponential phase using a 2∶1 multiplicity of infection (MOI; yeast/macrophage ratio) for 3 h. The wells were then washed three times with PBS to eliminate non-phagocytized cells, BMDMs were lysed with sterile H2O and the remaining phagocytized yeast cells were collected to determine total phagocytosis, using a Neubauer's Chamber. For the experiments, P. brasiliensis cells treated with DNase were also used. Briefly, after collecting and washing, P. brasiliensis yeast cells grown to the exponential phase were incubated at 37°C for 1 h with 20 µg of DNaseI (Ambion). As control, heat-inactivated DNase (70°C for 1 h) was used.
Cytokine levels were measured in serum collected from infected animals or in supernatants of infected cell cultures by capture enzyme-linked immunosorbent assay (ELISA) (eBioscience). The ELISA procedure was performed according to the manufacturer's protocol, and absorbances were measured with a Bio-Rad 680 Micro-plate Reader.
Quantification of neutrophils and mononuclear cells influx to the peritoneal cavity during the i.p. infection with P. brasiliensis was performed by flow cytometry (FCM) on a BD LSR II flow cytometer. Cells collected from the peritoneal cavity were stained, using specific antibodies for CD11b, CD11c, GR-1 and Ly6G to distinguish neutrophils and mononuclear cells populations, and a minimum of 100,000 cells per sample was acquired at low/medium flow rate. Offline data was analyzed with the flow cytometry analysis software package FlowJo 7.6.1.
Real-time polymerase chain reaction (RT-PCR)
Total RNA (1 µg) was isolated according to TRIzol methodology (Invitrogen). For liver samples, small portions were homogenized in between microscopy slides, suspended in 1 mL of TRIzol reagent and kept at −80°C for later analysis. For the i.p. experiment around 5×105 cells from the peritoneal exudates were used, suspended in 250 uL of TRIzol reagent and stored at −80°C for later analysis. RNA integrity was checked by the presence of clear 18S and 28S rRNA bands in agarose gel electrophoresis. The absence of DNA contamination in the samples was confirmed by the absence of PCR amplification of the ubiquitin gene in the isolated RNA. Total RNA (1 µg) was reverse transcribed using the iScript cDNA Synthesis kit (Bio-Rad) following manufacturer's instructions and 1 µL of cDNA used as a template for real-time quantification using the SsoFast EvaGreen SuperMix (Bio-Rad) following manufacturer's instructions. Real-time quantification was carried out on a CFX96 Real-Time System (Bio-Rad) using threshold cycle (Ct) values for ubiquitin transcripts as the endogenous reference. The primer sequences were designed and synthesized by TIB Mol. Biol. and were as follows: UBQ forward, TGG CTA TTA ATT ATT CGG TCT GCA T; UBQ reverse, GCA AGT GGC TAG AGT GCA GAG TAA; IL-10 forward, TTT GAA TTC CCT GGG TGA GAA; IL-10 reverse, GCT CCA CTG CCT TGC TCT TAT T; IL-17 forward, CTC AGA CTA CCT CAA CCG TTC CA; IL-17 reverse, TTC CCT CCG CAT TGA CAC A; TNF forward, GCC ACC ACG CTC TTC TGT CT; TNF reverse, TGA GGG TCT GGG CCA TAG AAC; MIP2 forward, CTC AGT GCT GCA CTG GT; MIP2 reverse, AGA GTG GCT ATG ACT TCT GTC T; IL-6 forward, TCG TGG AAA TGA GAA AAG AGT TG; IL-6 reverse, TAT GCT TAG GCA TAA CGC AC TAG. All measurements were performed in triplicate. A single melting peak was obtained for each gene analyzed in all samples.
Data is reported as the mean ± standard error of the mean (SEM) and all assays were repeated at least three times. All statistical analysis was performed using the GraphPad Prism Software version 5.01. For the experiments comparing two groups (see Fig. 1A and B, 4 and 5), a two-tailed unpaired Student t test was performed. Welch's correction was applied when making multiple comparisons. The One Way ANOVA test was performed in data presented in Fig. 1C using Turkey's multiple comparison post-test. The survival curves, representative of three independent experimental infections (Fig. 2, n = 21 mice), are represented using the Kaplan-Meier estimator, and Gehan-Breslow-Wilcoxon test was applied. For all data analysis statistical significance was considered at the level of 0.05 (2-tailed, 95% confidence interval).
