OUP user menu

Toll-like receptors and fungal infections: the role of TLR2, TLR4 and MyD88 in paracoccidioidomycosis

Vera L. G. Calich, Adriana Pina, Maíra Felonato, Simone Bernardino, Tania A. Costa, Flávio V. Loures
DOI: http://dx.doi.org/10.1111/j.1574-695X.2008.00378.x 1-7 First published online: 1 June 2008


The aim of this minireview is to present a concise view of the most important pattern recognition receptors used by the innate immune system to sense and control pathogen growth into host tissues. A brief review of the role of Toll-like receptors (TLRs) in fungal infections followed by some recent results on the function of TLR4, TLR2 and the MyD88 adaptor molecule in the pathogenesis of paracoccidioidomycosis are presented.

  • toll-like receptors
  • fungal pathogens
  • paracoccidioidomycosis
  • TLR2
  • TLR4
  • MyD88

Innate immunity and pattern recognition receptors

Cells of the innate immune system constantly sense the presence of invading microorganisms using several kind of conserved, transmembrane or intracytoplasmatic receptors called pattern recognition receptors (PRRs). These receptors recognize conserved molecular structures shared by groups of microorganisms and known as pathogen-associated molecular patterns (PAMPs). Because PAMPs are produced by pathogens but not by the host cells, their recognition by PRRs allows for self–nonself discrimination (Janeway & Medzhitov, 2002). The most important types of PRR are the Toll-like receptors (TLRs), the non-TLRs such as intracellular nucleotide-binding oligomerization domain (NOD)-like proteins and the C-type lectin receptors (CLRs) (Gordon, 2002; Brown & Gordon, 2003; Inohara & Nunez, 2003; Akira et al., 2006). The interaction between PAMPs and PRRs leads to the activation of cells of the innate immune system and subsequent production of mediators that are used to eliminate the invading pathogen and to control the adaptative immune responses. TLRs are crucial for many aspects of microbial elimination, including recruitment of phagocytes to the site of infection, microbial killing and activation of dendritic cells (DCs), which become immunogenic and endowed with a unique ability to induce full activation of T cells (Reis e Sousa, 2004). Interestingly, recent findings indicate that direct recognition of PAMPs by DCsis critical for priming appropriate T-cell responses. TLR signaling resulted in activated DCs that primed an effective T helper 1 (Th1) or Th2 response, whereas indirect activation by inflammatory mediators alone (proinflammatory cytokines) induced DCs that, although supporting expansion of CD4+ T-cell clones, did not promote Th1 or Th2 effector differentiation (Spörri & Reis e Sousa, 2005).

Thus far, 13 TLRs have been described that recognize a wide variety of pathogen structures including triacyl lipopeptides (TLR1 in association with TLR2), lipoteichoic acid and lipoproteins of Gram-positive bacteria (TLR2), double-stranded RNA (TLR3), lipopolysaccharide of Gram-negative bacteria (TLR4), bacterial flagellin (TLR5), diacyl lipopeptides (TLR6 in association with TLR2), single-stranded RNA (TLR7) and nonmethylated CpG of bacterial DNA (TLR9) (Takeda & Akira, 2005). Intracellular NOD proteins sense the presence of intracellular muropeptides (Inohara & Nunez, 2003). Upon ligand binding, innate immune receptors engage intracellular signaling pathways that result in the activation of conserved transcription factors for cell activation. One key transcription factor used by TLR-mediated innate immunity is NFkB. Almost all TLRs signal via MyD88, an adaptor protein, for NFkB activation, subsequent inflammatory cytokine production and control of adaptative immunity (Takeda & Akira, 2005; Akira et al., 2006). Besides their well-described role in the immunorecognition of conserved pathogen motifs, TLRs can be used to enhance microbial pathogenicity. Thus, Yersinia pestis was shown to evade the host's immune system by TLR2 activation and subsequent production of IL-10, one important macrophage deactivating cytokine (Sing et al., 2002). Furthermore, studies with whole pathogens or purified components have shown distinct patterns of TLR usage: TLR2 and TLR4 control in vivo Brucella abortus infection whereas only TLR4 is activated by the purified bacterial lipopolysaccharide (Campos et al., 2004).

