Published online before print May 2, 2007
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Institute for Infectious Diseases and Molecular Medicine, University of Cape Town, Cape Town, South Africa
1 Correspondence: Institute for Infectious Diseases and Molecular Medicine, University of Cape Town, Anzio Road, Observatory 7925, Cape Town, South Africa. E-mail: gordon.brown{at}mweb.co.za
ABSTRACT
During fungal infection, a variety of receptors initiates immune responses, including TLR and the β-glucan receptor Dectin-1. TLR recognition of fungal ligands and subsequent signaling through the MyD88 pathway were thought to be the most important interactions required for the control of fungal infection. However, recent papers have challenged this view, highlighting the role of Dectin-1 in induction of cytokine responses and the respiratory burst. Two papers, using independently derived, Dectin-1-deficient mice, address the role of Dectin-1 in control of fungal infection. Saijo et al. [1 ] argue that Dectin-1 plays a minor role in control of Pneumocystis carinii by direct killing and that TLR-mediated cytokine production controls P. carinii and Candida albicans. By contrast, Taylor et al. [2 ] argue that Dectin-1-mediated cytokine and chemokine production, leading to efficient recruitment of inflammatory cells, is required for control of fungal infection. In this review, we argue that collaborative responses induced during infection may partially explain these apparently contradictory results. We propose that Dectin-1 is the first of many pattern recognition receptors that can mediate their own signaling, as well as synergize with TLR to initiate specific responses to infectious agents.
Key Words: Toll-like receptor • Candida albicans • Pneumocystis carinii • Aspergillus fumigatus
INTRODUCTION
During fungal infection, a variety of receptors initiates immune responses, including the mannose receptor, complement receptor 3 (CR3), TLRs, and the β-glucan receptor (βGR), Dectin-1 [3 ]. TLR recognition of fungal ligands such as phospholipomannan and subsequent signaling through the MyD88 pathway were thought to be the most important interactions required for the control of fungal infection [4 , 5 ]. However, recent papers have challenged this view. In this review, we will discuss how Dectin-1 contributes to controlling fungal infection, highlighting the collaborative responses induced during infection and the apparent contradictory results in the recent literature. Last, we propose that Dectin-1 is the first of many pattern recognition receptors (PRRs), which mediate their own signaling and can synergize with TLRs to initiate specific responses to infectious agents.
DECTIN-1 DISCOVERY, LIGANDS, AND EXPRESSION
Dectin-1 was first described as a dendritic cell (DC)-specific receptor, which interacted with an endogenous ligand expressed on T cells [6 ]. By screening a macrophage cDNA expression library, Dectin-1 was shown to bind the β-glucan-rich particle zymosan as well as live Candida albicans [7 ]. Dectin-1 is a nonclassical, C-type, lectin-like receptor—a family of receptors, which bind predominantly protein ligands—so it came as a surprise that it bound β-glucan carbohydrate polymers. Subsequent studies identified the ligand as β-1,3-glucan, which makes up to 50% of fungal cell walls [8 ]. These glucans are not readily exposed in hyphae but are present at bud scars and on the surface of C. albicans conidia and Aspergillus fumigatus-maturing conidia, allowing direct interaction with Dectin-1 on the cell surface or in phagosomal compartments [9 10 11 ]. Although initially described as expressed only on DC, mouse Dectin-1 is highly expressed on neutrophils, alveolar macrophages, and inflammatory macrophages, with lower expression on resident macrophages, DC, and some T cells [12 ]. Human Dectin-1 (or the βGR) is expressed additionally on B cells, although the functional significance of this remains to be explored [13 ].
Fungal β-1,3-glucans can be recognized by a number of receptors apart from Dectin-1, such as CR3, lactosylceramide, and scavenger receptors [3 ]. Blocking experiments using Dectin-1 antibody and low molecular mass, soluble β-1,3-glucans indicate, however, that Dectin-1 is the major βGR on myeloid cells [14 ]. Moreover, blockade of Dectin-1 abrogated production of TNF by macrophages. These data were confirmed using the Dectin-1-deficient mouse: Binding of unopsonized zymosan and TNF cytokine production by macrophages was defective in vitro [2 ]. Dectin-1 is therefore the major βGR on myeloid cells and mediates the biological effects of β-glucans.