(A) - Protein levels after stimulation with 5 µg of P. brasiliensis DNA for 24 h. Asterisks represent significant differences between WT and TLR9 TLR9−/− mice (**P<0.01, *P<0.05). IL-6 was also found with higher expression levels in dendritic cells from TLR9−/− mice when compared to dendritic cells from WT mice (data not shown) (B) – Protein levels after stimulation with wild-type P. brasiliensis ATCC 60855 yeast cells grown either to the exponential or late-stationary phase for 24 h using a MOI of 2∶1, and lysed cells. Asterisks represent significant differences between WT and TLR9−/− mice. IL-6 levels were below detection limits; (C) - Percentage of total phagocytosis of wild-type P. brasiliensis ATCC 60855 yeast cells grown to the exponential phase using a MOI of 2∶1. Asterisks represent significant differences between macrophages from WT mice stimulated with P. brasiliensis yeast cells and either macrophages from WT mice stimulated with P. brasiliensis yeast cells pre-treated with DNase (WT+DNase) (*P<0.05) or macrophages from TLR9−/− mice stimulated with P. brasiliensis yeast cells (TLR9−/−) (*P<0.05). Pre-treatment of P. brasiliensis yeast cells with DNase does not affect phagocytosis by TLR9−/− macrophages (data not shown). When cells were treated with heat inactivated DNase, the phagocytic profile of WT macrophages is restored (WT−/−+DNase*), while for TLR9−/− macrophages it is not altered (TLR9−/−+DNase*).
Representative survival curves using the Kaplan-Meier estimator of an experimental intravenous infection carried out in WT and TLR9−/− C57BL/6 mice (n = 21) with 1×106 wild-type P. brasiliensis ATCC 60855 yeast cells grown to the exponential phase. The estimated mean survival for WT mice is 13.6±2.0, while for TLR9−/− mice is 6.8±1.4 (Gehan-Breslow-Wilcoxon test was applied). Data is expressed as percentage of live animals. The observed differences during the first 48 h post- infection were statistically significant (***P<0.001).
Triggering of TLR9 by P. brasiliensis modulates macrophage cytokine production and phagocytosis
Since previous studies have reported the recognition of fungal DNA by TLR9 , we questioned if P. brasiliensis triggers TLR9-mediated responses in macrophages. For that, we stimulated BMDMs generated from wild-type (WT) and TLR9−/− mice with purified P. brasiliensis DNA. Our results showed that purified P. brasiliensis DNA induced the secretion of TNF-α and IL-6 by BMDMs in a TLR9 dependent way (Fig. 1A). Next, we tested if TLR9 could also recognize P. brasiliensis DNA in the yeast cellular context and whether this depended on the cellular physiological stage and integrity. We stimulated WT and TLR9−/− BMDM with either P. brasiliensis cells from exponential or late-stationary growing cultures or with lysed cells. We found that BMDMs stimulated with P. brasiliensis yeast-form produced TNF-α in a TLR9 dependent manner (Fig. 1B), although the response was higher for late-stationary growing and lysed cells.
In addition to cytokine expression, TLR triggering also associates with the onset of phagocytosis , . We next investigated if TLR9 activation by the yeast-form of P. brasiliensis impacted phagocytosis. The percentage of phagocytosis of P. brasiliensis by BMDMs was significantly reduced when P. brasiliensis yeast cells were treated with DNase or when TLR9−/− BMDMs were used (Fig. 1C). When heat inactivated DNase was used, the percentage of phagocytosis by WT macrophages was similar to that observed when no treatment was performed, whereas TLR9−/− macrophages maintained the phagocytic profile, indicating that DNase is not interfering with phagocytosis. Thus, our data indicate that TLR9 activation is required for maximal P. brasilisensis phagocytosis.
TLR9 has a protective role in the early phase of P. brasiliensis infection
Given the in vitro impact of P. brasiliensis recognition by TLR9, we next sought to investigate a role for this receptor during the course of an in vivo experimental infection. WT and TLR9−/− mice were intravenously infected with P. brasiliensis and the survival rates followed over time. We found that 24 days post-infection 100% of TLR9−/− mice had succumbed, whereas for WT mice this was only observed 31 days post-infection (Fig. 2). Furthermore, the estimated mean survival for WT mice was of 13.6±2.0 days, while for TLR9−/− mice it was reduced to 6.8±1.4 days (p<0.001; Fig. 2). Even more striking was the observation that during the first 48 h post-infection, TLR9−/− infected mice showed severe impaired mobility, with a significant number of animals being humanely sacrificed as a result. As shown in Fig. 2, during this early period the mortality of the TLR9−/− mice was approximately 43%, while WT mice showed no signs of disease. Thus, our data highlight an unexpected protective role of TLR9 during the early phases of P. brasiliensis infection.
