Research Article: Turning Up the Heat: Inflammasome Activation by Fungal Pathogens

Date Published: July 23, 2015

Publisher: Public Library of Science

Author(s): Aldo Henrique Tavares, Pedro Henrique Bürgel, Anamélia Lorenzetti Bocca, Joseph Heitman.

http://doi.org/10.1371/journal.ppat.1004948

Abstract

Partial Text

Since its first description in 2002 [1], the inflammasome has been implicated in the mechanisms underlying a growing number of infectious, autoimmune, and metabolic diseases [2]. Regarding infectious processes, several studies have shown the involvement of this critical component of innate immunity in the outcome of infection with nearly every class of microbe, including fungi [3]. Innate immunity is the frontline of defense against infection and relies on the ability of its main players (phagocytes and epithelial barriers) to detect conserved components of microbes or pathogen-associated molecular patterns (PAMPs). In fungi, the carbohydrate polymers of the cell wall, such as chitin, β-glucan, and mannan are the major PAMPs recognized by the host’s innate immune cells; this recognition occurs via germline-encoded receptors termed pattern recognition receptors (PRRs) [4]. In addition to PAMPs, endogenous molecules associated with damaged host cells, or damage-associated molecular patterns (DAMPs), are released during tissue injury and activate PRRs. This innate detection system includes the Toll-like receptors (TLRs), C-type lectin receptors (CLRs), RIG-I-like receptors (RLRs), NOD-like receptors (NLRs), and AIM2-like receptors (ALRs). Although the main fungal-recognition PRRs (CLRs and TLRs) are bound to the cytoplasmic membrane of innate immune cells [4], fungal sensing by PRRs located in the cytosol, such as the NLRs and ALRs, is becoming increasingly evident.

Among the inflammasomes, the NLRP3 inflammasome is the main one associated with fungal infection. In contrast to its counterparts, which only respond to a few specific PAMPs, the NLRP3 inflammasome is activated by a diverse array of unrelated triggers that include PAMPs from every class of pathogen, environmental irritants, and DAMPs. Although the precise mechanism of NLRP3 inflammasome activation is unclear, there is evidence suggesting that it is a two-step process [6]. The first, or priming, step is an NF-κB-dependent pathway that triggers expression of pro-IL-1β, pro-IL-18, and optimal NLRP3. In the second, or activation, step, assembly of the inflammasome complex leads to caspase-1 activation to promote cleavage of the immature cytokines. Priming is most frequently achieved via PRR recognition of PAMPs. In this manner, fungal PAMPs are recognized by several CLRs and TLRs that can potentially activate NF-κB [4]. However, dectin-1-dependent signaling is emerging as the key pathway involved in fungus-induced NLRP3 priming (Fig 1) [7]. In addition, this PRR is necessary to activate a caspase-8-dependent inflammasome (see below). Dectin-1, the major β-glucan receptor, uses an immunoreceptor tyrosine-based activation motif to couple itself to Syk kinase for downstream signaling to NF-κB via CARD9-Bcl10-MALT1 (CBM) scaffold, resulting in cytokine production. In addition, phagocytosis and reactive oxygen species (ROS) production result from dectin-1 engagement [8]. The dectin-1 receptor is required for the production of pro-IL-1β in murine and human myeloid cells infected with Candida albicans, Microsporum canis, and Malassezia spp. [9–11]. Consistent with these results, mice deficient in dectin-1 and orally infected with C. albicans presented significantly reduced serum IL-1β levels [12]. Notably, mice lacking dectin-1 or CARD9 and humans with mutations in these proteins are susceptible to candidiasis [12–14]. Although the direct role of dectin-1 has not been evaluated, Syk-dependent NLRP3 priming also occurs in Paracoccidioides brasiliensis-, acapsular Cryptococcus neoformans- and Aspergillus fumigatus-infected cells [15–17]. Interestingly, glucuronoxylomannan, the main capsule component of C. neoformans, inhibits Syk signaling and ultimately NLRP3 inflammasome activation, which may facilitate the intracellular parasitism of this fungus [16]. Considering that the CLRs dectin-2 and mincle share the same downstream pathway as dectin-1, further studies will be necessary to assess the role of these receptors in NLRP3 priming by fungal pathogens.