SIGNALING THROUGH DECTIN-1, LEADING TO NF-
B ACTIVATION AND PHAGOCYTOSIS
What distinguishes Dectin-1 from other non-TLR PRRs is the presence of a functional, tyrosine-based, activation-like motif in its cytoplasmic tail [14 15 16 ]. Antigen receptors contain multiple activation motifs comprising two YXXL/I sequences with a relatively defined spacing between them [17 ]. Receptor engagement leads to tyrosine phosphorylation by Src family kinases, and the phosphorylated sequences provide docking sites for Syk kinases by interacting with the two Src homology 2 (SH2) domains in Syk. The spacing between the YXXL sequences is important to engage both SH2 domains of Syk family kinases, thus contributing to enzymatic activation [18 ]. However, this is not the case with Dectin-1, where only the membrane-proximal YXXL sequence is required to initiate responses [14 15 16 ]. This raises the intruguing possibility that the two SH2 domains of Syk bridge separate Dectin-1 monomers, thereby contributing to Syk activation in a novel manner [16 , 19 ].
Following ligation of Dectin-1, Syk kinase is activated and phosphorylates a number of substrates, such as the adaptor SH2 domain-containing leukocyte protein of 76 kDa (SLP-76) in macrophages (unpublished data). Phosphorylated SLP-76 is known to recruit phospholipase C
(PLC), contributing to its activation and production of inositol trisphosphate and diacylglycerol (DAG) [17
]. Syk also mediates activation of the PI-3K pathway, as measured by phosphorylation of protein kinase B (PKB), a downstream substrate of PI-3K (unpublished data). During antigen receptor signaling, DAG or PI-3K products contribute to the activation of PKC isoforms. To date, it is not clear which PKC isoforms are activated during Dectin-1 signaling, but inhibition of PKC enzymes blocks phagocytosis, indicating that at least one PKC isoform is activated [20
].
Specific PKC isoforms phosphorylate the transmembrane adaptor caspase recruitment domain 11 (CARD11)/CARD-containing membrane-associated guanylate kinase protein-1 during antigen receptor signaling, thereby inducing a conformational change leading to recruitment of the downstream effector molecules BCL10, mucosa-associated lymphoid tissue 1 (MALT1), and TNF receptor-associated factor 6 (TRAF6) [21
]. The TRAF6 complex couples to the IB kinases, leading to phosphorylation and degradation of IB and subsequent NF-B activation. This cascade is also triggered during Dectin-1 signaling (Fig. 1
). However, instead of using CARD11, there appears to be specificity for CARD9 following Dectin-1 ligation [22
]. It is not clear whether CARD9 is a general substitute for CARD11 in myeloid cells and is also used by other activating receptors to activate NF-B in these cells. However, it is clear that the cascades leading to NF-B activation following Dectin-1 and antigen receptor ligation are similar, differing in the manner of Syk family kinase activation and use of CARD9 or CARD11 (Fig. 1)
.
 View larger version (29K): [in a new window] |
Figure 1. Dectin-1 and TLR2 signaling pathways leading to NF-B activation and cytokine production. Dectin-1 uses Syk kinase and the adaptor CARD9, whereas TLR2 uses the adaptors MyD88 and MyD88 adaptor-like (MAL) and IL-1 receptor-associated kinase (IRAK) kinases to couple to the TRAF6 complex and to NF-B activation. Zymosan stimulation of DC leads to IL-10 production via the Dectin-1-Syk pathway alone, whereas IL-12p40 is induced by the TLR2-MyD88 pathway. Dectin-1 and TLR2 collaborate, possibly by signal integration, to induce TNF and IL-12p40 in macrophages and IL-2 in DC.
|
Rs, and the same was expected to be the case with Dectin-1-mediated phagocytosis. It is surprising, however, that Syk is not required for Dectin-1-mediated phagocytosis in macrophages and is largely dispensible in DC [16
, 20
, 23
]. However, the cytoplasmic, activation-like motif in Dectin-1 is still required for phagocytosis, as are PKC isoforms and the exchange factors RAC1 and CDC42, which regulate actin polymerization [20
]. A caveat in interpreting data about Dectin-1 is that many different ligands have been used. In our hands, highly purified, soluble β-glucans, which vary as much as 20-fold in molecular mass, do not stimulate cytokine responses through Dectin-1, whereas β-glucan-rich particles such as zymosan, curdlan, and highly purified particulate β-glucan from Saccharomyces cervisiae are stimulatory [14 , 24 ]. The reason for this may lie in the ability of particulate ligands to separate relatively small Dectin-1 molecules from larger inhibitory phosphatases such as CD45 on the cell surface during stimulation. Such size-based segregation would allow stimulatory signals to continue uninhibited, as proposed in the "kinetic segregation model" [25 ]. In this review, we therefore place greater emphasis on Dectin-1 responses induced by particulate ligands, which we believe more closely mimick stimulation with fungal conidia.