Absence of TLR9 associated with increased immunopathology early during P. brasiliensis infection
In view of the key protective role observed for TLR9 during the early stages of P. brasiliensis infection, we next investigated several parameters that could be associated with the premature death of TLR9−/−- infected animals. Since during the initial period of infection (the first 48 h) differences in the lungs of infected animals are difficult to assess, we performed a histological analysis of the liver of the infected animals. We found that whereas the livers of infected WT animals showed a normal structure, those of TLR9−/− hosts presented small areas of granulocytes/neutrophil infiltrates and of necrosis (Fig. 3). To dissect further if the histological differences observed were associated to the intensity of the immune response between WT and TLR9−/− animals, we assessed cytokine expression in the liver and blood of infected animals. In the liver, the expression of both pro- and anti-inflammatory cytokines, namely TNF-α, IL-6 and IL-10, was increased in the absence of TLR9 (Fig. 4A–E). Likewise, a significant increase of circulating levels of IL-6 was observed in TLR9−/−-infected mice (Fig. 4G). Moreover, a trend towards higher levels of circulating TNF-α was detected (Fig. 4F) whereas those of IL-10 were below detection limit (data not shown).
Histological sections of liver stained with HE from TLR9−/− C57BL/6 mice (A to F) and WT C57BL/6 mice (G and H) intravenously infected with 1×106 wild-type P. brasiliensis ATCC 60855 yeast cells grown to the exponential phase. Magnification: ×10 (A, B, D, E, G and H); ×40 (F) and ×60 (C). Squares represent either tissue with inflammatory infiltrates (A and B) or with necrotic areas (D and E). Figures C and F represent magnifications of B and E, respectively. Results are from one representative experiment of two independent experiments. White bars represent 200 µm.
To further validate if the absence of TLR9 correlated with an exacerbated immune response to P. brasiliensis, and to study differential patterns of cellular recruitment in WT versus TLR9−/− mice, we used a model of i.p. infection The analysis of the peritoneal exudates revealed that the cytokine expression was increased in the absence of TLR9 (Fig. 5A–E). Of notice, a marked increased expression of both MIP-2 and IL-17, known to be associated with neutrophil recruitment , , , was found in the absence of TLR9 (Fig. 5C, D). Consistently, in TLR9−/− mice we found an increased recruitment to the peritoneal cavity of both mononuclear cells (macrophages and dendritic cells) and neutrophils, with a special significance of the latter ones (Fig. 5F–H). Thus, the data obtained for peritoneal infection recapitulates that obtained for intravenous infection.
Levels of cytokines detected upon intravenous infection of WT and TLR9−/− C57BL/6 mice (n = 21) with 1×106 wild-type P. brasiliensis ATCC 60855 yeast cells grown to the exponential phase. Samples were taken from dying mice during the first 48 h post-infection. As control, samples from the same number of healthy WT mice (n = 9) were collected. (A–E) – Expression profile of TNF-α, IL-6, MIP-2, IL-17, and IL-10 at mRNA level in liver from the mice. Asterisks represent significant differences between WT and TLR9−/− mice (*P<0.05); (F and G) - Protein levels of TNF-α and IL-6 detected via ELISA in the blood serum of challenged mice. Asterisks represent significant differences between WT and TLR9−/− mice (**P<0.01). There is also a trend towards higher TNF-α levels in TLR9−/− mice, though no statistically significant differences were detected. Bars represent means and standard deviations.
brasiliensis. Intraperitoneal infection of WT and TLR9−/− C57BL/6 mice (n = 10) with 1×106 P. brasiliensis yeast cells grown to the exponential phase. (A–E) - Expression profile of TNF-α, IL-6, MIP-2, IL-17, and IL-10 at mRNA level in cells collected from the peritoneal cavity of WT and TLR9−/− mice. Peritoneal lavages were performed 6 h after infection and gene expression levels were accessed by RT-PCR. Asterisks represent significant differences between WT and TLR9−/− mice (*P<0.05); (F–H) - Influx of neutrophils and mononuclear cells to peritoneal cavities of WT and TLR9−/− mice. Peritoneal lavages were performed 6 h after infection, and total and differential leukocyte counts were done. An average of 2.1×106±1.4×106 cells was collected. Asterisks represent significant differences between WT and TLR9−/− mice (*P<0.05, **P<0.01). Bars represent means and standard deviations.
Neutrophilia mediates the early death of P. brasiliensis-infected TLR9−/− mice
Considering the high influx of neutrophils to the peritoneal cavity of TLR9−/− mice, which correlated with the high expression of MIP-2 and IL-17 found in the liver and peritoneal cavity, we next evaluated if this could be the detrimental factor during the infectious process of TLR9−/− hosts. For this purpose, we depleted neutrophils in WT and TLR9−/− mice prior to infection with P. brasiliensis, using a specific monoclonal antibody (MAb RB6-8C5). I.p. injection of MAb RB6-8C5 abrogated neutrophils from 6 h to up to 48 h post-injection (data not shown, ). As shown in figure 6, depletion of neutrophils reverted the highly susceptible phenotype observed in TLR9−/− mice during the first two days of infection. Depletion of neutrophils in WT mice had no influence on the mice survival during the first 48 h post-infection (Fig. 6). Therefore, a high neutrophil recruitment to the site of infection appears to be responsible for the death observed in the absence of TLR9 during the early stages of P. brasiliensis infection.