TLR-activated DCs can induce the differentiation of distinct T-cell-mediated effector mechanisms. Lipopolysaccharide-stimulated DCs produce low levels of IL-10 but high levels of IL-12 and tumor necrosis factor-α (TNF-α), favoring Th1 immunity. In contrast, peptidoglycan-activated DCs secrete low levels of IL-12 associated with prevalent production of IL-10, resulting in prevalent Th2 immunity (Qi et al., 2003). Interestingly, the recognition of microbial products by TLRs was found to block the suppressive effect of T regulatory (Treg) cells on pathogen-specific adaptative immune response and this effect was partially due to the synthesis of IL-6 by TLR-stimulated DCs (Pasare & Medzhitov, 2003).

Dendritic cell-specific intercelullar adhesion molecule-3-grabbing nonintegrin (DC-SIGN), mannose receptor (MR) and dectin-1 are C-type lectins that recognize glycoproteins and carbohydrates in pathogen cells. This interaction controls phagocytosis, microbicidal activity and signaling processes that direct cell adhesion and migration. As fungal cell walls are carbohydrate-rich structures, these PRRs are directly involved in the host's innate immunity to these pathogens. DC-SIGN as well as MRs are primarily activated by IL-4 and associated with the Th2 pattern of the immune response (Brown & Gordon, 2003; Cambi et al., 2005; Koppel et al., 2005). Dectin-1 activation by Candida albicans or curdlan, a dectin-1-specific β-glucan, was recently shown to induce the preferential secretion of tumor growth factor (TGF)-β and IL-6 and the subsequent activation of Th17 lymphocytes. This T-cell subset secretes IL-17, induces chemokine secretion at sites of infection, causes recruitment of neutrophils and is important in defense against extracellular pathogens, including Candida albicans (Leibundgut-Landmann et al., 2007; Palm & Medzhitov, 2007).

Upon ligand binding, innate immunoreceptors engage intracellular signaling pathways that converge to the activation of conserved transcription factors and subsequent cell activation. Almost all TLRs signal via MyD88 for NFkB activation, resulting in Th1 immunity associated with the prevalent IL-12 secretion. TLR2 activation, however, induces high levels of IL-10, and expansion of regulatory T cells (Treg) which further control T effector lymphocytes (Gordon, 2002; Akira et al., 2006; Sutmuller et al., 2006). Interestingly, recent work has shown that the preferential activation of dectin-1 by the selective agonist curdlan, and possibly dectin-2 by Candida albicans hyphae, is mediated by the caspase recruitment domain (CARD9) adaptor protein, resulting in increased IL-23 secretion and preferential induction of Th17 cells (Gross et al., 2006; Leibundgut-Landmann et al., 2007; Palm & Medzhitov, 2007). The fungal morphotype is also recognized by different PRRs and this fact was suggested to be exploited as an escape mechanism by fungal cells. For instance, Candida hyphae are recognized only by TLR2, inducing prevalent secretion of anti-inflammatory cytokines, whereas Candida blastoconidia interact with TLR4, Dectin-1 and TLR2, resulting in a complex pattern of cell activation (Romani, 2004; Netea et al., 2006).

In summary, PRR activation orchestrates the development of innate and adaptative immune responses, which are necessary for protection against infection, reinfection or containment of chronic infections. However, if activation of innate immune receptors is excessive, high levels of proinflammatory mediators [IFN-γ, TNF-α, nitric oxide (NO)] are secreted and can exert a deleterious effect to the host. Septic shock induced by lipopolysaccharide and TLR4 activation by Gram-negative bacteria is a good example of inadequate activation of immunity, which results in severe host pathology.

Recognition of fungi by TLR

Netea (2002) were the first to describe the use of TLRs by a fungal pathogen, Candida albicans. C3H/HeJ mice, which express a defective TLR4 gene, present an increased susceptibility to disseminated candidiasis and impaired recruitment of neutrophils to the site of infection when compared with normal, C3H/HeN mice. In addition, the chemokines keratinocyte-derived chemokine (KC) and macrophage inflammatory protein (MIP-2) were shown to be released in lower amounts by TLR4-defective macrophages. Following this pioneer work, other groups reported their studies on the role of TLRs and the MyD88 adaptor protein in Candida albicans infections. The main biological effects observed in diverse experimental approaches are summarized in Table 1. As can be seen, some discrepant findings were obtained with TLR2- and TLR4-deficient hosts (Netea et al., 2002, 2004; Villamón et al., 2004b; Murciano et al., 2006), although MyD88 deficiency appears consistently to lead to impaired protection or phagocyte– fungus interaction. This adaptor protein was shown to be involved in the induction of protective immune responses by DCs (Bellocchio et al., 2004a), as well as in the phagocytosis, killing and synthesis of cytokines by Candida-infected cells (Marr et al., 2003; Villamón et al., 2004a).