The mechanisms by which the NLRP3 inflammasome senses its vast array of activators seem to converge on several cellular disturbances that are probably nonexclusive, such as potassium (K+) efflux, calcium influx, ROS production, and the occurrence of cytosolic cathepsins derived from lysosomal disruption [6]. Although there are several controversies surrounding the actual role of each factor in inducing NLRP3 inflammasome assembly, it has been shown that fungi generally induce ROS, K+ efflux, and lysosomal rupture-dependent IL-1β release (Fig 1). ROS, which are a conserved danger signal, and K+ efflux are required by nearly all fungal species found to activate the NLRP3 inflammasome to date. Interestingly, dectin-1/Syk-mediated release of ROS was demonstrated to be critical in inflammasome activation in C. albicans-infected murine and human phagocytes [18]. Thus, Syk-coupled dectin-1 signaling is implicated in both priming and activation of the NLRP3 inflammasome. Recently, secreted aspartic protease (Sap) 2 and Sap6 from Candida were the first fungal proteins demonstrated to provide a caspase-1-dependent NLRP3 inflammasome activation signal (Fig 1) [20]. The inflammasome activation required Sap internalization via a clathrin-dependent mechanism, followed by induction of K+ efflux and ROS production. Additionally, these proteases activate, through the induction of type I IFN production, caspase-11 that cooperates with caspase-1 to maximize IL-1β maturation [21]. In view of the elevated secretion of Saps during vaginal candidiasis, studies are necessary to assess the in vivo relevance of these data.

In addition to the NLRP3 inflammasome, a NLR-independent caspase-8-dependent inflammasome and other canonical caspase-1-dependent inflammasomes are activated when stimulated with fungi (Fig 1). Gringhuis et al. [29] found that human dendritic cells stimulated with certain strains of C. albicans did not utilize caspase-1 to process IL-1β. This finding led to several experiments demonstrating a NLR-independent dectin-1/Syk-dependent inflammasome activation route, with the assembly of the CBM scaffold and processing of IL-1β mediated by recruitment of MALT-1/caspase-8 and ASC into this complex. This route was also activated by other species of Candida and different A. fumigatus strains. Moreover, the dectin-1-mediated activity, in contrast to NLRP3 inflammasome priming, did not require phagocytosis, suggesting a direct extracellular sensing mechanism. Later, Ganesan et al. [30] demonstrated that caspase-8, dectin-1, and CR3, another receptor implicated in β-glucan sensing, are necessary for IL-1β processing by murine dendritic cells infected with C. albicans. Interestingly, caspase-8 activation in these studies raises questions about Candida-induced programmed cell death pathways, as caspase-8 also has a significant role in initiating apoptosis [31].

The depletion of inflammasome components, inflammasome products, and relevant first-signal receptors has been linked to susceptibility of the host to several bacterial infections. Although the available data are scarce, it seems that this phenomenon extends to fungal infections. In C. albicans, it has been demonstrated that in a disseminated model of infection, loss of NLRP3 leads to increased mortality and an increased fungal burden in several organs, such as the kidney, lung, and liver [18,22]. van de Veerdonk et al. [35] further showed an essential role for ASC and caspase-1 in regulating adaptive antifungal immune responses and host survival during disseminated candidiasis through the induction of Th1 and Th17 development. NLRP3 inflammasome components were also studied in a model of mucosal Candida infection, with NLRP3-, ASC-, or caspase-1-deficient mice being more susceptible to invasive disease, given that they presented a higher fungal burden in the kidneys and the digestive system [9]. Furthermore, loss of the IL-1 receptor (IL-1R) and the priming-associated receptors dectin-1 and TLR-2 aggravated the infection. Using bone marrow chimeras, it was demonstrated that NLRP3 and NLRC4 have a protective but tissue-specific role in oropharyngeal candidiasis [32]. In contrast to NLRP3, whose activation is protective in both the hematopoietic and the stromal compartments, NLRC4 is essential in only the stromal compartment, where its activity is necessary for the induction of neutrophil influx into infected tissues and the avoidance of fungal dissemination, particularly early in infection. Nevertheless, these results should be interpreted with caution since it is also possible that mice lacking NLRC4 possess a diverse microbiota from other controls. In fact, microbiota differences are observed in inflammasome knockout mice [36], resulting in microbiota-driven differences in phenotypes (i.e., microbiota would track with the stromal compartment in bone marrow transplants). Recently, using an invasive pulmonary aspergillosis model, Karki et al. [33] showed that redundant activation of the NLRP3/AIM2 platform is essential for host protection, as it significantly limits dissemination of A. fumigatus hyphae from inflammatory foci. In addition, AIM2 and NLRP3 activity in both the hematopoietic and the stromal compartments is required to protect the host against aspergillosis. It was also shown that loss of key inflammasome-related cytokines (e.g., IL-1β and IL-18) is detrimental for the host. Regarding cryptococcosis, regardless of the route of infection (e.g., intraperitoneal or intranasal), mice lacking NLRP3 or ASC presented poorer survival compared with wildtype mice [23]. In fact, even for intranasal infection with a low virulence acapsular strain of C. neoformans, NLRP3 was required for effective lung leukocyte infiltration and fungal clearance. Finally, our group demonstrated that the presence of the IL-1R-dependent signaling, and NLRP3 is required to control the intracellular growth of P. brasiliensis within macrophages [15].

 

Source:

http://doi.org/10.1371/journal.ppat.1004948