Representative survival curves of an experimental intravenous infection carried out in WT and TLR9−/− C57BL/6 mice (n = 21) with 1×106 wild-type P. brasiliensis ATCC 60855 yeast cells grown to the exponential phase. Mice were treated i.v. with the neutrophil-depleting MAb RB6-8C5 6 h pre-infection. Grey lines represent the survival curve of the experimental intravenous infections carried out in WT and TLR9−/− C57BL/6 mice (n = 21) with 1×106 wild-type P. brasiliensis ATCC 60855 yeast cells grown to the exponential phase.
Since the discovery of TLRs and their major role in host recognition of conserved molecular structures from microorganisms, particularly those of invading pathogens, enormous advances have been made in comprehending how the immune system responds to pathogenic organisms. It is well established that the first phase of an immune response involves innate immune mechanisms , namely those triggered by TLR activation , , . These receptors are present in neutrophils, monocytes and macrophages, cells that besides their phagocytic activity are crucial for the signaling and amplification of the response against the pathogen .
The knowledge on PRR recognition and activation of an efficient immune response against P. brasiliensis has progressively increased, mainly in what concerns TLR2, TLR4, and Dectin-1 , . However, a role for TLR9 during P. brasiliensis infection has only been addressed in the context of vaccination , despite the evidences for the involvement of this TLR in other fungal infections , , , . Considering the high DNA content and number of unmethylated CpG oligonucleotides of this multinucleated fungus , a role for this receptor during P. brasiliensis infection is likely. We herein demonstrate that P. brasiliensis purified DNA activated TLR9 in macrophages, leading to the expression of cytokines. This is in line with previous studies showing that fungal DNA is recognized by TLR9. A. fumigatus DNA stimulates the production of pro-inflammatory cytokines in mouse and human dendritic cells . Similarly, murine dendritic cells express IL-12p40 and CD40 upon stimulation with DNA from C. neoformans . Human monocytes and macrophages from TLR9−/− mice were described to produce less IL-10 than cells from control mice when stimulated with C. albicans .
Despite TLR9 activation by purified P. brasiliensis DNA, upon macrophage stimulation with P. brasiliensis yeast cells grown to the exponential phase, TLR9 was activated in a lesser extent. This finding is likely due to the low exposure of DNA in this case, as when yeast cells on late stationary phase or lysed cultures were used, TLR9 recognized P. brasiliensis more prominently, leading to the production of pro-inflammatory cytokines by macrophages. Even though intact P. brasiliensis yeast cells did not fully activate TLR9 to induce the production of pro-inflammatory cytokines, in the absence of this receptor macrophages showed a decreased ability to phagocyte P. brasiliensis. Our findings are in line with previous studies showing that P. brasiliensis DNA activates macrophages, promoting their capacity to phagocyte P. brasiliensis . Altogether, our data indicate that TLR9 triggering affected the overall responsiveness of macrophages to P. brasiliensis. Furthermore, during infection, the release of DNA from P. brasiliensis is expected to occur following fungal cell death, thus suggesting that TLR9 activation in vivo is very likely to happen.
To assess the role of TLR9 in P. brasiliensis infections, we used in vivo models of infection. As our results show, TLR9 is crucial for mice survival in early times of infection (first 48 h). Remarkably, we found that, in mice showing severe signs of disease, lack of TLR9 increased the expression of pro-inflammatory cytokines in the liver. Earlier studies on immune responses to other organisms have revealed that an excessive inflammatory response, mainly due to high levels of TNF, can lead to premature death of the host cells , , , . In the context of P. brasiliensis infection, it was demonstrated that an excessive inflammatory response may be detrimental rather than protective. A more severe disease development in mice susceptible to P. brasiliensis was associated with the presence of increased IL-12 and IFN-γ levels in the lungs, suggesting that the production of pro-inflammatory mediators does not always correlate with immunoprotection . In addition to high TNF expression, we also found increased expression of MIP-2 and IL-17 in the liver and cells from the peritoneal cavity of P. brasiliensis-infected TLR9−/− mice. A detrimental role of IL-17 during P. brasiliensis infection was previously described as TLR2 and TLR4 deficiency associate with an increase of Th17 responses, lung pathology and more severe forms of infection , . Both IL-17 and MIP-2 have been previously associated with neutrophil recruitment , , , . In line with this, we observed a high influx of granulocytes/neutrophils into the peritoneal cavity of TLR9−/− i.p. infected animals. This enhanced neutrophil recruitment could thus be contributing to the detrimental response observed in P. brasiliensis-infected TLR9−/− animals. Indeed, upon transient neutrophil depletion, TLR9−/− mice survived during the first 48 hours post-infection, resembling the phenotype of WT mice. Therefore, the susceptibility profile observed for TLR9−/− mice early after infection with P. brasiliensis likely associates with an exacerbated neutrophil recruitment to the site of infection and/or with a particular detrimental phenotype of neutrophils. Although neutrophils are crucial during acute inflammatory response and subsequent resolution of fungal infection, in some situations, due to excessive release of oxidants and proteases, these cells may be responsible for injury to organs and fungal sepsis , . In addition to fungal infection, neutrophilia can also be harmful to the host in the context of other infections, such as M. tuberculosis and P. aeruginosa , , , .