View this table:
Table 1

The role of TLRs and MyD88 adaptor protein in some experimental models of Candida albicans infection

PRR deficiencyBiological effectReference
TLR4High susceptibility; normal PMN and macrophage killing activity; normal TNF-α, low KC, MIP-2 and impaired PMN influxNetea et al. (2002)
TLR4No increased susceptibility to disseminated infection. TLR4-deficient mice mount Th1 immunityMurciano et al. (2006)
TLR2High resistance; TLR2 induces IL-10, Treg cells and suppressed immunityNetea et al. (2004)
TLR2High susceptibility; low TNF-α, MIP-2; decreased PMN influx; no effect phagocytosis and NO productionVillamón et al. (2004a)
MyD88Hyphae: impaired phagocytosis, killing and cytokine secretionMarr et al. (2003)
MyD88High susceptibility, impaired production of cytokine, low type-1 CD4+ and CD8+ T cellsVillamón et al. (2004b)
MyD88, TLR4, TLR2, TLR9MyD88: high susceptibility. TLR signaling: depends on morphotypes, route of infection. Th1 response: DC-MyD-dependentBellocchio et al. (2004a)

TLR2 usage was shown to have protective or detrimental effects in models of Candida albicans infection; the conflicting results, however, could be attributed to the use of different experimental protocols (Netea et al., 2004; Villamón et al., 2004a, b), but brought important contributions to the understanding of immunopathology of infectious processes. The deleterious effect of TLR2 signaling during infection was associated with increased synthesis of IL-10 and an enhanced survival of CD4+CD25+ Treg cells, resulting in deficient T-cell immunity and impaired fungal clearance (Netea et al., 2004).

As observed with Candida albicans, the role of TLR in Cryptococcus neoformans infection needs to be further explored (Table 2). Glucuronoxylomannan, the major component of the polysaccharide capsule of Cryptococcus neoformans, is shed from the fungus and circulates in blood and cerebrospinal fluid of infected hosts. This polysaccharide was reported to activate cells transfected with CD14 and TLR4 but this interaction results in incomplete activation of cells, and no secretion of TNF-α (Shoham et al., 2001). In vivo, MyD88 and TLR2 but not TLR4 were shown to be required to induce protection against Cryptococcus neoformans infection (Yauch et al., 2004; Biondo et al., 2005). A more recent report, however, suggests that TLR2 and TLR4 do not or only marginally contribute to the host response to this pathogen (Nakamura et al., 2006).

View this table:
Table 2

The role of TLRs and MyD88 adaptor protein in some experimental models of Cryptococcus neoformans infection

PRR deficiencyBiological effectReference
TLR2, TLR4, CD14Glucoronoxylomannan stimulates cells via CD14 and TLR4; no TNF-α synthesisShoham et al. (2001)
TLR-2, TLR4, MyD88, CD14MyD88 has a major role in protection; CD14 and TLR2, minor roles; TLR4, no effectYauch et al. (2004)
TLR-2, TLR4, MyD88MyD88 and TLR2 have a major role in protection; TLR4 not importantBiondo et al. (2005)
TLR2 and TLR4Not important to protectionNakamura et al. (2006)

TLR2-, TLR4- and MyD88-dependent activation of host cells were shown to play a role in cytokine secretion, polymorphonuclear neutrophil (PMN) activation and susceptibility to infection by another opportunistic fungal pathogen, Aspergillus fumigatus (Wang et al., 2001; Marr et al., 2003; Meier et al., 2003; Netea et al., 2003; Bellocchio et al., 2004a, b; Braedel et al., 2004; Dubordeau et al., 2006). As described for Candida albicans, the germination from conidia to hyphae was proposed as an escape mechanism of A. fumigatus as conidia cells are recognized by TLR4 and TLR2, resulting in the production of proinflammatory cytokines, while hyphae stimulate production of IL-10 using a TLR2-dependent mechanism (Netea et al., 2003). Although some experimental approaches have revealed the important role of MyD88 adaptor protein in cell signaling and protective responses (Mambula et al., 2002; Bellocchio et al., 2004a), other reports claimed that MyD88 signaling and activation of NFκB are not important for fungal clearance (Marr et al., 2003; Dubordeau et al., 2006) (Table 3).