RESPONSES INDUCED BY DECTIN-1
Control of fungal infection occurs by phagocytosis/killing by cells of the innate immune system, predominantly neutrophils and macrophages, and the generation of protective Th1 responses driven by initial IL-12 production by DC [26 ]. One mechanism whereby phagocytes kill phagocytosed conidia is by induction of the respiratory burst, which is dependent on Syk in macrophages [23 ] but only partially dependent on Dectin-1 in macrophages and neutrophils [2 ]. Although Dectin-1 triggers the respiratory burst, there are clearly other receptors, which also trigger the respiratory burst efficiently during fungal infection and compensate in the absence of Dectin-1.
Given that Dectin-1 binds to fungal β-glucans and induces NF-B, we expected that IL-12p40 would be induced strongly following Dectin-1 ligation. However, the major cytokine induced in a Dectin-1- and Syk-dependent manner in DC is the anti-inflammatory IL-10 [16
]. Although IL-12p40 is induced in DC during stimulation with zymosan, IL-12p40 production is largely dependent on the TLR-MyD88 signaling pathway and not on the Dectin-1-Syk pathway in that system [16
]. Although there appears to be some redundancy in IL-12p40 production, these data suggest that TLR are indeed the most important for control of fungal infection. This interpretation is supported further by studies using C. albicans with defects in cell wall components known to be recognized by TLRs [5
]. However, a number of recent findings argue against this view. Production of early proinflammatory cytokines and chemokines such as TNF and MIP-2 require Dectin-1 and TLR2 [11
, 14
, 15
]. It is surprising that defects in the TLR-MyD88 pathway abolish Dectin-1-mediated production of TNF following stimulation with zymosan, and Dectin-1 deficiency severely diminishes the TNF response to zymosan and C. albicans conidia [2
, 14
]. The (inefficient) IL-12p40 production in macrophages also requires engagement of Dectin-1 and TLR2, again indicative of receptor collaboration or synergy [15
]. Although the mechanisms underlying receptor collaboration remain obscure, it is feasable that Dectin-1 and TLR pathways integrate at the level of TRAF6, a common component in both pathways, to enhance NF-B-dependent responses (Fig. 1)
. Such a mechanism may allow for redundancy in IL-12p40 production by Dectin-1 and TLR2, which could explain why the TLR-MyD88 pathway is not absolutely essential for the development of a protective Th1 response during A. fumigatus infection [27
]. It is more important that the debate over which receptors are most important to control fungal and other infections should be seen in light of receptor collaboration and not receptors functioning in isolation.
DECTIN-1 FUNCTION IN CONTROL OF A. FUMIGATUS, C. ALBICANS, AND PNEUMOCYSTIS CARINII
The first evidence for a role of Dectin-1 in fungal infection in vivo used blockade of Dectin-1 during intratracheal A. fumigatus infection. Alveolar macrophage inflammatory responses, including production of TNF and MIP-2, were reduced by a number of blocking agents and correlated with reduced inflammatory cell recruitment and modestly increased fungal burden [11
]. Production of IFN-, which is indicative of protective Th1 responses, was dependent on Dectin-1. This suggests that Dectin-1 functions to initiate innate and adaptive responses and that both arms of the immune system contribute to control infection in this system.