Our data implicating TLR9 in the phagocytic activity of macrophages raises the hypothesis that, in the absence of TLR9, less P. brasiliensis cells are phagocytized. Thus, in the absence of this receptor, the extracellular P. brasiliensis cells may contribute to an exacerbated recruitment of neutrophils, which can result in a deregulated immune response. Since a significantly higher recruitment of dendritic cells and neutrophils is observed upon infection of TLR9−/− hosts, it is possible that the higher cytokine expression observed in this scenario results from the fact that more producing cells are present. Therefore, a fine tuned balance is required during infection with P. brasiliensis, in order to protect the host from infection. Several mechanisms must operate to achieve this balance, as is the case reported for TLR2/TLR4 activation . It is also important to refer that P. brasiliensis can be triggering TLR9-independent mechanisms in macrophages, as supported in the literature , . One cannot rule out the hypothesis of a parallel activation of other receptors together with TLR9. Several studies refer that Dectin-1 interaction with TLR9 results in a sinergistyc induction of IL-10, TNF-α, IL-2, IL-6 and IL-23 and down-regulation of IL-12 , . Studies with A. fumigatus reported a link between TLR2-mediated recognition and the phagocytic response , whereas internalization of TLR2 with A. fumigatus phagossome was demonstrated . As it has been shown for TLR2/TLR4, it is also possible that TLR9 signaling is involved in the instruction of appropriate regulatory T cell responses, a hypothesis currently under investigation.
Overall, this study highlights the relevant role of TLR9/neutrophils for the nature of the immune response to P. brasiliensis, and paves the way to the development of new preventive/therapeutical strategies, as resistance patterns of the host can be of great value on the comprehension of susceptibility and pathogenesis of P. brasiliensis infections.
We thank Joana Guedes and Nuno Araújo for technical assistance on the in vivo experiments.
Conceived and designed the experiments: M. Saraiva, A. Carvalho, A.G. Castro, F. Rodrigues. Performed the experiments: J.F. Menino, A.G. Gomes-Alves, D. Lobo-Silva, M. Sturme, G.H. Goldman, J. Gomes-Rezende, A.L. Saraiva, J. Pedrosa. Analyzed the data: J.F. Menino, C. Cunha, A. Carvalho, L. Romani, A.G. Castro, F. Rodrigues. Contributed reagents/materials/analysis tools: F. Rodrigues. Wrote the paper: J.F. Menino, M. Saraiva, A.G. Castro, F. Rodrigues.
- 1. Brummer E, Castaneda E, Restrepo A (1993) Paracoccidioidomycosis: an update. Clin Microbiol Rev 6: 89–117.
- 2. Garcia AM, Hernandez O, Aristizabal BH, De Souza Bernardes LA, Puccia R, et al. (2010) Gene expression analysis of Paracoccidioides brasiliensis transition from conidium to yeast cell. Med Mycol 48: 147–154.
- 3. Menino JF, Osorio NS, Sturme MH, Barros D, Gomes-Alves AG, et al. (2012) Morphological heterogeneity of Paracoccidioides brasiliensis: relevance of the Rho-like GTPase PbCDC42. Med Mycol 50: 768–774.
- 4. Almeida AJ, Cunha C, Carmona JA, Sampaio-Marques B, Carvalho A, et al. (2009) Cdc42p controls yeast-cell shape and virulence of Paracoccidioides brasiliensis. Fungal Genet Biol 46: 919–926.
- 5. McEwen JG, Bedoya V, Patino MM, Salazar ME, Restrepo A (1987) Experimental murine paracoccidiodomycosis induced by the inhalation of conidia. J Med Vet Mycol 25: 165–175.
- 6. Franco M (1987) Host-parasite relationships in paracoccidioidomycosis. J Med Vet Mycol 25: 5–18.
- 7. Benard G (2008) An overview of the immunopathology of human paracoccidioidomycosis. Mycopathologia 165: 209–221.
- 8. Shikanai-Yasuda MA, Telles Filho Fde Q, Mendes RP, Colombo AL, Moretti ML (2006) [Guidelines in paracoccidioidomycosis]. Rev Soc Bras Med Trop 39: 297–310.