View this table:
Table 3

The role of TLRs and MyD88 adaptor protein in some experimental models of Aspergillus fumigatus infection

PRR deficiencyBiological effectReference
TLR4, CD14, TLR2TLR4 but not TLR2 recognizes hyphaeWang et al. (2001)
TLR4, TLR2TLR4 and TLR2 recognize conidia and hyphae, and induce TNF-α and MIP-2; impaired PMN influxMeier et al. (2003)
TLR4, TLR2Fungal antigens recognized by TLR4 and TLR2; enhanced phagocytosis and cytokine synthesis; activation and maturation of DCsBraedel et al. (2004)
TLR4, TLR2, TLR3, etcIndividual TLRs activate human PMNs for specialized antifungal effector functionsBellocchio et al. (2004b)
TLR4, TLR2Conidia use TLR4 and TLR2 to induce proinflammatory cytokines; hyphae use only TLR2 to produce IL-10Netea et al. (2003)
MyD88, TLR2, TLR4MyD88 and TLR2 required for optimal signaling responses of cellsMambula et al. (2002)
MyD88, TLR2, TLR4TLR2, TLR4 and MyD88 signaling dispensable for fungal clearanceDubordeau et al. (2006)
MyD88Normal phagocytosis and killing of conidia; normal cytokine secretionMarr et al. (2003)
MyD88 TLR4, TLR2, TLR9MyD88: high susceptibility. TLR signaling: depends on morphotypes, route of infection. Th1 response: DC-MyD-dependentBellocchio et al. (2004a)

As a whole, several reports have illustrated the use of different TLRs by a single fungal species, resulting in diverse biological activities. Studies with purified components of fungal cell walls revealed the major PRR and signaling pathways used by host cells to recognize fungal PAMPs; however, this picture is less clear when whole pathogens are used to infect normal or PRR-deficient hosts. The final activation, although influenced by the missing receptor, is mediated by the remaining PRRs, which can compensate or not the deficient receptor.

PRR and Paracoccidioides brasiliensis infection

Paracoccidioides brasiliensis, the causative agent of human paracoccidioidomycosis, is primarily a respiratory pathogen, infecting the host through inhalation of airborne spores. The great majority of infected subjects develop an asymptomatic pulmonary infection, although some individuals present clinical manifestations, which give rise to the localized (benign) or disseminated (severe) forms of the disease. Clinical and experimental evidence indicates that cell-mediated immunity plays a significant role in host defense against P. brasiliensis infection, whereas high levels of specific antibodies are associated with the most severe forms of the disease (Borges-Walmsley et al., 2002; Calich et al., 2008). Our laboratory developed a murine pulmonary model of infection in which A/Sn mice developed a chronic benign, pulmonary-restricted paracoccidioidomycosis whereas B10.A mice developed a progressive disseminated disease. The main immunological characteristics of this model are described elsewhere (Calich & Blotta, 2005).

Although the importance of innate immunity in resistance to fungal infection is well recognized (Roeder et al., 2004; Romani, 2004), the molecular mechanisms underlying recognition of P. brasiliensis by innate immune cells are not well known (Calich & Blotta, 2005; Calich et al., 2008). C3b receptors (CR3, CD11b/CD18) are membrane integrins that recognize iC3b of the complement system, several β-glucans and other cell-wall components expressing high mannose content (Brown & Gordon, 2003). We were the first to demonstrate that P. brasiliensis interaction with peritoneal macrophages was enhanced by iC3b opsonization of yeast cells (Calich et al., 1979). Studying murine macrophages, Jimenez (2006) verified that CR3 and MR were involved in the phagocytosis of P. brasiliensis spores (conidia). In addition, gp43, the immunodominant antigen of P. brasiliensis, was shown to bind to MR and to inhibit the phagocytic and fungicidal ability of peritoneal macrophages from resistant and susceptible mice (Popi et al., 2002). This finding led the authors to postulate the expression or secretion of gp43 as an escape mechanism of fungal cells.