In more recent study, we draw much the same conclusions using Dectin-1-deficient mice on the 129/Sv background [2
]. Using a C. albicans i.v. infection model, we demonstrate that Dectin-1-deficient mice are more susceptible than wild-type mice. To dissect early innate immune responses, we then used a peritoneal infection model. Early activation of macrophages, measured by egress of resident macrophages, was defective in Dectin-1-deficient mice. This defect correlated with defective cytokine and chemokine responses including production of TNF and MIP-1
, as well as defective recruitment of inflammatory cells to the peritoneal cavity. It is surprising that direct killing by macrophages and neutrophils of unopsonized or opsonized C. albicans in vitro was only reduced modestly in Dectin-1-deficient mice. This suggests that differences in control of fungal infection between wild-type and deficient mice may not be explained solely by a lack of direct killing by innate immune cells in deficient mice. Although this study does not provide new information about how Dectin-1 couples to adaptive immunity, it compliments the earlier A. fumigatus study in demonstrating that Dectin-1 is required for optimal activation of innate immune cells, leading to efficient cytokine production and subsequent recruitment of inflammatory cells to the site of infection. These studies are supported further by the finding that CARD9, which couples Dectin-1 to NF-B activation and hence, cytokine production (in collaboration with TLR), is required for control of C. albicans in C57BL/6 mice [22
]. Taken together, these three papers provide a compelling argument that collaborative responses of Dectin-1 and TLR are required for control of fungal infection by the collaborative induction of cytokines and chemokines.
Iwakura and colleagues [1 ], using Dectin-1-deficient mice, derived independently on the C57BL/6 background, provide a different perspective about the function of Dectin-1 in vivo. Using an intranasal P. carinii infection system, these authors argue that Dectin-1 plays a minor role in P. carinii infection. Although deficient mice demonstrated a transiently higher fungal burden than wild-type mice, there were no differences in fungal burden at later time-points or differences in survival. Early differences in fungal burden correlated with Dectin-1-dependent induction of the respiratory burst, complimenting a previous report [28 ]. TNF, IL-12, or IL-10 production in response to P. carinii was not mediated through Dectin-1 but instead, through the TLR-MyD88 pathway [1 ]. The authors thus conclude that Dectin-1 plays an early, minor role by direct killing of P. carinii and that cytokine production leading to protective responses is mediated solely through TLR.
These authors also determined whether Dectin-1 controls C. albicans infection, with the surprising finding that there were no differences in survival after i.v. challenge or in fungal burden after intratracheal challenge in their system [1 ]. Again, TNF, IL-12, and IL-10 production was not mediated through Dectin-1 but through the TLR-MyD88 pathway. The respiratory burst in response to C. albicans was also not dependent on Dectin-1. All three of these findings contradict directly those in the report by Taylor et al. [2 ]. An obvious difference between the two studies about Dectin-1-deficient mice is the background of the mice. However, CARD9-deficient mice, also on the C57BL/6 background, are similarly susceptible to C. albicans infection [22 ]. Another potential difference may be in the level of β-glucan exposure in the different strains of C. albicans that were used and the way they were cultured. However, it appears that resting conidia of C. albicans and A. fumigatus has different levels of β-glucan exposure, and yet, Dectin-1 is required for control of both in our hands [2 , 11 ].
As mentioned previously, deficiency of MyD88 not only prevents signaling through many TLR but can also block Dectin-1-mediated, collaborative responses. By simply characterizing responses in isolation as MyD88- or Dectin-1/Syk-dependent may be oversimplifying how Dectin-1 functions during fungal infection, by failing to account for collaboration between Dectin-1 and TLR. For example, the documented production of MIP-2 chemokine during P. carinii infection is dependent on Dectin-1 [28 ]. However, like TNF during stimulation with C. albicans, such production is enhanced by coligation of TLR. Such collaborative responses may provide a mechanism whereby Dectin-1 contributes to inflammatory cell recruitment during P. carinii and C. albicans infection, which would not be detected in MyD88-deficient mice, and is therefore attributed mistakenly as dependent solely on TLR. Although the differences between these studies may be explained partially by the need for collaborative responses to control fungal infection, the reasons for the differences in susceptibility of the two Dectin-1-deficient strains to C. albicans require further investigation.
RECEPTORS WITH STRUCTURAL OR FUNCTIONAL SIMILARITY TO DECTIN-1
Given the similarity of other C-type, lectin-like receptors to Dectin-1, it is likely that there will be many more non-TLR PRRs, which mediate their own signals and play roles in the control of infectious agents (Table 1 ). Examples include CLEC2, which contains an activation-like motif similar to that of Dectin-1 and has been proposed to be a receptor for HIV-1 [29 ]. Similarly, the receptor MICL is an activating receptor under certain conditions [30 ], and the receptors CLEC9A and CLECSF8 may be activating receptors, the latter presumably by coupling to a transmembrane adaptor [31 ]. Dectin-2, which despite its name, differs considerably from Dectin-1, binds high mannose structures on fungal hyphae and reportedly, associates with a transmembrane adaptor to initiate activating signals upon fungal stimulation [32 , 33 ]. Last, carcinoembryonic antigen-related cell adhesion molecule 3 (CD66d) expressed on neutrophils contains a classical activation motif and functions as an activating receptor when stimulated with certain opacity antigens expressed by Neisseria gonorrhoeae [34 ]. Whether any of these receptors also collaborate with TLR to tailor signals to specific, infectious agents remains to be tested.