- 9. Musatti CC, Rezkallah MT, Mendes E, Mendes NF (1976) In vivo and in vitro evaluation of cell-mediated immunity in patients with paracoccidiodomycosis. Cell Immunol 24: 365–378.
- 10. Janeway CA Jr, Medzhitov R (2002) Innate immune recognition. Annu Rev Immunol 20: 197–216.
- 11. O'Neill LA (2006) How Toll-like receptors signal: what we know and what we don't know. Curr Opin Immunol 18: 3–9.
- 12. Akira S (2001) [Toll-like receptors and innate immune system]. Tanpakushitsu Kakusan Koso 46: 562–566.
- 13. Medzhitov R (2001) Toll-like receptors and innate immunity. Nat Rev Immunol 1: 135–145.
- 14. Kawai T, Akira S (2007) TLR signaling. Semin Immunol 19: 24–32.
- 15. Villamon E, Gozalbo D, Roig P, O'Connor JE, Fradelizi D, et al. (2004) Toll-like receptor-2 is essential in murine defenses against Candida albicans infections. Microbes Infect 6: 1–7.
- 16. Braedel S, Radsak M, Einsele H, Latge JP, Michan A, et al. (2004) Aspergillus fumigatus antigens activate innate immune cells via toll-like receptors 2 and 4. Br J Haematol 125: 392–399.
- 17. Shoham S, Huang C, Chen JM, Golenbock DT, Levitz SM (2001) Toll-like receptor 4 mediates intracellular signaling without TNF-alpha release in response to Cryptococcus neoformans polysaccharide capsule. J Immunol 166: 4620–4626.
- 18. Netea MG, Van der Graaf C, Van der Meer JW, Kullberg BJ (2004) Recognition of fungal pathogens by Toll-like receptors. Eur J Clin Microbiol Infect Dis 23: 672–676.
- 19. Netea MG, Van Der Graaf CA, Vonk AG, Verschueren I, Van Der Meer JW, et al. (2002) The role of toll-like receptor (TLR) 2 and TLR4 in the host defense against disseminated candidiasis. J Infect Dis 185: 1483–1489.
- 20. Calich VL, da Costa TA, Felonato M, Arruda C, Bernardino S, et al. (2008) Innate immunity to Paracoccidioides brasiliensis infection. Mycopathologia 165: 223–236.
- 21. Carvalho A, Pasqualotto AC, Pitzurra L, Romani L, Denning DW, et al. (2008) Polymorphisms in toll-like receptor genes and susceptibility to pulmonary aspergillosis. J Infect Dis 197: 618–621.
- 22. Gonzalez A, Yanez A, Gozalbo D, Gil ML (2008) MyD88 is dispensable for resistance to Paracoccidioides brasiliensis in a murine model of blood-borne disseminated infection. FEMS Immunol Med Microbiol 54: 365–374.
- 23. Loures FV, Pina A, Felonato M, Feriotti C, de Araujo EF, et al. (2011) MyD88 signaling is required for efficient innate and adaptive immune responses to Paracoccidioides brasiliensis infection. Infect Immun 79: 2470–2480.
- 24. Bonfim CV, Mamoni RL, Blotta MH (2009) TLR-2, TLR-4 and dectin-1 expression in human monocytes and neutrophils stimulated by Paracoccidioides brasiliensis. Med Mycol 47: 722–733.
- 25. Loures FV, Pina A, Felonato M, Calich VL (2009) TLR2 is a negative regulator of Th17 cells and tissue pathology in a pulmonary model of fungal infection. J Immunol 183: 1279–1290.
- 26. Loures FV, Pina A, Felonato M, Araujo EF, Leite KR, et al. (2010) Toll-like receptor 4 signaling leads to severe fungal infection associated with enhanced proinflammatory immunity and impaired expansion of regulatory T cells. Infect Immun 78: 1078–1088.
- 27. Nakamura K, Miyazato A, Xiao G, Hatta M, Inden K, et al. (2008) Deoxynucleic acids from Cryptococcus neoformans activate myeloid dendritic cells via a TLR9-dependent pathway. J Immunol 180: 4067–4074.
- 28. van de Veerdonk FL, Netea MG, Jansen TJ, Jacobs L, Verschueren I, et al. (2008) Redundant role of TLR9 for anti-Candida host defense. Immunobiology 213: 613–620.
- 29. Ramaprakash H, Ito T, Standiford TJ, Kunkel SL, Hogaboam CM (2009) Toll-like receptor 9 modulates immune responses to Aspergillus fumigatus conidia in immunodeficient and allergic mice. Infect Immun 77: 108–119.
- 30. Ramirez-Ortiz ZG, Specht CA, Wang JP, Lee CK, Bartholomeu DC, et al. (2008) Toll-like receptor 9-dependent immune activation by unmethylated CpG motifs in Aspergillus fumigatus DNA. Infect Immun 76: 2123–2129.