TLR4 and P. brasiliensis infection

Comparative studies of in vivo susceptibility of different mouse strains to P. brasiliensis intraperitoneal infection led us to verify that TLR4-deficient (C3H/HeJ) mice were more resistant than TLR4 normal (C3H/HePas) animals (Calich et al., 1985). Recent findings from our laboratory (F.V. Loures, unpublished data) demonstrated that, compared with the normal strain, macrophages from TLR4-deficient mice had a lower phagocytic ability, which appears to influence the decreased number of viable P. brasiliensis yeasts recovered after cocultivation for a 72 h period. Deficient macrophages secrete lower levels of NO, IL-12 and monocyte chemoattractant protein-1 (MCP-1) but produced equivalent amounts of TNF-α. In contrast, IL-10 was synthesized in higher amounts by TLR4-deficient macrophages. Consistent with in vitro results, 96 h after in vivo pulmonary infection, TLR4-deficient mice presented decreased fungal loads in the lungs associated with lower levels of NO and proinflammatory cytokines [IL-12, and granulocyte macrophage colony-stimulating factor (GM-CSF)]. Similar results were observed at week 11 after infection of TLR4-mutant mice: decreased CFU counts associated with low IL-12 levels but high IFN-γ secretion. Paralleling its mild infection, the deficient strain secreted low levels of IgG1, IgG2b and IgM P. brasiliensis-specific isotypes (F.V. Loures, unpublished data).

Cytospin preparations of lung-infiltrating leukocytes at week 2 of infection showed an increased number of PMN neutrophils associated with decreased numbers of lymphocytes and monocytes. Diminished expression of CD25+, CD86+ and CD86+IAk+ cells were also observed by fluorescence-activated cell sorter (FACS) analysis of lung lymphocytes. In addition, by week 2 of infection no differences in the lymphoproliferative activity of TLR4-normal and TLR4-deficient spleen cells were detected. Although differences in CFU counts and synthesis of some inflammatory mediators occur in TLR4-deficient mice, such differences were not sufficient to alter their mean survival times (Loures; F.V. Loures & V.L.G. Calich, unpublished data).

TLR2 and P. brasiliensis infection

We have also comparatively analyzed P. brasiliensis infection of TLR2-normal (WT) and TLR2-gene knockout (KO) mice in a C57BL/6 background (F.V. Loures & V.L.G. Calich, unpublished data). In vitro infection of KO macrophages resulted in lower phagocytic indexes, decreased recovery of viable yeasts after 72 h of cocultivation and low levels of NO and MCP-1 in culture supernatants. Compared with WT mice, 48 h after infection KO mice presented diminished pulmonary fungal loads and NO, but these findings were not associated with differences in the levels of pro- and anti-inflammatory cytokines. Analysis of bronchoalveolar lavage fluids obtained 72 h after i.t. infection showed an increased influx of neutrophils to the airspaces of TLR2 KO mice in comparison with WT animals. Cytospin preparations of lung-infiltrating leukocytes at weeks 2 and 4 of infection showed a decreased proportion of macrophages but an increased number of PMN neutrophils and lymphocytes. In addition, a decreased number of IAk+ macrophages and CD4+ CD25+ T cells was detected in TLR2 KO mice. Further studies are needed, however, to characterize this T-cell subset more completely, which can exert effector or regulatory functions. Indeed, in a previous report Netea (2004) showed that TLR2 KO mice are less susceptible to Candida albicans infection due to the decreased presence of regulatory T cells and a more efficient fungal-specific immunity. At week 11 of infection TLR2-deficient and normal mice presented similar humoral immunity but the former strain presented increased fungal burden in the lungs. Despite this difference, both mouse strains exhibited equivalent mortality rates.

MyD88 adaptor molecule and P. brasiliensis infection

When macrophages were in vitro infected with P. brasiliensis yeasts for 72 h, an increased number of viable fungi was recovered from MyD88 KO macrophages in comparison with normal cells. This diminished fungicidal ability paralleled a decreased synthesis of NO and IL-12. The early in vivo infection reproduced the in vitro findings: higher fungal loads were found in MyD88 KO mice associated with lower levels of pulmonary NO and IL-12. This appears to indicate that absence of MyD88 molecule causes profound effects in cell activation, resulting in more severe infection. Mortality studies confirmed the higher susceptibility of MyD88 KO mice to P. brasiliensis infection, as their mean survival time was significantly lower than that of WT controls (F.V. Loures & V.L.G. Calich, unpublished data).