|
View this table: [in a new window] |
Table 1. Established and Putative PRRs That Contain Tyrosine-Based Signaling Motifs
|
CONCLUSION AND FUTURE PESPECTIVE
The discovery of Dectin-1 has changed the way that we think about non-TLR PRRs. Although receptors such as CD14 and CD36 collaborate with TLR by presenting ligands to TLR, Dectin-1 mediates its own signaling and induces specific responses such as IL-10 production in DC and the respiratory burst in macrophages and neutrophils. It is more important that TLR and Dectin-1 collaborate to induce many responses to fungal stimuli, such as TNF and IL-12p40 production in macrophages and IL-2 production in DC. Given the similarity of other C-type, lectin-like receptors to Dectin-1, it is likely that there will be many more non-TLR PRRs, which mediate their own signals and play important roles in controlling pathogens. The field of signaling non-TLR PRRs has been opened by the discovery of Dectin-1 and is likely to yield many more surprises in the near future.
ACKNOWLEDGEMENTS
This work is supported by the Wellcome Trust (UK), National Research Foundation of South Africa, National Institutes of Health (1R01HL080317), and CANSA South Africa. G. D. B. is a Wellcome Trust Senior Fellow in Biomedical Science in South Africa.
Received December 22, 2006; revised April 5, 2007; accepted April 6, 2007.
REFERENCES
B Immunity 25,701-715[CrossRef][Medline] chain to induce innate immune responses J. Biol. Chem. 281,38854-38866This article has been cited by other articles:
 |
|
 |
|
  M. J. Robinson, F. Osorio, M. Rosas, R. P. Freitas, E. Schweighoffer, O. Gross, J. S. Verbeek, J. Ruland, V. Tybulewicz, G. D. Brown, et al. Dectin-2 is a Syk-coupled pattern recognition receptor crucial for Th17 responses to fungal infection J. Exp. Med., August 31, 2009; 206(9): 2037 - 2051. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. S. Giles, T. R. T. Dagenais, M. R. Botts, N. P. Keller, and C. M. Hull Elucidating the Pathogenesis of Spores from the Human Fungal Pathogen Cryptococcus neoformans Infect. Immun., August 1, 2009; 77(8): 3491 - 3500. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Kankkunen, J. Rintahaka, A. Aalto, M. Leino, M.-L. Majuri, H. Alenius, H. Wolff, and S. Matikainen Trichothecene Mycotoxins Activate Inflammatory Response in Human Macrophages J. Immunol., May 15, 2009; 182(10): 6418 - 6425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. K. Means, E. Mylonakis, E. Tampakakis, R. A. Colvin, E. Seung, L. Puckett, M. F. Tai, C. R. Stewart, R. Pukkila-Worley, S. E. Hickman, et al. Evolutionarily conserved recognition and innate immunity to fungal pathogens by the scavenger receptors SCARF1 and CD36 J. Exp. Med., March 16, 2009; 206(3): 637 - 653. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Lin, E. De Stanchina, X. K. Zhou, Y. She, D. Hoang, S. W. Cheung, B. Cassileth, and S. Cunningham-Rundles Maitake Beta-Glucan Enhances Umbilical Cord Blood Stem Cell Transplantation in the NOD/SCID Mouse Exp Biol Med, March 1, 2009; 234(3): 342 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Ramaprakash, T. Ito, T. J. Standiford, S. L. Kunkel, and C. M. Hogaboam Toll-Like Receptor 9 Modulates Immune Responses to Aspergillus fumigatus Conidia in Immunodeficient and Allergic Mice Infect. Immun., January 1, 2009; 77(1): 108 - 119. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. L. Chaffin Candida albicans Cell Wall Proteins Microbiol. Mol. Biol. Rev., September 1, 2008; 72(3): 495 - 544. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