- 31. Bellocchio S, Montagnoli C, Bozza S, Gaziano R, Rossi G, et al. (2004) The contribution of the Toll-like/IL-1 receptor superfamily to innate and adaptive immunity to fungal pathogens in vivo. J Immunol 172: 3059–3069.
- 32. Mansour MK, Tam JM, Vyas JM (2012) The cell biology of the innate immune response to Aspergillus fumigatus. Ann N Y Acad Sci 1273: 78–84.
- 33. Souza MC, Correa M, Almeida SR, Lopes JD, Camargo ZP (2001) Immunostimulatory DNA from Paracoccidioides brasiliensis acts as T-helper 1 promoter in susceptible mice. Scand J Immunol 54: 348–356.
- 34. Rutz M, Metzger J, Gellert T, Luppa P, Lipford GB, et al. (2004) Toll-like receptor 9 binds single-stranded CpG-DNA in a sequence- and pH-dependent manner. Eur J Immunol 34: 2541–2550.
- 35. McEwen JG, Restrepo BI, Salazar ME, Restrepo A (1987) Nuclear staining of Paracoccidioides brasiliensis conidia. J Med Vet Mycol 25: 343–345.
- 36. Almeida AJ, Matute DR, Carmona JA, Martins M, Torres I, et al. (2007) Genome size and ploidy of Paracoccidioides brasiliensis reveals a haploid DNA content: flow cytometry and GP43 sequence analysis. Fungal Genet Biol 44: 25–31.
- 37. Amaral CC, Garcia IP, Fernandes GF, Almeida SR, Camargo ZP, et al. (2005) Adjuvant effect of synthetic oligodeoxyribonucleotides (CpG-ODN) from the Paracoccidioides brasiliensis gp43 gene on the Th2-Th1 immunomodulation of experimental paracoccidioidomycosis. Scand J Immunol 62: 325–333.
- 38. Oliveira MS, Fraga AG, Torrado E, Castro AG, Pereira JP, et al. (2005) Infection with Mycobacterium ulcerans induces persistent inflammatory responses in mice. Infect Immun 73: 6299–6310.
- 39. Appelberg R, Castro AG, Gomes S, Pedrosa J, Silva MT (1995) Susceptibility of beige mice to Mycobacterium avium: role of neutrophils. Infect Immun 63: 3381–3387.
- 40. Appelberg R, Castro AG, Silva MT (1994) Neutrophils as effector cells of T-cell-mediated, acquired immunity in murine listeriosis. Immunology 83: 302–307.
- 41. Neves BM, Silvestre R, Resende M, Ouaissi A, Cunha J, et al. (2010) Activation of phosphatidylinositol 3-kinase/Akt and impairment of nuclear factor-kappaB: molecular mechanisms behind the arrested maturation/activation state of Leishmania infantum-infected dendritic cells. Am J Pathol 177: 2898–2911.
- 42. Zhang X, Majlessi L, Deriaud E, Leclerc C, Lo-Man R (2009) Coactivation of Syk kinase and MyD88 adaptor protein pathways by bacteria promotes regulatory properties of neutrophils. Immunity 31: 761–771.
- 43. van de Veerdonk FL, Kullberg BJ, van der Meer JW, Gow NA, Netea MG (2008) Host-microbe interactions: innate pattern recognition of fungal pathogens. Curr Opin Microbiol 11: 305–312.
- 44. Sanjuan MA, Dillon CP, Tait SW, Moshiach S, Dorsey F, et al. (2007) Toll-like receptor signalling in macrophages links the autophagy pathway to phagocytosis. Nature 450: 1253–1257.
- 45. Blander JM (2008) Phagocytosis and antigen presentation: a partnership initiated by Toll-like receptors. Ann Rheum Dis 67 Suppl 3: iii44–49.
- 46. Ohtsuka Y, Lee J, Stamm DS, Sanderson IR (2001) MIP-2 secreted by epithelial cells increases neutrophil and lymphocyte recruitment in the mouse intestine. Gut 49: 526–533.
- 47. Kolls JK, Linden A (2004) Interleukin-17 family members and inflammation. Immunity 21: 467–476.
- 48. Burdon PC, Martin C, Rankin SM (2005) The CXC chemokine MIP-2 stimulates neutrophil mobilization from the rat bone marrow in a CD49d-dependent manner. Blood 105: 2543–2548.
- 49. Beutler B (2004) Innate immunity: an overview. Mol Immunol 40: 845–859.
- 50. Romani L (2011) Immunity to fungal infections. Nat Rev Immunol 11: 275–288.
- 51. Levitz SM (2010) Innate recognition of fungal cell walls. PLoS Pathog 6: e1000758.
- 52. Calich VL, Pina A, Felonato M, Bernardino S, Costa TA, et al. (2008) Toll-like receptors and fungal infections: the role of TLR2, TLR4 and MyD88 in paracoccidioidomycosis. FEMS Immunol Med Microbiol 53: 1–7.