Concluding remarks

An increasing number of reports document the primary importance of innate immunity not only by providing the first line of defense against invading pathogens but also by controlling essential mechanisms that induce and regulate adaptative immunity. Our results on the role of TLRs in paracoccidioidomycosis suggest P. brasiliensis yeasts use TLR2 and TLR4 to gain entry into macrophages and infect mammalian hosts. Indeed, P. brasiliensis yeasts appear to be recognized by TLR2 and TLR4, resulting in increased phagocytic ability, NO secretion and fungal infection of macrophages. These data appear to be paradoxical, but the killing activity usually associated with NO secretion was not able to reduce the fungal growth provided by the presence of TLRs. Thus, interaction with TLRs could be considered a pathogenicity mechanism of P. brasiliensis, which would use host receptors of innate immunity (TLR2 and TLR4) to infect cells and to guarantee its own multiplication.

The in vivo infection of TLR-deficient mice resulted in decreased fungal burdens, again suggesting that TLRs are used by P. brasiliensis yeasts to infect hosts. The opposite result, however, was seen with MyD88-deficient macrophages and mice as more severe infections were observed probably due to the intact fungal recognition mediated by the expression of normal PRR but impaired ability of cell activation resulting in diminished fungicidal ability.

As a whole, TLR deficiency caused less severe infections associated with altered secretion of NO, cytokines and chemokines, resulting in altered cellular influx to the site of infection. In both PRR-deficient strains, the lung inflammatory infiltrate was composed of a diminished number of macrophages associated with an increased presence of PMN neutrophils. The phagocytic and killing abilities of the latter cells perhaps contribute to the decreased fungal inoculum at the site of infection. Furthermore, the low level of MCP-1 was parallel to the decreased number of lung-infiltrating monocytes.

An interesting observation was the decreased number of CD4+CD25+ T cells in the lungs of TLR2 KO mice, although an equivalent situation was not observed in TLR4-deficient animals.

Mortality studies have shown that TLR deficiency was not able to change the late course of infection as no differences were observed between TLR-deficient and TLR-normal mice. Compensatory mechanisms appear to abolish the immunological differences caused by PRR deficiencies. The same was not true for MyD88 deficiency, which causes higher mortality of infected mice.

In summary, our studies suggest that MyD88 deficiency is more important than TLR2 or TLR4 deficiency and P. brasiliensis yeasts appear to use TLRs as a virulence mechanism, which facilitates the access of fungal cells into murine macrophages. Despite their TLR-mediated activation, macrophages are not able to control fungal growth, both in vitro and in vivo. However, the final balance between fungal growth and activation of the immune system appears to control disease outcome as the low fungal loads and impaired immunity of TLR-deficient mice and the high fungal burdens and enhanced immunity of normal-TLR mice result in equivalent survival times. Furthermore, our studies have also suggested that, besides TLR, other PRRs play a role in the host immune response against by P. brasiliensis infection.

In conclusion, studies aimed to characterize the role of TLRs in fungal infections are firmly demonstrating their important participation in the effector and regulatory mechanisms of innate and adaptative immunity against these pathogens. TLR2, TLR4 and the adaptor molecule MyD88 appear to control the phagocytic rates, cell migration and activation, cytokine and chemokine secretion as well as the expression of costimulatory molecules that affect dendritic cell activation and their competence as antigen-presenting cell (APC) to naïve T cells. The interaction between TLR and other PRRs, in a synergistic or antagonistic way with fungal agonists, can result in different effector (Th1, Th2 and Th17) and regulatory responses (Treg), which ultimately determine disease outcome. Despite the important information in the current literature, additional investigation is needed to characterize further the influence of TLRs in the immunopathology of fungal infections.


We are grateful to Dr Shizuo Akira for providing TLR and MyD88 KO mice. This work was supported by grants from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) e Conselho Nacional de Pesquisas (CNPq).


  • Editor: Willem van Leeuwen


View Abstract