- 53. Wang JP, Lee CK, Akalin A, Finberg RW, Levitz SM (2011) Contributions of the MyD88-dependent receptors IL-18R, IL-1R, and TLR9 to host defenses following pulmonary challenge with Cryptococcus neoformans. PLoS One 6: e26232.
- 54. Kulkarni AB, Huh CG, Becker D, Geiser A, Lyght M, et al. (1993) Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death. Proc Natl Acad Sci U S A 90: 770–774.
- 55. Cerami A (1993) Tumor necrosis factor as a mediator of shock, cachexia and inflammation. Blood Purif 11: 108–117.
- 56. Rink L, Kirchner H (1996) Recent progress in the tumor necrosis factor-alpha field. Int Arch Allergy Immunol 111: 199–209.
- 57. Rock KL, Kono H (2008) The inflammatory response to cell death. Annu Rev Pathol 3: 99–126.
- 58. Gonzalez A, Sahaza JH, Ortiz BL, Restrepo A, Cano LE (2003) Production of pro-inflammatory cytokines during the early stages of experimental Paracoccidioides brasiliensis infection. Med Mycol 41: 391–399.
- 59. Henningsson L, Jirholt P, Lindholm C, Eneljung T, Silverpil E, et al. (2010) Interleukin-17A during local and systemic Staphylococcus aureus-induced arthritis in mice. Infect Immun 78: 3783–3790.
- 60. Pelletier M, Maggi L, Micheletti A, Lazzeri E, Tamassia N, et al. (2010) Evidence for a cross-talk between human neutrophils and Th17 cells. Blood 115: 335–343.
- 61. Bellocchio S, Moretti S, Perruccio K, Fallarino F, Bozza S, et al. (2004) TLRs govern neutrophil activity in aspergillosis. J Immunol 173: 7406–7415.
- 62. Zelante T, De Luca A, Bonifazi P, Montagnoli C, Bozza S, et al. (2007) IL-23 and the Th17 pathway promote inflammation and impair antifungal immune resistance. Eur J Immunol 37: 2695–2706.
- 63. Eruslanov EB, Lyadova IV, Kondratieva TK, Majorov KB, Scheglov IV, et al. (2005) Neutrophil responses to Mycobacterium tuberculosis infection in genetically susceptible and resistant mice. Infect Immun 73: 1744–1753.
- 64. Cruz A, Fraga AG, Fountain JJ, Rangel-Moreno J, Torrado E, et al. (2010) Pathological role of interleukin 17 in mice subjected to repeated BCG vaccination after infection with Mycobacterium tuberculosis. J Exp Med 207: 1609–1616.
- 65. Lowe DM, Redford PS, Wilkinson RJ, O'Garra A, Martineau AR (2012) Neutrophils in tuberculosis: friend or foe? Trends Immunol 33: 14–25.
- 66. Ras GJ, Theron AJ, Anderson R, Taylor GW, Wilson R, et al. (1992) Enhanced release of elastase and oxidative inactivation of alpha-1-protease inhibitor by stimulated human neutrophils exposed to Pseudomonas aeruginosa pigment 1-hydroxyphenazine. J Infect Dis 166: 568–573.
- 67. Takaoka A, Wang Z, Choi MK, Yanai H, Negishi H, et al. (2007) DAI (DLM-1/ZBP1) is a cytosolic DNA sensor and an activator of innate immune response. Nature 448: 501–505.
- 68. Ishii KJ, Kawagoe T, Koyama S, Matsui K, Kumar H, et al. (2008) TANK-binding kinase-1 delineates innate and adaptive immune responses to DNA vaccines. Nature 451: 725–729.
- 69. Gerosa F, Baldani-Guerra B, Lyakh LA, Batoni G, Esin S, et al. (2008) Differential regulation of interleukin 12 and interleukin 23 production in human dendritic cells. J Exp Med 205: 1447–1461.
- 70. Dennehy KM, Ferwerda G, Faro-Trindade I, Pyz E, Willment JA, et al. (2008) Syk kinase is required for collaborative cytokine production induced through Dectin-1 and Toll-like receptors. Eur J Immunol 38: 500–506.
- 71. Luther K, Torosantucci A, Brakhage AA, Heesemann J, Ebel F (2007) Phagocytosis of Aspergillus fumigatus conidia by murine macrophages involves recognition by the dectin-1 beta-glucan receptor and Toll-like receptor 2. Cell Microbiol 9: 368–381.
- 72. Chai LY, Kullberg BJ, Vonk AG, Warris A, Cambi A, et al. (2009) Modulation of Toll-like receptor 2 (TLR2) and TLR4 responses by Aspergillus fumigatus. Infect Immun 77: 2184–2192.